At 12:44 15/1/01 -0800, Ioan Mitrea wrote: >I would like to know if there is any general idea about when Ant 1.3 >would be released .
I would hope soon .. we need a week or so to stabilize - after that I may raise a vote to see if it gets done ;) >The same question about Ant 2.0 no idea - was going to be mid-year sometime but that was before the "discussion" started. >>From what I can see on the list there is not much consensus about what >Ant 2.0 will be. I think there is consensus on what Ant2 will do but not consensus on how underlying data-model will work. >But is there a consensus way of dealing with backwards compatibility? try to keep it unless backwards compatability is wrong (at least for Ant2.0) - we are trying to keep full backward compatability for Ant1.3 thou. >Will old build.xml files still run or will there be tools converting >build.xml files ? For ant2 there will be a converter. >Is this a major concern in the design of Ant 2 or the benefits coming >from the architectural change >are more important than the compatibility concerns ? As I said there will be an easy path to updating I would guess (via xslt sheet I assume) so when the time comes there will be concern for compatability but we are not there yet. >I'm wondering about what are the risks of converting a complex Makefile >build system to ant and implementing testing on top of it before Ant 2.0 >comes out. Really depends on your environment. I call ant from makefiles to do certain things. If your code base is mostly java then stick with ant - however if you include C++/C and don't have time to build your own ant tasks then Make is still the way to go. Cheers, Pete *-----------------------------------------------------* | "Faced with the choice between changing one's mind, | | and proving that there is no need to do so - almost | | everyone gets busy on the proof." | | - John Kenneth Galbraith | *-----------------------------------------------------*
