> > Anyone know of a site that will validate remotely ...
> ... I don't think so ...
> There used to be such a service on the OTA pages, and we could (I
> suppose) add one to the new TEi pages. But....
> 1. What dtd would you expect us to use to validate it?
Ummm... the TEI DTD as specified in the DOCTYPE declaration of the
instance, perhaps? Whatever it does, it should not presume TEI-Lite!
> 2. SGML or XML?
I have to admit I was only thinking SGML. Is TEI ready for prime time
in XML yet? As pointed out by Dave Pawson on this list last month
there are TEI elements for which no XML content model exists.
> 3. What entities?
Good question. I guess none unless the user specifies some entity
sets in the internal subset.
> Is it really so hard for people to get hold of a quick fast
> validating parser like rxp or sp ? Hey, I believe you can even get
> an xml parser that works from Microslop these days...
As one of the people responsible for the ports of SGMLS to VM and to
the Macintosh, I'm probably not the person to answer this. On the
ohter hand, my gut instinct is that it's very easy to get a hold of a
validator, but another thing entirely to get it installed and running
on any given system.
> > [The pizza pages are] http://www.uic.edu/orgs/tei/sgml/pizza.html
> > and http://www.hcu.ox.ac.uk/TEI/pizza.html; they are not the
> > same, and each has advantages and disadvantages.
> Please expound on what you consider these to be ... since the
> former has no visible means of support, I'd like to be sure we've
> snarfed all the advantages from it before it disappears!
I prefer the Chicago page for both minor and major reasons.
I'm used to it. Also, because it has less explanatory prose with
useful links, and does not have the links to the Guidelines from
each tagset, it has fewer "links" for me to navigate through (I
often tab to links / radio buttons / check boxes, rather than use
the mouse). The Oxford page has approximately 103 such spots
before I get to generate my DTD, whereas Chicago has around 84.
The Chicago page is valid HTML; the Oxford page has errors. So as
not to bore people, I'll send a full report directly to Lou.
The page says "chapter six" where it means "chapter seven".
Not only can I not tell the Oxford page what FPI I want to use
for my extension files, I can't get it to use my extension files
at all. Some browsers do not give me anywhere to type in their
URLs at all, or don't allow me to type into the box they've
given me, and only allow me to browse for a file on my local
system. In general not only are my extension files not on my
local system, I don't normally run a web server (nor even an
FTP server, actually) on my desktop. Some browsers allow me
to enter a URL, but the page doesn't seem to use them properly.
Whether I enter an HTTP or an FTP URL, I get something like this
as the subset:
<!DOCTYPE TEI.2 SYSTEM "tei2.dtd" [
<!ENTITY % TEI.prose "INCLUDE">
<!ENTITY % TEI.linking "INCLUDE">
<!ENTITY % TEI.figures "INCLUDE">
<!ENTITY % TEI.analysis "INCLUDE">
<!ENTITY % isolat1 PUBLIC "ISO 8879:1986//ENTITIES
isolat1//EN" > %isolat1;
<!ENTITY % TEI.extensions.dtd SYSTEM "ext.dtd">
Here the system identifier for extensions.dtd file is inadaquate
(I had specified "http://mama.stg.brown.edu/~syd/ext.dtd"), and
the entire declaration for extensions.ent is missing. Other
browsers don't allow me to specify anything at all, but that
is a limitation of the browser, not the page.
Furthermore, if I do not check off any ISO entity sets, the
following invalid declaration and reference appear in my subset:
<!ENTITY % PUBLIC "ISO 8879:1986//ENTITIES //EN" > %;
> > Note that the "- O", "- -" and " and required an end-tag" parts
> > would be dropped if you're using XML instead of SGML.
> Here's a gotcha. You MUST include these, whether or not you're
> aiming for XML, if you plan to put this through the pizzachef,
> since the pizzachef only speaks SGML right now (even tho it outputs
Thanks for letting us know. Probably should include this information
on the page, probably along with the "don't refer to external para-
meter entities from your extension files" warning.
> > Also, I am not sure what, if anything, should be done with the
> > TEIform= attribute in a case like this.
> If there were a pre-existing TEI element of which the SECURITY
> element might reasonably be regarded as a specialization, ...
Indeed, and I agree with you. But what I didn't know what to do
with is the TEIform= on the modified OPENER element.
 TEI-L posting of 2001-01-11T15:31:17Z. I believe that the URL
will get you there.
 iCab (pre 2.1), Internet Explorer (5), and Netscape Communicator (4.7). I
suspect this behavior can be attributed to the specificatoin of type="file"
on the INPUT element.
 emacs w3, Charlotte.
 WannaBe 1.0a46, MacLynx beta 1, Lynx (2.8.3, I think).