This is the mail archive of the
xsl-list@mulberrytech.com
mailing list .
Re: format-number underspecified (Was: XSLT 1.1 comments from Steve Muench)
On Wed, Feb 14, 2001 at 10:23:17AM -0700, Uche Ogbuji wrote:
> > >Oracle has both Java and C implementations. I did not hear the developers
> > >of our C implementation complaining about format-number(). They just
> > >implemented the behavior as noted in the spec.
>
> I missed this message, and apparently all others from yesterday when our ISP
> had a network blackout.
>
> This is in response to Steve Muench.
>
> You represent a company who tells everyone for miles around how you *saved*
> one billion dollars on your IT last year. Lord knows how much money you have
> to throw at developers who are happy to write whatever C code you ask them.
>
> I represent a tiny company who can't even get an ISP with a redundant Internet
> connection. I'm not talking to you WRT format-number() from speculation about
> what "the developers" experienced. I am telling you that with my own fingers
> I struggled with format-number() in C to put together an efficient
> implementation for 4Suite. Jeremy Kloth polished up the job, and between the
> two of us, we spent an inordinate amount of time considering other ways we
> could have provided the same functionality if the XSLT WG had not decided to
> off-load the whole matter to the Java spec.
>
> And Jeremy and I are good C programmers. I have about a decade of experience
> in the language.
>
> It all wouldn't be so bad if Java had some extraordinarily elegant and mature
> mechanism for numerical formatting, but Java's DecimalFormat stuff is very
> poorly designed and even more poorly specified.
>
> I did not set out to implement a bunch of Java quirks: I set out to implement
> an XSLT processor. If the W3C is proud of the fact that open-source authors
> implement their specifications, they will do very well to consider that few
> open-source authors have the resources of Oracle.
I simply concur (except for the lack of redundant connexion) ...
It seems a number of XSLT implementations were actually implemented by one
or a couple of people, and are significant in term of deployment . This is
great, please keep XSLT relatively simple and reduce the few annoying areas
for developers like format-number() rather than expanding them.
(Also remember that opensource, small project implementations are often
keys to pass the Last Call and Candidate REC stage, sometimes large companies
don't invest much resources at that point prefering to wait for a finished
spec.)
Daniel
--
Daniel Veillard | Red Hat Network http://redhat.com/products/network/
veillard@redhat.com | libxml Gnome XML toolkit http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list