Re: Re: [ANNOUNCEMENT] Net_DNSBL-1.0.0 (stable) Released.

On Sun, 2004-11-28 at 16:36, Sebastian Nohn wrote: > Ryan King wrote: > > > I don't believe that a first release is allowed to be 'stable' according > > to the naming conventions. > > Honestly I can't believe that prima donna behaviour for a small and > concise package that has been proven to work in productive enviroments. > > Anyway. What do the naming conventions say about when and under which > circumstances a beta release can be made stable? No need to say such things, it's just common sense that it isn't stable release when you release for the first time, it's only for your own good really, I mean what if you find out something that requires you to break BC right after you release that first release ? You'd have to release Net_DNSBL2, which IMHO isn't very professional :-) I mean that would mean none release between first release and a major version. Even though this class (or it's only the method ?) has been used in some big blog application, it's all the same really, never been a release in PEAR so needs some testing obviously from the people that use PEAR, or well you get my drift, it's much better to release a beta, stick around it for few weeks, see if any bugs creep in (look also at the download statistics) and then you can decide if you want to go stable. Also I _think_ it's now in some documents that a stable package has to have end user docs, but I'm not sure. Also think the tests folder shouldn't have capital T (don't know if it matters for the installer, maybe Greg can answer :P If it doesn't matter then ignore this part :-)) Anyway don't call it primadonna behavior if we're actually trying to follow our own rules and possibly saving you from a big headache :) - Regards Helgi