| {"repo": "GoogleChrome/lighthouse", "instance_id": "GoogleChrome__lighthouse-5917", "base_commit": "3ad529d937a4aa5af4e68575c6780676edd23248", "problem_statement": "No domContentLoaded event in trace\nI'm using lighthouse for automatic page test, and I've encountered this warning, which makes the performance score very slow, could anyone help to figure out why is this happening?\n\n```\n~/dev/node-test/chrome lighthouse https://h5.m.taobao.com/trip/home/index.html\n ChromeLauncher Waiting for browser. +0ms\n ChromeLauncher Waiting for browser... +1ms\n ChromeLauncher Waiting for browser..... +515ms\n ChromeLauncher Waiting for browser.....✓ +4ms\n status Initializing… +770ms\n status Loading page & waiting for onload URL, Viewport, ViewportDimensions, ThemeColor, Manifest, ChromeConsoleMessages, ImageUsage, Accessibility, EventListeners, AnchorsWithNoRelNoopener, AppCacheManifest, DOMStats, JSLibraries, OptimizedImages, PasswordInputsWithPreventedPaste, ResponseCompression, TagsBlockingFirstPaint, WebSQL +346ms\n statusEnd Loading page & waiting for onload +14s\n status Retrieving trace +1ms\n status Retrieving devtoolsLog and network records +2s\n status Retrieving: URL +14ms\n status Retrieving: Viewport +1ms\n status Retrieving: ViewportDimensions +5ms\n status Retrieving: ThemeColor +6ms\n status Retrieving: Manifest +2ms\n status Retrieving: ChromeConsoleMessages +1ms\n status Retrieving: ImageUsage +2ms\n status Retrieving: Accessibility +59ms\n status Retrieving: EventListeners +262ms\n status Retrieving: AnchorsWithNoRelNoopener +230ms\n status Retrieving: AppCacheManifest +2ms\n status Retrieving: DOMStats +3ms\n status Retrieving: JSLibraries +15ms\n status Retrieving: OptimizedImages +9ms\n status Retrieving: PasswordInputsWithPreventedPaste +161ms\n status Retrieving: ResponseCompression +2ms\n status Retrieving: TagsBlockingFirstPaint +6ms\n status Retrieving: WebSQL +4ms\n status Loading page & waiting for onload ServiceWorker, Offline, StartUrl +811ms\n statusEnd Loading page & waiting for onload +230ms\n status Retrieving devtoolsLog and network records +47ms\n status Retrieving: ServiceWorker +7ms\n status Retrieving: Offline +0ms\n status Retrieving: StartUrl +34ms\n status Loading page & waiting for onload HTTPRedirect, HTMLWithoutJavaScript +333ms\n statusEnd Loading page & waiting for onload +179ms\n status Retrieving devtoolsLog and network records +0ms\n status Retrieving: HTTPRedirect +6ms\n status Retrieving: HTMLWithoutJavaScript +7ms\n status Disconnecting from browser... +5ms\n status Analyzing and running audits... +30ms\n status Evaluating: Uses HTTPS +0ms\n status Evaluating: Redirects HTTP traffic to HTTPS +8ms\n status Evaluating: Registers a service worker +1ms\n status Evaluating: Responds with a 200 when offline +0ms\n status Evaluating: Has a `<meta name=\"viewport\">` tag with `width` or `initial-scale` +1ms\n status Evaluating: Contains some content when JavaScript is not available +2ms\n status Evaluating: First meaningful paint +1ms\n status Evaluating: Page load is fast enough on 3G +125ms\n load-fast-enough-for-pwa:warn Caught exception: No domContentLoaded event in trace +34ms\n status Evaluating: Perceptual Speed Index +0ms\n status Evaluating: Screenshot Thumbnails +2s\n status Evaluating: Estimated Input Latency +187ms\n status Evaluating: No browser errors logged to the console +5ms\n status Evaluating: Keep server response times low (TTFB) +1ms\n status Evaluating: First Interactive (beta) +1ms\n first-interactive:warn Caught exception: No domContentLoaded event in trace +0ms\n status Evaluating: Consistently Interactive (beta) +0ms\n consistently-interactive:warn Caught exception: No domContentLoaded found in trace. +0ms\n status Evaluating: User Timing marks and measures +0ms\n status Evaluating: Critical Request Chains +11ms\n status Evaluating: User can be prompted to Install the Web App +3ms\n status Evaluating: Configured for a custom splash screen +1ms\n status Evaluating: Address bar matches brand colors +0ms\n status Evaluating: Manifest's `short_name` won't be truncated when displayed on homescreen +0ms\n status Evaluating: Content is sized correctly for the viewport +1ms\n status Evaluating: Uses Images with appropriate aspect ratio +0ms\n status Evaluating: Avoids deprecated APIs +2ms\n status Evaluating: Site works cross-browser +0ms\n status Evaluating: Page transitions don't feel like they block on the network +0ms\n status Evaluating: Each page has a URL +0ms\n status Evaluating: `[accesskey]` values are unique. +0ms\n status Evaluating: `[aria-*]` attributes match their roles. +1ms\n status Evaluating: `[role]`s have all required `[aria-*]` attributes. +0ms\n status Evaluating: Elements with `[role]` that require specific children `[role]`s, are present. +0ms\n status Evaluating: `[role]`s are contained by their required parent element. +0ms\n status Evaluating: `[role]` values are valid. +0ms\n status Evaluating: `[aria-*]` attributes have valid values. +0ms\n status Evaluating: `[aria-*]` attributes are valid and not misspelled. +0ms\n status Evaluating: `<audio>` elements contain a `<track>` element with `[kind=\"captions\"]`. +0ms\n status Evaluating: Buttons have an accessible name. +0ms\n status Evaluating: The page contains a heading, skip link, or landmark region. +1ms\n status Evaluating: Background and foreground colors have a sufficient contrast ratio. +0ms\n status Evaluating: `<dl>`'s contain only properly-ordered `<dt>` and `<dd>` groups, `<script>` or `<template>` elements. +0ms\n status Evaluating: Definition list items are wrapped in `<dl>` elements. +0ms\n status Evaluating: Document has a `<title>` element. +0ms\n status Evaluating: `[id]` attributes on the page are unique. +0ms\n status Evaluating: `<frame>` or `<iframe>` elements have a title. +0ms\n status Evaluating: `<html>` element has a `[lang]` attribute. +0ms\n status Evaluating: `<html>` element has a valid value for its `[lang]` attribute. +0ms\n status Evaluating: Image elements have `[alt]` attributes. +0ms\n status Evaluating: `<input type=\"image\">` elements have `[alt]` text. +1ms\n status Evaluating: Form elements have associated labels. +0ms\n status Evaluating: Presentational `<table>` elements avoid using `<th>`, `<caption>` or the `[summary]` attribute. +0ms\n status Evaluating: Links have a discernible name. +0ms\n status Evaluating: Lists contain only `<li>` elements and script supporting elements (`<script>` and `<template>`). +0ms\n status Evaluating: List items (`<li>`) are contained within `<ul>` or `<ol>` parent elements. +0ms\n status Evaluating: The document does not use `<meta http-equiv=\"refresh\">`. +0ms\n status Evaluating: `[user-scalable=\"no\"]` is not used in the `<meta name=\"viewport\">` element and the `[maximum-scale]` attribute is not less than 5. +1ms\n status Evaluating: `<object>` elements have `[alt]` text. +0ms\n status Evaluating: No element has a `[tabindex]` value greater than 0. +0ms\n status Evaluating: Cells in a `<table>` element that use the `[headers]` attribute only refer to other cells of that same table. +0ms\n status Evaluating: `<th>` elements and elements with `[role=\"columnheader\"/\"rowheader\"]` have data cells they describe. +0ms\n status Evaluating: `[lang]` attributes have a valid value. +0ms\n status Evaluating: `<video>` elements contain a `<track>` element with `[kind=\"captions\"]`. +0ms\n status Evaluating: `<video>` elements contain a `<track>` element with `[kind=\"description\"]`. +0ms\n status Evaluating: Avoids enormous network payloads +0ms\n status Evaluating: Offscreen images +7ms\n offscreen-images:warn Caught exception: No domContentLoaded event in trace +24ms\n status Evaluating: Serve images as WebP +0ms\n status Evaluating: Optimize images +1ms\n status Evaluating: Enable text compression +1ms\n status Evaluating: Properly size images +0ms\n status Evaluating: Avoids Application Cache +17ms\n status Evaluating: Avoids an excessive DOM size +0ms\n status Evaluating: Opens external anchors using `rel=\"noopener\"` +2ms\n status Evaluating: Avoids requesting the geolocation permission on page load +0ms\n status Evaluating: Reduce render-blocking stylesheets +1ms\n status Evaluating: Avoids `document.write()` +0ms\n status Evaluating: Avoids Mutation Events in its own scripts +1ms\n status Evaluating: Avoids front-end JavaScript libraries with known security vulnerabilities +71ms\n status Evaluating: Avoids WebSQL DB +0ms\n status Evaluating: Avoids requesting the notification permission on page load +1ms\n status Evaluating: Allows users to paste into password fields +0ms\n status Evaluating: Reduce render-blocking scripts +0ms\n status Evaluating: Uses HTTP/2 for its own resources +0ms\n status Evaluating: Uses passive listeners to improve scrolling performance +2ms\n status Generating results... +1ms\n Printer domhtml output written to /Users/dickeylth/dev/node-test/chrome/h5.m.taobao.com_2017-11-17_17-15-36.report.html +20ms\n CLI Protip: Run lighthouse with `--view` to immediately open the HTML report in your browser +0ms\n ChromeLauncher Killing Chrome instance 93287 +1ms\n```\n\nYou could see there's\n\n```\n load-fast-enough-for-pwa:warn Caught exception: No domContentLoaded event in trace +34ms\n first-interactive:warn Caught exception: No domContentLoaded event in trace +0ms\n consistently-interactive:warn Caught exception: No domContentLoaded found in trace. +0ms\n offscreen-images:warn Caught exception: No domContentLoaded event in trace +24ms\n```\n\nand the score is really slow:\n\n\n\n\nwhich could not reflect the real performance in my opinion.\n\nLooking forward to any help! Thank you.\n\n## Relevant pre-fix discussion\n\n### Issue #3838\n\n#### Comment at 2018-02-12T02:43:28Z\n\nHere is the cause, browser trys to open anther url before all tracing events are completed. after opening, the trace data is polluted\n\n\n#### Comment at 2018-03-30T12:30:00Z\n\nI've got the same issue when original URL shows an authentication screen then redirects to authentication service then redirects back. \n\nSo, there's 2 redirects and 2 trace recordings:\n1. origin loads authentication screen, then redirects to authentication service\n2. (redirect 1) authentication service validates the request and sends credentials to load the origin\n3. origin starts to load\n\nIssue in Lighthouse 2.9.3 is `Something went wrong with recording the trace over your page load. Please run Lighthouse again. (NO_DCL)` \n\n", "created_at": "2018-08-28T00:26:14Z", "version": "v3.0.3", "environment_setup_commit": "3ad529d937a4aa5af4e68575c6780676edd23248", "image_assets": "{\"problem_statement\": [\"assets/GoogleChrome__lighthouse-5917/001.png\", \"assets/GoogleChrome__lighthouse-5917/002.png\", \"assets/GoogleChrome__lighthouse-5917/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "PrismJS/prism", "instance_id": "PrismJS__prism-2509", "base_commit": "bf4f323391d546d4b2e1a388f05c512a27491e9c", "problem_statement": "Fix hello-document support in shell-session language\n**Information:**\n- Prism version: 1.20.0\n- Plugins: none\n- Environment: Browser\n\nDoes the latest version of Prism from the [download page](https://prismjs.com/download.html) also have this issue?\n\nyes\n\n**Description**\n\nFirst, thanks so much for Prism.\n\nPutting together the regular expressions to pull this off must be painstaking!\n\nConsider the following code block using the `shell-session` language.\n\n```\n$ cp /etc/ipsec.secrets /etc/ipsec.secrets.backup\n\n$ cat << \"EOF\" > /etc/ipsec.secrets\n: RSA vpn-server-a.key\nEOF\n\n$ cat << \"EOF\" > /etc/ipsec.secrets\n: RSA vpn-server-a.key\n# : RSA vpn-server-b.key\nEOF\n```\n\nIssues:\n\n- Everything after `cat` is part of the output token. I believe the above code block should render three command tokens with no output.\n\n- Lines that start with `#` (or `$`) inside hello-documents become new commands. I believe commands inside of hello-documents should be ignored as they are part of a parent command.\n\nSee following screenshot.\n\n\n\n## Relevant pre-fix discussion\n\n### Issue #2503\n\n#### Comment at 2020-08-06T17:11:55Z\n\n@sunknudsen You mean like this, right?\n\n\n\nThe problem is that our here doc strings assumed that there's nothing after the identifier.\n\nIs it correct to assume that every after the identifier (so e.g. `> /etc/ipsec.secrets` but not the lines below) is a command and not part of the string? I'm asking because, as you can see, bash isn't highlighting the strings correctly and I don't know the correct interpretation.\n\n#### Comment at 2020-08-06T19:42:33Z\n\n@RunDevelopment I am fairly new to here documents, but my gut feeling is only `cat` should be considered a command and everything else after it (including the delimiting identifier) a string.\n\nThis is how it is illustrated on Wikipedia (see https://en.wikipedia.org/wiki/Here_document) and how Visual Studio Code handles it in markdown codeblocks and shell scripts (coloring the delimiting identifier is a nice addition).\n\n\n\n", "created_at": "2020-08-06T21:20:46Z", "version": "v1.20.0", "environment_setup_commit": "bf4f323391d546d4b2e1a388f05c512a27491e9c", "image_assets": "{\"problem_statement\": [\"assets/PrismJS__prism-2509/001.png\", \"assets/PrismJS__prism-2509/002.png\", \"assets/PrismJS__prism-2509/003.png\", \"assets/PrismJS__prism-2509/004.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "apache/echarts", "instance_id": "apache__echarts-10173", "base_commit": "243b22f4d3c1c2e32b06517d6295067fdd076acf", "problem_statement": "Bar chart issue: Generates merged bar chart if we reduces width of zoom bar to its certain width.\nDescription:\n1. Created bar chart with following given option object:\n2. I reduced width of zoom bar to its certain width, as given in option object: \"minSpan\": 0.5 \n3. Have a look at attached screenshot.\n```js\noption = {\n xAxis: {\n type: 'category',\n data: ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']\n },\n yAxis: {\n type: 'value'\n },\n series: [{\n data: [120, 200, 150, 80, 70, 110, 130],\n type: 'bar'\n }],\n \"dataZoom\": {\n \"type\": \"slider\",\n \"show\": true,\n \"minSpan\": 0.5,\n \"filterMode\": \"empty\",\n \"start\": 87.65476305836627,\n \"end\": 94.3056847654145,\n \"xAxisIndex\": [\n 0\n ]\n },\n};\n```\n\nResult:\n\n\n\n## Relevant pre-fix discussion\n\n### Issue #9882\n\n#### Comment at 2019-01-31T17:33:44Z\n\nHere's a gif that explains the issue better:\n\n\n\n#### Comment at 2019-03-26T17:59:54Z\n\nThis also affects stacked column charts\n\n", "created_at": "2019-03-27T16:46:47Z", "version": "4.2.0-rc.2", "environment_setup_commit": "243b22f4d3c1c2e32b06517d6295067fdd076acf", "image_assets": "{\"problem_statement\": [\"assets/apache__echarts-10173/001.png\", \"assets/apache__echarts-10173/002.png\", \"assets/apache__echarts-10173/003.gif\", \"assets/apache__echarts-10173/004.gif\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "apache/echarts", "instance_id": "apache__echarts-12850", "base_commit": "6f78c0d5191378e604fc88eb07b7ffabdb69aa3e", "problem_statement": "myChart.setOption({}) 时 lines.data 丢失\n### Version\n4.8.0\n\n### Steps to reproduce\nmyChart.setOption({}) 时, lines类型的线数据会丢失, 其他类型的数据(如line)依然存在.\n这个方法的行为是不是应该统一?\n\nhttps://gallery.echartsjs.com/editor.html?c=xeqTwJtMgx&v=1\n\n- before:\n\n\n- after:\n\n\n### What is expected?\n\nmyChart.setOption({}) 要么数据全丢, 要么全保留\n\n### What is actually happening?\nlines类型的线数据丢失\n\n<!-- This issue is generated by echarts-issue-helper. DO NOT REMOVE -->\n\n## Relevant pre-fix discussion\n\n### Issue #12836\n\n#### Comment at 2020-06-22T14:58:12Z\n\n@myshzzx \n代码，在本地运行，如图所示，可以显示lines类型的线数据\n\n", "created_at": "2020-06-23T03:42:03Z", "version": "4.8.0", "environment_setup_commit": "6f78c0d5191378e604fc88eb07b7ffabdb69aa3e", "image_assets": "{\"problem_statement\": [\"assets/apache__echarts-12850/001.png\", \"assets/apache__echarts-12850/002.png\", \"assets/apache__echarts-12850/003.png\", \"assets/apache__echarts-12850/004.