Community
Participate
Working Groups
- Add http://localhost:8080/never as a repo (or any other authenticating repo) - Help>Check for Updates - you get prompted for authentication. Cancel it. - update check proceeds (good). - you get prompted again for authentication. ?? The reason here is that neither the UI nor the repo manager remember authentication failures. Bug 274447 discusses whether we want/need to do this. It's not clear that we want to remember/suppress login generically, but there are some UI workflows where the user will get multiple prompts because we don't. The question in this bug is how to handle the prompts in the UI if we do not remember it in the manager. It would be possible in certain workflows to have the UI keep track. In this example, the UI could track repo failures and ensure that those repos aren't used when trying to plan the updates.
*** Bug 276803 has been marked as a duplicate of this bug. ***
User will still be hit with prompts if artifacts themselves are protected - i.e. a repository can load without issues even if password is not known. The scope of credentials-failure-memory needs to be controlled by the UI (or certain jobs at least) - but the actual implementation is best made close to Credentials handling. I can see how credentials "scope" could be turned on/off. When on, it would remember the authentication state per host, and not prompt the user.
This can actually cause a pretty serious problem. If you have a site that you don't remember the password to, and you try and update, you continually get prompted for the password. However, the dialog is modal, so you can't even disable the repo. The best you can do is kill eclipse, start it again, remove the repo, try and update.
(In reply to comment #3) > This can actually cause a pretty serious problem. If you have a site that you > don't remember the password to, and you try and update, you continually get > prompted for the password. However, the dialog is modal, so you can't even > disable the repo. The best you can do is kill eclipse, start it again, remove > the repo, try and update. > I am working on a solution to this in Bug 256281 (which is also about other things). The solution is to remember the login status during the run and automatically fail subsequent attempts to talk to a repository the user does not know the password for. I.e. user gets three attempts per location rather than three attempts per artifact :)
Ian, if you cancel the password prompt it should continue with the update.
(In reply to comment #5) > Ian, if you cancel the password prompt it should continue with the update. > The bug here is that the normal flow of password prompts will happen twice instead of once. But yes, the update will continue. So if you cancel the site, you'll get two cancels (one for the update check, adn another for the planning). If you try and fail 3 times (3 strikes you're out), you'll get 6 prompts, 3 for the update check, 3 for the planning. What Henrik is proposing is a way for us to scope an operation for credentialing so that the UI could "start" the scope, check for updates, do the planning, and then end the scope.
(In reply to comment #6) > (In reply to comment #5) > > Ian, if you cancel the password prompt it should continue with the update. > > > If you try and fail 3 times (3 strikes you're out), you'll get 6 prompts, 3 for > the update check, 3 for the planning. I can't be sure if I canceled 6 times, but I did hit cancel more than once. (I eventually gave up and typed in the password). I did get through the update check, but I'm pretty sure I hit cancel a bunch of times at the planning stage.
This bug hasn't had any activity in quite some time. Maybe the problem got resolved, was a duplicate of something else, or became less pressing for some reason - or maybe it's still relevant but just hasn't been looked at yet. If you have further information on the current state of the bug, please add it. The information can be, for example, that the problem still occurs, that you still want the feature, that more information is needed, or that the bug is (for whatever reason) no longer relevant.
This works okay for me in the latest release. Only a single prompt that I can cancel...