2= 014-02-06 15:34 GMT+01:00 Vishesh Handa <me@vhanda.in>:

On Thursday 06 February 2014 14:37:= 49 Frank Reininghaus wrote:
> Hi,
>
> the build is broken again. Attached patch fixes it for me (I haven'= ;t
> tested it with Baloo installed yet though).
>
> 2014-02-06 10:41 GMT+01:00 Vishesh Handa:
> > On Thursday 06 February 2014 09:34:59 Frank Reininghaus wrote: > >> Hi,
> >>
> >> 2014-02-05 Christoph Feck:
> >> > Hi,
> >> >
> >> > It would be nice to have these changes merged in the nex= t two weeks
> >> > (i.e. one week before the freeze, see
> >> > http://techbase.kde.org/Schedules/KDE4/4= .13_Release_Schedule )
> >>
> >> that's fine. From my point of view, the Baloo code is bas= ically ready
> >> (once the foreach loop is re-added to prevent that 322348 com= es back).
> >
> > Fixed
> >
> >> I'll try to have another look at the code later, and I al= so encourage
> >> others to do so. If nothing else comes up today, I'd say = that the
> >> branch can be merged (if Vishesh thinks that the commit histo= ry of
> >> that branch is worth keeping) or alternatively, master can be= merged
> >> into feature/baloo and then the diff be applied to master as = a single
> >> commit.
> >
> > Lets keep the history.
>
> Just out of curiosity, may I ask why the history is important? The
> reason I'm asking is that I often use "git log" and &quo= t;git blame" to
> find out when and why some code was changed, and to understand why the=
> code is as it is (often very useful when I find out which line causes<= br> > a bug and I'm wondering why the line is there at all). This techni= que
> is obviously much less useful if some lines/files get changed very
> often.
>
> If the branch contained a sequence of commits with valid intermediate<= br> > states which could help to find the cause of regressions in the
> future, or which could help to understand the structure of the new
> code, I could see that keeping all those commits makes sense. However,=
> from my point of view, most of the commits will just add noise to the<= br> > output of "git log" and "git blame" and make findi= ng out about the
> origin of some lines quite painful.
>
> But I don't want to start bikeshedding about this issue - if you c= an
> tell any reason why the history will be useful to you in the future, > I'll shut up ;-)
>

Nah. It's cool. No history it is.

I'll push the changes later today.

--
Vishesh Handa