[PATCH 4 of 4 stable] largefiles: don't mute/obfuscate http errors when putlfile fails

[Pierre-Yves David]

On 10/22/2015 07:12 AM, Mads Kiilerich wrote:
> On 10/22/2015 12:35 PM, Augie Fackler wrote:
>> On Wed, Oct 21, 2015 at 09:16:07PM +0200, Mads Kiilerich wrote:
>>> # HG changeset patch
>>> # User Mads Kiilerich <madski at unity3d.com>
>>> # Date 1445454338 -7200
>>> #      Wed Oct 21 21:05:38 2015 +0200
>>> # Branch stable
>>> # Node ID 672fe15eacfa6cb600febf2e0ad48e9f9fd81c18
>>> # Parent  45a74ff314b78e2539049bbf3e90d59e9a6c90dd
>>> largefiles: don't mute/obfuscate http errors when putlfile fails
>> Looks good, but these seem a bit new-feature-y and not
>> regression-fix-y for stable to me. mpm?
>
> I agree they not are regression fixes and thus could be postponed; it
> fixes stuff that was introduced by this 'various' dude who introduced
> largefiles.
>
> I do however consider it a bug that store corruption could go unnoticed
> and be propagated to the working directory where it caused an
> "impossible" "--clean leave working dir with [incorrect] stuff to
> commit". The fix for that is appropriate for stable.
>
> Incorrect and misleading reporting of bugs also seems like a bug to me.

Should we get this patch someone now that the freeze is over?

-- 
Pierre-Yves David