Bug#370505: xdvi: typing the search string after hitting Ctrl-F needs mousing
Frank Küster <frank@debian.org> wrote:
> I currently don't have an overview of how the patches once were
> organised, were supposed to be organized, and are in fact disorganized.
Okay...
>>From the names I'd say: patch-xdvi-370505 is not a good idea, IIRC the
> number is the xdvi bug number on sourceforge. patch-tmp makes sense
Er, but there is currently a patch-xdvi-300109 and the bug number refers
to the *Debian* BTS.
> since it's already applied upstream. On the other hand we'll never get
> a new xdvi source from teTeX, so creating a new patch won't hurt and
> maybe make things clearer.
That's what I am going to do.
[ About the "A file may only be changed in _one_ of them." sentence ]
> No, either I was wrongly assuming that quilt couldn't do it; or this is
> just a leftover from the handwritten patch system, and I didn't properly
> adjust it when I added the remark about quilt.
Then, I'll remove that sentence and also add a reference to quilt.html.
> I didn't know, or maybe README.patches started to exist before we used
> quilt. I think it would probably be better to document it in the
> patch.
There are pros and cons (for putting the comments in the individual
patch files):
pros: when you look at the patch file, it's easier to understand
what's happening, why, etc.
cons: README.patches (if up-to-date) gives the reader an overview
about the various patches applied to the package. Without it,
you'd have to do "less debian/patches/*" or similar to achieve
the same goal.
In fact, almost every patch in debian/patches already has a comment
before the actual patch, containing diffstat output.
> Anyway, feel free to change what you think is appropriate; tetex-bin is
> not mine ;-)
Yeah, but you've acquired nonetheless some authority over it through
your continuous work. Heh.
(OK, authority over a dead package ;-) ... :-|)
For the moment, due to the pros and cons outlined above, I'll leave the
structure with debian/patches/README.patches. If someone feels it's
better to document the patches in the individual patch files, fine, he
can do that.
--
Florent
Reply to: