On Sat April 11 2009 8:04:20 am Sven Burmeister wrote:
Am Samstag, 11. April 2009 12:31:39 schrieb Carlos E. R.:
It's a tricky situation, this. We are told that we'll get kde3 from the OBS from now on, but it turns out that if there are bugs there they are rejected.
KDE3 will be community maintained, so if there is a KDE3 community, it will fix bugs. I think it makes sense that openSUSE focuses its resources on the products that are part of the 11.2 release.
Sven
I think it makes sense that openSuSE focuses its resources on the CURRENT release, not a release that is sometime in the future. It also seems that by correcting bugs in the CURRENT release, the fixes can be ported/included in the future one. This idea of saying we will support the CURRENT release until EOL and then saying 'well, we have no resources because we are concentrating on the future unreleased version' is quite backwards. Fix the current bugs, incorporate the fixes in the development versions and you've doubled your resource productivity. Bugs are mistakes in logic or programming and if you or a dev made a mistake in (say11.1), even if the development version isn't going to use that exact code, if the features of (say 11.2) are going to be incorporated in any form, then it follows that early mistakes in logic will be repeated. Fixing the logic/code for the current version will help prevent bugs in future versions caused by making the same mistake over and over and over. Fixing it and applying the fix/knowledge/logic to the future nips those bugs before most of them can happen. It also lends credibility to openSuSE's promise to maintain the CURRENT version. The promise to maintain the current version doesn't only include the promise to fix security bugs, but ALL bugs in the release. The order the bugs are attacked surely depend on their severity, but all are important to kill. As it is now, a few months from now, 11.2 will be released and that will be it's death sentence in terms of fixing bugs because attention will turn to 11.3 development and of course, leftover bugs in earlier versions will have little chance of being fixed because the limited resources are then again focused on the future version. Don't BACKPORT fixes, FORWARD PORT fixes from the CURRENT release(s) to the future releases under development. As to the issue of peripheral packages like KDE. Well, I seem to recall that when the TIRES on a certain manufactures cars started failing and causing accidents, BOTH the tire manufacturer AND the auto manufacturer were sued and the manufacturer of the car was just as responsibleas its supplier in getting the issue fixed. As openSuSE incorporates KDEn as one of its' more popular and even essential to many, features, it is incumbant on openSuSE to assume responsibility for ensuring bugs in released versions are fixed and not by saying the equivilent of 'buy a new car, the new tires we use solved the problem...but the new car won't be available for X months and you'll have to live with your old one if you don't want to buy the new one.'. That's bogus IMO. At a minimum, poor management decisions at the highest levels. The grunts do as they are told and could do it right if the management would let/make them. As openSuSE is the 'beta test' platform for SLE*, bugs not being addressed/fixed in openSuSE's current release(s) will often make it into the SLE* distributions that have a higher level of support requirements. Bugs reaching these levels will be more costly, both in terms of manpower and in terms of bottom line profit. It also erodes the credibility of the product Novell is selling unnecessarily. Putting the focus back where it belongs is the first step in improving not only the current release, but the future ones AND ultimately the bottom line profits of Novell. But first, management at all levels has to change *their* attitude and redirect their limited resources even at the expense of a 'release a month' type of scheduling. OpenSuSE used to be known for quality. Its' reputation is rapidly becoming very tarnished, and this thread illustrates a very large reason why. When Steve and I owned and distributed QuickBBS, we definitely had limited resources (only two of us) and we both felt fixing bugs was the most important thing before we added features to the product. I believe I am speaking from experience and lessons learned from that venture. While QuickBBS predated most of the internet, without it and programs like it, I'm not sure that communications for and by the public would be where it is today. It and (now because of my age) I are being or have been replaced by better and newer but I still get mail from old users (customers) of that product and in some places, it is actually still being used and some of the ideas and features I see in Websites (internet equivelent of a BBS in many cases) are quite literally modifications of features we had in QuickBBS. Like the devs of today, both Steve and I had DAY JOBS, Martin-Maretta, now Lockheed Martin Corp, so we understand allocating our time the way the devs who also have day jobs in most cases must do also. When Adam Hudson wrote the original version of QuickBBS, his job was school, then he discovered 'girls' so Steve and I had a head start with the code, to be sure, but that is much like open source software where the devs have mountains of code and libraries to work with. Still, he and I were working in a cutting edge environment as is Linux and openSuSE. That never prevented fixing bugs from coming before developing new features....we could do both and the bugs tended to be new ones, not old ones just hanging around because our resources wouldn't permit devoting the time to finding them. I told this story mostly to try and show I am not completely unknowledgable about BOTH sides of the QC issues, allocation of resources, dealing with the public (customers), long hours debugging (no fun) where we'd rather be coding and experimenting (fun). We learned early on about WHERE the priorities needed to be. I am long since being a whiney kid and openSuSE and its' included packages are not in the language I (or Steve) are familiar with. C++ is very different from Pascal and so is PHP, HTML, Java and many other support 'languages' used in or by openSuSE. So, now I am a USER of the product and not a developer, but I still know which end I'm supposed to wipe when I get up off the pot (assuming I make it that far :)) and I still know the focus of bug fixing is WRONG now. My company and its' employees were never allowed to forget either. I think our many customers that we (and now that I am retired), they supported over the years benefitted. My company started with installation and support of Altair 8800 computers in hospitals and attorneys offices, expanded to include newer and better equipment, added custom software design and support and I don't think it would have grown if we did business by putting the priorities in the wrong places like I think Novell and openSuSE are seemingly doing. I hope openSuSE will rethink its' priorities concerning bugfixing vs development and make adjustments. If openSuSE, as a tool for Novell product development, starts costing (or failing to save) money to the parent company, guess who is going to lose? -- Richard -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org