Re: solution for those depending on PEAR packages with non-PEAR

Pierre-Alain Joye wrote

What S. Schmidt said is the perfect example of what I do not like to see: "it would be great, if I could use it to provide patched versions any just any package" You do understand that each patched package would be a separate url, and if a higher version of the original package is released, it would automatically overwrite the patched version? This is not easy - maintaining a continually patched version of a package would be a tremendous headache for a developer. I'm talking about providing the ability to release a version of a package that contains a patch in CVS that is known to work for a small situation.

I understand Lukas point of view quit well. He is sane enough to apply the patch exception for (nearly) only bugs. But what you propose is: Hey guys, let patches, forks, whatever you want to do, any times you do not like something or want something. And this is scary. And not what I propose :). Read the lines (and also between them): the solution I propose makes it somewhat difficult for developers to make patched versions, but easy for users to install them. This is the right way to do this, even if the implementation details need working out. Unlike channels, which should be easy for developers to set up, making a patched version should be kind of annoying, so that people are discouraged from doing it for trivial problems. However, making it possible is a different story from making it difficult.

That is, unless PEAR really is useless. If that's the case, then I think we need to talk :).

Is it possible to stop to be a demagogue, giving wrong effects to non existent causes. That brings nothing and does not help to give a honest and clear explanation. You talking to me or to yourself?

Greg