On 6/1/05, Gert Driesen <[EMAIL PROTECTED]> wrote: > Guess we do need to think about it a little more, like: > > - do we want if/unless/failonerror attributes on each individual "subtask" > or only on the <registry> task itself ?
Hmm. Maybe we should go for a mix solution: if any of these attributes is specified in the <registry> task, subtasks inherit its value. If a subtask has any of this attributes specified, it overrides the <registry> task setting. What about this? > - do we want a subkey attribute on these subtasks like the Ant task has ? Yes, I like this design. I had only a quick look at those Ant tasks, but I liked the general idea. > - do we only support string type values ? In the first version: yes. If the interface and the codebase is stable, we may add other types of values. > - if we get rid of support for mulitple hives, then can this task eventually > replace <readregistry> too, or would people consider this to be a less of > functionality (they can still use a foreach task to accomplish the same) ? Yup, that's a problem. I'd rather replace the current <readregistry> task so that this functionality is not duplicated. Many people use the <foreach> task nowadays, so that shouldn't be an obstacle. > and last but not least: do "we" have time to work on this ;-) I suppose that when you write "we", you think about me :). I can't promise anything of course, but I think I can handle development of this stuff. Best regards! Marcin ------------------------------------------------------- This SF.Net email is sponsored by Yahoo. Introducing Yahoo! Search Developer Network - Create apps using Yahoo! Search APIs Find out how you can build Yahoo! directly into your own Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005 _______________________________________________ NAntContrib-Developer mailing list NAntContrib-Developer@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nantcontrib-developer