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "ariakit/ariakit", "instance_id": "ariakit__ariakit-4251", "base_commit": "e6d5c8e1891c22a03da72922955bcc6dbbd64eea", "problem_statement": "Menu Show / Hide Change in Behavior in 0.14.4 vs 0.14.3\n### Current behavior\n\nI have a top navigation menu system that is all Ariakit.\n \n This works perfectly in 0.14.3, but in .4, I see the menus overlap awkwardly and show multiple at a time. I see that showOnHover is set to true, but they aren't coming down properly in 0.14.4.\n\n\n\n### Expected behavior\n\nLike it did in 0.14.3\n\n### Workaround\n\nhttps://github.com/ariakit/ariakit/issues/4247#issuecomment-2506083953\n\n### Possible solutions\n\nI think this has something to do with the transition animation, which I set on my mui components to \"unset\" previously to stop this in earlier versions.\n\n\n\n## Relevant pre-fix discussion\n\n### Issue #4247\n\n#### Comment at 2024-11-23T16:31:58Z\n\nI tried raking off all of the render props (which were using Mui ListItem and Mui/Buttons) and it still fails. I suspected that a Mui transition is deep in there. There is just something going on with the way they come down in .4. I have a simple case where I loop and call menu.hide() on them and they just come down slowly and overlap.\n\n\n\n#### Comment at 2024-11-24T13:44:33Z\n\n", "created_at": "2024-11-26T17:43:46Z", "version": "@ariakit/core@0.4.13", "environment_setup_commit": "e6d5c8e1891c22a03da72922955bcc6dbbd64eea", "image_assets": "{\"problem_statement\": [\"assets/ariakit__ariakit-4251/001.png\", \"assets/ariakit__ariakit-4251/002.png\", \"assets/ariakit__ariakit-4251/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "ariakit/ariakit", "instance_id": "ariakit__ariakit-520", "base_commit": "d5e1e8f070c435acdbab92a95a938be23e5803ae", "problem_statement": "Some components don't render the passed id prop\n## 🐛 Bug report\n\n### Current behavior\n\nI used the codesandbox link from https://github.com/reakit/reakit/blob/master/README.md and tested a situation i discovered when using the `id` attribute in combination with components that use the `Hidden` component.\n\nThe moment i declare an `id`, it failes to render.\n\n### Steps to reproduce the bug\n\nProvide a repo or sandbox with the bug and describe the steps to reproduce it.\n\n1. Open sandbox: https://codesandbox.io/s/crazy-curran-2x3wc (you can use it as a base)\n2. remove/add the `id` attribute for `Hidden` or any other component depending on it.\n3. See console logs\n\n### Expected behavior\n\nShould render without failure.\n\n### Possible solutions\n\nI currently do not have an propper understanding of the new `Id` component and how it works to so\ni cannot give a possible solution.\n\n### Environment\n\nPlease, run the command below inside your project directory.\n```sh\n$ npx envinfo --system --binaries --browsers --npmPackages \"{react*,reakit*}\"\n\n System:\n OS: Linux 4.18 Ubuntu 18.10 (Cosmic Cuttlefish)\n CPU: (8) x64 Intel(R) Core(TM) i7-8565U CPU @ 1.80GHz\n Memory: 796.86 MB / 15.32 GB\n Container: Yes\n Shell: 4.4.19 - /bin/bash\n Binaries:\n Node: 12.10.0 - ~/.asdf/installs/nodejs/12.10.0/bin/node\n Yarn: 1.19.0 - ~/.asdf/installs/nodejs/12.10.0/.npm/bin/yarn\n npm: 6.10.3 - ~/.asdf/installs/nodejs/12.10.0/bin/npm\n Browsers:\n Chrome: 78.0.3904.97\n Firefox: 68.0\n npmPackages:\n react: ^16.8.6 => 16.9.0 \n react-benchmark: ^2.1.0 => 2.1.1 \n react-docgen: ^4.1.1 => 4.1.1 \n react-dom: ^16.8.6 => 16.9.0 \n react-is: ^16.8.1 => 16.9.0 \n react-known-props: ^2.4.2 => 2.4.4 \n react-test-renderer: ^16.11.0 => 16.11.0 \n react-window: ^1.8.1 => 1.8.5 \n reakit: https://pkg.csb.dev/reakit/reakit/commit/35fb93af/reakit => 1.0.0-beta.13 \n reakit-utils: 0.7.2 => 0.7.2 \n\n\n```\n\n## Relevant pre-fix discussion\n\n### Issue #518\n\n#### Comment at 2019-12-17T10:14:31Z\n\nHi @namjul! I can't reproduce it. If I set `id=\"test\"` on `Dialog`, it properly renders the component with the correct id.\n\n\n#### Comment at 2019-12-17T13:46:32Z\n\nThanks for the reproduction! The problem there is that you're setting an invalid id (`123`). IDs can't start with numbers.\n\nThe code is throwing on this line:\nhttps://github.com/reakit/reakit/blob/d5e1e8f070c435acdbab92a95a938be23e5803ae/packages/reakit/src/Hidden/__utils/useWarningIfMultiple.ts#L12\n\nThis will throw on any website:\n\n\nWe could avoid throwing that by wrapping that part of the code within `try`/`catch`.", "created_at": "", "version": "", "environment_setup_commit": "d5e1e8f070c435acdbab92a95a938be23e5803ae", "image_assets": "{\"problem_statement\": [\"assets/ariakit__ariakit-520/001.png\", \"assets/ariakit__ariakit-520/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "bpmn-io/bpmn-js", "instance_id": "bpmn-io__bpmn-js-1442", "base_commit": "f3394441ad73b84fdc23790556c4aadcdcbd4b82", "problem_statement": "Pasting Multiple Shapes Onto Connection Results In Unexpected Connections\n__Describe the Bug__\n\n\n\n\n\n__Steps to Reproduce__\n\n1. Copy two tasks\n2. Paste the tasks onto a sequence flow\n\n__Expected Behavior__\n\n<!-- A clear and concise description of what you expected to happen. -->\n\nThe [behavior](https://github.com/bpmn-io/bpmn-js/blob/develop/lib/features/modeling/behavior/DropOnFlowBehavior.js) doesn't kick in for multiple shapes __or__ it connects them in a way that _makes sense_.\n\n__Environment__\n\n - Library version: 8.3.1\n\n## Relevant pre-fix discussion\n\n### Issue #1440\n\n#### Comment at 2021-04-30T11:08:33Z\n\nRoot-Cause:\n\nThe elements are added in order from left to right, but on the same Sequenz flow. Because of that, the second element comes first, see the gif below:\n\n\n__Solution Ideas__\nI see two options:\n1. Add the Elements back-to-front.\n2. Determine the target-Flow again after each element is created\n\nHowever, both have Problems:\n1. I have made a draft implementation in the linked diagram-js PR. However, reversing the order in which the elements are added in the `CreateElementsHandler` makes assumptions about the structure of the arguments hand feels like a hack.\n__EDIT__: This will actually break the Nested elements, such as Participants/Sub-Processes. We would probably have to change the Tree-destructuring from https://github.com/bpmn-io/diagram-js/blob/master/lib/features/copy-paste/CopyPaste.js#L296\n2. The mouse event targets only one flow, so we do not have references to the next target flow. I'm not 100% sure if this approach can be achieve this without a bigger refactoring.\n\n@philippfromme , I'd be interested in your opinion on this. Would changing the order when copying the shapes (opposed to when creating the new ones) make this less hacky?\n\n#### Comment at 2021-04-30T12:08:07Z\n\nHonestly, I think we should disable this behavior when more than one shape is created. We simply cannot know what connections to create if we are creating multiple shapes at the same time. The behavior can also easily blow up completely:\n\n\n\nThe behavior was created at a time when only single elements could be created.\n\n#### Comment at 2021-04-30T12:11:28Z\n\nMoving the elements is rejected so pasting should be rejected, too:\n\n", "created_at": "", "version": "", "environment_setup_commit": "f3394441ad73b84fdc23790556c4aadcdcbd4b82", "image_assets": "{\"problem_statement\": [\"assets/bpmn-io__bpmn-js-1442/001.gif\", \"assets/bpmn-io__bpmn-js-1442/002.gif\", \"assets/bpmn-io__bpmn-js-1442/003.gif\", \"assets/bpmn-io__bpmn-js-1442/004.gif\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "chakra-ui/chakra-ui", "instance_id": "chakra-ui__chakra-ui-5651", "base_commit": "d5807a4ef73baf0aa7687a144ca4a4474abe953b", "problem_statement": "Chakra-UI + NextJS - useBreakpointValue not working with NextJS!\n### Description\n\n When I use the useBreakpointValue hook inside a component using NextJS, the application breaks and I get the error: Warning: Did not expect server HTML to contain a div in div.\n\n### Link to Reproduction\n\nhttps://codesandbox.io/s/keen-chaplygin-n7zcw\n\n### Steps to reproduce\n\n1. Go to https://codesandbox.io/s/keen-chaplygin-n7zcw\n2. Wait for the application to load and see the console\n\n### Chakra UI Version\n\n@chakra-ui/core = 0.8.0 / @chakra-ui/react = 1.8.1 \n\n### Browser\n\nGoogle Chrome 97.0.4692.99\n\n### Operating System\n\n- [ ] macOS\n- [X] Windows\n- [ ] Linux\n\n### Additional Information\n\n@chakra-ui/react: latest (1.8.1)\n@emotion/react: latest (11.7.1)\n@emotion/styled: latest (11.6.0)\n@types/node: 17.0.12\n@types/react: 17.0.38\nframer-motion: latest (6.2.3)\nnext: 12.0.9\nreact: 17.0.2\nreact-dom: 17.0.2\ntypescript: 4.5.5\n\n\n\n\n\n## Relevant pre-fix discussion\n\n### Issue #5506\n\n#### Comment at 2022-02-18T10:27:01Z\n\nIsn't solved for me in version 1.7.3 or 1.8.0. It's only on initial page load - if I navigate within the app everything is defined correctly, but when I reload the browser the values are not being applied Also if I modify the code then the hot-reload applies everything correctly.\n\nThis is some strange behaviour I've noticed, the discrepancy between pink and blue in applying the colours vs the content of the variable. Not sure why this isn't redrawing:\n\n### On browser refresh\n\n\n\n### After in-app navigation or hot reload:\n\n", "created_at": "2022-02-23T22:54:03Z", "version": "@chakra-ui/babel-plugin@1.0.2", "environment_setup_commit": "d5807a4ef73baf0aa7687a144ca4a4474abe953b", "image_assets": "{\"problem_statement\": [\"assets/chakra-ui__chakra-ui-5651/001.png\", \"assets/chakra-ui__chakra-ui-5651/002.png\", \"assets/chakra-ui__chakra-ui-5651/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "chartjs/Chart.js", "instance_id": "chartjs__Chart.js-10766", "base_commit": "7410a3c30e5b06d91ccc6a621136c6b061ee9b5d", "problem_statement": "minBarLength Stacked bar chart not working\nHello there,\n\nThis is an old annoying bug, that still exists in latest version, the minBarLength doesn't work as expected in case of Stacked bar,\nthe minBarLength doesn't apply to each dataset, but instead it applies to the whole bar, which cause overlapping, between bars.\n\nPlease fix this, it's really unusable for many usecases.\n\nYou'll find bellow a reference with the illustrated problem.\n\nThank you for your hard work.\n\n\n=======================================\n_Originally posted by @tomzag in https://github.com/chartjs/Chart.js/issues/5741#issuecomment-750866201_\n========================================\n\nThere seems to be a problem with stacked bar charts.\n\nHere for example a horizontal stacked bar chart:\n\n\nThe `minBarLength` works but the stacked version isn't respecting it and the bars overlap.\n\n## Relevant pre-fix discussion\n\n### Issue #10664\n\n#### Comment at 2022-10-04T09:00:13Z\n\n> Hi @nabenzine, would you please kindly share a minimal chart configuration to reproduce this issue? I would like to see if I can help fix this problem!\n\nHi @Kit-p, thank you for your message, I am a bit hopeless, I created a codesandbox that illustrate the problem [codesandbox](https://codesandbox.io/s/chartjs-stacked-bars-forked-7cqu7v)\n\n\n\nThank you in advance !\n\n#### Comment at 2022-10-06T20:16:39Z\n\n> @nabenzine I have made some progress in this. Here is a [codesandbox](https://codesandbox.io/s/chartjs-stacked-bars-forked-40bu32) with my version of Chart.js.\n> \n> I would appreciate if you could help play around with it and report any problem you notice, thank you!\n\nThat's amazing, thank you very much, It's seems to work fine except for label, I couldn't make them work as usual with the **chartjs-plugin-datalabels**\n\n\n\nLabels are not centered as in my initial codesandbox, maybe I am doing something wrong, Here is the \n[codesandbox](https://codesandbox.io/s/chartjs-stacked-bars-forked-zh5y6t)", "created_at": "", "version": "", "environment_setup_commit": "7410a3c30e5b06d91ccc6a621136c6b061ee9b5d", "image_assets": "{\"problem_statement\": [\"assets/chartjs__Chart.js-10766/001.png\", \"assets/chartjs__Chart.js-10766/002.png\", \"assets/chartjs__Chart.js-10766/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "chartjs/Chart.js", "instance_id": "chartjs__Chart.js-3192", "base_commit": "b33aff20919ca65e7914aa368fb99961e36e89be", "problem_statement": "Padding inside the canvas with a line chart without scale\nHello,\nHow can i make the line graph stick to the exact border of the canvas ?\nI 've made a line graph without scales and there are always a gap between the left and the bottom borders of the canvas and the beginning of the charts... I try to play with the responsive and the maintainAspectRatio options without success... here is my configuration .\n\nThx\n\n```\nnew Chart(ctx, {\n // responsive: false,\n type: 'line',\n // maintainAspectRatio:false,\n data: {\n labels: labels,\n datasets: [{\n fill:true,\n backgroundColor: \"#47afbd\",\n data: data\n }]\n },\n options: {\n defaultFontSize:10,\n animation:{\n onProgress:function(){\n if(self.$('#engagement-chart-wrap').hasClass('isLoading')){\n self.$('#engagement-chart-wrap').removeClass('isLoading');\n }\n }\n },\n legend: {\n display: false\n },\n tooltips:{\n enabled:false\n },\n elements:{\n point:{\n radius:0\n }\n },\n scales: {\n\n yAxes: [{\n\n ticks :{\n display:false,\n padding:0\n },\n gridLines:{\n drawTicks:false,\n drawBorder:false\n },\n scaleLabel:{\n display:false\n }\n\n }],\n xAxes: [{\n ticks :{\n display:false,\n padding:0\n\n },\n gridLines:{\n display:false,\n drawTicks:false,\n drawBorder:false\n }\n\n }]\n }\n\n }\n });\n```\n\n## Relevant pre-fix discussion\n\n### Issue #2872\n\n#### Comment at 2016-06-29T15:19:50Z\n\nI'm also getting padding issues without a scale. But mine are out of range.\nHow can we add padding to the canvas? (Chart v2)\n\nWith scale. e.g.\n\n```\nxAxes: [{\n display: true\n```\n\n\n\nWithout e.g. \n\n```\nxAxes: [{\n display: false\n```\n\n\n\nOnly seems to happen when both x and y axis display is equal to false.", "created_at": "2016-08-23T00:10:25Z", "version": "v2.2.1", "environment_setup_commit": "b33aff20919ca65e7914aa368fb99961e36e89be", "image_assets": "{\"problem_statement\": [\"assets/chartjs__Chart.js-3192/001.png\", \"assets/chartjs__Chart.js-3192/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "chartjs/Chart.js", "instance_id": "chartjs__Chart.js-3584", "base_commit": "efced4780ce2697d0d44bf1e32a4e2efebaa18e2", "problem_statement": "[BUG] Retina scaling no longer working in 2.4.0 (master branch)\nTesting out the latest using the master branch (since I needed a new feature) and now my charts are twice as big and blurry.\n\nI believe it is related to [this commit](https://github.com/chartjs/Chart.js/commit/16bcd6adc579cb3deae16ea915680bc219924cdc).\n\n## Expected Behavior\nUsing a width/height of 640x480, a canvas is created like so:\n```\n<canvas width=\"1280\" height=\"960\" style=\"display: block; width: 640px; height: 480px;\"></canvas>\n```\n\n## Current Behavior\nWith `responsive` set to `false`, the canvas is created as:\n```\n<canvas width=\"1280\" height=\"960\" style=\"display: block;\"></canvas>\n```\n\nWith `responsive` set to `true`, the canvas is created as:\n```\n<canvas width=\"2560\" height=\"1920\" style=\"display: block; width: 1280px; height: 960px;\"></canvas>\n```\n\n## Relevant pre-fix discussion\n\n### Issue #3575\n\n#### Comment at 2016-11-09T21:24:32Z\n\nAny example should do the trick (such as using the [Getting Started](http://www.chartjs.org/docs/#getting-started-creating-a-chart) example in the documentation). I would put together a JSFiddle but I don't think I can have it use 2.4.0 since it isn't released yet. If this is possible, let me know and I certainly will!\n\nBut here are a couple screenshots:\n\nUsing 400x400 with responsive set to false:\n\n\n\nWith responsive set to true:\n\n\n\nBoth charts appear as 800x800 and are twice as big as they should be.", "created_at": "", "version": "", "environment_setup_commit": "efced4780ce2697d0d44bf1e32a4e2efebaa18e2", "image_assets": "{\"problem_statement\": [\"assets/chartjs__Chart.js-3584/001.png\", \"assets/chartjs__Chart.js-3584/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "chartjs/Chart.js", "instance_id": "chartjs__Chart.js-4774", "base_commit": "52145de5db5dc3bae093880429b35954f0034993", "problem_statement": "[2.7] Lost responsive after destroy recreate\nIf some options have changed, I destroy the chart then create a new one.