That's ok for me, that's why I said that we shouldn't have another snap build with 0.9.0 version.
-----Original Message-----In pure version number terms, 0.9.0 *is* beta 1. It will not be on any other release.
From: Rapha�l Enrici [mailto:[EMAIL PROTECTED] Sent: 08 August 2003 15:28
To: Dave Page
Subject: Re: [pgadmin-hackers] Package naming conventions
If you look to what I did in debian packages,
the actual beta is 0.9.0-0.1
if we build a snapshot release of a new 0.9.0 (what we shouldn't as it would be incompatible with what we said before), it will be named 0.9.0-0.1+cvsYYYYMMDD.1 which is greater than 0.9.0-0.1 so it's ok. if we build a snapshot release of a new devel branch i.e 0.9.1 (what should be the right way of handling this), it will be named 0.9.0-0.0[AND NOT 1]+cvsYYYYMMDD.1 which is also greater than 0.9.0-0.1, so upgrade is also ok.
I just add that we may have to release new packages to correct some breaks coming from the package itself for example and that this will be followed by the minor number incremented in the package number. (0.9.0-0.1, 0.9.0-0.2,... and finally 0.9.0-1.0 if the package become an official member of debian... May be one day... Who knows ?)
Snapshots will now be 0.9.1 + date, and then beta 2 will be 0.9.2, then snapshots will be 0.9.3 + date and so on.Still ok for me :) but the date is just part of the package release for snapshots.
Do we need anything more complicated?No that's what I tried to told in my previous mail (I surely badly explained it), with only some considerations regarding packages. But that's not what is actually done for some packages on the ftp site: I just wanted to cath your eyes on this.... So shall the rpm packages, and may be some other, be rebuilt for this beta release ? If so, Jean Michel do you need some help concerning this work ?
Regards,
Rapha�l
Regards, Dave.
PS. In pgAdmin II we didn't use dates, but each build incremented the build number (z in x.y.z). That relied on VB to autoincrement the number though :-(
---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
