It's just the way they present it "Python 3.10.0" is in reality "Python3 Version 10.0".

As to the question at hand. I don't know s**t about the python ecosystem, but at some point in the not too distant future things requiring the new features of 3.10 are likely to show up.

Given Slackware's slow release cadence I don't think holding back is an option.

Well, I voted too. Thanks to ponce for the nudge on #slackbuilds channel

Py3.10 being in -current now has given me the opportunity to slack off a bit from wearing the maintainer's hat when it comes to testing progs (and deps) on SBo-git. upstream has to play catch-up to Py3.10 as well... So I naturally intend to use this happenstance block of free time (is there even such a thing anymore these dayS?) to dive into the wonders, and curses, of esp32 programming, soldering, blogging, whilst trying not to burn down the flat.

See you in November.

Yep! Let's remember that this choice is not for the immediate satisfaction that pyCrapOne compiles and works today.

No matter which version is chosen, the Slackware 15.0 release will be stuck with it for the next 7-8 years - and in several years, you will wish to have 3.10 instead.

Thinking on long term, probably the best one is 3.10 because more and more programs will require it.

And no. The Debian and its bastard offspring Devuan aren't the best examples to follow, because those guys are famous that they use old software releases with tons of changes made by themselves, because reasons.

I tell'ya as long term Ubuntunian.

If Patrick would have wanted to follow this path he would have done that twenty years ago. But I wouldn't make any assumption about the span of life of Slackware 15.

If everything works with python 3.9 but only something with python 3.10, then it is essentially our hopes things will be fixed in the future. As well we can stuck with broken python packages. How long it will take to developers to move on? And on side of my paranoia if Fedora goes with python 3.10 means me for my paranoia stay away as far as possible from it. Nice distro to put on computer to make it blow up time to time.

Point is: how much is in decision a thought that life is too short so let it be?

I dislike change. I voted 3.10 because the change has been made, things only move forward, other upstream programs will support 3.10 soon and any issues can always be solved. Also people will figure out ways to live with both versions.

Wael Hammoudeh

3.9 for 15.0

Those knowledgable enough to want the particular improvements in 3.10 are also those most able to deal with any issues caused by it.

Proven with no surprises for a stable release please

We already know that python3.10 builds on slackware 15. What we don't know is what packages it broke that need to be updated. And when the upstream for each package updates, we won't know if that update will require dependency FOO, which might not get into slackware 15 either, if it goes stable shortly. We might have to resort to patching code and that's not ideal. In another words, it is a no-win battle sticking in python3.10 now just hoping that it is best for long term. We don't know that. If we wait until python3.10 is well tested, it will probably take some time to do that, then some other package updates and people will want it too. But that breaks more things. No win battle.

I'd say that people that want python3.10 will be able to install it anyway (we know it builds), and will be able to deal with the breakage when it is found. python3.9 should be sufficient for a release now, and it is better tested anyway.

My opinion: Set a Slackware 15.0 release date (you don't have to debate this with us). Name the versions of packages that are pending, get them in, and get -stable out. You can still make as many packages as possible compliant with python3.10 in the source tree in the meantime anyway, so we can get the best of both worlds, without worrying about breakage too much. The only caveat is I still don't know how inter-twined python can be, particularly with the build tool chain (meson, ninja, etc). I really need a stable tool chain, and need something that doesn't move long enough so I can get some real work done. In that sense, python3.9 ought to be fine for our tool chain, if it is not actually, then we got bigger problems beyond what anyone can do about it here, since who can ever depend on python if it breaks things that easy every year?

Does anyone know of anything that doesn't have an update that works with 3.10?

In fact, the big question is:

Is something from Slackware 15.0 RC1 having issues with Python 3.10 ?

We aren't supposed to be The World Guardians, right?

So, let's leave the saving of World in the hands of Greta Thunberg, OK? ;)

Hell if I know...
 

I honestly don't know - other than having python3 as a dependency for me I suppose it does not matter. It probably matters more to those who write python scripts - which also brings me to that other thread about whether python2 should be outright removed on the grounds that it is EOL and a potential security risk. I wish there was a third option because I seriously do not know if it should be 3.9 or 3.10 - given that it is just the 3.x branch and either way security patches will apply to either version. I vote "I just don't know, whatever version is fine by me."

Python is and always will be a moving target.

It's simultaneously a programming language and an interpreter for that language, and the developers keep making changes to both between releases, giving us the worst of both worlds.

As there is no perfect solution, we might as well go with the latest version. Otherwise, Slackware 15 is pretty much guaranteed to be stuck with a non-supported version with security vulnerabilities before the next release.

My sixth sense told me our BDFL already made decision. No need to keep this thread going on. Is about 2:1 for python 3.10. If democracy matters - which we know it does not - then it should be 3.10. For me the only question is where doubts come from? Our post are useless from point of decision-making. Is it about kind of risk being cause by introducing 3.10? But if we would be able to evaluate - which we are not - even the high risk is not decisive. There are other things which may justify to take high risk. Good manager is making decision not only because of all information - but because of interacting with cosmic forces and true nature of everything.

Kinda unfortunate if you ask me. I am by no means knowledgeable about programming in general - I mean I know things like C++ has new standards added, but I bet it is not as fast as how Python adopts new standards, maybe it can since it is an interpreted language vs C++ which is a compiled language , and maybe that makes all the difference I do not know; however it seems that it is catching those who write python are being caught off guard. I can understand perhaps the transition from 2 to 3 - then again even the maintainers gave ample warning beforehand to get their python2 code transitioned over to python3, then again easier said than done depending how complex their projects are. Maybe perhaps the devs at python should slow down a bit? Again I don't know, but from what I infer is that languages such as C or C++ have a slower rate of changes vs python.

Tough choice really. If we choose 3.9 does Slackware 15 ship earlier? If yes then I vote for 3.9.
if not then should we care? Any breakages will be fixed via upstream eventually.