coreutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: coreutils release plans


From: Pádraig Brady
Subject: Re: coreutils release plans
Date: Mon, 03 Mar 2014 10:22:57 +0000
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130110 Thunderbird/17.0.2

On 03/03/2014 09:55 AM, Ondrej Vasik wrote:
> On Mon, 2014-03-03 at 09:11 +0000, Pádraig Brady wrote:
>> So coreutils-8.23 is shaping up to be a bug fix release
>> which I hope could be released within 3-4 weeks.
>> There are a few outstanding items including various df fixes
>> and sort multithreaded fixups to include. We should concentrate
>> on any of the outstanding non feature bugs in this time.
> 
> What df fixes in particular? Based on Fedora-7 beta reports (rhbz
> #920806, #1024730 , #1042840 ...) I know there are still the
> deduplication issues caused by /etc/mtab symlink (fuse fs, bindmounts)
> which might be relevant to backport to released Fedoras.

Thanks for the bug references. I've about 5 df patches in this area, including:

  fix handling symlinks in all cases
  fixing deduplication of virtual file systems (like tmpfs, which can have 
storage)
  avoiding warnings when with namespace stuff in /proc/mounts (default on F20 
at least),
  ...

I will update the bugs as needed (I've noted 1024730 at least should be fixed).

>> After that the ground would be clear for a more
>> invasive update to add prelim multibyte support
>> and also the unified join/uniq/sort key selection.
> 
> Do you plan just the first few utilities (expand/unexpand) from patches
> written by Ondrej Oprala or something more? Anyway - sounds like a
> another test for downstream i18n patch maintainance :) - both parts will
> affect it heavily.

Correct. Initially expand/unexpand will be updated at least.
As time progresses we can handle each utility in turn,
resulting in the easy exclusion of entire utilities from the downstream i18n 
patch.
Note some utils will probably need handling together (like join,sort,uniq).

>> So I'm thinking:
>>
>> 4 weeks:  8.23 bugfix stable
>> 10 weeks: 9.0  feature unstable
>> 14 weeks: 9.1  bugfix stable
> 
> Sounds pretty challenging - to have the key-selection and the first
> multibyte items in 6 weeks.

Well the patches have already gone through a couple of iterations,
so I'm optimistic :) Mainly I want to put a little pressure on (myself)
to get these merged as they've lingered too long.

thanks,
Pádraig.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]