Re: Re: documentation revisioning (was Re: sockets extension...pecl it)

On Thu, Sep 12, 2002 at 01:29:41PM -0700, Brian Lalor wrote : > Markus Fischer <mfischer@guru.josefine.at> writes: > > > As long as the extension is marked EXPERIMENTAL so are the > > docs and so they may be out of date for current version and > > also up to date with current versions and not reflect the > > behaviour of any of the older versions. > > *sigh* > > I should have made this more clear. The sockets is but an *example* of why > the documentation un-versioned is a bad idea. Yes, the sockets code is > experimental, but the MySQL stuff (for example) is not. If someone changes > the documentation in CVS and that gets put up on the website, then the poor > unsuspecting user going to look up the spec for mysql_connect() is going to be > scratching their head trying to figure out why the heck the code they're > writing that conforms to the documentation doesn't work. You made yourself clear enough. Basically: Currently we do not version the manual in the way you want it to. And: Good Luck going for it :) Currently we do not have resources to keep the complete manual updated enough for every release. So I don't see how we could manage versioned manuals. Also, if you make a snapshot of the manual there's no guarantee that every function is up to date documented. So you end up with a manual say for 4.0.6 which has mysql_whatever() only documented until 4.0.4 and then you have to go anyway for more current docs. IMHO, that *is* more confusing. -- GnuPG Key: http://guru.josefine.at/~mfischer/C2272BD0.asc - It's not a fug, it's a beature. -