- From: Tony Graham <tgraham@mentea.net>
- Date: Tue, 12 Feb 2013 14:21:58 -0000 (GMT)
- To: public-ppl@w3.org
Firstly, even if I'm repeating myself by saying it, it's good to see active discussion taking place here again. I would like to see that we can channel the current energy into something of long-term use and value. Secondly, I think that there's a few things that we need or need to think about before embarking on spec changes or API specifications: - Buy-in from more of the CG members -- we've heard from a few people so far, but doing things properly will require a lot of email traffic and focus for quite a while, so if people we haven't heard from have objections or other things they'd like to see done, now would be a good time to air them - A copy of the WD -- "We" don't have a copy of the XSL 2.0 WD. Several of us could lay our hands on the last internal editor's draft that hasn't been made public, but we, the CG, don't have a clear-cut right to use it, and nor do we yet have a CG member-accessible source code repository to keep it in. Since the XPPL WG is/was a closed group, we would strictly speaking need Liam or someone within the W3C to make the last draft public and/or bless our use of it, and we'd need the infrastructure maintainers to make a Mercurial repository for us. - Review of existing requirements -- we should know to what extent the existing requirements cover what we're talking about - A common understanding of the area tree model -- the makeup of the area tree is described here and there in the text of the spec. I don't know that there is a comprehensive description of the area tree that you'd use as the model -- the infoset, if you like, to deliberately use a loaded term -- for designing the objects and methods of an API. Several of us have a mental model of how the area tree fits together, and some even have one in running code, but we, collectively, don't know whether we all have the same model in mind, and we've already had one question today about where the spec says you should write out the area tree. (IMO, it doesn't say you have to, but others may disagree.) Having a model of the area tree would also be useful if and when we get to other XSL 2.0 requirements such as multi-volume indexes. - Use cases -- another aspect of knowing that we're all talking about the same thing Regards, Tony.
Received on Tuesday, 12 February 2013 14:22:21 UTC