\n```\nif (chart) chart.destroy()\nchart = new Chart(...)\n```\nSince 2.7, the div which handle resize detection is removed but is not re-created. \nNo problem with the iframe in 2.6\n\n## Relevant pre-fix discussion\n\n### Issue #4737\n\n#### Comment at 2017-09-11T10:32:15Z\n\n@simonbrunel your jsfiddle reproduce the issue perfectly\nInitial : \n\nAfter click on recreate :\n", "created_at": "2017-09-18T20:19:23Z", "version": "v2.2.2", "environment_setup_commit": "52145de5db5dc3bae093880429b35954f0034993", "image_assets": "{\"problem_statement\": [\"assets/chartjs__Chart.js-4774/001.png\", \"assets/chartjs__Chart.js-4774/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "ckeditor/ckeditor5", "instance_id": "ckeditor__ckeditor5-14413", "base_commit": "88c7c7890f65c45f80344b4095d1defec1375284", "problem_statement": "Can't set the style on <div> using Styles dropdown\n## 📝 Provide detailed reproduction steps (if any)\n\n* Add GHS configuration:\n\n```javascript\nhtmlSupport: {\n allow: [\n {\n name: /.*/,\n attributes: true,\n classes: true,\n styles: true\n }\n ] \n }\n```\n\n* Add Styles configuration:\n\n```javascript\nstyle: {\n definitions: [\n { name: 'Message: Success', element: 'div', classes: ['alert', 'alert-success'] },\n { name: 'Message: Info', element: 'div', classes: ['info-boxalert', 'alert-info'] }\n ]\n}\n```\n\n* Call `editor.setData( '<div>Test</div>' )` and put the selection inside of it.\n* Try to apply defined styles.\n\n### ✔️ Expected result\n\n_It's possible to apply styles to the element._\n\n### ❌ Actual result\n\n_It's impossible to apply styles to the element._\n\n## 📃 Other details\n\nMight be related - #12625.\n\nThe same scenario works fine for the example `<section>` element.\n\n---\n\nIf you'd like to see this fixed sooner, add a 👍 reaction to this post.\n\n## Relevant pre-fix discussion\n\n### Issue #13341\n\n#### Comment at 2023-05-04T14:13:56Z\n\n@wimleers requested I upload some screenshots from Drupal 9 + CKEditor 4 config vs. same config in CKE5 to demonstrate the basics of the issue. (We have plugins to make adding style dropdowns easier than writing the JS demonstrated in previous comments of this issue).\n\nI should note that the issue here isn't limited to DIV, but pretty much any block-level element appearing in the styles dropdown, like UL, OL, DL, TABLE, H1, H2, etc. I havent done extensive testing on which block level items are affected.\n\n\n\nI’ve explicitly allowed the `<div>` tag in the \"manually editable tags\" settings as well, but they’re still disabled in the dropdown even when cursor placed inside an empty `<div>` manually added in Source mode.\n\n\n\n\nCKE4: \n\n\n\nCKE5:\n\n\nI should note that one can still manually switch to Source mode, and enter the HTML code manually and it works. Eg `<div class=\"alert success\"> </div>`\n\nCKE5, with manual edits via src mode:\n", "created_at": "2023-06-19T14:35:23Z", "version": "v38.0.0-alpha.0", "environment_setup_commit": "88c7c7890f65c45f80344b4095d1defec1375284", "image_assets": "{\"problem_statement\": [\"assets/ckeditor__ckeditor5-14413/001.png\", \"assets/ckeditor__ckeditor5-14413/002.png\", \"assets/ckeditor__ckeditor5-14413/003.png\", \"assets/ckeditor__ckeditor5-14413/004.png\", \"assets/ckeditor__ckeditor5-14413/005.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "ckeditor/ckeditor5", "instance_id": "ckeditor__ckeditor5-14706", "base_commit": "6bbde59ef34a9aa2ce4613f50dff74f24dec7c8b", "problem_statement": "[Firefox] Error when clicking on label elements inside editor\n## 📝 Provide detailed reproduction steps (if any)\n\nThere are multiple ways to reproduce this ( see screencasts below ), example steps:\n1. Open Firefox and go to https://ckeditor.com/ckeditor-5/demo/\n2. Open image alternative balloon\n3. Click several times on label `Text alternative` \n\n### ✔️ Expected result\n\nNothing Happens\n\n### ❌ Actual result\n\n```js\nUncaught CKEditorError: Permission denied to access property Symbol.toStringTag\nRead more: https://ckeditor.com/docs/ckeditor5/latest/framework/guides/support/error-codes.html#error-Permission denied to access property Symbol.toStringTag\n tr istext.js:17\n domSelectionToView domconverter.js:515\n _handleSelectionChange selectionobserver.js:145\n observe selectionobserver.js:104\n fire emittermixin.js:199\n i emittermixin.js:247\n attach emittermixin.js:202\n listenTo emittermixin.js:61\n observe selectionobserver.js:103\n attachDomRoot view.js:288\n init classiceditorui.js:104\n create classiceditor.js:193\n promise callback*create/< classiceditor.js:193\n create classiceditor.js:188\n A app-bb5ed7289d.js:33\n P app-bb5ed7289d.js:33\n q app-bb5ed7289d.js:33\n init app-bb5ed7289d.js:33\n default app-bb5ed7289d.js:44\n init app-bb5ed7289d.js:16\n EventListener.handleEvent* app-bb5ed7289d.js:16\n n app-bb5ed7289d.js:1\n <anonymous> app-bb5ed7289d.js:1\n <anonymous> app-bb5ed7289d.js:1\nckeditorerror.js:64:2\n```\n## Screencasts\n\n\nVideo key frames derived from `assets/ckeditor__ckeditor5-14706/001.mp4`:\n\n\n\n\n\nVideo key frames derived from `assets/ckeditor__ckeditor5-14706/002.mp4`:\n\n\n\n\n\nVideo key frames derived from `assets/ckeditor__ckeditor5-14706/003.mp4`:\n\n\n\n\n\n\n## 📃 Other details\n\n* Browser: Firefox\n* Happens since version `24.0.0`\n* It happens when I click inside forms, specifically on label elements, but only those that are inside editor - for example I cannot reproduce this when opening media embed toolbar dropdown and clicking on the label.\n\n---\n\nIf you'd like to see this fixed sooner, add a 👍 reaction to this post.\n\n## Relevant pre-fix discussion\n\n### Issue #9635\n\n#### Comment at 2021-05-06T11:54:46Z\n\nTo me it looks like some sort of weird issue in Firefox.\n\nFrom a quick research, it looks like it is connected with selection and inputs. As far as I can tell, the error does not occur when you click on texts in the version item (for example, author name).\n\nThe issue happens when you click on place next to an input. I think that Firefox messes up with selection then.\n\nWhat happens is, Firefox set selection in a way that the selection is in the input but when you try to get a range from the selection and then start and end container of the range, Firefox returns \"Restricted\" object, which throws when any of its property is accessed:\n\n\n\nWhen you check selection anchor and focus, correct DOM element is returned, but in line 512 \"Restricted\" object is returned.\n\nAFAICS, this error does not have an impact on editor, i.e. editor works. Am I right?\n\n\n\nWhat we could do is:\n\n1. Try to reproduce this using the most basic HTML possible and report the error to Firefox.\n2. Until it is fixed, we could either use try-catch or just use selection anchor and offset in this part of code (since this is just one range, using selection anchor and offset should be safe?).\n3. However, I guess there are other places where we use `isText()` that also might be affected by that.\n\n#### Comment at 2021-05-26T01:19:18Z\n\nI just found another way to reproduce this, but it doesn't involve interacting with CKEditor *at all*. All you need is a CKEditor instance on the page. Ignore it. Leave it there. Focus on the rest of the form. Click randomly on any other `<label>` element on the page. Eventually, the error pops up. Very weird, right?\n\n\n\nIn spite of the error, the editor works fine. This error doesn't break anything, as far as I can see.", "created_at": "2023-08-01T15:50:15Z", "version": "v39.0.1", "environment_setup_commit": "6bbde59ef34a9aa2ce4613f50dff74f24dec7c8b", "image_assets": "{\"problem_statement\": [\"assets/ckeditor__ckeditor5-14706/001.mp4\", \"assets/ckeditor__ckeditor5-14706/002.mp4\", \"assets/ckeditor__ckeditor5-14706/003.mp4\", \"assets/ckeditor__ckeditor5-14706/004.png\", \"assets/ckeditor__ckeditor5-14706/005.png\", \"assets/ckeditor__ckeditor5-14706/006.png\", \"assets/ckeditor__ckeditor5-14706/derived/001/frame-01.png\", \"assets/ckeditor__ckeditor5-14706/derived/001/frame-02.png\", \"assets/ckeditor__ckeditor5-14706/derived/001/frame-03.png\", \"assets/ckeditor__ckeditor5-14706/derived/002/frame-01.png\", \"assets/ckeditor__ckeditor5-14706/derived/002/frame-02.png\", \"assets/ckeditor__ckeditor5-14706/derived/002/frame-03.png\", \"assets/ckeditor__ckeditor5-14706/derived/003/frame-01.png\", \"assets/ckeditor__ckeditor5-14706/derived/003/frame-02.png\", \"assets/ckeditor__ckeditor5-14706/derived/003/frame-03.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "ckeditor/ckeditor5", "instance_id": "ckeditor__ckeditor5-8015", "base_commit": "db9d1eb7c4afc13ab07495bc14c4cdb8714cee63", "problem_statement": "Link balloon displayed in wrong place for selection after a styled link\n## 📝 Provide detailed reproduction steps (if any)\n\n1. Create a link like this: `<$text linkHref=\"aaa\">foo</$text><$text linkHref=\"aaa\" bold=\"true\">bar</$text>[]`\n2. Click the link toolbar button\n\n### ✔️ Expected result\n\nThe link Baloon should appear close to the caret.\n\n### ❌ Actual result\n\nThe link balloon appears in an incorrect place.\n\n\n\n## 📃 Other details\n\n* Browser: …\n* OS: …\n* CKEditor version: …\n* Installed CKEditor plugins: …\n\n---\n\nIf you'd like to see this fixed sooner, add a 👍 reaction to this post.\n\n## Relevant pre-fix discussion\n\n### Issue #7926\n\n#### Comment at 2020-08-24T14:22:57Z\n\nStill I'm unable to repro it on Win (which simply might be not affected). @ckeditor/qa-team can you check it on your end both win/mac?\n\nMeanwhile here's the screencast from my attempt:\n\n\n\n#### Comment at 2020-08-24T14:26:13Z\n\nI can reproduce this, can be either bold or italics at the end of the link. And it needs to be active when you place the cursor: \n \n", "created_at": "2020-09-03T09:00:11Z", "version": "v22.0.0", "environment_setup_commit": "db9d1eb7c4afc13ab07495bc14c4cdb8714cee63", "image_assets": "{\"problem_statement\": [\"assets/ckeditor__ckeditor5-8015/001.png\", \"assets/ckeditor__ckeditor5-8015/002.gif\", \"assets/ckeditor__ckeditor5-8015/003.gif\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "cypress-io/cypress", "instance_id": "cypress-io__cypress-7742", "base_commit": "7c5819393d5792c8d1bb79668823e423c4ec7527", "problem_statement": "cy.wait() does not resolve when the XHR contains static resource-like text in the XHR's query params or hash (like `.js`, `.html`, `.css`)\n### Current behavior:\n\nDiscovered this while trying to test for ad requests for my employer. When a URL with file extension, like .html, and a question mark (for a query string parameter) is visited, the request that is listened for is no longer detected. Remove the question mark, and the request is detected.\n\n### Desired behavior:\n\nBefore promoting a build in production, we test the build using a query-string parameter, so the wish is for this to work when a query-string parameter is present in the visited URL.\n\n### Test code to reproduce\n\n```js\ndescribe(\"Oregon Live Ads\", () => {\n it(\"test cy.route()\", () => {\n cy.viewport(1500, 1024);\n\n Cypress.on('uncaught:exception', () => {\n return false;\n });\n\n cy.server();\n\n cy.route({\n url: /gampad\\/ads\\?/,\n }).as(\"waitForAd\");\n\n cy.visit(\n \"https: |
| {"repo": "cypress-io/cypress", "instance_id": "cypress-io__cypress-9254", "base_commit": "40dd31102c686ad89da1153f7abb7b41b8595e29", "problem_statement": "Cypress Doesn't Finish Typing Strings Before Continuing to Next Field\n### Current behavior:\n\nWhen typing a long string into an input field, such as an email address to login in our case, Cypress 3.5.0 appears to no longer type the entire string. It bails on typing the complete string after the period in the email and never completely types the string. For example, `test@webdomainhere.com` would be typed as `test@webdomainhere.` in our app, failing the login.\n\nI have also seen similar things with usernames using mixed case, such as `associateEnabled` would be typed as `associate`, stopping before the capital letter in the string.\n\nDue to this bug we have had to roll back to using Cypress 3.4.1, where this worked fine.\n\nExample screenshot showing a failed login in Cypress 3.5.0 where the email should've had 'com' typed after the period:\n\n\n\n### Desired behavior:\n\nEntire string passed to a cy.type() function should be typed into the input field.\n\n### Steps to reproduce: (app code and test code)\n\nType a long string, such as a long email address into a form field and then move on to a password field with a .click() and see that the end of the first string isn't entered.\n\n### Versions\nCypress v3.5.0 (v3.4.1 works fine, we have had to roll back)\nMac OSX 10.14.6\nElectron 61 / Chrome 78\n\n## Relevant pre-fix discussion\n\n### Issue #5480\n\n#### Comment at 2019-10-28T19:11:47Z\n\n>Simply use cy.type() on an input field with a string with mixed case and it will reproduce the problem\n\nRecreating what you said does *not* reproduce the problem. There are likely other variables involved with the input or your test that we cannot isolate without you providing a completely reproducible example. \n\n**Index.html**\n\n```html\n<!doctype html>\n<html lang=\"en\">\n<body>\n <input id=\"my-input\">\n</body>\n</html>\n```\n\n**Spec.js**\n\n```js\nit('works', () => {\n cy.visit(\"index.html\")\n cy.get('#my-input').type('cAmElCaSe')\n .should('have.value', 'cAmElCaSe')\n})\n```\n\n\n\n\nPlease provide a fully reproducible example so that we can fix your issue as quickly as possible. Thanks!\n\n#### Comment at 2019-10-31T17:46:41Z\n\n@TheWanderingWalnut I was not able to reproduce any failure in the provided test case in 3.4.1 or 3.5.0. \n\nI did replace `partnerAccountActive` with a string `'partnerAccountActive'` and added a visit. This is passing for me in both versions. \n\n```js\nit('types in both fields', () => {\n cy.visit('https://cypressprodtesting.checkfront.com/')\n cy.get('#cf_id').click().type('partnerAccountActive')\n cy.get('#cf_pw').click().type('partnerAccountActive')\n cy.get('#cf_id').should('have.value', 'partnerAccountActive')\n cy.get('#cf_pw').should('have.value', 'partnerAccountActive')\n})\n```\n\n\n\nMaybe there is some other variable I am missing. I tried in Canary 80, Chrome 78, Chromium 74 and Electron 73 on my osMac Mohave machine using `cypress open. \n\nPlease also try updating to our recently released 3.6.0, since it fixed some type bugs that were introduced in 3.5.\n\n#### Comment at 2019-11-04T19:13:31Z\n\nHave confirmed, this fails occasionally, especially after rerun in 3.6.0 in both Chrome and Electron.\n\nThis is irrespective of any characters being capitalized. \n\n```js\nit(\"Logs in as a partner\", () => {\n cy.visit(\"http: |
| {"repo": "cypress-io/cypress", "instance_id": "cypress-io__cypress-9281", "base_commit": "197f3097f7beaa09cd4dd1e2a0e9a39610f5ed69", "problem_statement": "Error during screenshot capture: RangeError: Index out of range | The value of \"offset\" is out of range.\n### Current behavior:\nExecuting `cy.screenshot()` on large pages causes error. Stack trace:\n\n```bash\n RangeError: Index out of range\n at checkInt (buffer.js:1180:11)\n at Buffer.writeUInt32BE (buffer.js:1271:5)\n at new Jimp (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/jimp/index.js:137:30)\n at stitchScreenshots (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/lib/screenshots.js:199:17)\n at /Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/lib/screenshots.js:295:21\n at tryCatcher (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/util.js:16:23)\n at Promise._settlePromiseFromHandler (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/promise.js:510:31)\n at Promise._settlePromise (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/promise.js:567:18)\n at Promise._settlePromise0 (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/promise.js:612:10)\n at Promise._settlePromises (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/promise.js:691:18)\n at Async._drainQueue (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/async.js:133:16)\n at Async._drainQueues (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/async.js:143:10)\n at Immediate.Async.drainQueues (/Users/rbayliss/Library/Caches/Cypress/3.0.1/Cypress.app/Contents/Resources/app/packages/server/node_modules/bluebird/js/release/async.js:17:14)\n at runCallback (timers.js:781:20)\n at tryOnImmediate (timers.js:743:5)\n at processImmediate [as _immediateCallback] (timers.js:714:5)\n```\n<!-- images, stack traces, etc -->\n\n### Desired behavior:\nScreenshots are captured.\n\n### Steps to reproduce:\nCall `cy.screenshot()` on a page that exceeds 16,000,000 square pixels (in my case, the element is 1195px by 13518px). Reproduction test case: https://github.com/rbayliss/cypress-test-tiny/tree/index_error\n\n<!-- Issues without reproducible steps might get closed. *Tip* You can fork https://github.com/cypress-io/cypress-test-tiny repo, set up a failing test, then tell us the repo/branch to try. -->\n\n### Versions\n\n```\nCypress package version: 3.0.1\nCypress binary version: 3.0.1\nChrome 67.0.3396.87 on OSX 10.13\n```\n\n## Relevant pre-fix discussion\n\n### Issue #2034\n\n#### Comment at 2020-06-18T13:23:52Z\n\nAnother reproducible example: https://github.com/masakudamatsu/line-height-picker/blob/visual-testing/cypress/e2e/visual-tests.js\n\nCypress v4.8.0\n\n\n\nThe element to be screenshotted has SVG inside\n\n\n\n\n\nSeems this only happens when there is vertical scroll necessary to get the full image, even if the element itself is small\n\n#### Comment at 2020-07-28T08:50:48Z\n\nNarrowed this issue down to a smaller repro so we don't have to rely on outside repos/URLs. This is reproducible outside of any snapshot plugins. \n\nIf I open the DevTools attached to the window then the test passes without this error which is super fun. 🤦 \n\n`index.html`\n\n```html\n<html>\n<body>\n <svg width=\"778pt\" height=\"308pt\" style=\"background-color: green;\"></svg>\n <svg id=\"blue\" width=\"778pt\" height=\"308pt\" style=\"background-color: blue;\"></svg>\n <div style=\"height: 300px; background-color: red;\"></div>\n</body>\n</html>\n```\n\n`spec.js`\n\n```js\nit('works', () => {\n cy.visit('index.html')\n cy.get('#blue').screenshot()\n})\n```\n\n```\nRangeError: The value of \"offset\" is out of range. It must be >= 0 and <= 20553340. Received -11304\n at boundsError (internal/buffer.js:72:9)\n at Buffer.readUInt32BE (internal/buffer.js:295:5)\n at Jimp.<anonymous> (/Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/node_modules/@jimp/plugin-crop/dist/index.js:43:37)\n at scan (/Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/node_modules/@jimp/utils/dist/index.js:53:9)\n at Jimp.scanQuiet (/Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/node_modules/@jimp/core/dist/index.js:1262:32)\n at Jimp.cropQuiet (/Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/node_modules/@jimp/plugin-crop/dist/index.js:42:12)\n at Jimp.<computed> [as crop] (/Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/node_modules/@jimp/core/dist/index.js:1176:23)\n at crop (/Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/lib/screenshots.js:161:24)\n at /Users/jennifer/Library/Caches/Cypress/4.11.0/Cypress.app/Contents/Resources/app/packages/server/lib/screenshots.js:435:17\n at processImmediate (internal/timers.js:439:21)\nFrom previous event:\n at $Cypress.automation (http://localhost:51135/__cypress/runner/cypress_runner.js:164254:14)\n at automate (http://localhost:51135/__cypress/runner/cypress_runner.js:157822:20)\n at automateScreenshot (http://localhost:51135/__cypress/runner/cypress_runner.js:157832:10)\n at scrollAndTake (http://localhost:51135/__cypress/runner/cypress_runner.js:157901:12)\nFrom previous event:\n at takeScrollingScreenshots (http://localhost:51135/__cypress/runner/cypress_runner.js:157904:18)\n at takeElementScreenshot (http://localhost:51135/__cypress/runner/cypress_runner.js:158019:10)\n at http://localhost:51135/__cypress/runner/cypress_runner.js:158123:14\nFrom previous event:\n at takeScreenshot (http://localhost:51135/__cypress/runner/cypress_runner.js:158115:19)\n at Context.screenshot (http://localhost:51135/__cypress/runner/cypress_runner.js:158229:14)\n at Context.<anonymous> (http://localhost:51135/__cypress/runner/cypress_runner.js:167810:21)\n at http://localhost:51135/__cypress/runner/cypress_runner.js:167243:15\nFrom previous event:\n at runCommand (http://localhost:51135/__cypress/runner/cypress_runner.js:167222:8)\n at next (http://localhost:51135/__cypress/runner/cypress_runner.js:167368:14)\nFrom previous event:\n at next (http://localhost:51135/__cypress/runner/cypress_runner.js:167368:34)\nFrom previous event:\n at http://localhost:51135/__cypress/runner/cypress_runner.js:167409:37\nFrom previous event:\n at run (http://localhost:51135/__cypress/runner/cypress_runner.js:167402:19)\n at $Cy.cy.<computed> [as visit] (http://localhost:51135/__cypress/runner/cypress_runner.js:167854:11)\n at Context.runnable.fn (http://localhost:51135/__cypress/runner/cypress_runner.js:168079:21)\n at callFn (http://localhost:51135/__cypress/runner/cypress_runner.js:103224:21)\n at Test.../driver/node_modules/mocha/lib/runnable.js.Runnable.run (http://localhost:51135/__cypress/runner/cypress_runner.js:103211:7)\n at http://localhost:51135/__cypress/runner/cypress_runner.js:173313:28\nFrom previous event:\n at Object.onRunnableRun (http://localhost:51135/__cypress/runner/cypress_runner.js:173301:20)\n at $Cypress.action (http://localhost:51135/__cypress/runner/cypress_runner.js:164070:61)\n at Test.Runnable.run (http://localhost:51135/__cypress/runner/cypress_runner.js:171914:13)\n at Runner.../driver/node_modules/mocha/lib/runner.js.Runner.runTest (http://localhost:51135/__cypress/runner/cypress_runner.js:103883:10)\n at http://localhost:51135/__cypress/runner/cypress_runner.js:104009:12\n at next (http://localhost:51135/__cypress/runner/cypress_runner.js:103792:14)\n at http://localhost:51135/__cypress/runner/cypress_runner.js:103802:7\n at next (http://localhost:51135/__cypress/runner/cypress_runner.js:103704:14)\n at http://localhost:51135/__cypress/runner/cypress_runner.js:103770:5\n at timeslice (http://localhost:51135/__cypress/runner/cypress_runner.js:97696:27)\n```\n\n\n\n`DEBUG=cypress:server:screenshot` logs\n\n\n```\n cypress:server:screenshot (s4) capturing screenshot { titles: [ 'works' ], testId: 'r2', capture: 'viewport', clip: { x: 8, y: -0.34375, width: 992, height: 410.65625 }, viewport: { width: 1025, height: 894 }, scaled: false, blackout: [], startTime: '2020-07-28T09:06:53.193Z', current: 1, total: 1, specName: 'spec.js' } +0ms\n cypress:server:screenshot (s4) capture and check { tries: 1, totalDuration: 0 } +2ms\n cypress:server:screenshot (s4) received screenshot data from automation layer data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAACAIAAAb8CAYAAACEPaEyAAABKWlDQ1BTa2lhAAAokX2RP0jDQBTGfx +416ms\n cypress:server:screenshot (s4) read buffer to image 2050 x 1788 +200ms\n cypress:server:screenshot (s4) helper pixels\n cypress:server:screenshot {\n cypress:server:screenshot topLeft: { r: 255, g: 255, b: 255, a: 255, isNotWhite: false },\n cypress:server:screenshot topLeftRight: { r: 255, g: 255, b: 255, a: 255, isWhite: true },\n cypress:server:screenshot topLeftDown: { r: 255, g: 255, b: 255, a: 255, isWhite: true },\n cypress:server:screenshot bottomLeft: { r: 255, g: 255, b: 255, a: 255, isWhite: true },\n cypress:server:screenshot topRight: { r: 255, g: 255, b: 255, a: 255, isWhite: true },\n cypress:server:screenshot bottomRight: { r: 255, g: 255, b: 255, a: 255, isBlack: false }\n cypress:server:screenshot } +0ms\n cypress:server:screenshot (s4) pixelConditionFn { pixelRatio: 2, subject: 'app', hasPixels: false, expectedPixels: false } +1ms\n cypress:server:screenshot (s4) no previous image to compare +0ms\n cypress:server:screenshot (s4) resolving with image { tries: 1, totalDuration: 0 } +0ms\n cypress:server:screenshot (s4) pixel ratio is 2 +0ms\n cypress:server:screenshot (s4) multi-part 1/1 +0ms\n cypress:server:screenshot (s4) dimensions before are { x: 8, y: -0.34375, width: 992, height: 410.65625 } +0ms\n cypress:server:screenshot (s4) dimensions for cropping are { x: 16, y: -0.6875, width: 1984, height: 821.3125 } +0ms\n cypress:server:screenshot (s4) crop: from 16, -0.6875 +1ms\n cypress:server:screenshot (s4) to 1984 x 821.3125 +0ms\n cypress:server:screenshot (s4) clearing 0 cached multipart images +9ms\n```\n\n#### Comment at 2020-10-14T08:11:22Z\n\nThis bug is really coming in the way of my success with visual testing with Cypress(cypress-image-snapshot/plugin). I get the out of range error only when I run with headless chrome, inside a container! I setup the viewport to (1600,900) in cypress.json. Dont use cy.viewport() anywhere in the test. I get the snapshots with cy.matchImageSnapshot() for the test- locally, in both headed and headless chrome. Also headless, locally. But not on headless-inside container!\n\nI am on 3.8.3", "created_at": "", "version": "", "environment_setup_commit": "197f3097f7beaa09cd4dd1e2a0e9a39610f5ed69", "image_assets": "{\"problem_statement\": [\"assets/cypress-io__cypress-9281/001.png\", \"assets/cypress-io__cypress-9281/002.png\", \"assets/cypress-io__cypress-9281/003.png\", \"assets/cypress-io__cypress-9281/004.png\", \"assets/cypress-io__cypress-9281/005.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "directus/directus", "instance_id": "directus__directus-20185", "base_commit": "9d80bd899be567042410e7254fd1287853e9580c", "problem_statement": "Detail Group crashing pages\n### Describe the Bug\n\nAfter upgrading to 10.7.0 some of my collections getting frozen in studio app, I discovered that If you have detail group inside detail group it's a problem, before it was all good, also no outside/console errors are visible, page simply freezing and then in few second browser tab crashes.\n\n### To Reproduce\n\nCreate detail group and then create another detail group and drop inside the previous one.\n\n### Directus Version\n\nv10.7.0\n\n### Hosting Strategy\n\nSelf-Hosted (Docker Image)\n\n## Relevant pre-fix discussion\n\n### Issue #20145\n\n#### Comment at 2023-10-24T09:14:37Z\n\nI am using the raw group instead together with other grouping extensions (tabs). This is an example of what's causing the freeze on mine:\n\n\n\n#### Comment at 2023-10-24T11:19:17Z\n\n> @paulvonber Would it be possible for you to share your data model? While I think I'm able to reproduce it in another way, it doesn't seem to trigger a freeze on my side when simply having a detail group inside another.\n\nso here is what I have:\n\n\nAll of the detail groups by default are \"closed\", so when I go to collection entry and try to open the very top parent one it crashes, if I take out those four details group out of the parent then all works as expected.\nIf I set parent group by default to \"opened\" then I can't even enter item page it crashes in the collection list view (table page)\n\n#### Comment at 2023-10-24T14:36:39Z\n\nThis is a new directus self hosted created using `pnpm dlx create-directus-extension@latest` therefore on v10.7.0\n\nSo `rgroup_...` is `raw groups`, `dgroup_...` is `detail groups` in the image below?\n\n\n\nthe above works fine, but as soon as I move the raw group into the detail group then everything goes sideways:\n\n\n\n#### Comment at 2023-10-24T14:40:18Z\n\nand raw groups takes more I think: \n", "created_at": "", "version": "", "environment_setup_commit": "9d80bd899be567042410e7254fd1287853e9580c", "image_assets": "{\"problem_statement\": [\"assets/directus__directus-20185/001.png\", \"assets/directus__directus-20185/002.png\", \"assets/directus__directus-20185/003.png\", \"assets/directus__directus-20185/004.png\", \"assets/directus__directus-20185/005.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "directus/directus", "instance_id": "directus__directus-26045", "base_commit": "f7ae717e6ec7086571d9bb2040bb0b32c6a087b4", "problem_statement": "Missing data if the distinct count is failing\n### Describe the Bug\n\nThe collection page will not show any data if the distinct count (its result shown at the top right corner of the page) is not loaded in. If the distinct count fails, it won't show the data at all.\n\n\n\nI would consider the `x-y of z` visual a _nice to have_, but not critical for it to block the data visualisation all together. For larger datasets, this feature keeps most collections empty for longer than needed, given that the actual query for the data itself is faster than the distinct count.\n\n### To Reproduce\n\nCreate a large table if 1M entries. Open the collection through the CMS and observe the load times.\n\nTo 'break' the distinct count query can be a bit harder. Either reduce the performance of the database, or increase the number of entries.\n\n### Directus Version\n\nv11.6.1\n\n### Hosting Strategy\n\nSelf-Hosted (Docker Image)\n\n### Database\n\nPostgreSQL 15.13\n\n## Relevant pre-fix discussion\n\n### Issue #25984\n\n#### Comment at 2025-10-18T21:49:52Z\n\nHere are some additional screenshots of the issue: the page fails to display the data, even though it has been successfully retrieved via the corresponding API call.\n\n\n\n\n#### Comment at 2025-10-21T14:25:43Z\n\nI've setup the staging environment with all extensions and flows removed. I also had the opportunity to manually vacuum the table. The latter reduced the loading time for the distinct count (which was needed anyways).\n\n\n\nIn regards to the data visualisation, the following steps occurred:\n\ntime (s) | data shown | `x-y of z items` shown\n--- | --- | ---\n0.0 | ❌ | ❌ \n0.045 | ❌ | ❌\n1.9 | ✔️ | ❌\n3.88 | ✔️ | ✔️\n\nHere $`3.88 \\approx 2(1.9 - 0.045)`$, which questions; why do we need the distinct count twice?", "created_at": "", "version": "", "environment_setup_commit": "f7ae717e6ec7086571d9bb2040bb0b32c6a087b4", "image_assets": "{\"problem_statement\": [\"assets/directus__directus-26045/001.png\", \"assets/directus__directus-26045/002.jpg\", \"assets/directus__directus-26045/003.jpg\", \"assets/directus__directus-26045/004.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "directus/directus", "instance_id": "directus__directus-26739", "base_commit": "54a35009d15177d6a9dde22c67ac2b9c5036a74e", "problem_statement": "Flow operation does not display templated field value without template mode enabled\n### Describe the Bug\n\nIf a field value has been set to a template (.e.g. `{{$trigger.payload.card}}`) it will not show in the input field unless template mode is enabled\n\n### To Reproduce\n\n1. Create a Flow with a `webhook` and `read data` operation\n2. Add a template of `{{$trigger.payload.id}}` for the `IDs` section of the read operation\n3. Observe it showing in the Read Operation card preview\n4. Attempt to edit the Read operation and observe the value is not showing unless template mode is enabled\n#### Reproduction Video\nVideo key frames derived from `assets/directus__directus-26739/001.mov`:\n\n\n\n\n### Directus Version\n\nv11.10.2\n\n### Hosting Strategy\n\nSelf-Hosted (Docker Image)\n\n### Database\n\n_No response_\n\n## Relevant pre-fix discussion\n\n### Issue #25732\n\n#### Comment at 2025-09-12T15:47:29Z\n\nBecause the \"IDs\" field there is a \"Tags interface\" when editing the \"raw value\" it is expecting an array. \n\n\n\n", "created_at": "", "version": "", "environment_setup_commit": "54a35009d15177d6a9dde22c67ac2b9c5036a74e", "image_assets": "{\"problem_statement\": [\"assets/directus__directus-26739/001.mov\", \"assets/directus__directus-26739/002.png\", \"assets/directus__directus-26739/003.png\", \"assets/directus__directus-26739/derived/001/frame-01.png\", \"assets/directus__directus-26739/derived/001/frame-02.png\", \"assets/directus__directus-26739/derived/001/frame-03.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "eslint/eslint", "instance_id": "eslint__eslint-11799", "base_commit": "cb1922bdc07e58de0e55c13fd992dd8faf3292a4", "problem_statement": "envs in overrides was not applied\n<!--\n ESLint adheres to the [JS Foundation Code of Conduct](https://js.foundation/community/code-of-conduct).\n\n This template is for bug reports. If you are here for another reason, please see below:\n\n 1. To propose a new rule: https://eslint.org/docs/developer-guide/contributing/new-rules\n 2. To request a rule change: https://eslint.org/docs/developer-guide/contributing/rule-changes\n 3. To request a change that is not a bug fix, rule change, or new rule: https://eslint.org/docs/developer-guide/contributing/changes\n 4. If you have any questions, please stop by our chatroom: https://gitter.im/eslint/eslint\n\n Note that leaving sections blank will make it difficult for us to troubleshoot and we may have to close the issue.\n-->\n\n**Tell us about your environment**\n\n* **ESLint Version:** v5.16.0\n* **Node Version:** v10.15.3\n* **npm Version:** 6.9.0\n\n**What parser (default, Babel-ESLint, etc.) are you using?**\n\n**Please show your full configuration:**\n\n<details>\n<summary>Configuration</summary>\n\n<!-- Paste your configuration below: -->\n```js\n \"eslintConfig\": {\n \"extends\": [\n \"eslint:recommended\"\n ],\n \"overrides\": [\n {\n \"files\": [\n \"./lib |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| \ninterface DataTransfer {\n...\n```\n\nAR\n---\n\nNo JSDoc description and no link to online docs (mdn)\n\n\n\n## Relevant pre-fix discussion\n\n### Issue #268776\n\n#### Comment at 2025-09-30T09:35:33Z\n\n@aeschli I tested this via TS playground and it works in pure TS:\n\n\n\n[here](https: |
