pcdm: autologin broke #1455
Comments
current laptop and previous BE, autologin worked here |
Previous BE? You mean autologin worked for you before and after your last system update? Are you running STABLE or UNSTABLE? What have you set timeout to? |
virtualbox is broke upstream in FreeBSD at the moment, so previous BE, within virtualbox, UNSTABLE autologin worked. Current UNSTABLE, autologin works for me. never paid attention to time out, so whatever the default is |
Its a race condition, and Ken has already fixed this: beanpole135/pcdm@de4d11d |
I thought I'd re-open this because its been over a month since Ken said he fixed it and I'm sure UNSTABLE has been rebuilt since then, right? |
UNSTABLE let me re-quantify my answer. on initial bootup on metal and virtualbox, autologin works. on metal, loggin out and back in , autologin DOES NOT work |
I'm not sure what Ken's thoughts are on the issue, but enabling auto-login when a user manually logs out seems to be opposite of the intention of 'logging out'. If you have more than one user on a system, there would be no way to EVER change users, because as soon as User A logs out they get logged back in, meaning User B can never log in. |
i agree. but on my systems and in virtualbox. autologin works as advertised. was looking at outside the box approach to the "autologin" question. i am unable to duplicate on any of my systems |
I believe the default autologin timeout is 10 seconds, and as Rod says this does work (under UNSTABLE, at least) but setting the autologin timeout to either 0 or 1 seconds fails. I have not tried 2-9 but I want a 0 second (instant) timeout, ideally. Some display managers, such as the Mint one, do offer an autorelogin timeout for when you have just logged out but I personally don't like nor do I understand the need for such a feature. I only want instant (0 second) autologin on initial boot to work. |
Since the old code only allowed timeouts of >1 second... can you set it to 2 and see if that works for you? If it works when you set it to two seconds... I'm going to guess you're not running the latest code. My system is a complete mess right now, due to testing another patch so me testing it isn't the cleanest assessment. There's a joke in there somewhere, about the state of how busy your life is... if an extra 2 seconds at login is a problem for your daily schedule... but i'm to tired to try and find it. lol |
Yes, 2 second timeout does autologiin OK Is there any reason we can't have a 0 second timeout? |
If my understanding of the issue is correct... this is the scenario. It's a classic race condition. The 'timeout' you see and are presented with in lumina-config are just a way for the user to set a variable amount of time before logging in automatically. Ken's patch did the latter. It inserted a wait method until things had settled. If you No option will get you a faster login, because it's not possible to start the desktop until other processes have settled. Would you mind running an MD5 hash on your PCDM binary? Just drop to a shell and run |
62a25273a40370b7e6664e4fa4af601d |
OK, I think I might have another bug to report then. Why does the GUI update manager tell me I'm up-to-date? |
probably because the version in github isn't the version we are shipping. We don't recompile our packages and push them to the repo every day. The version of PCDM you have 'should' have the fix, but I'll have to check to see if PCDM is an older package. |
Tested/verified that autologin with a delay of 0 seconds works fine with the version of PCDM from source. |
pcdm autologin isn't working for me with both Lumina and MATE under both the latest STABLE and UNSTABLE versions with the timeout set to 0 seconds. I think it tries but fails and you end up having to login manually.
The text was updated successfully, but these errors were encountered: