POST
|
Hi @GILBERTNG1 thanks for reporting, this sounds like RDP performance issues that are unrelated to the JavaScript SDK. Does the host machine have a GPU and is this a shared machine? No GPU can have a significant affect on rendering performance. And, multiple users can cause variations in performance. There are many interrelated factors that can affect RDP session performance such as screen resolution, connectivity speeds (like you mentioned), number and type of users on the machine, GPU, etc. 4K RDP resolution requires a high-bandwidth connection with consistent low-latency. Here's an informative Microsoft article on the topic: https://learn.microsoft.com/en-us/azure/virtual-desktop/rdp-bandwidth and here are some network guidelines for display resolutions: https://learn.microsoft.com/en-us/windows-server/remote/remote-desktop-services/network-guidance#display-resolutions.
... View more
2 weeks ago
|
0
|
1
|
37
|
POST
|
The Angular bug has been fixed in zone.js version 0.14.5. We recommend updating zone.js if you are having issues with the SDK's widgets. https://github.com/Esri/jsapi-resources/tree/main/esm-samples/jsapi-angular-cli#known-issues
... View more
2 weeks ago
|
0
|
0
|
53
|
POST
|
Hi @RoelGarcia please provide a repro app: https://developers.arcgis.com/javascript/latest/troubleshooting/#minimal-reproducible-application. This error is likely related to webpack shimming the node crypto library. Without a repro app we can't really offer specific recommendations. You might search for "webpack crypto is not defined" on the internet for ideas.
... View more
2 weeks ago
|
0
|
0
|
244
|
POST
|
This might be an issue with the Angular syntax that you are using. Here's an example codepen using the ArcGIS ESM CDN (for prototyping-only) that shows getElementById() working: https://codepen.io/andygup/pen/poBxxxw?editors=1000. With Angular try using "@viewChild": https://angular.io/api/core/ViewChild. And, here's an example usage of viewChild with "@arcgis/core": https://github.com/Esri/jsapi-resources/blob/main/esm-samples/jsapi-angular-cli/src/app/app.component.ts#L28 @ViewChild('mapViewNode', { static: true }) private mapViewEl!: ElementRef;
... View more
a month ago
|
0
|
0
|
58
|
POST
|
Hi @SaurabhUpadhyaya I see you have two recent questions that mention errors related to WebGL contexts. This most likely is related to your ArcGIS code and not an issue with React. Your best option is to provide a vanilla JS codepen using the ESM CDN: https://github.com/Esri/jsapi-resources/tree/main/esm-samples/jsapi-esm-cdn. That way we can all take a look and provide suggestions. Note, the ESM CDN is for testing and prototyping, only. It doesn't offer the performance of a locally built ESM app.
... View more
a month ago
|
0
|
0
|
45
|
POST
|
I'm not seeing those skips on my local build. Please post a link to a repro sample.
... View more
a month ago
|
0
|
1
|
57
|
POST
|
According to Ionic Stencil team, this is a known warning and it should not be affecting any functionality: https://github.com/ionic-team/stencil/issues/5427#issuecomment-1977185144. The warning is showing up in your Angular builds because "@arcgis/core" has a dependency on "@esri/calcite-components" which has a dependency on Stencil. If you find any Calcite or "@arcgis/core" functionality that is being affected then let us know and provide a repro sample using codepen (for vanilla JS), StackBlitz or a GitHub repository.
... View more
a month ago
|
0
|
3
|
256
|
POST
|
We haven't seen that error before with Angular/esbuild. Can you provide a repro app so we can examine it in more detail?
... View more
a month ago
|
0
|
6
|
293
|
POST
|
Hi @ForrestLin you'll need to provide a repro app. We recommend using our esm-samples as a starting point: https://github.com/Esri/jsapi-resources/tree/main/esm-samples/jsapi-angular-cli. Here's our troubleshooting guide for local builds for more info: https://developers.arcgis.com/javascript/latest/troubleshooting/.
