JVM and Saxon are two separate things. There is no real benefit to
using the new version of Saxon with as XSLT 1.0 stylesheet and the older
versions of Saxon should still work with the new version of JVM. There
are some speed improvements in Saxon after version 6.53, but other than
that, the only real benefit is if you are parsing an XSLT 2.0 stylesheet.
Michele R Combs wrote:
> Something very weird is going on and I hope someone can tell me why. We've been using an adapted eadcbs6.xsl stylesheet for upwards of 2 years with no problems. We're using saxon to run it to produce HTML from our EAD. Earlier this week (at the request of our IT department) we switched to the latest version of Saxon, due to security concerns with the Java Virtual Machine required by our older version of Saxon. We tested it, everything looked good.
> However, I have just discovered that for some reason with the new version of Saxon (220.127.116.11, I believe -- the Java version, not the .NET version) the entire section in eadcbs6.xsl where the <meta> tags are generated (roughly line 225 on) produces nothing!! This is the section of code starts with "<!-- Generates HTML meta tags that are inserted into the HTML ouput -->"
> So where our output HTML used to include for example:
> <meta http-equiv="Content-Type" name="dc.title" content="George William Curtis Letters"> or
> <meta http-equiv="Content-Type" name="dc.subject" content="Journalists, United States.">
> there is now zip, nada, bupkis.
> I tested it by running some EAD files through both the old and the new version of Saxon and comparing the output, and quite clearly it's being caused by the new Saxon since that's the only variable. I don't think it has to do with the saxon:omit-meta-tag attribute, as that seems to apply only to the first one, the one that specifies the file type as text/html. Does it have something to do with this saxon being XSL2.0? Our stylesheet is still XSL1.0 and saxon throws up a warning about that, but the transformation completes successfully so I figured it didn't matter. Thoughts??
> Michele Combs
> Manuscripts Librarian
> Special Collections Research Center
> Syracuse University Libraries
> 222 Waverly Ave.
> Syracuse, NY 13244
> [log in to unmask]