- From: John Lumley <john@saxonica.com>
- Date: Fri, 15 Dec 2023 18:40:56 +0000
- To: Norm Tovey-Walsh <norm@saxonica.com>
- Cc: "C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>, Steven Pemberton <steven.pemberton@cwi.nl>, public-ixml@w3.org
- Message-Id: <CD9F20FA-86DE-4C53-AB2F-5E64F95935FB@saxonica.com>
Why not use % - we don’t use it for anything else. And what about considering the prolog to be a property map, eg. %ixml version: “1.2”, foo: “bar” % John Lumley Sent from my iPad > On 15 Dec 2023, at 18:26, Norm Tovey-Walsh <norm@saxonica.com> wrote: > > "C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com> writes: >> Actually, I think there is a reason for the "ixml": human readers who >> are trying to figure out what format the file they are looking at is in > > Indeed. It’s useful for iXML to have an identifiable “magic number”[1]. > >> My first thought, looking at the examples, was that we could do a lot >> worse than using the syntax of XML processing instructions with >> pseudo-attributes, and use >> >> <?ixml version="1.1"?> > > That’s clever, but isn’t it also inviting a certain amount of confusion? > If you type <?xml instead of <?ixml that’s going to be quite different. > Given that iXML and XML rub shoulders, I’m not sure we want to make the > distinction “blurry”. > > We already use parentheses for grouping, so I’m not an immediate fan of > wrapping the prolog in them. I think I’d favor something that wasn’t > already in the syntax. Luckily, there’s no real pressure here for it to > be a single character. So this could work: > > ::: The prolog goes here ::: > > If I hadn’t *just* said that I favor something that isn’t already in the > syntax, I’d be tempted to suggest Python-style multiline strings: > > """ > The prolog goes here > """ > > Be seeing you, > norm > > [1] https://en.wikipedia.org/wiki/File_format#Magic_number > > -- > Norm Tovey-Walsh > Saxonica
Attachments
- application/octet-stream attachment: signature.asc
Received on Friday, 15 December 2023 18:41:11 UTC