James Duncan Davidson <[EMAIL PROTECTED]> wrote: > On 12/18/00 5:28 AM, "Conor MacNeill" <[EMAIL PROTECTED]> wrote:
>> How would you say the current Ant is radically different from what >> you want it to be? > > Its more complicated in syntax.. How so? Because we've added nested elements instead of comma separated lists in attributes? > Its less flexible in picking up tasks.. Quite the opposite, taskdef can load tasks from a different classpath now, while the user was forced to include everything in the system CLASSPATH in Ant 1.1 and earlier. And then nobody disagrees on making that system better, I think you can go back to June or July to see the first proposals on how to make that better. The main reasons why it is not there yet is (1) there has been a huge demand for a 1.2 release as it fixed tons of bugs (not all leftovers from the former presidency) and we didn't want to put the new features in without bigger tests and (2) development has more or less stalled here when you announced your Revolution. > And it doesn't lend itself to integration well. Which has been addressed as well, especially with Sim's "Proposed API refactoring" thread that predates AntEater as well. > And it's a pain in the arse to deal with all those scripts which > were intended to go away. Granted, add that to the spec document. AFAIK three of the four proposals work more or less without scripts. >> So who gets to define what is Ant? Is that you? > > Bluntly, yes. Sorry, but I disagree. You have as much power as any other committer. I'm not saying you should go away, quite the opposite, but I don't see room for a "primus inter pares" either. And if you want to enforce it > With the help of a lot of people. I'm afraid that won't be the same people that would be here to help otherwise. Stefan
