Austin Acton <[EMAIL PROTECTED]> writes: [...]
> 1. Throwing in new features right before a release (post-alpha) and not > giving them time to mature (many drak's in the v8 and now v9 releases > for example). Why do you have to wait so late in introduce new > utilities or features? don't you have a clue why? it's quite simple, it's due to time limits (esp. the dreaded fact that days are only 24 hours) redhat has many new tools in its 8.0 beta (limbo), many don't even run. So at least, they're not doing better. > Look how upset we are about rpmdrake! Can you > make us happy by release-time? guess what... we're trying ;p > 2. Rushing the beta testing and/or releasing the final before WE think > it's ready. I realize that WE are not the experts, but WE are the > people buying, supporting, and promoting something that we KNEW had bugs > before it was released. I don't want this to turn into a Debian style > turtle-speed-distro (hehe), but already I feel like 9.0 is going to be > released before it sounds "polished," much like 8.1. we have announced our deadlines for some time now. We won't postpone them unless something really wrong happens. postponing is *bad*, a freeze must be short otherwise: - at mandrakesoft, developpers grow tired of testing. testing is not the most fun part. - people stop testing a frozen distro (guess how many developpers use debian's unstable and not stable) - bug-fixing in nearly gold distro is time consuming and leads to ugly hacks. Since to lessen the side-effect of a modification, you duplicate code, hack locally instead of changing the architecture more globally. Of course those hacks will get in your way in future :-( > 3. Difficulty in reporting bugs. This is the least serious, and it is > getting better. This list is great. Bugzilla... not so great. Know that the difficulty also comes from the bug reporters. Bugzilla used to be more fully opened resulting in many dumb bug reports. Taking care of this is time consuming.
