What's stopping you using the java classes? You could even use it from Jython. Isn't that non-XML enough for you? If you really want to go the non-XML route, you could always use Scons.
Anything else would be YASL - and, personally, I think that's a waste of time and energy. Stick to the XML - it may not be pure but it's what made Ant the success it is today. Roger On Tue, 20 Jul 2004 08:37:41 -0500, Dominique Devienne <[EMAIL PROTECTED]> wrote: > > From: Nicolas Mailhot [mailto:[EMAIL PROTECTED] > > > > On mar, 2004-07-20 at 09:56 +0200, Johan Vromans CPWR wrote: > > > The current situation is that we have the worst of all: it's not a > > > programming language, it's not a scripting language, it's not even > > > XML anymore. > > > > The way dependency lists are put inside a text array in an attribute > > never was XML-ish (better use several depend children). Or the use of > > flat-text property files (though pure XML is gaining ground). > > > > Anyway, a lot of the shortfalls of ant syntax disappear when you couple > > it to some xslt processing, but then you get bitten by the parts that > > look like XML but aren't really (which can be workarounded without too > > much effort, but is a PITA nevertheless). > > If some of you guys out there are hung on the XML syntax (which I personally > don't have a problem with), nothing ever prevents you from introducing a new > ProjectHelper instead of the default XML-based one. As someone (Jan or > Matt?) recently wrote on this (dev?) list, if you guys find a flaw in the > logic that allows to implement custom ProjectHelpers, the Ant team is > committed to fix it to make ProjectHelper useful and implement-able. This is > different than endorsing these alternative syntaxes, but good enough, no? > > As far as the critique about Ant using non-XMLish constructs like the > depends attribute of targets (true enough), again, nothing prevents a group > to define a more XML-friendly 'Canonical Ant' schema with an XSL to go in > either directions. This would also work for non-XML syntax, which could > transform their own grammars into this canonical Ant XML. RelaxNG showed > it's possible with the XML and compact syntax. I'll just note that the .rnc > syntax is great for simple schemas, but that it quickly gets worse than the > XML (IMHO) to support more RNG features. The good thing is that is has the > two syntaxes, so that ends the debate ;-) > > In short, some people have the itch to introduce an alternate non-XML syntax > to Ant. Are you going to do anything about it? --DD > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > -- http://www.badstep.net --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
