Not logged in
Log in now
Create an account
Subscribe to LWN
An unexpected perf feature
LWN.net Weekly Edition for May 16, 2013
A look at the PyPy 2.0 release
PostgreSQL 9.3 beta: Federated databases and more
LWN.net Weekly Edition for May 9, 2013
Other desktop support?
Posted Jan 16, 2013 10:54 UTC (Wed) by micka (subscriber, #38720)
And unlike most other outdated applications, you don't need to rebuild the package with one or two rdeps, you must build a whole pile of pieces.
Posted Jan 16, 2013 13:14 UTC (Wed) by juliank (subscriber, #45896)
Posted Jan 16, 2013 13:27 UTC (Wed) by micka (subscriber, #38720)
And if there is not enough people to test testing, maybe it just means that there is not point in releasing stable (and in freezing testing) ?
Posted Jan 16, 2013 13:43 UTC (Wed) by mpr22 (subscriber, #60784)
Posted Jan 16, 2013 13:53 UTC (Wed) by micka (subscriber, #38720)
Posted Jan 16, 2013 18:56 UTC (Wed) by vonbrand (subscriber, #4458)
Because if they only froze testing, all its users would migrate to unstable...
Posted Jan 17, 2013 7:57 UTC (Thu) by DavidS (subscriber, #84675)
Because "unstable" is the place where stuff goes that should be released with the next stable. See Developer Reference 22.214.171.124:
> This development cycle is based on the assumption that the unstable
> distribution becomes stable after passing a period of being in testing.
Posted Jan 17, 2013 8:05 UTC (Thu) by micka (subscriber, #38720)
Posted Jan 18, 2013 6:30 UTC (Fri) by DavidS (subscriber, #84675)
Uploading to unstable publishes packages as "destined for stable". In the case of libraries, that also means that users of this library will start linking against the new binary. This again implies that the two packages now must go to testing together. This quickly leads to situations where significant parts of unstable cannot progress to testing due to issues in central packages (think gtk).
Building new packages against libraries from testing would reduce the formal requirements for testing propagation, but would lead to untested combinations in testing (as unstable users had a different version installed).
Posted Jan 18, 2013 8:06 UTC (Fri) by micka (subscriber, #38720)
> Building new packages against libraries from testing would reduce the formal requirements for testing propagation, but would lead to untested combinations in testing
But I'd prefer that. I still think that those who care about testing should be the ones testing it.
Posted Jan 16, 2013 13:47 UTC (Wed) by pboddie (subscriber, #50784)
It all sounds like yet another cunning plan to second-guess the herd of users and make them do something they don't really want to, instead of making it easy for those who are interested to do something, and encouraging (rather than coercing) others to join them.
As for KDE being the most desirable desktop environment, that's quite a depressing statement on the topic, but not entirely unsurprising given my recent experiences from renewed exposure to KDE 4 (sorry, KDE Plasma), Trinity, GNOME and Unity.
Posted Jan 16, 2013 14:49 UTC (Wed) by cortana (subscriber, #24596)
Posted Jan 16, 2013 15:15 UTC (Wed) by pboddie (subscriber, #50784)
People being able to try out testing would be one fewer reason to freeze unstable, I think.
Posted Jan 16, 2013 15:58 UTC (Wed) by cortana (subscriber, #24596)
Posted Jan 16, 2013 21:06 UTC (Wed) by pkern (subscriber, #32883)
Hence we require updates to flow through unstable. We are not reviewing all what's uploaded to unstable, although many of the uploads there are now voluntarily reviewed to see if they fit the release criteria for migration to testing at this point. Any upload to unstable still ages there, too, just like in non-freeze times. This means that the worst bugs that are introduced by bug fixes are likely to be found by unstable's user base before they hit testing. In turn this makes testing more stable and solid to be used at this point before it's finally released. It also allows some more intrusive RC bug fixes to be accepted under the premise that we'll soon find out if they're flawed.
Copyright © 2013, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds