Re: grouping-size

I suggest you post on the xsl-list at mulberrytech.com to make contact with people who know the XSL-FO scene.

Michael Kay
Saxonica

> On 22 Dec 2021, at 17:04, Victor Mote <vic@outfitr.com> wrote:
> 
> Hi Michael:
> 
> Thanks for your insight. That all makes sense.
> 
> I knew that XSL-FO 2.0 was stalled, but didn't realize it was dead. If there is interest in reviving it, I would be interested in helping. I have worked on two open-source Java implementations, and continue to work on one of those now. I have doubts about how useful that experience would be in helping to move a standard forward.
> 
> Victor Mote
> 
> On Tue, Dec 21, 2021 at 5:23 PM Michael Kay <mhk@mhk.me.uk <mailto:mhk@mhk.me.uk>> wrote:
> Personal response: as far as I'm aware there is no active development or maintenance of the XSL-FO specs. You will be aware that work on XSL-FO 2.0 was never completed; if I recall correctly this was because there weren't enough volunteers to complete the work. Updating the specs to work with XSLT 3.0 would surely involve many more small changes like this one, and the reality is that it isn't going to happen.
> 
> I'll observe that XSLT 1.0 and 2.0 didn't say what happened if grouping-size was set to something other than a positive integer, so the chances are that implementations will vary. During 3.0 development we spotted this omission and made the decision that it should be an error. XSL-FO implementers might like to take that as guidance on the preferred interpretation.
> 
> Regards,
> 
> Michael Kay
> Saxonica
> 
> > On 22 Dec 2021, at 00:04, Victor Mote <vic@outfitr.com <mailto:vic@outfitr.com>> wrote:
> > 
> > Hi XSL-FO Editors:
> > 
> > There is what seems to me to be a small flaw in the XSL-FO 1.1 definition of the property grouping-size, Section 7.26.3, which defines the value to be a <number>, and references the XSLT Recommendation. This does track with the XSLT 2.0 specification, but the XSLT 3.0 document changes that to <integer>, which makes much more intuitive sense to me. Both documents have the definition very near the top of Section 12.
> > 
> > Since only positive integers are usable for grouping-size, please consider clarifying whether a numeric value other than those should be treated as an error, or whether some rounding scheme should be employed.
> > 
> > My apologies if this has been discussed before. It looks like the Bugzilla site is decommissioned, and I do not find a pertinent discussion in the mailing list archives.
> > 
> > Thanks very much.
> > 
> > Victor Mote
>