[Fwd: SV: [jdom-interest] Line numbers]

Phill_Perryman at Mitel.COM Phill_Perryman at Mitel.COM
Wed Aug 27 09:16:03 PDT 2003

Some of us would disagree (see Jason's earlier comment). I process computer
generated files with many thousands of elements. They don't have parsing
errors so why would I want a larger and slower application. And if it
should give me an error it would be on line 1 as he files don't have line
breaks (more waste of space).

IS Dept, Software Engineer.
phill_perryman at mitel.com
Tel: +44 1291 436023
----- Forwarded by Phill Perryman/Cal/Mitel on 27/08/2003 17:05 -----
                      Mike Brenner                                                                                                     
                      <mikeb at mitre.org>        To:       jdom-interest at jdom.org                                                        
                      Sent by:                 cc:                                                                                     
                      jdom-interest-admi       Subject:  [Fwd: SV: [jdom-interest] Line numbers]                                       
                      n at jdom.org                                                                                                       
                      27/08/2003 16:47                                                                                                 

----- Message from Mike Brenner <mikeb at mitre.org> on Wed, 27 Aug 2003
10:19:36 -0400 -----
      To: Paul Cantrell                
          <cantrell at pobox.com>         
 Subject: Re: SV: [jdom-interest] Line 

Paul Cantrell wrote:
> To the JDOM developers: This seems like it would be a really useful
> feature for JDOM to support, and it looks like something you could do
> fairly generically, without compromising JDOM's design.  If you have a
> feature request tracking system, well, consider this a request.

In general, the weakest link in existing xml tools is the
inability to identify exactly what went wrong in terms
the user can immediately fix. XSL, SVG, OWL, EBXML,
MATHML, etc., are particularly odious offenders.

Four bytes of memory per element would be a trivial price
to pay for knowing which line of which file has which problem.
I think it would actually be more bytes, but it would be
great for JDOM to offer the ability to tell the user
exactly what went wrong, and why, and where.

More information about the jdom-interest mailing list