Careful with systemd update

[devil]

Systemd and udev 239-1 both have a bugreport. If I was you, I would wait these out:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=902287
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=902185

Systemd und udev 239-1 haben beide einen Fehler. Ich würde abwarten…

[seasons]

Why do I feel like a lab rat whenever a new major version of systemd hits sid? Can't they push to experimental first?

Oh, and thanks for the links. I don't think the second bug affects us though, unless someone is trying to update udev and hold systemd back for some reason.

[vilde]

Why do I feel like a lab rat whenever a new major version of systemd hits sid? Can't they push to experimental first?

Oh, and thanks for the links. I don't think the second bug affects us though, unless someone is trying to update udev and hold systemd back for some reason.

Nobody is delivering a really tested product anymore. For example if you buy a new car or a motorbike it can come with one of maybe 6 different brakes. When the cars come in to the mandatory services the brakes will be evaluated and they will after a while change all brakes on all the cars to the system that they find best, the costumers/owners will never see or know this. This is cheaper than long evaluating times and we customers are unknown made to lab rats or guinea pigs.
Sorry for the ot 

[melmarker]

Hihihi - it might be that someone hasn't noticed it right now - but we build on the development distribution of debian - also known as Sid. What do you expect? Should they introduce new versions in stable?

[vilde]

Hihihi - it might be that someone hasn't noticed it right now - but we build on the development distribution of debian - also known as Sid. What do you expect? Should they introduce new versions in stable?

Maybe they will soon, to make the user test, but on the other hand we are the users who tests, but that's a part of the fun and so far I have never manage to find a better dist.

[seasons]

Hihihi - it might be that someone hasn't noticed it right now - but we build on the development distribution of debian - also known as Sid. What do you expect? Should they introduce new versions in stable?

Yes, I understand that, but there is a difference between sid and experimental. My point is that upgrade-stopping bugs seem to occur way too often with systemd packages. Sometimes, they are upstream bugs and sometimes they are packaging bugs. It's gotten to the point where I distrust major systemd updates. I will wait two days and check here when I see them. I feel that systemd could use some time to bake in experimental because of how important and far-reaching the package is.

[melmarker]

Ok - and again - Sid/unstable is the development distribution of debian - and the bad bad debian boys and girls really use it for development purposes. The situation is harmless - compared to the situation some years ago, but please do yourself a favour and don't expect a system that is bug free from the moment the packages hit debian/sid. It's just a normal development process and sid should be read as recursive acronym 'still in development' - and it is damn rigtht that systemd hit sid asap after release. And it does no matter who make the mistakes and introduced bugs (upstream or package maintainers) - systemd is just to important to bitrot in experimental.

[seasons]

please do yourself a favour and don't expect a system that is bug free from the moment the packages hit debian/sid.

I don't, nor did I say I did. Once again, I'm saying systemd is a particularly bad offender. Maybe it's just my perception.

systemd is just too important to bitrot in experimental.

That's why they should test pre-release versions in experimental.

[melmarker]

systemd is just too important to bitrot in experimental.

That's why they should test pre-release versions in experimental.

Who should test - and why - sid is the development distribution already? And i'm sure that the maintainers don't put a new version into sid without tests. Anyways - maybe it's only a misunderstanding - if i upload new versions to experimental that has only one reason - having all packages ready and compiling to start a transition - in these cases it helps to know that all packages build, can be installed  without problems and can go in together, without delays from uploading - i didn't think that someone test these packages beside of me. In case no transition is needed i upload directly to sid - as usual. Thats the reason for Sid.

[sunrat]

My cousins work in a lab. We have lab rat in our genes.

[piper]

@sunrat

Thanks for making my day

[seasons]

Who should test - and why - sid is the development distribution already? And i'm sure that the maintainers don't put a new version into sid without tests. Anyways - maybe it's only a misunderstanding - if i upload new versions to experimental that has only one reason - having all packages ready and compiling to start a transition - in these cases it helps to know that all packages build, can be installed  without problems and can go in together, without delays from uploading - i didn't think that someone test these packages beside of me. In case no transition is needed i upload directly to sid - as usual. Thats the reason for Sid.

Fair enough (and thanks for explanation). I guess I'll just continue to be cautious with systemd updates. Now, if you'll excuse me, I have to go run on this wheel.

[Geier0815]

Has someone tested the 239-3 of udev and systemd? Works everything as axpected?
[Edit] Ok, udev seems to be fixed, systemd not. I will wait a little bit longer [/Edit]

[melmarker]

i didn't test anything - i just installed and run it - works fine for me (tm)

[eQuacky]

it runs for me either - up to now no problems with 239-3