On 2015-10-24 20:17, Richard Gaskin wrote:
This is part of the reason I raised this thread. We've seen some rather broadly varying ideas about what Open Language means, but I share your concerns and I'm not sure that's what Mark intended, even if some are excited at the prospect.
I was long concerned about what would happen if the ability to add syntax was 'opened up' to all. However I ceased being concerned about it after I worked out how it could be done.
People should be free to experiment and develop their own syntax for their own ends - why should we restrict that?
As it stands LiveCode syntax is *heavily* based around the idea of it being a domain-specific language for GUI development. However, I think the ideas present within LiveCode (and predecessor xTalks) have far far greater applicability than that.
At the end of the day, any syntax extensions will be under heavy constraint due to the existing syntactic forms which exist in the engine already and in terms of what potential users will expect.
Market forces, will in the end, sort out the 'good' from the 'bad' - and as Geoff Canyon pointed out even if a 'bad' syntax extension becomes popular - it will only do so because it provides facilities that no-one else has provided and are important... In which case it might be 'distasteful' to use, but pragmatically that doesn't matter if it gets the job done.
I believe the core of the issue is that to date all xTalks have required comma-delimited arguments for custom commands. The good news is that this is pretty much how most programming languages work, so it's not particularly onerous. But the bad news is that it means that the libraries we share bear no relationship syntactically with the build-in commands.
Indeed - it has been a long standing problem. Something which has been often opined about by a great many people within our community (including yourself, I believe, at certain points - although I might be wrong on that ;)).
However, I should perhaps point out that Open Language evolved from me trying to work out how to actually service the great many feature requests we get for adding syntax. The current implementation of the parser in the engine, whilst it works, is incredibly difficult to modify. This not only means that it is not at all straightforward to add English-like syntax to new features, but also that it is far too expensive (time-wise) to experiment with new syntactic ideas. Unfortunately, syntax is one of those things that you really do need to try before you buy - thought experiments really don't cut it in many cases.
I *believe* (emphasis added to note that I'd he happy to be corrected by Mr. Waddingham if this isn't correct) that Open Language was proposed primarily (perhaps only) to allow library scripters to define syntax that fits in more closely with the flavor of the rest of the language. For example, today I might write: CreateDocument "Window Title", tFilePath ...but with Open Language I could write: create document titled "Window Title" using file tFilePath Mr. Waddingham, is that correct?
Pretty much. Syntax is about patterns. There are already a great many patterns which exist in LiveCode Script, even if they are not directly apparant.
That said, I admit I'm rather enamored of this request: <http://quality.runrev.com/show_bug.cgi?id=3157>
That would be probably about a 8 line definition file to implement when we have Open Language - I'll look forward to making it available on April 1st the year Open Language arrives :)
Warmest Regards, Mark. -- Mark Waddingham ~ m...@livecode.com ~ http://www.livecode.com/ LiveCode: Everyone can create apps _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
