Useless save password prompt for company intranet

323 views
Skip to first unread message

Pierre-Antoine LaFayette

unread,
Oct 7, 2010, 12:53:10 PM10/7/10
to Chromium-dev
My company has an intranet that is secured and triggers a browser login dialog. Every time I login I get the "Do you want Google Chrome to save your password?" prompt but it is useless. If I select yes, it doesn't save anything and I still have to enter my credentials every time. If I select never, the prompt still appears each time the login period times out and I login.

So everyday I see this save password prompt that is absolutely useless in this context. I'm not sure what the correct behavior should be but this has been annoying me for a long time. I just want to make sure people agree this is a bug. If anyone knows what's going on under the hood, I'd love to hear it.

Thanks.

--
Pierre.

Chris Bentzel

unread,
Oct 7, 2010, 1:04:11 PM10/7/10
to pierre.l...@gmail.com, Chromium-dev
Hi Pierre,

This is a known issue and has been fixed. 


--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev

Pierre-Antoine LaFayette

unread,
Oct 7, 2010, 1:57:08 PM10/7/10
to Chris Bentzel, Chromium-dev
I'm guessing this code has not made it into the canary since I can still reproduce it. I'll try a nightly build.
--
Pierre.

Chris Bentzel

unread,
Oct 8, 2010, 11:59:06 AM10/8/10
to Pierre-Antoine LaFayette, Chromium-dev
We discussed this more off-list. The site in question was cycling through different authentication realms, which caused the infobar to reappear even though the "never save the password" option was chosen. Arguably we could choose to store settings on a [server, scheme] key rather than a [server, scheme, realm] key but I think the latter is preferable in general.

Mike Mammarella

unread,
Oct 8, 2010, 8:54:03 PM10/8/10
to cben...@chromium.org, Pierre-Antoine LaFayette, Chromium-dev
Removing the realm from the key would also fix this bug:
http://code.google.com/p/chromium/issues/detail?id=45379

...which is caused by a device that puts a second-granularity
timestamp (!) into the realm. I'm not necessarily advocating removing
the realm from the key, but just pointing out a relevant data point.

Chris Bentzel

unread,
Oct 9, 2010, 10:17:23 AM10/9/10
to Mike Mammarella, Pierre-Antoine LaFayette, Chromium-dev
Thanks Mike, that's exactly the case reported by Pierre-Antoine.

I'll continue discussion there.
Reply all
Reply to author
Forward
0 new messages