Re: [xsl] Re: Does anyone know how to make IE less useless for XSLT developement?

["Terence Kearns" <terence.kearns@xxxxxxxxxxxxxxx>]

Dimitre,

It was interesting to see how many of the processors that you tested
with actually opted to report the error. It's food for thought.

Thankyou also for taking the time to report your findings on what
makes good programmers and bad programmers.


On 22/02/06, Dimitre Novatchev <dnovatchev@xxxxxxxxx> wrote:
> > I have run this through 3 other processors and none of them complained
> > about the above template. See the error IE gives in my original post
> > below.
>
> I wonder how you could find so many incompliant XSLT processors.
>
> Of course, "@name" is not a valid QName and a compliant XSLT processor
> raises an error.
>
> From all 12 XSLT processors I tested with:
>
>     MSXML3
>     MSXML4
>     MSXML6
>     .NET XslTransform
>     JD
>     XalanJ 2.4.1
>     XalanC 1.5
>     Gexslt,
>     Saxon 6.5.3
>     Saxon 8.5.1
>     Saxon 8.6.1
>     Saxon.NET
>
> only two didn't raise an error (JD and XalanC 1.5)
>
> From my observations, good programmers rarely blame the programming
> environment for every issue, but check very thouroughly that they
> indeed know the subject they are dealing with.
>
> Jeni Tennison, M.D. Peterson and Chris Bayes, who all have great
> XSLT-based web sites, haave give never blamed IE for any temporary
> problems they had (and who hasn't such problems) in the development of
> their sites. This is why they were successful.
>
> Judging from the text of your last message, you still don't know with
> which version of MSXML you actually perform the transformation -- let
> me help here, it is MSXML3.
>
>
> --
> Cheers,
> Dimitre Novatchev
> ---------------------------------------
> The significant problems we have cannot be solved at the same level of
> thinking with which we created them.
>
>
>
> On 2/22/06, Terence Kearns <terence.kearns@xxxxxxxxxxxxxxx> wrote:
> > OK, for anyone interested, I eventually traced down the problem.
> > Internet Explorer (MSXML) cannot handle the following template. And
> > now I have another question.
> >
> > <xsl:template match="HtmlAttribute">
> >    <xsl:attribute name="@name"><xsl:value-of
select="text()"/></xsl:attribute>
> > </xsl:template>
> >
> > So is this a bug in MSXML?
> >
> > I have run this through 3 other processors and none of them complained
> > about the above template. See the error IE gives in my original post
> > below.
> >
> > BTW. to uncover this, I had to poke and prod with NUMEROUS commenting
> > outs to track down the problem. OK, this is normal for a developer
> > (sometimes), but IE is a huge pain because it does not read the
> > stylesheet changes unless you exit all instances of IE and start it up
> > again. I had to do this each time I commented something out to test
> > it. Furthermore, it STILL says "the XML source is unavailable for
> > viewing". As far as I am concerned, this is a REAL WORLD issue and
> > weighs heavily against IE in the XML/XSLT development stakes if one
> > wants to evaluate and compare browser platforms - I don't care how
> > inferior Transformiix is supposed to be under certain conditions, at
> > least I don't have to restart Mozilla for every single code change.
> >
> > Anyway, I think the code above has shown MSXML to have a bug when
> > Transformiix does not. So it's 6 of one, half a dozen of the other.
> > They both have shortcomings. One allows you to develop with it - in
> > business time speeds.
> >
> > I'm not sure how I'm gonna work around this bug. I need this functionality
:/
> >
> >
> >
> > On 10/02/06, Terence Kearns <terence.kearns@xxxxxxxxxxxxxxx> wrote:
> > > Unfortunately there are still some traditionalists out there who
> > > insist on using IE. So far, my web application is only working in
> > > Firefox. I load it up in IE, and I just get a lame error with no line
> > > number and "the source is unavailable". Honestly, I don't know how
> > > people put up with developing apps exclusively under IE. It's like
> > > having both arms and legs chopped off once you've developed with FF.
> > > IE says "This name may not contain the '@' character: -->@<--name
> > > Error occurred during compilation of included or imported sty..."
> > > My server-side transform processor doesn't compain about any errors
> > > either so I'm assuming it's not just FF being lenient. I don't want to
> > > transform server-side unless I absolutely have to.
> > >
> > > Surely there must be some sort of extension that you can get for IE
> > > that allows you to debug with it.