[gentoo-dev] Re: bzr.eclass into Portage

[Steve Long]

Ulrich Mueller wrote:

>>>>>> On Mon, 06 Oct 2008, Jorge Manuel B S Vicetto wrote:
>
>> No objections here, just a question. Do you know if the issue with the
>> lp:// sources has been fixed in bzr?
>
No objections, a minor point wrt bash:

EBZR_OPTIONS="${EBZR_OPTIONS:-}" (and similar variants)
doesn't do anything (beyond waste lex and yacc time.) I can understand the
maintenance argument, but I don't think it really flies, given the
inordinate lengths considered in the past to avoid sourcing an ebuild.

The same consideration applies to all those "constant values" 'and indeed'
${foo} as opposed to $foo, though first time I raised that I got sworn at,
so not expecting miracles here.

[[ -z ${EBZR_REPO_URI} ]] && die ..
Here's how I'd write that:
[[ $EBZR_REPO_URI ]] || die ..

I've heard the "be explicit" argument before (hey antarus;) and here's why I
disagree:
If you don't know test (''help test'') and what its default is, then you
really don't know the basics of shellscript (you possibly only think you
do.) If you don't know shell, and can't begin to understand what that might
do, then you shouldn't consider coding as a career, and I'd expect you to
take quite a while to go through the #bash crucible; if you ever make it
I'd have a lot of time for you.

(Since you use || elsewhere, I don't expect to hear the "|| is cryptic even
if we say OR in speech" argument.)

I appreciate that appears like 3 or 4 points: they all come under the 'clean
bash' heading: it runs faster, as well as being easier to read and write.

> Looks like this is working fine with bzr-1.5, so I'll change the
> dependency.
>
Given that, is there any reason not to use 1.6 if installed, and fallback to
an earlier version if not? Personally I'd just use an unversioned dep in
the latter case, given that 1.5 is stable and 1.7.1 is ~arch (amd64).
Doesn't sound like it's going to be long to get 1.6.

I'm thinking: "maximise utility before you unleash it on the tree".