| {"repo": "microsoft/vscode", "instance_id": "microsoft__vscode-321227", "base_commit": "fc8f90cdb7b4b5c8446d6e5212021c8a19d87163", "problem_statement": "Integrated browser tab unexpectedly paints on top of modal\nVersion: 1.125.0-insider (Universal)\nCommit: d420356ab9998a711c36d7c51a2d7d2c91c73a87\nOS: Darwin arm64 24.6.0\n\nSteps to Reproduce:\n\n1. Open a new integrated browser tab\n2. Open \"Language Models\" modal of the GitHub Copilot\n3. Hit the \"+ Add Models...\" button\n\nExpected result: A drop down list shows up below the \"+ Add Models...\" button\n\nActual result: The drop down list that is expected to show up actually shows up. But, in addition, the integrated browser tab (which until this point was behind the modal) suddenly gets painted on top of the modal ! (See screenshot)\n\n\n\nNote: I have not been able to reproduce this consistently - after i took the screenshot above, i used \"Toggle Developer Tools\" to open the dev tools to just check what might be the issue. But ever since i did that, the issue stopped happening anymore magically ! But my hope is you might be able to use this issue as a reason to review the relevant subsystems (integrated browser, modals) for latent bugs.\n\n## Relevant pre-fix discussion\n\n### Issue #321088\n\n#### Comment at 2026-06-12T20:55:25Z\n\nAn important fact is that the issue only happens when the \"Add Models...\" dropdown doesn't overlap with the browser rectangle:\n\n\nWhen it does overlap, it doesn't repro\n", "created_at": "", "version": "", "environment_setup_commit": "fc8f90cdb7b4b5c8446d6e5212021c8a19d87163", "image_assets": "{\"problem_statement\": [\"assets/microsoft__vscode-321227/001.png\", \"assets/microsoft__vscode-321227/002.png\", \"assets/microsoft__vscode-321227/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "mrdoob/three.js", "instance_id": "mrdoob__three.js-33438", "base_commit": "5f343712b4f88b4bc827759c9357d4c17c75d6e7", "problem_statement": "SVGLoader: Shape generation breaks with large stroke width.\n### Description\n\nThe original issue was reported at the forum: https://discourse.threejs.org/t/glitches-with-svg-loader/46803\n\nSVGs that exceed a certain `stroke-width` are not rendered correctly. As visible in the below screenshot, the resulting shape is rendered with additional triangles on its back side. When decreasing the stroke width (in the below SVG try 200), the issue goes away.\n\n I suspect there is a problem in [SVGLoader.pointsToStrokeWithBuffers()](https://github.com/mrdoob/three.js/blob/4a6cf23c5f69ecde00e79171be7235deb39dff79/examples/jsm/loaders/SVGLoader.js#L2412) which generates triangles based on the stroke definition from the SVG.\n\n### Reproduction steps\n\n1. Load the following SVG with the loader to reproduce the glitch.\n```\n<?xml version=\"1.0\"?>\n<svg viewBox=\"0 0 600 792\" width=\"600\" height=\"792\">\n <path stroke=\"#64C36E\" fill=\"none\" id=\"link_2_21_ok\" d=\"M30,463.61904761904765C150,463.61904761904765,150,571.0476190476189,270,571.0476190476189\" stroke-width=\"290\" stroke-opacity=\"0.6\"/>\n</svg>\n```\n\n### Code\n\n- See fiddles.\n\n### Live example\n\n* [jsfiddle-latest-release](https://jsfiddle.net/f5veyqst/2/)\n* [jsfiddle-dev](https://jsfiddle.net/f5veyqst/)\n\n\n### Screenshots\n\n\n\n\n### Version\n\nr148\n\n### Device\n\nDesktop, Mobile, Headset\n\n### Browser\n\nChrome, Firefox, Safari, Edge\n\n### OS\n\nWindows, MacOS, Linux, Android, iOS\n\n## Relevant pre-fix discussion\n\n### Issue #25326\n\n#### Comment at 2023-06-11T08:16:28Z\n\n@Mugen87 I've implemented your idea, by holding two contour points arrays (left and right ones) and combining them at the end, giving the full stroke contour.\n\nIt works mostly correctly in the use case. I hoped it worked in the general case but you can see that the Tiger.svg has some artifacts. It is caused by the contour I generate, which folds over itself on some corners. Earcut does not like that on complex paths. As before, solving this would require checking self-intersections, but now in the linear contour, not in a triangle soup. It is doable, but only would work if the original path doesn't self-intersect.\n\nDetail of the contour folding over itself:\n\n\n\nTiger.svg:\n\n\n\nUse case:\n\n\n\n\nUse case detail:\n\n\n\n\nThe algorithm is smaller and cleaner (751 lines down to 495). Performance seems similar as before. The drawback is that the two contour arrays must be allocated. This can be optimized by providing reusable user arrays.\n\nAs the whole ```pointsToStrokeWithBuffers()``` is modified, I don't know very well how to integrate this change. For now I've just duplicated the function with the name ```pointsToStrokeContourWithBuffers()```.\n\nThe function ```pointsToStroke()``` accepts a boolean parameter ```useEarcut``` (default false) and selects which function to call.\n\nIt seems ```THREE``` doesn't export ```Earcut``` so I've copied ```Earcut.js``` in ```/jsm``` just for testing.\n\nI've tried to make a live preview but failed. You can read the modified loader and example here: https://github.com/yomboprime/three.js/commit/563aea9172ada3e7dfa862e3125334415b40f336\n\nExpecting your feedback to see if this can be useful.\n\n#### Comment at 2023-06-18T23:21:11Z\n\nUnfortunately I've not been able to solve it for the general case. I've spent many hours on this.\n\nThe problem is when the path points do auto-intersect like this (the contour gets flipped over itself):\n\n\nI think this problem needs more thought than an amateur mathematician (me) can give.", "created_at": "2026-04-21T13:38:49Z", "version": "r184", "environment_setup_commit": "5f343712b4f88b4bc827759c9357d4c17c75d6e7", "image_assets": "{\"problem_statement\": [\"assets/mrdoob__three.js-33438/001.png\", \"assets/mrdoob__three.js-33438/002.png\", \"assets/mrdoob__three.js-33438/003.png\", \"assets/mrdoob__three.js-33438/004.png\", \"assets/mrdoob__three.js-33438/005.png\", \"assets/mrdoob__three.js-33438/006.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "mui/material-ui", "instance_id": "mui__material-ui-39071", "base_commit": "ba4c5596cdbbdf07e78cb10ca7231db9968812db", "problem_statement": "[material] Invalid `color` prop has no effect\n- [X] I have searched the existing issues\n- [X] I have tested the latest version\n\n### Steps to reproduce 🕹\n\nLink to live example: [CodeSandbox fork](https://codesandbox.io/s/mui-invalid-color-is-silently-swallowed-gtwdg2) based on the [Typography demo](https://mui.com/material-ui/react-typography/) from the docs\n\n1. Open [CodeSandbox fork ](https://codesandbox.io/s/mui-invalid-color-is-silently-swallowed-gtwdg2)\n2. Observe invalid `color` prop to `mui.Typography` has no effect\n\n### Current behavior 😯\n\nInvalid `color` prop has absolutely no effect:\n- no warnings or errors\n- no type checking\n- no invalid CSS (at least it would serve as an indicator to the developer)\n\n### Expected behavior 🤔\n\nIn `NODE_ENV=development` or using an optional flag to `mui.createTheme`, etc.\n- we should tell the developer there's an invalid `color` prop\n\n# Proposal\n\n<details>\n <summary>Here's what we use internally</summary>\n \n```ts\nimport * as mui from \"@mui/material\"\n\nimport { palettes } from \"../options/palette\"\n\nlet validColors: string[] | undefined\n/**\n * @__NO_SIDE_EFFECTS__\n */\nexport const isColorValid = /* @__PURE__ */ (color?: unknown) => {\n if (process.env.NODE_ENV === `production`) return\n if (typeof color !== `string`) return\n\n if (!validColors) {\n const tones = Object.keys({\n main: true,\n light: true,\n dark: true,\n contrastText: true,\n } satisfies Record<keyof mui.SimplePaletteColorOptions, true>)\n\n const colors = Object.keys({\n primary: true,\n secondary: true,\n error: true,\n warning: true,\n info: true,\n success: true,\n } satisfies Record<ColorWithTones, true>)\n\n const text = Object.keys({\n disabled: true,\n primary: true,\n secondary: true,\n } satisfies Record<keyof mui.TypeText, true>)\n\n const background = Object.keys({\n default: true,\n paper: true,\n ground: true,\n } satisfies Record<keyof mui.TypeBackground, true>)\n\n /**\n * Sometimes, we want to let the user to a color that is not in the palette (theme)\n */\n const validStaticColors = [`white`]\n\n /**\n * A user can use a literal color, by using \"mui.useTheme\" and then pass a literal color\n */\n const literalThemeColors = Object.keys(palettes).flatMap((paletteName) => {\n const palette = palettes[paletteName]\n const literals = new Set<string>() // to avoid duplicates\n for (const key of Object.keys(palette)) {\n const value = palette[key]\n if (typeof value === `string`) {\n literals.add(value)\n continue\n }\n\n for (const valueKey of Object.keys(value)) {\n const nestedValue = value[valueKey]\n if (typeof nestedValue === `string`) {\n literals.add(nestedValue)\n continue\n }\n }\n }\n return [...literals]\n })\n\n validColors = [\n ...validStaticColors,\n ...literalThemeColors,\n `primary`,\n `secondary`,\n ...background.map((tone) => `background.${tone}`),\n ...text.map((tone) => `text.${tone}`),\n ...colors.flatMap((color) => tones.map((tone) => `${color}.${tone}`)),\n ]\n }\n\n if (!validColors.includes(color)) {\n throw new Error(\n `Invalid color: \"${color}\"\\n` +\n `Valid colors are: ${validColors.join(`, `)}`,\n )\n }\n}\n```\n</details>\n\n## Relevant pre-fix discussion\n\n### Issue #38478\n\n#### Comment at 2023-09-04T19:39:38Z\n\nHey @o-alexandrov, thanks for the report!\n\nThis is a bug. I think what's happening is that \"background\" is a key in the [theme palette](https://mui.com/material-ui/customization/default-theme/?expand-path=$.palette), so we try to use that value, but it's an object (with `\"paper\"` and `\"default\"` keys). The expected behavior here is that `\"background\"` goes through to CSS. That way, the developer gets notice that it's not valid. This is the current behavior for other not valid colors, for example, when providing `\"not-a-color\"` in the codesandbox:\n\n\n\nAdding the ready-to-take label.\n\n#### Comment at 2023-09-15T06:02:42Z\n\nI debugged this a bit and found that it's actually setting the color value to the `background` property found in the theme object. But the problem as you pointed out correctly is that the `background` property is an object. So it's setting it as -\n\n```\ncolor: {\n paper: '#value',\n default: '#value',\n}\n```\n\nWhen this goes through emotion, it's style gets generated as -\n\n```\n.emotion-client-render-bd0kf1-MuiTypography-root color{paper:#fff;default:#fff;}\n```\n\nCodesandbox inspect element -\n\n\n\nWhich is correct as per emotion. So we should be checking the value to be a primitive type (string/number/boolean etc). If it's not, we should not set the property's value itself and also log an error to console in dev mode.\n\nThe testing can be made easier by adding this test in `Typography.test.js`\n\n\n it.only('should', () => {\n const { container, debug } = render(\n <Typography variant=\"h6\" color=\"background\">\n Hello\n </Typography>,\n );\n console.log(document.documentElement.innerHTML);\n debug(container);\n });\n\nThe logged html can be checked for correctness of generated CSS.\n\n@DarhkVoyd Let me know if you'll be taking this up. If not, I'll fix it.", "created_at": "", "version": "", "environment_setup_commit": "ba4c5596cdbbdf07e78cb10ca7231db9968812db", "image_assets": "{\"problem_statement\": [\"assets/mui__material-ui-39071/001.png\", \"assets/mui__material-ui-39071/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "mui/material-ui", "instance_id": "mui__material-ui-40865", "base_commit": "4e797158fe59df81bc604f57d2620726e93a82a8", "problem_statement": "[Select] Outlined component's label has no background fill (looks \"strikethrough\")\nNormally a textfield has a border. All select fields does not have topborder.\n\n- [x ] The issue is present in the latest release.\n- [x ] I have searched the [issues](https://github.com/mui-org/material-ui/issues) of this repository and believe that this is not a duplicate.\n\n## Current Behavior 😯\nTextfields have a border everywhere:\n\n\nBut select fields does not have a top-border:\n\n\n## Expected Behavior 🤔\nTopborder also on select fields\n\n## Steps to Reproduce 🕹\nAnd that is a problem. I tried to reproduce this behavior, but I dont get this error it in codesandbox:\n[Codesandbox example](https://codesandbox.io/s/material-demo-forked-9psn4?file=/demo.tsx)\n\nThe next strange thing is in codesandbox it is also not correct. In the codesandbox example the label \"gender\" is strikethrough.\n\n<details>\nMaterial-ui v5.0.0 beta 2\n</details>\n\n## Relevant pre-fix discussion\n\n### Issue #27533\n\n#### Comment at 2021-08-17T06:11:45Z\n\n@eps1lon \nCan i pls add some more problems to this ticket:\n\nThere is another codesandbox example:\n[Next example](https://codesandbox.io/s/material-demo-forked-ilgze)\n\n1. I changed my initial codesandbox example to the example on material-ui documentation page. But why do we have two labels?\n\n\n2. Second problem: Now when you click on the select box the menuItems are not shown when you have a Select native.\n\nThank you for the effort\n\n#### Comment at 2021-08-18T14:52:37Z\n\nI am also running into issues with the `label` prop on `Select` elements. \n\nIssue 1: When I set a `label` prop, it doesn't show up in the output.\n\n\nIssue 2: When I use an `InputLabel`, the label appears struck-out:\n\n\nCode example:\n```\nconst useStyles = makeStyles(theme => ({\n formControl: {\n minWidth: 120\n }\n}));\n\nexport const Base = () => {\n const classes = useStyles();\n return (\n <FormControl className={classes.formControl} variant=\"outlined\">\n <InputLabel>My Label</InputLabel>\n <Select onChange={action('change')}>\n <MenuItem value={1}>1</MenuItem>\n <MenuItem value={2}>2</MenuItem>\n <MenuItem value={3}>3</MenuItem>\n </Select>\n </FormControl>\n )\n};\n```\n\nWorkaround: Using both a `label` prop _and_ an `InputLabel` appear to render something that looks good. I just think there is a better way out there somehow.\n\n\nCode:\n```\nconst useStyles = makeStyles(theme => ({\n formControl: {\n minWidth: 120\n }\n}));\n\nexport const Base = () => {\n const classes = useStyles();\n return (\n <FormControl className={classes.formControl} variant=\"outlined\">\n <InputLabel>My Label</InputLabel>\n <Select label=\"My Label\" onChange={action('change')}>\n <MenuItem value={1}>1</MenuItem>\n <MenuItem value={2}>2</MenuItem>\n <MenuItem value={3}>3</MenuItem>\n </Select>\n </FormControl>\n )\n};\n```\n\nUpdate: I see on the codesandbox of the reference documentation that both `InputLabel` element _and_ `label` prop on the `Select` are used to achieve the correct display. Kind of confusing!\n\n\n#### Comment at 2021-11-08T20:03:36Z\n\nI am having the same issue. Not sure what the work around is for this currently.\n\n\n\n\n\n#### Comment at 2021-12-14T10:29:12Z\n\n@michaldudak I must have forgotten to save the sandbox after editing. It's not fixed yet.\n\n.\n\n#### Comment at 2022-06-17T19:08:08Z\n\nIn my case, I had to assign a label value to the `Select`'s input component as well...\n\n\n\n#### Comment at 2023-05-09T05:07:45Z\n\nSame problem here as [AlexanderHOtt](https://github.com/AlexanderHOtt): \n\n\n#### Comment at 2024-01-25T17:30:11Z\n\nanyone found the right solution for this yet? I've tried all the workarounds, but I don't get the desired result... The label is strikethrough if I don't select anything inside of it \nhow it looks:\n\nhow it should always look:\n\n\nhere's a code sandbox do demonstrate this behaviour: https://codesandbox.io/p/sandbox/material-ui-outlined-select-label-is-not-rendering-properly-forked-3nx92r - as you can see, when you set the select to be displayed empty, the label gets strikethrough until you click inside of it. does anyone know how to fix this? :|", "created_at": "", "version": "", "environment_setup_commit": "4e797158fe59df81bc604f57d2620726e93a82a8", "image_assets": "{\"problem_statement\": [\"assets/mui__material-ui-40865/001.png\", \"assets/mui__material-ui-40865/002.png\", \"assets/mui__material-ui-40865/003.png\", \"assets/mui__material-ui-40865/004.png\", \"assets/mui__material-ui-40865/005.png\", \"assets/mui__material-ui-40865/006.png\", \"assets/mui__material-ui-40865/007.png\", \"assets/mui__material-ui-40865/008.png\", \"assets/mui__material-ui-40865/009.png\", \"assets/mui__material-ui-40865/010.png\", \"assets/mui__material-ui-40865/011.png\", \"assets/mui__material-ui-40865/012.png\", \"assets/mui__material-ui-40865/013.png\", \"assets/mui__material-ui-40865/014.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "openlayers/openlayers", "instance_id": "openlayers__openlayers-13013", "base_commit": "a715b1f5c6a64f80298fd424306f2bafb9752e4c", "problem_statement": "Black tiles with DataTileSource and WebGLTileLayer\n**Describe the bug**\n\nThe`ol.layer.WebGLTile` doesn't render data that were loaded with `ol.source.DataTile` using a custom tile grid:\n\nExtent: `[0, -23828, 23903, 0]`\nSizes: `[[1, 1], [7, 7], [ 25, 25 ], [100, 100]]`\nTile sizes: `[[770, 768], [240, 240], [240, 240], [240, 240]]`\n\n\n\nThe following warning message gets logged:\n\n```\nWebGL warning: texImage: Desired upload requires more data than is available: (767 rows plus 770 pixels needed, 767 rows plus 258 pixels available)\n```\n\nI didn't previously experience this issue and wonder whether it may be due to either https://github.com/openlayers/openlayers/pull/11532 or https://github.com/openlayers/openlayers/pull/12024.\n\n**Expected behavior**\n\nNot sure whether this is a bug or whether my understanding of the tile grid is wrong.\n\n## Relevant pre-fix discussion\n\n### Issue #12995\n\n#### Comment at 2021-11-13T15:42:16Z\n\n> Does [770, 767] work? `770 * 767 + 258 = 577 * 1024` suggests the data has been truncated and the final row is not complete.\n\nThe warning goes away, but the data still doesn't get rendered.\n\nI attached a listener to `tileloadend` and that is the output:\n\n\n\nThe length of the data seems to be correct.\n\nWhat's interesting, however, is that there are additional tiles requested at coordinates `[0, -1, 0]` and `[0, 1, 0]` and the data for those tiles is `undefined`:\n\n\n\nThe reason the data is `undefined` is because my loader returns `undefined` for tile coordinates that don't exist. However, if I instead return white tiles (i.e., create an empty array of the given size and fill all pixels with 255), the tiles are still rendered black.\n\n", "created_at": "2021-11-16T17:23:53Z", "version": "v6.9.0", "environment_setup_commit": "a715b1f5c6a64f80298fd424306f2bafb9752e4c", "image_assets": "{\"problem_statement\": [\"assets/openlayers__openlayers-13013/001.png\", \"assets/openlayers__openlayers-13013/002.png\", \"assets/openlayers__openlayers-13013/003.png\", \"assets/openlayers__openlayers-13013/004.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "openlayers/openlayers", "instance_id": "openlayers__openlayers-15685", "base_commit": "7a88544d0f02253088bedaf8c0adf3a73d24739b", "problem_statement": "Problem with setFill() for styling VectorTile\nIn ol@9.1.0 when I am creating new ol.layer.VectorTile({}) and styling it, I am having issue with setFill() which makes additional semitransparent layer at the top of the other layers.\n\nSteps to reproduce the behavior:\n\n1. Use style \n ```js\n let selectionStyle = new ol.style.Style({\n text: new ol.style.Text({\n text: text,\n overflow: true,\n font: `12px sans-serif`,\n padding: [5, 10, 5, 10],\n textAlign: 'center',\n justify: 'center',\n textBaseline: 'middle',\n fill: new ol.style.Fill({\n color: '#fff',\n }),\n backgroundFill: new ol.style.Fill({\n color: '#454545',\n }),\n backgroundStroke: new ol.style.Stroke({\n color: '#fff',\n width: 1,\n }),\n }),\n stroke: new ol.style.Stroke({\n color: '#000',\n width: 2,\n }),\n fill: new ol.style.Fill({ color: 'rgba(255, 255, 255, 0.5)' }),\n });\n ```\n2. Create layer\n ```js\n const myLayer = new ol.layer.VectorTile({\n properties: { name:'Test' },\n visible: true,\n renderMode: 'vector',\n source: new ol.source.VectorTile({\n format: new ol.format.MVT({\n idProperty: 'id',\n }),\n tileLoadFunction: ...,\n url: `some_url`,\n }),\n style: selectionStyle,\n });\n ```\n3. **If I remove fill** from the style **it will be ok** and without that additional layer but I need Fill to style myLayer\n\nIf I use\n```js\n selectionStyle.setFill(\n new ol.style.Fill({\n color: 'rgba(252, 220, 215, 0.5)'\n })\n );\n```\nit will as well make that additional ghost layer on some VectorTiles. Additionally on the web side, on mouse hover will cause map to flicker when moving cursor around.\n\n**Expected behavior**\nFill in Style should Fill VectorTile and should not create any additional layers above. Please check notes on attached image. \n\n\n\n\n## Relevant pre-fix discussion\n\n### Issue #15681\n\n#### Comment at 2024-03-28T18:13:34Z\n\nCheck image and try this code. Zoom 3-4 times and move to Spain for example. Line 45 paint countries to red color, but also render on some text labels some ghost boxes. If you remove that line (fill: new Fill({color: 'rgba(252, 220, 215, 0.5)'}),) boxes disappears but countries are not filled with color. :(\n\nhttps://codesandbox.io/p/sandbox/vector-tile-selection-forked-2l6llr\n\n\n\nIf you move map you will notice some labels have those ghost boxes and that is random, not everywhere.\n\n\n\nThis is not just cosmetic issue, but in more complex situation where you have mouse hover it start flickering, like changing randomly labels and VectorTiles background transparency.", "created_at": "2024-03-28T21:50:47Z", "version": "v9.1.0", "environment_setup_commit": "7a88544d0f02253088bedaf8c0adf3a73d24739b", "image_assets": "{\"problem_statement\": [\"assets/openlayers__openlayers-15685/001.png\", \"assets/openlayers__openlayers-15685/002.png\", \"assets/openlayers__openlayers-15685/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "outline/outline", "instance_id": "outline__outline-12569", "base_commit": "8464d995899c84917aad8d085a967776568748fc", "problem_statement": "API MARKDOWN ROUND-TRIP ISSUE FOR TABLES, INLINE CODE, AND LATEX\n### Is there an existing issue for this?\n\n- [x] I have searched the existing issues\n\n### This is not related to configuring Outline\n\n- [x] The issue is not related to self-hosting config\n\n### Current Behavior\n\nShort report:\n> When a document contains special Markdown formatting such as inline code, LaTeX/math, and pipe-containing expressions inside table cells, `documents.info` and `documents.update` may not be fully round-trip safe.\nRe-importing exported Markdown can change formatting, split table rows, leak content outside the table, or truncate the original content. The API export does not always appear to be a reliable re-import payload for the same document.\n\n## OBSERVED BEHAVIOR\nThe problematic cases seem to appear when Markdown combines:\n\n- table cells\n- inline code formatting such as `` `...` ``\n- LaTeX/math blocks such as `$$...$$`\n- special characters such as `|`, `_`, `{`, `}`, and backticks\n\nExamples of failure modes observed through the API:\n\n1. A table cell containing expressions such as `` `|Clock_NTP_Camera1 - Clock_GPS_Camera1|` `` is not preserved exactly after API import/export.\n2. Math content containing pipe characters may be exported in one Markdown shape and then re-imported into a different internal structure.\n3. Re-importing Markdown through `documents.update` can split one logical table row into multiple fragments.\n4. Content that should stay inside a table cell can appear outside the table after re-import.\n5. Exact Markdown formatting is not preserved, even when the logical content is intended to remain unchanged.\n\nThis seems especially relevant for technical documentation using:\n\n- formulas\n- code-like expressions\n- identifiers with underscores/braces\n- pipe-delimited expressions\n- tables combining prose and notation\n\n### Expected Behavior\n\nIdeally, the following would hold:\n\n1. `documents.info` would return Markdown that can be sent back through `documents.update` without changing the document structure unexpectedly.\n2. If Outline intentionally normalizes Markdown, that normalization would still preserve the exact document structure and content boundaries.\n3. Special formatting inside table cells would remain stable and would not truncate content, split rows, or leak nodes outside the table.\n\n### Steps To Reproduce\n\nUse a document containing a Markdown table with a cell that mixes prose and special formatting.\n\nExample input:\n\n```md\n| Condition | Facts |\n| --- | --- |\n| Absolute time difference | The system checks `|Clock_NTP_Camera1 - Clock_GPS_Camera1|` and `|Clock_NTP_Camera2 - Clock_GPS_Camera2|`. |\n```\n\nThen:\n1. create or update the document through `documents.update`\n2. fetch it with `documents.info`\n3. send the fetched Markdown back again through `documents.update`\n4. fetch it once more\n\nObserved result:\n\n- the exported Markdown may no longer match the original input\n- replaying the exported Markdown may further alter the document\n- the table cell may be truncated or structurally split\n\n## IMPACT\nThis makes reliable external tooling much harder for clients that sync Markdown to Outline through the public API.\n\n### Environment\n\n```markdown\n- Outline:\n- Browser:\n```\n\n### Anything else?\n\n_No response_\n\n## Relevant pre-fix discussion\n\n### Issue #12510\n\n#### Comment at 2026-05-29T16:25:31Z\n\n# VERIFICATION REPORT: MARKDOWN TABLE-CELL IMPORT ISSUE PARTIALLY PERSISTS\n\n## SUMMARY\n\nI verified the behavior against the locally deployed instance at:\n\n- `https://local.outline.dev:3000/`\n\nThe result is:\n\n- **the issue is partially fixed, but not fully resolved**\n- the inline-code table case now renders correctly\n- the math-in-table case still renders incorrectly\n- the unescaped-math variant is even worse and gets truncated on import/export\n- exported Markdown appears stable on replay only after Outline has already normalized or corrupted the imported content\n\nIn short:\n\n> The recent fix appears to improve replay stability and to fix the simple inline-code table case, but it does not fix the math-in-table failures: one variant renders incorrectly and another is truncated.\n\n## WHAT I TESTED\n\nI tested two Markdown table-cell patterns:\n\n1. inline code containing literal pipe characters\n2. math content containing escaped pipe characters, `<br>`, braces, underscores, and a trailing inline code span\n\nThe checks were:\n\n1. create a document from the Markdown\n2. inspect the rendered result in the UI\n3. fetch the exported Markdown\n4. replay that exported Markdown through the API\n5. fetch the exported Markdown again and compare\n\nThis separates two different questions:\n\n1. **does the document render correctly after import?**\n2. **is the exported Markdown stable when replayed?**\n\nBoth matter, but render correctness is the primary user-facing requirement.\n\n## CASE 1: INLINE CODE WITH PIPE CHARACTERS INSIDE A TABLE CELL\n\n### Input Markdown\n\n```md\n| Condition | Facts |\n| --- | --- |\n| Absolute time difference | The system checks `|Clock_NTP_Camera1 - Clock_GPS_Camera1|` and `|Clock_NTP_Camera2 - Clock_GPS_Camera2|`. |\n```\n\n### Expected Behavior\n\nThe `Facts` cell should render as a single table cell containing prose and two inline code spans:\n\n- `` `|Clock_NTP_Camera1 - Clock_GPS_Camera1|` ``\n- `` `|Clock_NTP_Camera2 - Clock_GPS_Camera2|` ``\n\nThe leading and trailing pipe characters should remain part of each code span.\n\n### Actual Behavior\n\nThis case now renders correctly:\n\n- the code is still rendered as highlighted inline code\n- the pipe characters are preserved\n- the visual line breaks in the screenshot are consistent with normal wrapping inside a narrow table cell, not with a parsing error\n\nThis case should therefore be treated as a **passing regression check**, not as an active repro.\n\n### Evidence\n\nScreenshots:\n\n\n\n\n## ACTIVE REPRO: MATH / LATEX WITH PIPE CHARACTERS INSIDE A TABLE CELL\n\n### Input Markdown\n\n```md\n| Condition | Facts |\n| --- | --- |\n| Excessive absolute difference between primary and secondary time references | The system checks whether <br>$$\\|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}\\|$$<br> and $$\\|Clock_{NTP_{Camera2}} - Clock_{GPS_{Camera2}}\\|$$ remain below `Time_difference_treshold`. |\n```\n\nWhich should render in .md:\n\nThe system checks whether <br>$$|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}|$$<br> and $$|Clock_{NTP_{Camera2}} - Clock_{GPS_{Camera2}}|$$ remain below `Time_difference_treshold`.\n\n### Expected Behavior\n\nThe `Facts` cell should render as a single table cell containing:\n\n- the prose text\n- the intended `<br>` line breaks\n- two intact math expressions\n- the final inline code span `` `Time_difference_treshold` ``\n\nAt minimum, the imported result should preserve the intended meaning and structure of the cell content.\n\n### Actual Behavior\n\nThis case is still broken:\n\n- the math is shown as raw `$$...$$` text rather than rendering correctly\n- the result is visibly different from the source Markdown intent\n- the import path is still unreliable for this content pattern\n\n### Evidence\n\nScreenshots:\n\n\n\n## ADDITIONAL NOTE: UNESCAPED PIPE CHARACTERS INSIDE `$$...$$` CAUSE TRUNCATION\n\n### Input Markdown\n\n```md\n| Condition | Facts |\n| --- | --- |\n| Excessive absolute difference between primary and secondary time references | The system checks whether <br>$$|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}|$$<br> and $$|Clock_{NTP_{Camera2}} - Clock_{GPS_{Camera2}}|$$ remain below `Time_difference_treshold`. |\n```\n\n### Expected Behavior\n\nThis should display as LaTeX inside the table cell and preserve the full sentence content.\n\n### Actual Behavior\n\nThis variant fails more severely than the escaped version:\n\n- the imported content is truncated\n- the rendered cell stops at `The system checks whether $$`\n- the exported Markdown is already corrupted after the first import\n\nThe first exported form was reduced to:\n\n```md\n| Condition | Facts |\n|-----------|-------|\n| Excessive absolute difference between primary and secondary time references | The system checks whether<br>$$ |\n```\n\nThis means the API accepts the input, but the parser effectively treats the raw pipe inside `$$...$$` as a table delimiter and loses the rest of the cell content.\n\n### Evidence\n\nScreenshot:\n\n\n\nUnescaped math table truncation\n\n## IMPORT CORRECTNESS VS EXPORT STABILITY\n\nThe verification shows that two properties need to be tracked separately:\n\n1. **import correctness**\n2. **export stability on replay**\n\nCurrent status:\n\n- case 1: import looks correct, export replay is stable\n- case 2: import is wrong, export replay is stable\n\nThis means the recent fix appears to have improved the exporter / replay behavior and to have fixed the simpler inline-code case, but it has **not** fixed the remaining math-heavy table-cell case.\n\n## WHY THIS STILL MATTERS FOR SYNC CLIENTS\n\nThis is still a real problem for external tooling that sends Markdown to the API.\n\nPractical consequences:\n\n- a sync push can still produce a visibly wrong document for valid Markdown\n- clients still cannot safely assume that math-heavy content inside table cells will survive import intact\n- technical documentation remains exposed because it often combines tables, formulas, identifiers, braces, underscores, and inline code\n\nSo while the situation is better than before, the Markdown import path is still not reliable for the remaining repro pattern.\n\n## REMAINING HIGH-RISK PATTERN\n\nThe remaining failed case combines:\n\nMarkdown tables\nrich content inside a table cell\nLaTeX / math delimiters\nescaped pipe characters\nbraces and underscores\nprose mixed with formatted content in the same cell\nThis suggests the remaining bug is in or near the parsing / tokenization path for math-rich content inside table cells.\n\n## RECOMMENDED VERIFICATION CASES\n\n### Regression Check\n\n```md\n| Condition | Facts |\n| --- | --- |\n| Absolute time difference | The system checks `|Clock_NTP_Camera1 - Clock_GPS_Camera1|` and `|Clock_NTP_Camera2 - Clock_GPS_Camera2|`. |\n```\n\nExpected result: renders correctly and remains stable on replay.\n\n### Active Repro\n\n```md\n| Condition | Facts |\n| --- | --- |\n| Excessive absolute difference between primary and secondary time references | The system checks whether <br>$$\\|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}\\|$$<br> and $$\\|Clock_{NTP_{Camera2}} - Clock_{GPS_{Camera2}}\\|$$ remain below `Time_difference_treshold`. |\n```\n\nExpected result after a full fix: renders correctly and remains stable on replay.\n\nCurrent observed result: replay is stable, but the rendered import is still wrong.\n\n## CONCLUSION\n\n**Verification result:** the issue is **partially fixed but still reproducible**.\n\nWhat appears fixed:\n\n- exported Markdown replay stability for the tested samples\n- the simple inline-code table-cell case\n\nWhat still fails:\n\n- the math-heavy table-cell import/render case\n\nSuggested conclusion to send back:\n\n> **Partially fixed, not fully resolved.** The local patch improves round-trip stability and fixes the simple inline-code case, but the math-in-table import/render bug still reproduces.\n\n#### Comment at 2026-06-01T09:52:14Z\n\nHello Tom,\nSorry to bother you again with that, but I believe that I have \"nailed\" down the \"Table/LaTex\" issue.\nThis is, I believe important as this situation breaks a simple .md API export, even rendered in a regular GitHub flavoured md renderer.\nThis has nothing to do with round-trip here.\n\nHere is the breakdown:\nA LaTex formula entered in Outline is automatically outputted via API in its **DISPLAY** form e.g.\n $|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}|$ ->\n```\n$$\n|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}|\n$$\n```\nThe issue is that this form is not valid in a table as encoding must follow these rules:\n-) Formulas must be encoded **INLINE**: `$$....$$` -> `$....$`;\n-) Because of INLINE, the `$` must directly surround the formula;\n-) Pipe character must be escaped (only in tables)\n\nExemple:\n\n\n\nComes out as:\n```\n| Excessive absolute difference between primary and secondary time references | $$\n|Clock_{NTP_{Camera1}} - Clock_{GPS_{Camera1}}|\n$$The system checks `\\|Clock_NTP_Camera1 - Clock_GPS_Camera1\\|` and `\\|Clock_NTP_Camera2 - Clock_GPS_Camera2\\|`. If either value exceeds `Time_difference_treshold`, the candidate is rejected. The default value of `Time_difference_treshold` is **1000 ms**. |\n```\nWhich doesn't show correctly in VSCode:\n\n", "created_at": "", "version": "", "environment_setup_commit": "8464d995899c84917aad8d085a967776568748fc", "image_assets": "{\"problem_statement\": [\"assets/outline__outline-12569/001.png\", \"assets/outline__outline-12569/002.png\", \"assets/outline__outline-12569/003.png\", \"assets/outline__outline-12569/004.png\", \"assets/outline__outline-12569/005.png\", \"assets/outline__outline-12569/006.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "plotly/plotly.js", "instance_id": "plotly__plotly.js-1901", "base_commit": "b70ef1d7b482109cb6a85f83a4d4ca86093b9ab7", "problem_statement": "Histogram does not scale to the range of new data on restyle\nIf you set up a histogram...\n\n```\nvar data1 = [1, 2, 2, 3, 3, 3, 4, 4, 5];\nvar hist = document.getElementById('hist_div');\nPlotly.plot(hist, [{x: data1, type: 'histogram' }]);\n```\n\n... and then update this histogram with wider data followed by a restyle...\n\n```\nvar data2 = [10, 20, 20, 30, 30, 30, 40, 40, 50];\nPlotly.restyle(hist, {x: [data2]});\n```\n\n... the histogram keeps the old range (in this case 1 to 5) and does not scale to the range of the new data (which is the case when using scatter style plot). Here is a codepen showing this: http://codepen.io/rasmusab/pen/megKOo\n\nI have not found any documentation regarding whether this is the expected behavior. If it is, what would be the best practice for updating a histogram with new data?\n\n## Relevant pre-fix discussion\n\n### Issue #24\n\n#### Comment at 2017-07-20T16:45:00Z\n\nHa, for colorscales it turns out we [already do](https://github.com/plotly/plotly.js/blob/master/src/components/colorscale/calc.js#L57-L65) what I was about to propose here: if you explicitly specify `*auto: false`, then it will fill in the auto values but leave it looking like `*auto: false` so that future changes will not re-invoke the autoscaling; but if you don't specify an explicit `*auto` value it defaults to `true` since the explicit ranges are missing.\n\nSo for example:\n```\nvar x1 = [1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4];\nvar y1 = [1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4];\nPlotly.newPlot(gd, [{type: 'histogram2d', x: x1, y: y1, zauto: false}]);\n```\n\n\nand then:\n```\n// move the last point so we have one bin that gets 1 sample, another gets 3\nPlotly.restyle(gd,'x', [[1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 3]])\n```\n\nrange is still 1.5->2.5\n\nBut if you don't specify `zauto: false` in the first place, the initial plot is identical but after `restyle` you get:\n\nrecalculated autocolor: 1->3\n\nSo I think it's *only* in the context of histogram bins that this is a change - and therefore I'd consider it an inconsistency, a bug, not a breaking change, and I should implement the same logic with explicit `autobin*: false` values in histograms.", "created_at": "", "version": "", "environment_setup_commit": "b70ef1d7b482109cb6a85f83a4d4ca86093b9ab7", "image_assets": "{\"problem_statement\": [\"assets/plotly__plotly.js-1901/001.png\", \"assets/plotly__plotly.js-1901/002.png\", \"assets/plotly__plotly.js-1901/003.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "plotly/plotly.js", "instance_id": "plotly__plotly.js-4977", "base_commit": "c925cd4688df2fc5fc9aacce1d87f36877ae5196", "problem_statement": "hoverlabel not working properly with subplots when resizing window\nThere is something buggy going on with my subplots using x unified\n\nWhen creating the plot on its own, the hoverlabel is fine, but when I put it inside a dash app via dcc.Graph component, the labels don’t show. I created two versions in chart studio, one that is exported from the dash app and the second created using the `chart_studio` module.\n\nThe latter works fine, but the former works initially, but then all but the last plot’s hover labels no longer work when the window is resized.\n\nAny idea what is going on here?\n\nCreated using dash app via dcc.Graph: https://chart-studio.plotly.com/~mbkupfer/72/#/\n\nCreated using chart_studio (works good): https://chart-studio.plotly.com/~mbkupfer/70/#plot\n\nDependencies:\n\n```\ndash 1.13.3\ndash-core-components 1.10.1\ndash-html-components 1.0.3\ndash-renderer 1.5.0\ndash-table 4.8.1\nplotly 4.8.2\n```\n**[Update]**\nIt actually looks like this issue occurs in both created plots.\n\n## Relevant pre-fix discussion\n\n### Issue #4971\n\n#### Comment at 2020-06-30T22:38:42Z\n\n> Looks like the expected behaviour to me.\n>> For the hover data to not show?\n>>> For the hover not to show across subplots.\n\nOh, no...what I meant was that it simply didn't show. For example, in first screenshot, I get a unified hoverlabel on the first plot, but when I resize the window it no longer shows...all I get are the spikelines. The only subplot that shows hoverlabels at this point is the last one.\n\n\n\n", "created_at": "", "version": "", "environment_setup_commit": "c925cd4688df2fc5fc9aacce1d87f36877ae5196", "image_assets": "{\"problem_statement\": [\"assets/plotly__plotly.js-4977/001.png\", \"assets/plotly__plotly.js-4977/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "pnpm/pnpm", "instance_id": "pnpm__pnpm-4969", "base_commit": "e9946fcfa7e1892fb550a2164c2040fea2574cdd", "problem_statement": "Lockfile generated after `pnpm patch` cannot be used with `--no-frozen-lockfile` on CI for Windows\n### pnpm version: 7.4.1\n\n### Code to reproduce the issue:\n\nI [opened a PR](https://github.com/withastro/astro/pull/3747) to replace `patch-package` with `pnpm patch` and `pnpm patch-commit`, but [our CI is failing](https://github.com/withastro/astro/runs/7136424474?check_suite_focus=true) on **Windows** only.\n\nOur lockfile was generated on macOS, so I'm wondering if the paths saved to the lockfile https://github.com/withastro/astro/pull/3747/files#diff-32824c984905bb02bc7ffcef96a77addd1f1602cff71a11fbbfdd7f53ee026bbR3-R10 are the culprit here?\n\n### Expected behavior:\n\nCI environments defaulting to `--no-frozen-lockfile` should still work on Windows.\n\n### Actual behavior:\n\nOn Windows only, we're getting this error when running `pnpm install` in CI.\n\n```\nERR_PNPM_FROZEN_LOCKFILE_WITH_OUTDATED_LOCKFILE Cannot perform a frozen installation because the lockfile needs updates\n\nNote that in CI environments this setting is true by default. If you still need to run install in such cases, use \"pnpm install --no-frozen-lockfile\"\n```\n\n### Additional information:\n\n - `node -v` prints: 14, 16\n - Windows, macOS, or Linux?: **Windows** only\n\n## Relevant pre-fix discussion\n\n### Issue #4961\n\n#### Comment at 2022-07-01T03:59:16Z\n\nIf command is: `pnpm install --no-frozen-lockfile` in Windows. `pnpm-lock.yaml` file diff is:\n\n```diff\ndiff --git a/pnpm-lock.yaml b/pnpm-lock.yaml\nindex a6b1d0a9..434d9ae8 100644\n--- a/pnpm-lock.yaml\n+++ b/pnpm-lock.yaml\n@@ -2,10 +2,10 @@ lockfileVersion: 5.4\n\n patchedDependencies:\n '@changesets/cli@2.23.0':\n- hash: kcozqtpxuwjzskw6zg5royevn4\n+ hash: kuwjnalnvawb7qsar2xshwwayi\n path: patches/@changesets__cli@2.23.0.patch\n '@changesets/assemble-release-plan@5.1.3':\n- hash: t3ywtunilq53oqooixy7kemkoi\n+ hash: 6bkrbdwcq4wpuel6jqtb3kupwi\n path: patches/@changesets__assemble-release-plan@5.1.3.patch\n \n importers:\n@@ -35,9 +35,9 @@ importers:\n dependencies:\n '@astrojs/webapi': link:packages/webapi\n devDependencies:\n- '@changesets/assemble-release-plan': 5.1.3_t3ywtunilq53oqooixy7kemkoi\n+ '@changesets/assemble-release-plan': 5.1.3_6bkrbdwcq4wpuel6jqtb3kupwi\n '@changesets/changelog-github': 0.4.4\n- '@changesets/cli': 2.23.0_kcozqtpxuwjzskw6zg5royevn4\n+ '@changesets/cli': 2.23.0_kuwjnalnvawb7qsar2xshwwayi\n '@octokit/action': 3.18.1\n '@typescript-eslint/eslint-plugin': 5.30.0_5mtqsiui4sk53pmkx7i7ue45wm\n '@typescript-eslint/parser': 5.30.0_b5e7v2qnwxfo6hmiq56u52mz3e\n@@ -3995,7 +3995,7 @@ packages:\n semver: 5.7.1\n dev: true\n \n- /@changesets/assemble-release-plan/5.1.3_t3ywtunilq53oqooixy7kemkoi:\n+ /@changesets/assemble-release-plan/5.1.3_6bkrbdwcq4wpuel6jqtb3kupwi:\n resolution: {integrity: sha512-I+TTkUoqvxBEuDLoJfJYKDXIJ+nyiTbVJ8KGhpXEsLq4N/ms/AStSbouJwF2d/p3cB+RCPr5+gXh31GSN4kA7w==}\n dependencies:\n '@babel/runtime': 7.18.6\n@@ -4023,13 +4023,13 @@ packages:\n - encoding\n dev: true\n \n- /@changesets/cli/2.23.0_kcozqtpxuwjzskw6zg5royevn4:\n+ /@changesets/cli/2.23.0_kuwjnalnvawb7qsar2xshwwayi:\n resolution: {integrity: sha512-Gi3tMi0Vr6eNd8GX6q73tbOm9XOzGfuLEm4PYVeWG2neg5DlRGNOjYwrFULJ/An3N9MHtHn4r5h1Qvnju9Ijug==}\n hasBin: true\n dependencies:\n '@babel/runtime': 7.18.6\n '@changesets/apply-release-plan': 6.0.0\n- '@changesets/assemble-release-plan': 5.1.3_t3ywtunilq53oqooixy7kemkoi\n+ '@changesets/assemble-release-plan': 5.1.3_6bkrbdwcq4wpuel6jqtb3kupwi\n '@changesets/changelog-git': 0.1.11\n '@changesets/config': 2.0.0\n '@changesets/errors': 0.1.4\n@@ -4104,7 +4104,7 @@ packages:\n resolution: {integrity: sha512-5C1r4DcOjVxcCvPmXpymeyT6mdSTLCNiB2L+5uf19BRkDKndJdIQorH5Fe2XBR2nHUcZQFT+2TXDzCepat969w==}\n dependencies:\n '@babel/runtime': 7.18.6\n- '@changesets/assemble-release-plan': 5.1.3_t3ywtunilq53oqooixy7kemkoi\n+ '@changesets/assemble-release-plan': 5.1.3_6bkrbdwcq4wpuel6jqtb3kupwi\n '@changesets/config': 2.0.0\n '@changesets/pre': 1.0.11\n '@changesets/read': 0.5.5\n```\n\nIf you put these changes on macOS, the same error will be reported:\n\n\n#### Comment at 2022-07-01T08:45:59Z\n\nI have created a reproduction repo for this issue: https://github.com/BlackHole1/test-pnpm-4961\nThis is basically where the problem code is located:\nhttps://github.com/pnpm/pnpm/blob/b2d271d29ce9767e24e900df760e6c0b070ec759/packages/crypto.base32-hash/src/index.ts#L6\n\nSee: https://github.com/BlackHole1/test-pnpm-4961/actions/runs/2595432720\n\nmacOS: https://github.com/BlackHole1/test-pnpm-4961/runs/7146689132?check_suite_focus=true#step:5:11\n\n\nLinux: https://github.com/BlackHole1/test-pnpm-4961/runs/7146689233?check_suite_focus=true#step:5:11\n\n\n\nWindows: https://github.com/BlackHole1/test-pnpm-4961/runs/7146689367?check_suite_focus=true#step:5:11\n\n\n#### Comment at 2022-07-01T08:58:32Z\n\nWhen use: `Buffer.from(str).toString('base64');`, result diff:\n\n\n\nSo here's the problem: https://github.com/pnpm/pnpm/blob/b2d271d29ce9767e24e900df760e6c0b070ec759/packages/crypto.base32-hash/src/index.ts#L10\n\n#### Comment at 2022-07-01T09:22:20Z\n\nWhen file content to buffer. Windows system has additional `d0` hex.\n\n\nI think it's a nodejs bug", "created_at": "2022-07-02T12:06:33Z", "version": "v7.4.1", "environment_setup_commit": "e9946fcfa7e1892fb550a2164c2040fea2574cdd", "image_assets": "{\"problem_statement\": [\"assets/pnpm__pnpm-4969/001.png\", \"assets/pnpm__pnpm-4969/002.png\", \"assets/pnpm__pnpm-4969/003.png\", \"assets/pnpm__pnpm-4969/004.png\", \"assets/pnpm__pnpm-4969/005.png\", \"assets/pnpm__pnpm-4969/006.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "processing/p5.js", "instance_id": "processing__p5.js-8712", "base_commit": "9206f8d542b248d767ecdc652a26de988926952c", "problem_statement": "Points are always rendered in the center\n### Most appropriate sub-area of p5.js?\n\n- [ ] Accessibility\n- [ ] Color\n- [ ] Core/Environment/Rendering\n- [ ] Data\n- [ ] DOM\n- [ ] Events\n- [ ] Image\n- [ ] IO\n- [ ] Math\n- [ ] Typography\n- [ ] Utilities\n- [ ] p5.strands\n- [x] WebGL\n- [ ] DevOps, Build process, Unit testing\n- [ ] Internationalization (i18n)\n- [ ] Friendly Errors\n- [ ] Other (specify if possible)\n\n### p5.js version\n\n1.11.12\n\n### Web browser and version\n\nChrome 146.0.7680.165\n\n### Operating system\n\nWindows\n\n### Steps to reproduce this\n\nfunction setup() {\n createCanvas(600, 600, WEBGL);\n}\n\nfunction draw() {\n background(0);\n\n camera(0, 0, 200, 0, 0, 0, 0, 1, 0);\n\n stroke(255);\n strokeWeight(8);\n \n point(20, 20, 0);\n}\n\nthe behavior is very different from 1.11.11 and 2.0\n\n## Relevant pre-fix discussion\n\n### Issue #8701\n\n#### Comment at 2026-04-07T12:35:49Z\n\nI can't reproduce this issue on Linux. I get the same output for both versions. But I don't think it is a problem with the OS\n\n\n\n\n\n#### Comment at 2026-04-07T16:58:12Z\n\n\n\n\n\n#### Comment at 2026-04-07T17:01:23Z\n\n\n\n#### Comment at 2026-04-07T17:03:46Z\n\n\n\n#### Comment at 2026-04-08T02:45:38Z\n\nthe same picture in the last two examples, in the references section on the website\nhttps://p5js.org/reference/p5/point/\n", "created_at": "2026-04-08T11:14:31Z", "version": "v1.11.12", "environment_setup_commit": "9206f8d542b248d767ecdc652a26de988926952c", "image_assets": "{\"problem_statement\": [\"assets/processing__p5.js-8712/001.png\", \"assets/processing__p5.js-8712/002.png\", \"assets/processing__p5.js-8712/003.png\", \"assets/processing__p5.js-8712/004.png\", \"assets/processing__p5.js-8712/005.png\", \"assets/processing__p5.js-8712/006.png\", \"assets/processing__p5.js-8712/007.