Greg Green wrote: > > For my part, I have good intentions of downloading and testing the > beta versions, but time always slips away before I am able to get to > it. One problem is that it is rather painstaking to make sure that all > of the seperate pieces (Qt, Sip, Python, PyQt) all match while you are > trying to build them. I haven't come up with a good process there. If > the build process recognized that the versions were mismatched and > stopped that would probably help, but I don't know if that is a > practical thing to do. It would certainly help the volume on the > mailling list though, as a large percentage of the volume on the list > is people discovering the hard way that this matters.
I'm not sure I agree. If the version of Python matters then it's a bug. So far, SIP and PyQt have had the same version numbers so people make the connection. The configure process should handle the different Qt versions - if it doesn't then it's a bug. My perception of the majority of installation problems is that people have multiple versions of Qt installed. Trolltech's claims of forward and backward compatibility are definately overrated. PyQt, by its very nature, is always going to be the most complex Qt program ever written. It exercises the API like no other single application. People have multiple copies installed (obviously this only applies to Linux) because the distros are very, very slow in upgrading to new releases. From an ego point of view it's very nice to know that your code is being shipped with Red Hat, Suse etc, but their approach to keeping it up to date is completely wrong. There is zero communication between those distros and the contributors of the packages. If Red Hat, etc. were to tell me (under NDA if necessary) what their release schedule was, I'd be more than happy to make sure that they could ship with something that was reliable but as up to date as possible. I find it ironic, given that the open source model is supposed to produce better quality software, that Linux distros simply apply the test of time as the only QA measure. In other words, this version is 6 months old and doesn't seem to have many bug reports raised against it - so it must be good (or, rather, it can't be crap). In reality, it doesn't have bug reports against it because nobody is using that version any more. Sorry - I'll climb down from my soap-box now. Phil _______________________________________________ PyKDE mailing list [EMAIL PROTECTED] http://mats.gmd.de/mailman/listinfo/pykde
