Thread: Ready to roll beta
I plan to roll 1.2beta1 tomorrow unless anyone has a good reason not to. I know there is an outstanding issue with FC2, however I'm not convinced this is something that need holdup production of the source tarball. If anyone disagrees or has any other objections, please let me know ASAP. Regards, Dave.
----Message d'origine---- >Sujet: [pgadmin-hackers] Ready to roll beta >Date: Tue, 7 Sep 2004 21:35:57 +0100 >De: "Dave Page" <dpage@vale-housing.co.uk> >A: <pgadmin-hackers@postgresql.org> > >I plan to roll 1.2beta1 tomorrow unless anyone has a good reason not to. >I know there is an outstanding issue with FC2, however I'm not convinced >this is something that need holdup production of the source tarball. > >If anyone disagrees or has any other objections, please let me know >ASAP. So, I SHOUT! :) Can you let us a couple of day to manage all the build to be tested at least once. As discussed yesterday evening, I have some stuff to review (should not be too much) and I must take care of keyboard issuesunder debian. FC2 should be ok now thanks to Diego and Devrim but we should document the way --with-wx* configure options work as it wasjust the way to call them which was not correct. Can we wait until the WE ? Thx. Raphaël
> -----Original Message----- > From: blacknoz@club-internet.fr [mailto:blacknoz@club-internet.fr] > Sent: 08 September 2004 10:54 > To: Dave Page > Cc: pgadmin-hackers@postgresql.org > Subject: Re: [pgadmin-hackers] Ready to roll beta > > So, I SHOUT! :) <grin> > Can you let us a couple of day to manage all the build to be > tested at least once. > As discussed yesterday evening, I have some stuff to review > (should not be too much) and I must take care of keyboard > issues under debian. Thought that one was resolved... > FC2 should be ok now thanks to Diego and Devrim but we should > document the way --with-wx* configure options work as it was > just the way to call them which was not correct. From what I could gather, the configure script failed to run the symlinked wx-config, which I can only imagine is a shell issue, but yes, a doc would help. > Can we wait until the WE ? We can, but I can't hold off the postgresql-win32 beta any more, so we will have to ship the current cvs tip with that. Regards, Dave
blacknoz@club-internet.fr wrote: > > So, I SHOUT! :) > > Can you let us a couple of day to manage all the build to be tested at least once. > As discussed yesterday evening, I have some stuff to review (should not be too much) and I must take care of keyboard issuesunder debian. > FC2 should be ok now thanks to Diego and Devrim but we should document the way --with-wx* configure options work as itwas just the way to call them which was not correct. > > Can we wait until the WE ? > Thx. Well, this is an unfortunate situation. Beta1 win32 should roll ASAP to be included in pginstaller Beta2 (which is overdue), but we need time for *ix. I'd vote for making a code freeze now, bumping up the version to Beta1 now, but delaying cvs tagging until all configuration stuff is fixed. This way, we can get Beta1 for win32 out now, and have the time to tidy up for *ix beta. Regards, Andreas
> -----Original Message----- > From: Andreas Pflug [mailto:pgadmin@pse-consulting.de] > Sent: 08 September 2004 09:06 > To: blacknoz@club-internet.fr > Cc: Dave Page; pgadmin-hackers@postgresql.org > Subject: Re: [pgadmin-hackers] Ready to roll beta > > Well, this is an unfortunate situation. Beta1 win32 should > roll ASAP to be included in pginstaller Beta2 (which is > overdue), but we need time for *ix. > > I'd vote for making a code freeze now, bumping up the version > to Beta1 now, but delaying cvs tagging until all > configuration stuff is fixed. > > This way, we can get Beta1 for win32 out now, and have the > time to tidy up for *ix beta. Hmm, I'm not convinced the fc2 issue is one we can fix anyway (it's trying to execute the right file, just seems to be failing because it's a symlink - which shouldn't happen of course). The other issue is the 'f' key on Debian - are we simply waiting for confirmation that the newer compiler resolves that? Also, don't forget that everything at release is built from the source tarball, and not CVS (at least it should be). So once that is created, that's it. Regards, Dave.
