Bart Smaalders wrote:
> http://cr.opensolaris.org/~barts/16327/
I hate to bikeshed (really!), but can we drop the "require-" from the type
names? I don't think it adds anything, and just makes the manifests more
cumbersome and harder to talk about. Perhaps "any" could be "anyof" or
"any-of".
pkg.5.txt:
- I agree with Alan's comments about "values" here.
- line 229: you should specifically mention that "multiple dependent
packages" are indicated by multiple fmri attributes; a couple of words
of duplication between this and the fmri attribute is useful, I think.
- line 238: quotes around 'true'.
- line 246, 248: "Has an effect for 'conditional'", maybe, or "Affects
only 'conditional'". You should mention here that the value of
'predicate' is an FMRI.
depend.py:
- line 131: I'd rewrite this as "if not isinstance(fmri_string,
basestring)"
- line 255, 260: use same indent level as lines 252, 257.
- line 262: this clause should come before the definitions of pfmris and
installed_versions.
- line 288: what happens if building the fmri fails?
- line 318: "applies"
pkg_solver.py:
- line 897: need trailing comma?
- line 924: what happens on fmri creation failure?
- line 931: debugging statement?
- line 995: is there no way that conditional dependencies could end up
here?
- line 1008: double space after =
manifest.py:
- line 133: blank line before this
Danek
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