... View more
a month ago
|
0
|
9
|
318
|
POST
|
> What is the well-known limitation of ES modules you mention I mentioned it above. Basically, unbundled ES modules (ESM) such as "@arcgis/core" typically perform poorly in a production environment. We are constantly optimizing the ArcGIS JS Maps SDK, however ESM builds require a combination of treeshaking, lazy-load and code splitting (chunking) to create an optimal build output. This isn't specific to ArcGIS, for example you can read about why bundling is recommended in the Vite.js documentation: https://vitejs.dev/guide/why.html#why-bundle-for-production. > Hence, far and away the biggest improvement I could get in build performance, is an approach that doesn't always re-build the ArcGIS library dependency. Vite.js could really help with your use case of having fast(er) build times. Vite does dependency pre-building to dramatically speed up builds: https://vitejs.dev/guide/dep-pre-bundling. Try this sample app: https://github.com/Esri/jsapi-resources/tree/main/esm-samples/jsapi-vite-ts. On my 5 year old machine it takes 20 seconds for a production build. For dev builds, it's pretty close to zero on my machine - it takes about 150 - 170ms (milliseconds), and rebuilds are almost instant. And, also to reiterate checking out the new map-components. You can access those via CDN and they include TypeScript types: https://developers.arcgis.com/javascript/latest/components-get-started-cdn/.
... View more
04-08-2024
04:07 PM
|
0
|
0
|
130
|
POST
|
Do you require the Webpack-specific tooling? If not, you might take a look at Vite.js. > The pre-built AMD modules served that scenario, but per above, the TypeScript typings for that are being deprecated. If/when you migrate to (new) map web components, those are typed. > Why is the biggest library I reference (by a huge margin) the only one that doesn't come with a viable pre-compiled option? TL;DR - It stems from a well-known limitation of ES modules. Components will make life much easier in that respect. Currently for ESM there's no ECMA-standard equivalent to the AMD dojo build layers. Those are what makes the AMD ArcGIS CDN much more performant than the ESM ArcGIS CDN. For example, JavaScript module declarations have been in TC39 Stage 2 since 2022, and it could be quite a while before it gets finalized as a standard and then adopted in browsers.
... View more
04-05-2024
08:57 AM
|
1
|
2
|
170
|
POST
|
Hi @DavidDrennan it sounds like this could be a CSS issue. Did it work in previous versions of the API? Try to repro in a codepen using vanillaJS. Otherwise, you'll need to provide a simple, focused StackBlitz or GitHub repo.
... View more
04-04-2024
09:16 AM
|
0
|
1
|
107
|
POST
|
Hi @JonathanTiu Yes, we did figure it out. It's related to fine tuning the API and how it uses workers under-the-hood. The size of the workers is just one parameter out of many that affects the initial load performance metrics. Overall, for our baseline testing apps, those metrics improved between 4.27 and 4.29. With that said, if you have a simple, focused repro app that shows a major performance regression on the initial app load, then please open a Tech Support ticket so they can get this formally documented.
... View more
04-03-2024
03:46 PM
|
1
|
0
|
22
|
POST
|
@JasonDoingMapsto speed up your webpack/TypeScript dev builds, you might look into https://www.npmjs.com/package/esbuild-loader.
... View more
04-02-2024
09:26 AM
|
0
|
4
|
250
|
POST
|
Hi @AndresKasekamp you'll need to do further testing by removing '@arcgis/core' functionality and calcite elements one-by-one to try and find the root cause. Otherwise there's too much going on in the sample including the calcite react wrapper, calcite-panel, calcite-shell, calcite-shell-panel and various '@arcgis/core' modules.
... View more
04-01-2024
08:40 AM
|
0
|
0
|
153
|
Title | Kudos | Posted |
---|---|---|
1 | 03-15-2024 08:21 AM | |
1 | 04-03-2024 03:46 PM | |
1 | 04-05-2024 08:57 AM | |
1 | 11-13-2023 01:23 PM | |
1 | 03-26-2024 07:52 AM |
Online Status |
Offline
|
Date Last Visited |
Friday
|