I just upgraded Chrome to version 51 (51.0.2704.63 x64), and now none of my WAB apps, even AGOL-based apps, will fully load. They partially load and the progress thingy just keeps spinning. On my server, WD 1.2 and 1.3 will load into Chrome and show apps, but none of the apps will load fully when launched. WAB Dev 2.0 will only partially load into Chrome, and no apps are shown. I can see that I am logged in to my organization's Portal, so that is not the issue.
Does anyone else see this? Is there some way I can configure Chrome 51 settings to allow WD 2.0 and apps to load fully?
All is normal with Firefox 46.0.1 and IE11.
My organization's default enterprise browser is Chrome, so if anyone upgrades to current version, they won't be able to use any of my apps.
Chrome 51 includes many security fixes:
Solved! Go to Solution.
It looks like Google is aware of this, and if I am reading the bug report correctly, 52, with a hotfix, is being fast-tracked.
Yep, and the problems/service disruption with ArcGIS-related products are mentioned as among the most critical reasons for the fast-tracking of the fix.
It looks like the new Chrome version will be with v51, branch 2704, not v52, but it's hard to tell for sure.
Merge approved for M51 (branch 2704). +bhthompson, ChromeOS TPM, since this is labeled as a ChromeOS-only bug, please let me know if you have any concerns on getting this merged in. If not, nainar@, please merge this in ASAP if there are no outstanding issues, we are cutting another stable candidate for desktop early next week.
The WAB patch for AGOL just went out. The builder and apps including 2D and 3D should work again in Chrome 51. Thank all of you!
Jianxia
what about developer edition ?
The WAB Dev points at the hosted version of the ESRIJS API so you will need to wait for a patch on this and the WAB. However, I have managed to get it working with the following (hooky) steps with 2.0 in 2D (I don't know about 3D I don't think it's relevant as it doesn't use xstyle so long as you fix builder).
1. copy and paste the contents of main.js in to Javascript Beautifier and Formatter Online C:\WebAppBuilder\arcgis-web-appbuilder-v2.0\WebAppBuilderForArcGIS\client\builder\main.js
2. remove lines 55399-55428. This is a function which overrides xstyle. Sorry not prepared to paste the resulting file as it is copyright
To be honest from what I can see it then all works fine. I'm not sure if that means esri have patched their hosted versions of the API yet? If not then that does still mean it is pointing at the old version of xstyle. You have the option to override this by specifying that it uses the latest x-style in both the init.js (in builder and stem app). If it doesn't work you can also host the API locally. Let me know if you find any issues because I have other ideas for resolving this (my own apps I've hosted the api locally and use the latest xstyle)
When will the hotfix for Portal 10.3.1 be released? Yes, there's a "workaround", but have you ever tried to remind hundreds of non-technical users to use this browser and not that one, for hours every morning?
Tom,
Chrome 51.0.2704.79 was just released for all users with a fix for this issue. Esri worked directly with Chromium product management at Google and has gotten this fix incorporated quickly into the update going out now.
Go to "Help > About Google Chrome" to force an update check and upgrade to this latest patch level. The previous version was 51.0.2704.63 and the fixed version is 51.0.2704.79.
Obviously, this is a temporary solution for Portal users since Chrome will make such changes down the road. The Portal team is working on the patches for 10.3 to 10.4 versions.
Thanks,
Jianxia
Jianxia,
Does the new Chrome version render Junshan Liu's patch moot? Should that patch be applied to new WAB Dev 2.0 installations or not?
Best regards,
Barnaby
Edit: It appears to me that the patch is no longer needed......
It also appears that now Dev 2.0 runs either way after the Chrome release . . .
Yep, and Dev 1.2 and 1.3 also. So the 2.0 patch, and the two I made for Dev 1.2 and 1.3, are all moot as far as I can tell.