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "quarto-dev/quarto-cli", "instance_id": "quarto-dev__quarto-cli-13215", "base_commit": "77a82a631bcc6219c4f7892f7d538f5e2b58cd57", "problem_statement": "Incorrect rendering of URL logos supplied to _brand.yml when rendering to HTML\n### Bug description\n\nWhen brand logos are supplied via a URL, quarto renders the URL in the HTML using `\\` instead of `/`. This will appropriately show the logo when previewing the render locally, but when pushed to a GitHub Pages site the URL gets converted to a relative path based on the current website. For example, if I have a logo hosted at `https://mysite.github.io/branding/logos/example.png`, and I try to link to it in the `_brand.yml` at `https://mysite.github.io/mypage`, the URL embedded in the html will convert the logo URL to `https:\\mysite.github.io\\branding\\logos\\example.png`, which will try to redirect to `https://mysite.github.io/mysite.github.io/branding/logos/example.png` (real example shown in the screenshot)\n\n\n\n### Steps to reproduce\n\n````qmd\n---\ntitle: \"Reproducible Quarto Document\"\nformat: html\nbrand:\n logo: \n images:\n quarto-logo: \n path: \n alt: \"Quarto icon\"\n small: quarto-logo\n---\n\nThis is a reproducible Quarto document using `format: html`. It is written in Markdown and contains embedded Python code. When you run the code, it will produce a message.\n\n{{< lipsum 1 >}}\n\n{{< brand logo small>}}\n\nThe end.\n````\n\n### Actual behavior\n\nHTML renders the image as follows: `<img src=\"https:\\quarto.org\\quarto.png\" class=\"img-fluid\">`\nWhen shown in a local preview, this renders the image correctly. When hosted on an external website using GitHub Pages, this appends the image URL to the website URL\n\n### Expected behavior\n\nHTML renders the image with the full URL: ``\nWhen shown on an external website in a browser, this should correctly render the image.\n\n### Your environment\n\n- IDE: Positron 2025.05.0 build 58\n- OS: Windows 11\n\n### Quarto check output\n\n```\nQuarto 1.7.29\n[>] Checking environment information...\n Quarto cache location: C:\\Users\\Jonathan.Peake\\AppData\\Local\\quarto\n[>] Checking versions of quarto binary dependencies...\n Pandoc version 3.6.3: OK\n Dart Sass version 1.85.1: OK\n Deno version 1.46.3: OK\n Typst version 0.13.0: OK\n[>] Checking versions of quarto dependencies......OK\n[>] Checking Quarto installation......OK\n Version: 1.7.29\n Path: C:\\Users\\Jonathan.Peake\\AppData\\Local\\Programs\\Quarto\\bin\n CodePage: 1252\n\n[>] Checking tools....................OK\n TinyTeX: (external install)\n Chromium: (not installed)\n\n[>] Checking LaTeX....................OK\n Using: TinyTex\n Path: C:\\Users\\Jonathan.Peake\\AppData\\Roaming\\TinyTeX\\bin\\windows\\\n Version: 2025\n\n[>] Checking Chrome Headless....................OK\n Using: Chrome found on system\n Path: C:\\Program Files\\Google\\Chrome\\Application\\chrome.exe\n Source: Windows Registry\n\n[>] Checking basic markdown render....OK\n\n[>] Checking Python 3 installation....OK\n Version: 3.12.9 (Conda)\n Path: C:/Users/Jonathan.Peake/AppData/Local/miniforge3/python.EXE\n Jupyter: (None)\n\n Jupyter is not available in this Python installation.\n Install with conda install jupyter\n\n[>] Checking R installation...........OK\n Version: 4.4.2\n Path: C:/Users/Jonathan.Peake/AppData/Local/Programs/R/R-4.4.2\n LibPaths:\n - C:/Users/Jonathan.Peake/AppData/Local/Programs/R/R-4.4.2/library\n knitr: 1.50\n rmarkdown: 2.29\n\n[>] Checking Knitr engine render......OK\n```\n\n## Relevant pre-fix discussion\n\n### Issue #12643\n\n#### Comment at 2025-04-29T18:50:18Z\n\nI also created a live GitHub Pages site from the reprex quarto file: https://jonpeake.github.io/quarto-reprex/reprex.html\n\nWhen accessed from site url:\n\n\n\nWhen previewing locally:\n\n\n\n#### Comment at 2025-04-29T22:43:45Z\n\nI followed the formatting used in this doc, which suggests that a URL is not only allowed but supported:\nhttps://quarto.org/docs/authoring/brand.html\n\n\n\n", "created_at": "2025-08-13T15:05:52Z", "version": "v1.8.19", "environment_setup_commit": "77a82a631bcc6219c4f7892f7d538f5e2b58cd57", "image_assets": "{\"problem_statement\": [\"assets/quarto-dev__quarto-cli-13215/001.png\", \"assets/quarto-dev__quarto-cli-13215/002.png\", \"assets/quarto-dev__quarto-cli-13215/003.png\", \"assets/quarto-dev__quarto-cli-13215/004.png\", \"assets/quarto-dev__quarto-cli-13215/005.png\", \"assets/quarto-dev__quarto-cli-13215/006.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "quarto-dev/quarto-cli", "instance_id": "quarto-dev__quarto-cli-5018", "base_commit": "7f8bc06d2ddbc5ddd1b8ae21e63d4ed5ee0c5861", "problem_statement": "Error rendering code in callout (code listing in a callout)\nError:\n\n- The error appears to be caused by placing attributed code inside a callout \n- Issue could be that we're incorrectly turning this into a code listing\n- Issue could be that we don't properly support code listings in callouts\n\n\n```\nLaTeX Error: Not in outer par mode.\n\nSee the LaTeX manual or LaTeX Companion for explanation.\nType H <return> for immediate help.\n ...\n\nl.221\n```\n\n\nrepro:\n\n````markdown\n\n---\ntitle: Hello\nformat: pdf\nkeep-tex: true\n---\n\n::: {.callout-note}\nIf the `www-frame-origin` option is set to `same` or `any` instead of a URI, you must add the following setting with the actual URI:\n\n```{.ini filename=\"/etc/rstudio/rsession.conf\"}\ntutorial-api-client-origin=http://localhost:8080\n```\n:::\n\n````\n\n\nrelevant tex diff:\n\n**Good**\n```latex\n\\begin{tcolorbox}[enhanced jigsaw, left=2mm, rightrule=.15mm, arc=.35mm, toprule=.15mm, coltitle=black, bottomtitle=1mm, titlerule=0mm, opacityback=0, breakable, opacitybacktitle=0.6, toptitle=1mm, colbacktitle=quarto-callout-note-color!10!white, colback=white, title=\\textcolor{quarto-callout-note-color}{\\faInfo}\\hspace{0.5em}{Note}, bottomrule=.15mm, leftrule=.75mm]\n\nIf the \\texttt{www-frame-origin} option is set to \\texttt{same} or\n\\texttt{any} instead of a URI, you must add the following setting with\nthe actual URI:\n\n\\textbf{/etc/rstudio/rsession.conf}\n\n\\begin{Shaded}\n\\begin{Highlighting}[]\n\\DataTypeTok{tutorial{-}api{-}client{-}origin}\\OtherTok{=}\\StringTok{http://localhost:8080}\n\\end{Highlighting}\n\\end{Shaded}\n\n\\end{tcolorbox}\n```\n\n**Bad**\n```latex\n\\begin{tcolorbox}[enhanced jigsaw, leftrule=.75mm, arc=.35mm, toptitle=1mm, left=2mm, colback=white, breakable, colbacktitle=quarto-callout-note-color!10!white, bottomtitle=1mm, title=\\textcolor{quarto-callout-note-color}{\\faInfo}\\hspace{0.5em}{Note}, bottomrule=.15mm, titlerule=0mm, opacityback=0, colframe=quarto-callout-note-color-frame, toprule=.15mm, rightrule=.15mm, opacitybacktitle=0.6, coltitle=black]\n\nIf the \\texttt{www-frame-origin} option is set to \\texttt{same} or\n\\texttt{any} instead of a URI, you must add the following setting with\nthe actual URI:\n\n\\begin{codelisting}\n\n\\caption{\\texttt{/etc/rstudio/rsession.conf}}\n\n\\begin{Shaded}\n\\begin{Highlighting}[]\n\\DataTypeTok{tutorial{-}api{-}client{-}origin}\\OtherTok{=}\\StringTok{http://localhost:8080}\n\\end{Highlighting}\n\\end{Shaded}\n\n\\end{codelisting}\n\n\\end{tcolorbox}\n```\n\n## Relevant pre-fix discussion\n\n### Issue #5011\n\n#### Comment at 2023-03-29T03:39:27Z\n\nLet's talk about this live tomorrow.\n\nThe issue is the presence of `filename=\"/etc/rstudio/rsession.conf\"`. That causes us to attempt to render the code block as a listing, so that we can do the caption treatment on it.\n\nIn 1.2, we'd render the code block by itself like this:\n\n\n\nIn 1.3, we render the code block by itself like this:\n\n\n\nI did this by design, but only because I didn't consider that people would want to put a listing inside a callout.\n\nI think the new appearance is better, but I agree that we should have a way to disable this treatment since LaTeX callouts are necessarily relatively limited in functionality.\n\n#### Comment at 2023-03-29T03:41:56Z\n\nWithout the filename attribute, we render successfully in 1.3:\n\n\n\nThis suggests a workaround: move `/etc/rstudio/rsession.conf` to outside the listing (and filename declaration), and into plain markdown content before the code block, `**/etc/rstudio/rsession.conf**`:\n\n````\n---\ntitle: Hello\nformat: pdf\n---\n\n::: {.callout-note}\nIf the `www-frame-origin` option is set to `same` or `any` instead of a URI, you must add the following setting with the actual URI:\n\n**/etc/rstudio/rsession.conf**\n\n```{.ini}\ntutorial-api-client-origin=http://localhost:8080\n```\n:::\n````\n\n\n\n\n", "created_at": "2023-03-29T12:37:43Z", "version": "v1.3.294", "environment_setup_commit": "7f8bc06d2ddbc5ddd1b8ae21e63d4ed5ee0c5861", "image_assets": "{\"problem_statement\": [\"assets/quarto-dev__quarto-cli-5018/001.png\", \"assets/quarto-dev__quarto-cli-5018/002.png\", \"assets/quarto-dev__quarto-cli-5018/003.png\", \"assets/quarto-dev__quarto-cli-5018/004.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "react-hook-form/react-hook-form", "instance_id": "react-hook-form__react-hook-form-6774", "base_commit": "c124399551e15ebc2d19a5ba94f336db3f02564e", "problem_statement": "issue: \"isDirty\" and \"dirtyFields\" work incorrectly after calling \"reset\" or \"setValue\"\n### Version Number\n\n7.17.2\n\n### Codesandbox/Expo snack\n\nhttps://codesandbox.io/s/elegant-goldberg-ebrms\n\n### Steps to reproduce\n\n1st Use Case\n1. Go to https://codesandbox.io/s/elegant-goldberg-ebrms\n2. Click on 'SET VALUE' button\n3. See `isDirty === false` and `dirtyFields` is empty\n\n2nd Use Case\n1. Go to https://codesandbox.io/s/elegant-goldberg-ebrms\n2. Click on 'SET VALUE' button\n3. Change Text field\n4. See `isDirty === true` and `dirtyFields === {\"text\":true}`\n\n3rd Use Case\n1. Go to https://codesandbox.io/s/elegant-goldberg-ebrms\n2. Click on 'RESET' button\n3. See `isDirty === false` and `dirtyFields` is empty\n\n4rd Use Case\n1. Go to https://codesandbox.io/s/elegant-goldberg-ebrms\n2. Click on 'RESET' button\n3. Change Text field\n4. See `isDirty === true` and `dirtyFields === {\"text\":true}`\n\n### Expected behaviour\n\n1st Use Case\n`isDirty` should be `true` and `dirtyFields` should contain both fields\n\n2nd Use Case\n`isDirty` should be `true` and `dirtyFields` should contain both fields\n\n3rd Use Case\n`isDirty` should be `true` and `dirtyFields` should contain both fields\n\n4th Use Case\n`isDirty` should be `true` and `dirtyFields` should contain both fields\n\nExplanation: in all use cases the values are different from the default values, so the `isDirty` flag should be set to `true` and `dirtyFields` should contain all the fields that a different from the default values.\n\nI believe this issue relates to the recently closed issue https://github.com/react-hook-form/react-hook-form/issues/6721 and the related discussion https://github.com/react-hook-form/react-hook-form/discussions/6710\n\n### What browsers are you seeing the problem on?\n\nChrome\n\n### Relevant log output\n\n_No response_\n\n### Code of Conduct\n\n- [X] I agree to follow this project's Code of Conduct\n\n## Relevant pre-fix discussion\n\n### Issue #6764\n\n#### Comment at 2021-10-11T23:06:05Z\n\ncase 1 and 2\n\nDid you miss the following for `shouldDirty`?\n\n\n\ncase 4\n\n- you should consider providing a `defualtValues` for the reset state without react hook form would refer to the original `defaultValues`.\n\n```tsx\n<Button\n variant=\"contained\"\n onClick={() => {\n formMethods.reset(); // equal defaultValues\n }}\n>\n Reset defaults\n</Button>\n```\n", "created_at": "2021-10-12T06:12:20Z", "version": "v7.17.2", "environment_setup_commit": "c124399551e15ebc2d19a5ba94f336db3f02564e", "image_assets": "{\"problem_statement\": [\"assets/react-hook-form__react-hook-form-6774/001.png\", \"assets/react-hook-form__react-hook-form-6774/002.png\"], \"patch\": [], \"test_patch\": []}"} |
| {"repo": "storybookjs/storybook", "instance_id": "storybookjs__storybook-22336", "base_commit": "3e32fc62fdb93c78bb5c4ee5ce76a10d1b4cb7b6", "problem_statement": "preview > decorators causes: \"Rendered more hooks than during the previous render.\"\n**Describe the bug**\nWhen adding a decorator to `.storybook/preview.js` an error occurs: \"Rendered more hooks than during the previous render.\"\n\n```\nError\nRendered more hooks than during the previous render.\nCall Stack\n useHook\n vendors~main.iframe.bundle.js:11148:13\n useMemoLike\n vendors~main.iframe.bundle.js:11175:18\n useMemo\n vendors~main.iframe.bundle.js:11187:10\n withBackground\n vendors~main.iframe.bundle.js:3858:106\n undefined\n vendors~main.iframe.bundle.js:11056:21\n undefined\n vendors~main.iframe.bundle.js:12992:12\n undefined\n vendors~main.iframe.bundle.js:12999:14\n renderWithHooks\n vendors~main.iframe.bundle.js:249128:18\n mountIndeterminateComponent\n vendors~main.iframe.bundle.js:251954:13\n beginWork\n vendors~main.iframe.bundle.js:253192:16\n```\n\n\n\n\n**To Reproduce**\nRepro: TBD\n\nAdd the following code to your `.storybook/preview.js`\n\n```\nexport const decorators = [Story => <Story />];\n```\n\n\n**System**\n\n```\n System:\n OS: macOS 11.4\n CPU: (16) x64 Intel(R) Core(TM) i9-9880H CPU @ 2.30GHz\n Binaries:\n Node: 14.16.0 - ~/.nvm/versions/node/v14.16.0/bin/node\n Yarn: 1.22.10 - /usr/local/bin/yarn\n npm: 6.14.11 - ~/.nvm/versions/node/v14.16.0/bin/npm\n Browsers:\n Chrome: 91.0.4472.101\n Firefox: 87.0\n Safari: 14.1.1\n npmPackages:\n @storybook/addon-actions: ^6.2.9 => 6.2.9 \n @storybook/addon-essentials: ^6.2.9 => 6.2.9 \n @storybook/addon-links: ^6.2.9 => 6.2.9 \n @storybook/react: ^6.2.9 => 6.2.9 \n```\n\n## Relevant pre-fix discussion\n\n### Issue #15223\n\n#### Comment at 2023-05-01T17:41:56Z\n\nLooks like I'm somewhat mistaken. The internal decorators use custom look-alike hooks, so Storybook is not intercepting the hooks from user-provided decorators and stories that use React hooks, but preview-api has it's own custom React-like hooks that the built-in decorators use. The problem appears to be with how the jsxDecorator works.\nWithout our custom decorator the conditionally renders the story, these are the phases:\n\n\n\nNo errors, but also no UPDATE, only MOUNT. However, once our custom decorator is added, the biggest change is now the story has to go through the update phase, and the jsxDecorator gets out of sync. It never gets MOUNT called, but it gets UPDATE called twice at the end, and now we get the \"Rendered more hooks\" error:\n\n\n\njsxDecorator uses 1 hook, `useEffect`, but since it never got MOUNT and had the state saved off, when it gets to UPDATE it has no hooks registered and throws the error. It doesn't seem to matter that it got called twice at the end, the first call throws the error, the 2nd one also throws the error for good measure, but it appears having the conditional decorator interferes with jsxDecorator going through MOUNT, and causes it UPDATE to be a bit async compared to the others perhaps, as you will notice there are 2 sets of updates, and jsxDecorator doesn't participate in the first round, so I think it must get queued somehow, and then both of them drop at the very end.", "created_at": "", "version": "", "environment_setup_commit": "3e32fc62fdb93c78bb5c4ee5ce76a10d1b4cb7b6", "image_assets": "{\"problem_statement\": [\"assets/storybookjs__storybook-22336/001.png\", \"assets/storybookjs__storybook-22336/002.png\", \"assets/storybookjs__storybook-22336/003.png\"], \"patch\": [], \"test_patch\": []}"} |
|
|
