- From: Sandro Hawke <sandro@w3.org>
- Date: Wed, 30 Sep 2009 08:33:37 -0400
- To: Stella Mitchell <stellamit@gmail.com>
- cc: Axel Polleres <axel.polleres@deri.org>, "RIF (E-mail)" <public-rif-wg@w3.org>
> --001636ed6f6b7f38060474ca97ca > Content-Type: text/plain; charset=ISO-8859-1 > > > > > As for the rest of this thread... I don't care about "the" PS. It's > > just there as a shorthand for specifying the semantics. I'll consider > > it a huge failure of the community if people ever actually use it > > (instead of better PS's that come along). > > > > My comment was that since the RIF PS defines a variable name to be an > NCName, then, according to the specification, any variable name in an > admissible RIF XML document must be an NCName. Oh, right. I forgot about that issue. Yes, I'm pretty sure we decided to not restrict variable names a long time ago, and we can't easily change that at this point. What's the status of this in the current (nearly-CR) documents? We should probably throw in a test case with some variable names with spaces and non-ascii characters in them. Force people to escape them in their translators. (I expect people to do something like turn all non-alpha-numeric characters to the hex for the character, delimited by "_", or something like that.) - s
Received on Wednesday, 30 September 2009 12:33:42 UTC