Dave Page wrote: > > > > > > Also, don't forget that everything at release is built from the source > tarball, and not CVS (at least it should be). So once that is created, > that's it. I meant to release Beta1 win32 as binary only. Regards, Andreas
> -----Original Message----- > From: Andreas Pflug [mailto:pgadmin@pse-consulting.de] > Sent: 08 September 2004 09:29 > To: Dave Page > Cc: blacknoz@club-internet.fr; pgadmin-hackers@postgresql.org > Subject: Re: [pgadmin-hackers] Ready to roll beta > > Dave Page wrote: > > > > > > > > > > > > Also, don't forget that everything at release is built from > the source > > tarball, and not CVS (at least it should be). So once that > is created, > > that's it. > > I meant to release Beta1 win32 as binary only. Yes, but to do that (properly) I need to produce the tarball first. That's the best way to ensure consistent releases. Or are you saying that we ignore consistency in this instance? Regards, Dave.
Dave Page wrote: > > Yes, but to do that (properly) I need to produce the tarball first. > That's the best way to ensure consistent releases. Or are you saying > that we ignore consistency in this instance? I don't understand the problem. The binary is for pginstaller only, and its supporting files (languages) might differ slightly from the official pgAdmin3 beta release (if we allow it) which we wouldn't publish and announce until cvs is tagged. The version number of pgAdmin3 that's included in pgInstaller should reflect that it is functionally identical to pgAdmin3 Beta1. How should this affect consistency? Regards, Andreas
> -----Original Message----- > From: Andreas Pflug [mailto:pgadmin@pse-consulting.de] > Sent: 08 September 2004 09:46 > To: Dave Page > Cc: blacknoz@club-internet.fr; pgadmin-hackers@postgresql.org > Subject: Re: [pgadmin-hackers] Ready to roll beta > > Dave Page wrote: > > > > > Yes, but to do that (properly) I need to produce the tarball first. > > That's the best way to ensure consistent releases. Or are > you saying > > that we ignore consistency in this instance? > > I don't understand the problem. The binary is for pginstaller > only, and its supporting files (languages) might differ > slightly from the official > pgAdmin3 beta release (if we allow it) which we wouldn't > publish and announce until cvs is tagged. > > The version number of pgAdmin3 that's included in pgInstaller > should reflect that it is functionally identical to pgAdmin3 Beta1. > > How should this affect consistency? Because it will claim to be beta1 even though it does not reflect and cannot be rebuilt exactly the official beta1 tarball. Sod it though it's only beta 1... I'll bump the version number and build the win32 installer. Regards, Dave
Dave Page wrote: > > > > Because it will claim to be beta1 even though it does not reflect and > cannot be rebuilt exactly the official beta1 tarball. If we agree not to change anything for win32, it *can* be rebuilt from a later tarball, no? > Sod it though it's only beta 1... I'll bump the version number and build > the win32 installer. Right. If we really encounter problems, we simply push out another beta immediately. Regards, Andreas
----Message d'origine---- >Sujet: Re: [pgadmin-hackers] Ready to roll beta >Date: Wed, 8 Sep 2004 09:51:10 +0100 >De: "Dave Page" <dpage@vale-housing.co.uk> >A: "Andreas Pflug" <pgadmin@pse-consulting.de> >Copie à: <blacknoz@club-internet.fr>, <pgadmin-hackers@postgresql.org> > > > >> -----Original Message----- >> From: Andreas Pflug [mailto:pgadmin@pse-consulting.de] >> Sent: 08 September 2004 09:46 >> To: Dave Page >> Cc: blacknoz@club-internet.fr; pgadmin-hackers@postgresql.org >> Subject: Re: [pgadmin-hackers] Ready to roll beta >> >> Dave Page wrote: >> >> > >> > Yes, but to do that (properly) I need to produce the tarball first. >> > That's the best way to ensure consistent releases. Or are >> you saying >> > that we ignore consistency in this instance? >> >> I don't understand the problem. The binary is for pginstaller >> only, and its supporting files (languages) might differ >> slightly from the official >> pgAdmin3 beta release (if we allow it) which we wouldn't >> publish and announce until cvs is tagged. >> >> The version number of pgAdmin3 that's included in pgInstaller >> should reflect that it is functionally identical to pgAdmin3 Beta1. >> >> How should this affect consistency? > >Because it will claim to be beta1 even though it does not reflect and >cannot be rebuilt exactly the official beta1 tarball. > >Sod it though it's only beta 1... I'll bump the version number and build >the win32 installer. What about releasing beta1 as we always do so that win32 can be released. My bet is that all debian troubles are due to mistakeson my side. FC2 is solved with the correct configure line. Let's go for it. If I need to patch for debian, I'll patch (I already did this for 1.0.2 which did not build properly fromscratch...) and send feedback for the next beta so that we have a good stable release. Cheers, Raphaël