[tei-council] proposal: call for stylesheets

[Julia Flanders]

Sebastian, you're right, one stipulation would need to be that the 
donor of any stylesheet gives permission for general use in some 
standard way specified by the TEI. I assume we have such an 
arrangement where applicable for any TEI tool or stylesheet currently 
made available via our site, and that we would continue to do so. 
This is not an obstacle, just a thing to remember.

The point of the contest was not primarily to make a big deal of 
adjudicating among the contestants, but to provide a bit of fun to 
motivate people to contribute. So the questions of how you compare 
them and who are the judges is not really central here. It's just a 
front for a community project (a heap of style sheets that can serve 
as the basis for further development and improvement), a foolish game 
to motivate the activity. But if we don't like the idea of the 
competition, then there's no need to use it. Just requesting that 
people donate stylesheets would also be fine.

I  should also stress that I think having 10 different people 
contributing stylesheets on 10 different areas of the TEI is a good 
thing--this kind of diversity would be a good way to give people more 
options, more examples of possible approaches. Having 10 different 
coding styles might or might not be a problem; I don't know enough 
about XSLT to know whether this would be illuminating for comparison 
purposes, or just a hassle. However, since we're working with donated 
materials here, we're scarcely in a position to complain, unless 
"difference" means "broken". If there are particular problems or 
approaches that we want to exclude we could include this in the RFP.

Concerning your last point--although the TEI stylesheets, as you say, 
are designed for internal use, the fact that they're distributed with 
OxyGen has given them an unintended status. The goal of the 
solicitation proposal was to place these stylesheets in a larger 
context of possibilities, and to reduce people's reliance on them. At 
the moment, people use them for things they're not intended for, just 
because they're handy. It would be great if we had more options to 
show people. My personal feeling is that the TEI probably doesn't 
want to try to create "official" stylesheets--that the better 
approach is to offer people a bunch of samples, and to encourage 
people to create samples that are as good as possible. If people want 
to work on the current TEI stylesheets in an open-source spirit, 
modify them, etc., I think that would be great, but I personally 
wouldn't  be in favor of focusing on these stylesheets instead of 
calling for broader community input and samples.

Anyway--it sounds as if this is one vote against the competition but 
in favor of soliciting donated stylesheets...?

At 8:10 PM +0000 11/29/04, Sebastian Rahtz wrote:
>Julia Flanders wrote:
>
>>
>>  There's also the possibility that making this a sort of community 
>>project would motivate people to make and share improvements on the 
>>most promising samples. ("perfect, if only it did XYZ...")
>>
>>  So I wasn't imagining that people would whip up a stylesheet just 
>>for the competition, but that they would take something they 
>>already have and document and/or polish it for sharing purposes.
>
>If I wear my open source hat, I don't think this is very efficient 
>way to proceed. If we end up with 10 stylesheets
>by 10 different people, written in 10 different coding styles, 
>covering 10 different areas of the TEI,
>with 10 different copyright notices, what then? How do you compare 
>them? Who are the judges?
>It would be much better for people to join in a single community 
>project and contribute according to their ability,
>I think.
>
>I stress the issue of licensing. If a stylesheet written by A N 
>Other is posted on the TEI web site,
>it needs an open source license to be of practical uses to other 
>projects. Better, the copyright would
>be donated to the TEI, so that we can guard the IPR for future generations.
>
>Since I didn't take part in the conversation, I find it hard to 
>understand what people were thinking
>they really wanted here.  The minutes say "Council discussed the 
>problem of SR's style-sheets and the dichotomy
>that they serve both as an example and what the TEI itself uses".  I 
>suggest that this dichotomy
>is entirely illusory: the stylesheets are NOT intended as an example 
>of TEI XSLT programming at all,
>but as a serious contribution for real use. The idea of them as an 
>"example" is a piece of mischievous whimsy
>spread by Michael Beddow; they are _not_ written as a master class 
>in obscure XSLT to challenge
>future generations! The dichotomy is between their twin parentage of 
>what I provide and maintain
>for my local department, and what I wrote with NSF funding for the 
>TEI. Possibly in retrospect it was a bad
>idea to make the stylesheets for the TEIC depend on the ones for Oxford.
>
>It seems to me that there are two very different problems here:
>
>a) we  need more examples of TEI-processing software for people to 
>study and get ideas from.
>    we definitely want to encourage people to show their work. I 
>personally dislike the idea of a
>    competition, but its a possible way to proceed.
>
>b) we (the TEIC) own the copyright[1] of all the XSLT code which I 
>maintain on Sourceforge, and
>   we need it for our internal work (viz, producing the Guidelines); 
>but those stylesheets are not
>   well-documented, and their parameterized interface is fairly 
>off-putting as a result. We need to
>   decide whether the TEI really wants this sorted out, and take a 
>more positive fatherhood role over its
>   bastard child, or whether to leave them as is as internal tools 
>which the editorial process uses.
>  I don't think you solve problem b) with the solution to problem a). 
>So which do we care about more?
>
>Just to clear up a misunderstanding, by the way, the stylesheet 
>releases on the TEI Web site
>are the last version which I wanted to release  publicly. I have not 
>neglected this, it is
>intentional. I am currently in the throes of making the stylesheets 
>do the right thing
>with xml:id and @target, so they are in a mess right now.
>
>[1] its rather important that this be the case, as otherwise I'd 
>have problems with IPR at Oxford. long
>story.
>
>--
>Sebastian Rahtz
>Information Manager, Oxford University Computing Services
>13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
>
>OSS Watch: JISC Open Source Advisory Service
>http://www.oss-watch.ac.uk