[tei-council] TEI telco Agenda

[Martin Holmes]

Hi all,

Following Kevin's example, I'm presenting quick outlines of the things 
I'm supposed to be reporting on at the telco:

On 16/02/12 00:52, Martin Holmes wrote:

1.3. TEI and AdBlock (MH) (Would be good to get a recap of where
things stand and what we decided)

This is bug <http://purl.org/TEI/BUGS/3472477>. The original problem 
arose when we discovered that one section of the Guidelines was 
invisible to some readers. It turned out that its @id attribute, "msad", 
was on a list of ids being blocked by AdBlock Plus. AdBlock Plus 
behaviour is controlled by any of many different blocking lists, to 
which people subscribe. The two most common lists are EasyList and 
Fanboy's List. I filed a bug with EasyList, and their maintainer fixed 
the problem within 24 hours. I also filed a bug with Fanboy's List, but 
that has been ignored.

In preparation for the Telco, I've taken a look through a
sampling of about ten of the filters listed here:

http://adblockplus.org/en/subscriptions

None of them have the specific element id "msad", although the server
"msads.net" crops up frequently. It may be the case at this point that 
only Fanboy's list has the problem, but the maintainer is completely
unresponsive. It would be impractical to check all the lists on the page
manually, although it could be automated.

There is nothing to prevent a similar problem appearing in future, with 
any of the @xml:ids which give rise to @xhtml:ids in the web version of 
the Guidelines. In view of this, Sebastian suggested changing the output 
XSLT to prefix all ids with "tei_". Lou thought this was a good idea, 
but you warned that it would break some existing links to anchors on the 
web. It would not break links to pages, only to anchors within pages. 
Personally I think it's worth breaking the links, so we should ask 
Sebastian to make the change the stylesheets, but perhaps this is a 
change that should only be made when we move to 2.1, since it may 
partially break some existing links on the web. During the Telco, 
perhaps we can make a decision on this.

1.6. Attribute Datatypes (MH)

This relates to the following bug reports:

<http://purl.org/TEI/BUGS/3441933>
<http://purl.org/TEI/BUGS/3455686>

The basic issue was that some textual descriptions of attributes did not 
match their actual formal datatypes; in investigating this, we 
discovered that many attributes didn't actually have formal datatypes. 
All of the initial problems have been fixed. A table showing the current 
state of attributes is here:

<http://web.uvic.ca/lancenrd/martin/tei/valdescs.htm>

The remaining attributes without datatypes (92 and 409 in that table) 
were intrinsically problematic (the first deprecated, the second not 
fully defined). The latter was given its own ticket:

<http://purl.org/TEI/BUGS/3486352>

which is now closed, because that issue is fixed.

So the only remaining issue is the deprecated attribute quotation/@form, 
which has no datatype. For the sake of completeness, we could assign it 
one, or we could just ignore it.

1.8. TEI Release Process Document (MH)

During the last release, several of us worked to revise a detailed 
version of the Process Release Document "Building a TEI Release", which 
is a working paper here:

<http://www.tei-c.org/Activities/Council/Working/tcw22.xml>

We now believe this is pretty much as comprehensive as it can be, and 
should be able to guide anyone through the process of making a release, 
but please read through it and bring any comments to the Telco. There 
were also a couple of post-release problems with the db that handles 
Roma, which we hope are now fixed, but we won't know for sure until 
another release has been done. Gabriel has volunteered to be the next 
release engineer.


There is one more issue I'm scheduled to speak on:

2. @ref/key/cRef & Private URI Schemes (MH)

and I'm still working on an outline of that.

Cheers,
Martin