Re: please do not use CVS of PEAR for anything important until REST is worked out.

Hi Justin, Hi all, Dunno what is unfair. First of all the CVS is public so far I remember, and I am very well informed of the danger implied by its use (so far I broke everything several times and had a hard time to rebuild). Now, you should understand I'm quite choked by Greg's sudden change of view. If you remember, the mood was much more a call to testers, check the list. Indeed, I took that into account, since more as 3 monthes, each morning (happily UTC+1), when greg did make some changes, I have checkout the CVS and run several tests. As an example, I was the first to mention the eating of memory, what would have been a blocking factor for pear release. You can also check the bugs, and not everything went into the bug system. Now, I must mention, among the 3 bugs I created these last days: - one is a typo where I could have mailed Greg private, but I was in the system as: - one is a weird mix of PEAR::Error and PEAR::ErrorStack as old as November AFAIR - the last could be related to the current changes from Greg, but I'm unsure. I made also a report concerning the whole stable packages available in PEAR. I'm quite surprised to get no answer, but no more surprised when I see where the discussions go. To conclude, I will wait some sign from Greg to start again my reports, I will nevertheless continue my check-outs and my tests. à+ -- bertrand Gugger http://toggg.com/ Justin Patrin wrote: >On 5/5/05, bertrand Gugger <bertrand@toggg.com> wrote: > > >>Hi Greg ! >>I'm not sure to understand your message. >>I assume you mean not using the pear from cvs for production ... >>I will keep on testing it. >>Just what should we do with the bugs/features we find if we cannot use >>the bug system ? >>If your stuff is unachieved, then simply don't publish it untill you are >>happy about it. >> >> >> > >That's not quite fair, bertrand. CVS is for the developer to store >their changes in. Sometimes I will commit something to CVS, even >though it's not done, to make sure that I don't lose changes. I also >may commit if I'm going to try to refactor but I don't want to lose >the changes I made so far just in case I want to revert. > >Basically, Greg is giving us a small courtesy here by telling us not >to use PEAR's CVS because he expects it to be unstable. This is how we >should always think of CVS code unless told otherwise. > > > >>Greg Beaver wrote: >> >> >> >>>Hello all, >>> >>>I am trying to juggle several things at once in my life, and so >>>development has slowed a bit. It is nearly summer, and the dormant >>>PEAR developers will surface soon :). Please do not use CVS of PEAR >>>or report bugs in it until I/we have a chance to get the unit tests >>>working and develop a whole new set of tests for REST-based stuff. >>> >>>Any bugs you report will simply be fixed by these actions, and in the >>>mean time, 1.4.0a11 works very well for everyday actions. >>> >>>Thanks, >>>Greg >>> >>> >>> >>-- >>PEAR Development Mailing List (http://pear.php.net/) >>To unsubscribe, visit: http://www.php.net/unsub.php >> >> >> >> > > > >