xml-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brett McLaughlin <brett.mclaugh...@lutris.com>
Subject Re: What name?
Date Tue, 11 Jul 2000 21:18:22 GMT


Arnaud Le Hors wrote:
> 
> [I'm cross posting because having two threads on the two lists is simply
> not productive.]
> 
> Brett McLaughlin wrote on xerces-j-dev:
> >
> > Somewhere in here people went from a new project to Xerces 2.
> 
> Stefano started it. See
> http://xml-archive.webweaving.org/xml-archive-general/2127.html:
> 
> > James is just trying to do the same: start a new codebase to create the
> > itches to scratch.
> >
> >
> > Is it wrong to give it another codename? probably.

This I disagree with - I like Stefano, he's great, but I think this is
not the best idea. I certainly am all for Spinnaker /becoming/ Xerces 2
at some point - I really think a next generation parser is great, and
would love to see Xerces embrace it. What I'm emphasizing is that many,
many people will see Xerces 2, and leave. They have tried Xerces, they
see it's complex, they know what it looks like, and, for better or
worse, it feels like an IBM project.

Even if we all agree "Xerces 2 can start from scratch" people will think
otherwise. It's not right, but it's human nature, and we should cater to
that because we /know/ it will happen. Also, I think its premature to
say we are ready for Xerces 2 - Spinnaker won't even be functionally the
same as Xerces for quite a while, I'm sure - I mean, we're talking ints
vs Strings here. That's very very early. If we call this Xerces 2,
another tendency will be for people to stop forward progress on Xerces -
that is also a bad thing right now - still lots to do in Xerces 1.

For all these reasons, plus the allowance of code from places like
Crimson without an act of congress, plus people's conceptions of Xerces,
I am not comfortable with the name Xerces 2. At the same time, please
understand, I am 100% behind Xerces 2 /being/ Spinnaker if it works out.

> > Is it wrong to place it in another CVS module? probably.
> > Is it wrong to go off without asking first? probably.
> >
> >
> > So, let me start off another option:
> >
> >
> > 1) we forget about spinnaker
> > 2) we create a new CVS branch under xml-xerces where Xerces2 should
> > reside
> > 3) in case, we create a xerces2-dev mail list
> 
> To which I said: Sounds great!
> 
> Now you wrote:
> 
> > I don't
> > know that everyone is happy with Spinnaker starting out as Xerces 2 - I
> > know I am not. I think the right thing is let this thing evolve on its
> > own, without any presumptions about its final resting place. Leave room
> > for complete reinvention, not just upgrades.
> 
> The point is that we are interested in working for the next version of
> Xerces. Not just making an experiment. So it's a natural way to call it
> Xerces 2. If it's the number you don't like. Fine. Let's call it Xerces
> NG (Next Generation). But I really wonder why some of you aren't more

I just think that the name "Xerces" carries with it some negative
connotations for /some/ people. I'm not making this up, I swear ;-) I
get mails about this a lot on JDOM and at Enhydra. Let's just avoid that
- then when we make it Xerces 2 (if that is what happens), we can clear
those all away, because people are confident in the newer "code name."

> confident in the success of this. Everybody says Xerces is a great
> success. Even though we all know and recognize it has defiencies. How
> couldn't we do better?? We plan to commit just as much time to this new

We could - but right now, I think two tracks are needed. One to bring
Xerces 1 to a full-blown, completely bug-proof (as much as is
reasonable) parser. Without changing internals. But I think we need a
new start to get the benefit of being able to change internals. We also
need to lower the cost of entry - it is really high. The primary mails I
see for committers are:

OK, Joe Blow just joined us at IBM, can we get some +1s.

Then, 3 other IBM guys +1 him, and he's in. This isn't helpful for
people to see - it also proven by the low number of patches (almost
never) that go to the mailing lists. People ask questions, IBM fixes it.
This is fine, it's life - but let's at least try to improve that
political deficiency by giving people a chance, and by IBM saying
publicly, "Yes, we are willing to work at this, so if a new name helps,
we're for it." That's what I'm getting at - not that Xerces can't
improve. 

At the same time, I would maintain that a good, SAX, DOM, JDOM capable
parser can be done in under 500K. I don't believe Xerces can be pared
down 1/2 MB and still be anything but a completely rewritten thing - so
let's rewrite, and give it a code name. Big deal - if it helps, great!

> version of Xerces than we have commited to Xerces 1. So, having more
> experience and help from others, I don't see how we'd fail.

I agree - but what I'm afraid of is that Xerces 2 will /not/ garner help
from others, whereas a "perceived revolution" such as Spinnaker will.
Even if IBM has to play the "foil" or "bad guy" initially, isn't it
worth it if at the end Xerces 2, or whatever, is the best, most usable,
parser around?

-Brett

> 
> And even if it was a failure after all. Then what? We'd start a new
> version. What's the problem??? On the other hand, as I said in another
> message moving things in CVS is a major pain. So I'd much rather choose
> a place that's independent of its status.
> --
> Arnaud  Le Hors - IBM Cupertino, XML Technology Group
> 
> ---------------------------------------------------------------------
> In case of troubles, e-mail:     webmaster@xml.apache.org
> To unsubscribe, e-mail:          general-unsubscribe@xml.apache.org
> For additional commands, e-mail: general-help@xml.apache.org

-- 
Brett McLaughlin, Enhydra Strategist
Lutris Technologies, Inc. 
1200 Pacific Avenue, Suite 300 
Santa Cruz, CA 95060 USA 
http://www.lutris.com
http://www.enhydra.org

Mime
View raw message