On Tuesday, November 22, 2011 10:00 AM Anton Aylward wrote: [snip] [snip] [snip adinfinitum]
I'd go further. Compare with Microsoft.
Look at the delays they had bringing out Vista and 7 and look at the number of critical post-release bugs that had to be (still are being) fixed.
And they do have the resources and processes for QA.
Appears that the "A question" and "A reply" threads have somewhat overlapped. I guess this is as good a point as any to toss in my two cents . . . Of course the quantity & quality of resources are always a factor, as well as well-developed engineering and testing processes. But no matter what the resources or processes, nor how long the schedule, there will always be issues with any new major release. The probability is exponential given an OS's innumerable moving parts, inter-dependencies, and use cases - let alone when thousands of end-user apps are added to the mix (a challenge we shouldn't forget that MS largely offloads to the 3rd-parties). My frame of reference is having worked in both the lab and as a field engineer at one of the largest hardware manufacturers (who btw write much more sw than MS). It was our policy to discourage customers from immediately upgrading to a new release unless it contained a feature or fix which was truly imperative to their business. This policy was the same at our major competitors. Sometimes we preferred to backport a patch (if feasible) rather than compel an upgrade. Of course, our "setting expectations" policy resulted in push-back from sales and marketing, sometimes customers or even our own management. Nevertheless, it was critical for maintaining customer satisfaction. From some posts here I gather than openSUSE is sometimes used for internal or even external customers. In these cases, I would think that whomever is responsible for supporting the upgrades would, if permitted, communicate similar cautions and strongly encourage waiting until, at the minimum, the first major round of patches. This step is often signaled as a discrete step in the form of a "service pack". While desktop power-users understand the risks with any new release, most end- users do not. MS has of course played a major unhelpful role here, resulting in the avg desktop user having the highest expectations yet lowest ability to contend with unexpected problems. All the more reason for caution and setting expectations. Rhetorical questions: How many users read the Release Notes before upgrading? How many users read the critical bugs page? How many users know this info exists? How many users understand what the info means and/or its significance? How good a job do we do in encouraging all users to read and understand this info in advance? What advice or tools do we encourage users to consider for planning for an upgrade, whether a simulation via an upgrade Live-DVD or another method? Should we put more focus on these areas? One last question which I think relates: I wonder if we are compounding the challenge by trying to do too much? Both Fedora and Ubuntu have a more focused target audience. And both distributions (despite having more resources) deliver just one desktop environment. Additional environments are the province of spins or fork projects. Yet openSUSE delivers KDE, Gnome, Xfce, desktop and server, and more, all in a single offering. In short, perhaps the answer to a better new release experience, rather than more testers (a reactive solution) and the like, is better communication, better setting of expectations, easier ways of pre-determining the upgrade impact, and a tighter focus in terms of users, product structure, and deliverables. And, finally, what are the effects of all these decisions on project growth and sustainability? If you've read thru all this, thanks and please forgive the length. And where you disagree, no need to bring out the sledge hammer. --Dennis -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org