Ari Heitner wrote:
20010126125801.E447@andrew">
On Thu, Jan 25, 2001 at 11:07:05PM +0000, Gervase Markham wrote:
Heck, why not have separate domain names for all mozilla.org's
componentised technologies? Then, they would all be magically finished and
usable! ;-)

<goes off to register www.xpconnect.net>


Not to be negative towards the XPConnect owners ... but ...

It's not a matter of "magically they'll be finished if they become
independent projects". It's a matter of "maybe if they become independent
projects, the people who have been offering to fix all the dumb crap that's
broken now will be allowed to do so".
What in the name of all creatures great and small are you talking about?  How are you not "allowed" to fix bugs, for crying out loud?  Please fix!
20010126125801.E447@andrew">
I have been requesting the right to implement and submit for approval
"right"?  "submit"?  Yeesh.  This is *open source*, and mozilla.org is chartered (and I spend a good deal of my time, and have for almost three years) with getting quality submissions checked into cvs.mozilla.org.  You have every right to use the code according to its license, and (in my opinion) you have an interest, if not a duty, to help avoid forking.  So do I.  So let's go.
20010126125801.E447@andrew">
 (and
get guidance on along the way, since it's a lot of code i'm not that
familiar with) a couple of XPConnect features, most importantly support for
typelibs in multiple locations.
And jband has been your sometimes grumpy Obi-Wan.  What's the problem?
20010126125801.E447@andrew">
This is functionality for *mozilla* as a *component* -- part of mozilla's
core goal, and should easily fall under the "get a browser done quickly
rather than be all things to all people" part of mozilla's devel philosophy
(which i totally understand). IMHO it's a serious problem if things like
plugins and embedders can't add their own type info, and are therefore
forced to use their own private installations of mozilla (we're talking
about unix here; no braindamaged default install permissions to let any
user's programs write to the system components/ dir). For that matter,
component registration is broken too, but at least plugins and embedders can
register components at startup.

I've offered to write this about half a dozen times in various places.

I've stated clearly that I *do* have the time to do this.

I've stated clearly that IBM will *pay*! m! e to do this -- it's necessary for
SashXB.

I'm *still* not seeing and attitude on the part of mozilla developers that
says "we need to help outside developers contribute to mozilla".
I give up.  If jband's attitude bugged you, take it up with *him* or (better) *get over it*.  He's just one of many strong players who act as XPCOM peers and part-owners, although I must say he's one of our best, and he is also actually keeping his XPCOM code ownership commitments (unlike too many who have gone off to other jobs and other projects).

Your attitude is showing, too.  It would be a shame if we couldn't collaborate because we didn't show our animal underbellies to each other. Can we agree that I want your patches?  I'm not going to kiss up too much to get them, mind you (only a little ;-).  And you (and jband, and me, and others) will no doubt have a learning curve to ride, both code-wise and personally/ego-wise.  But we want your patches, and more -- we want your brain (well, some space in it, anyway).
20010126125801.E447@andrew">
Although I *am* seeing complaints on the part of mozilla developers that
they feel like no outside-the-project people have joined the fold, and that
they're carrying all the weight themselves.
Yeah, well -- it's a two way street.  I don't see any patches being rejected, or anyone being censored or deprived of inalienable rights.
20010126125801.E447@andrew">
This little discrepancy is starting to annoy me.
Then get to hacking!
20010126125801.E447@andrew">
(for the record, for the time being SashXB is going to follow the disgusting
hackish behaviour of installing its own private copy of mozilla. this is
*not* a good thing for SashXB, which is supposed to be *very* small -- it
would otherwise be < 1 mb. this is part of a large pattern of *braindamaged*
behaviour on mozilla's part wrt the difference between system- and
user-level information. this is imho because the mozilla developers honestly
didn't get this right the first time -- and they didn't get it right the
first time because they developed on windows, which is always a bad idea, in
this case do to continually broken access and installation idioms on
windows).
I happen to agree (and blame hyatt for the chrome-related parts of this brain-damage -- but without hyatt, we wouldn't have any of XUL, XBL, our UI, etc., so a little slack is in order).  How bad is the situation?  What are the bugs on file?  Let's nominate them for mozilla 1.0 and then see who can own them, and by what milestones they might be fixed.

Can everyone take a breather from the "politics"?  It would be refreshing to get back to code issues.
20010126125801.E447@andrew">
/me is *pissed*.
Hope I haven't pissed you off worse.  Really, thicker skins on all sides, and dogged persistence to cross the under-documented design/code barrier, are called for.

/be

(Sorry if some of you see this twice, NNTP barfed on an overlong line that I can't detect.)


Reply via email to