On 04/25/2017 01:23 PM, Tomas Chvatal wrote:
Richard Brown píše v Út 25. 04. 2017 v 12:01 +0200:
On 25 April 2017 at 11:37, Jan Engelhardt <jengelh@inai.de> wrote:
On Tuesday 2017-04-25 10:34, ianseeks wrote:
I remember people saying the same when we jumped to 42.x just over two years ago.
In many ways, they were wrong - I didn't spend as much time as I expected repeating that message. It's been over a year since I had to deal with a question on the topic. One year of having to answer questions about why 15? Sure, I can live with that.
Is spending man hours for a year trying to justify a mistake with firstly with Leap 42 and now going backwards a good use of time? It would be easier to swallow the old pride and say, ok, we got it wrong with Leap 42, lets just call it opensuse 15.x
Is going backwards, and the discussion that it entails, a good use of time? It would be easier to swallow the pride and admit that REPEATING a jump is even worse.
Come on Jan, you're not doing your public perception any favours here. As a key member of the Factory review team you're expected to be rational and considerate. You wouldn't like it if I publicly complained in terms of "good use of time" every time you get on your high horse and block other contributors contributions for reasons which you think are right, would you?
I'd advise you take a step back and consider that your own personal brand is an important thing to keep in good order. The Board can take its responsibility/praise for whatever harm/benefit this decision may have done to the openSUSE brand.
There is no 'pride' to swallow here. This was a collective decision of the openSUSE Board, the body elected by the community with the responsibility to make decisions on behalf of the community. We did our job, to the best of our ability, considering all of the factors we could. There are no new factors that we didn't consider that have been raised in this thread, so there is no grounds for reconsidering the decision.
And besides, lets imagine we DID reconsider - at this point we'd cause even more confusion and disruption given the world already knows that the next version is going to be Leap 15. Yeah, no, it would take a pretty compelling new piece of information to justify that cost now.
And as you're a Factory reviewer, also consider it from a packagers perspective
"suse_version for Leap 45 is changing from 1315 to 1500. What's the relationship between suse_version and the distro you're using? No clue. You just need to learn. You're new? f*ck you, good luck becoming an openSUSE packagers. You need to reference a specific minor version? the sle_version is changing from 120300 to 150000, the leap_version is changing from 420300 to 450000, but not if you're still building packages for Leap 42.1 that is still using sle_version 120100. You want to build for Tumbleweed, SLE, and Leap? haha, good luck."
The death of the rpm leap_version macro and the simplified nature of suse_version actually having some link to reality now should make things easier for our packagers.
"suse_version for Leap 15 and SLE 15 is 1500. They'll increment together from now on. sle_version includes the minor version for when you need to be precise, like 150100. We're making it as easy as possible to build for Tumbleweed, SLE, and Leap while still ensuring all 3 can diverge when they need to."
Which should mean you should have an easier time and less messed up nested if nonsense in spec files.
The whole 'backwards' complaint is nonsense given we did it with Leap already, in a far more technically invasive way.
We need to find the best way forward for the long term ease of maintainability, clarity for new contributors, and simple easy messaging for the big wide world about the unique things about Leap that differentiate it from every other distribution out there.
Leap 15 ticks all of those boxes. It's the right choice, and I'll probably keep thinking this way until someone actually provides substantive, practical, hands-on evidence of something which the Board overlooked in our decision.
I really have to back up Richard here.
We "as a board" considered all the options and actually going back to match with SUSE Linux Enterprise products seem like the best choice.
In tooling we will be matched up as per what Richard wrote, so packagers won't have to remember that leap is 1315 and defined macro is_opensuse and minor version is yet another macro.
Of course I have to remember it because I will surely build my Leap packages at least 6 more years.
From the simple perspective indeed 15 < 42 :)
Decrementing the version was the only thing you could do wrong, and you did it. Thanks, I'll get over it. But please stop trying to justify this. Nobody decrements version numbers. Show me one other project which did the same BS. Don't you get that you did something wrong?
But on the other hand if it lets us to clean up rest of the mess it is worth doing bit of hassle. Esp. from my packagers point of view I have to say this is quite improving the situation with the macros.
I wish you that all the upstream sources you are packaging would decrement their version. cu, Rudi -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org