Re: [wmvs] Open Issues for 0.7.0

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Bjoern Hoehrmann <derhoermi@gmx.net> wrote:

>When I proposed these steps, I thought of the remaining issues as those
>that turn the Validator in a state *worse than 0.6.7* not as those that
>are present in both the HEAD of that time and 0.6.7; #856 however lists
>many things of the latter kind, so maybe I did not communicate this well
>enough? What needs to be done to put HEAD into a state that is at least
>as good as 0.6.7?

The list of blockers for Bug #856 is a pretty all-inclusive affair; it lists
every bugfix, wishlist item, and feature addition that is a) relevant for
0.7.0, and b) not allready decided won't go in and is thus targetted at a
later release.

This doesn't mean everything on that list will make it into 0.7.0 (or any of
the following 0.7.x releases)! I imagine quite a few will get bumped 
for simple scheduling issues, and another chunk of them will get bumped
because they turn out to require more extensive changes than we feel
comfortable making for 0.7.0.

At its core, the only new features that _Block_ the 0.7.0 release, as far as
I'm concerned, are the Template system (HTML::Template) and the new Config
system (Config::General).

(
  If time and implementation cost allows, I'd also really like to test
  out Tabtastic for the results, but the threshold for bumping it is
  really low since the existing interface works well enough.
)

Both of these are "in", but they're not finished (read "Polished" if you like)
yet, and they'll both require quite a bit of testing to shake out all
remaining issues. They are also both fairly big and intrusive changes, meaning
any given minor problem in either is likely to create what our users perceive
as a fairly big bug in the application/service as a whole.

IOW, despite them being checked in and more-or-less-working, I'm not
comfortable releasing them yet; and the issues remaining are largely the kind
that only a moderate period of "baking" will take care of.

There is also probably at least one moderate rearranging left for both of them
before I'm happy; possibly more, but I hope not.


While this happens I intend to get in as many extra bug fixes and new features
as I can without further destabilizing the code or significantly prolonging
development time.

The existing candidates for this are the bugs listed as blockers on Bug #856.
I'm also adding new bugs as I go along; e.g. I added a RFE bug about
SetEnv'ing a base config path because I'd just checked in validator.conf with
my local devel paths instead of the generic «/usr/local/validator» path again
(which I've now done so often I'm sick of it and intend to fix it once and for
all).


If there is any one _specific_ bug you feel is inappropriate for 0.7.0, then
make a comment to that effect on Bug #856. I'd tell you to just remove the
blocker, but given your opinion on the matter (cf. IRC), I fear I'd find the
release bug closed and 0.7.0 running on v.w3.org:80 before I could blink! :-)

But the list of bugs is there, and it can be pared down by writing patches,
fixing bugs, writing documentation, or by evaluating a bug and finding it to
be better left for a later release. The release itself can be sped up by
testing the code and writing new (preferably detailed) bug reports about each
issue (e.g. there's some "Warnings" missing in the output that is not logged
in Bugzilla anywhere; and some CSS borkage in the same area, etc.).

I'm also not averse to backing out changes from CVS if I get feedback that a
checkin has negative side-effects. I very much appreciate corrective feedback
if I get too far afield!

And hopefully I don't have to mention that writing code is the best way to
help make this release happen sooner!



Does that answer your question? Comments?



- -- 
I'm [less] than thrilled by the [VM situation]; all sides of it. I [think]
we need a [fork] in that area so that you guys would stop stepping on each
others' toes.  I'm taking no part in your merry 5-way clusterfuck  -- sort
that mess out between yourselves.                -- Alexander Viro on lkml

-----BEGIN PGP SIGNATURE-----
Version: PGP SDK 3.0.3

iQA/AwUBQTnNAKPyPrIkdfXsEQKi2ACfU77obvvuB49DRkq3XgdfL61OUpcAn1N3
qdWJU3T/Mq21l6IefdghaAwT
=gkQx
-----END PGP SIGNATURE-----

Received on Saturday, 4 September 2004 14:11:20 UTC