[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Have JDOM / XOM / etc. failed? If so, why?

• To: Michael Champion <michael.champion@h...>
• Subject: Re: Have JDOM / XOM / etc. failed? If so, why?
• From: Jason Hunter <jhunter@a...>
• Date: Thu, 30 Mar 2006 11:23:55 -0800
• Cc: xml-dev <xml-dev@l...>
• In-reply-to: <BAY107-W3141174C75D52E999860B99D10@p...>
• References: <BAY107-W3141174C75D52E999860B99D10@p...>
• User-agent: Thunderbird 1.5 (Windows/20051201)

Michael Champion wrote:
> In some internal discussions of the XLinq 
> http://msdn.microsoft.com/netframework/future/linq/ project, I got a 
> very interesting bit of devil's advocacy that went something like this:  
> "People complain about the DOM, but they don't embrace alternatives.  
> For all the work that people have done to provide alternatives such as 
> JDOM, dom4j, XOM, etc. in the Java world, the typical users and the 
> major Java players still use DOM, warts and all."   I'm not at all 
> convinced this is true, but I don't have any information at my 
> fingertips to dispute it.  Would anyone care to present facts on one 
> side or the other?

There's a poll at servlets.com that asked this question:

http://www.servlets.com/polls/results.tea?name=doms

I intentionally didn't mail any DOM or JDOM list announcing it, so it 
should be a fairly random sampling.

> - Duh, the network effect.  A mediocre standard beats a better 
> non-standard every time. 

I think the attitude of people writing libraries/specs is, "We have to 
support DOM because it's the standard, and if we support one then why 
bother to support two, that just makes things more complicated."  The 
people who get to choose their Java-based XML API directly aren't 
choosing DOM if they realize there's alternatives.

> - Sun and IBM haven't included any of the alternatives in their 
> distributions, so the masses don't even know these things exist, or fear 
> being stranded in a backwater if they do adopt one.

One thing I discovered with JDOM was that we'd have been better served 
to have JDOM technically require parsers to do something in order to 
support it.  Xerces long ago had a discussion thread about this 
growingly popular JDOM thing and what did they need to do to enable it. 
  Answer: nothing, JDOM was designed to work easily with every parser. 
So we lost the ability to say "Supported in Xerces".  Also, Xerces 
didn't talk in their feature list about JDOM, it didn't come included 
with Xerces like it would have otherwise, and so on.  Compare to DOM 
which with pure interfaces needs support to do anything.

It's a similar story with JSPs.  Every web vendor needed to do something 
to enable JSPs to work, and you'd see each web vendor giving talks about 
their new JSP support.  Compare to Apache Velocity that just works, so 
it never gets mentioned by vendors.  Vendors can't differentiate using a 
third party plugin library that just works, so you get no vendor buzz, 
and without vendor buzz you lost the star quality.

Sun was smart in how they judged JSP success.  They counted vendors 
supporting it.  They didn't judge how much users liked it.

-jh-

• References:
  • Have JDOM / XOM / etc. failed? If so, why?
    • From: "Michael Champion" <michael.champion@h...>

• Prev by Date: Have JDOM / XOM / etc. failed? If so, why?
• Previous by thread: Have JDOM / XOM / etc. failed? If so, why?
• Index(es):
  • Date
  • Thread