Compulsive non-LTS Testing
Why don’t we make it a part of installing a non-LTS Ubuntu to perform hardware checks and report back details?
At the moment we have no clue what proportion of people are effected by hardware bugs and the data we have is biased towards people who perform the tests manually.
Shouldn’t we be made more aware of testing and getting involved in testing, even for our own good?
I suspect it’s less an issue of not having wide enough testing, and more an issue that there lacks the resources to process/analyze test results and turn them into actionable reports.
(And by resources, I mostly mean people-power)
Agreed.
Problem is that bugfixing appears to be low priority as soon as a new development cycle commences in Ubuntu.
Ceased reporting bugs for Ubuntu due to absolute lack of attention from the bugsquad. (Are there bugsquads?) That’s also my reason for abandoning Ubuntu.
Shuttleworth (and Canonical/Ubuntu) seems to be extremely project orientated. That’s all good – but abandonment isn’t.
100 paperclips was really good – but where are the 20 paperclips for each subproject?
Ubuntu lacks capability wrt converting projects into regular development/maintenance strategy. Done once – got the headlines – move on.
I see this as the most significant caveat in Ubuntu thus the most probable reason for Ubuntu having peaked.
Wish it wasn’t so…
Compulsory rather than compulsive?
But would it not be great to have this data in analyzable format? I was looking at all the “Ubuntu Freeze” threads on the Ubuntu forums, with so many hundreds of posts and NOT A CLUE what the troubled posters have in common.
Oh, certainly.
GPLv4 should add a clause saying in order to use the code, users must participate in the FOSS community (only Official FSF Approved communities, of course, don’t want any riff-raff getting help) at a rate of ten seconds for every day they use the software.
When did the idea of free software become non-free?
It takes some time to install the os anyways. I think it would be great if the tests could be done during the installation. However, I would also suggest a post installation test after the proprietary drivers were installed in case they break something.
A simple tool could auto-run on first boot and ask the user is they want to test the installation. Whatever checks are deemed necessary can be run and the result reported to the user. In order to run this tool the user must agree to some terms, that being the fact that anonymous test results will be sent to Ubuntu for analysis. As long as such tool returns compelling information to the user, good or bad, people will use it. Good for the user, good for Ubuntu.
I really disagree with the whole release-it-now-and-any-problems-won’t-be-fixed-until-the-next-release–if-then-even mentality which is all too reminiscent of Microsoft’s own philosophy.
This is why I:
1) ONLY run LTS+Point Releases of Ubuntu (ala 10.04.1LTS)
Right now I am on 8.04.4LTS
2) I do not upgrade distro versions, but rather do clean installs.
Ubuntu is a great starting point from which other distros can evolve, like Linux Mint, which has a MUCH better track record, no mandatory release cycle, better community support, and so forth.
Kudos to the Debian crew from which there would be no Ubuntu.
Ubuntu should also make the LTS releases from Debian stable vs Debian Testing. Their maintenance releases, (ala 8.10, 9.04, 9.10 et al) come from Debian Unstable…which is called “UNSTABLE” for a reason.
“what proportion of people are effected”
The common verb is “affect”, the common noun is “effect”. “I am deeply affected by Ubuntu’s effective sound effects”. The reverse (noun “affect” and verb “effect”) are rarely used.
Thanks for not giving up and going with the atrocious “impacted”, which should only be used for asteroids and wisdom teeth.