Ports Update : python-2.7 flagged as EOLed, so Thunderbird port will expire as well ?

Status
Not open for further replies.
heavy lifting to cling to Python2/Tauthon.

Heavy lifting to cling to Python2? For the time being only security updates are needed. How big is the PyPy community compared to Python? Pypy 2.7 is still maintained. PyPy has three versions.

Python 2 is murdered. And it rises from its grave to compete Python3 into oblivion.
 
1. Python devs are not going to commit to working on or backport security updates. Why look for security bugs in a product that isn't maintained. Keep your 0-day relatively few people care about and use it on outdated stuff. All that work is going to have to come from somewhere.

2. Switching to something "compatible" still requires testing and retooling. If you're going to do that, but you were planning on upgrading to Python3, why retool to PyPy when you weren't planning on doing that.
 
Actually the ongoing migration to Python 3 is well underway. The number of ports that depend on Python 2 is shrinking.

If someone is willing to spend time on Python improvement, that time is much better spent for that migration and for improving Python 3 support in the ports collection. Python 2 is dead (or let’s say “retired” if you prefer). Trying to keep it alive & working is just a waste of time – and even worse, it is dangerous because it might encourage people to continue using it after it goes EOS and starts to acquire security issues that won’t be fixed.

I have personally migrated dozens of Python programs this year, probably > 100 kLOC. It’s not really difficult (provided you’re somewhat familiar with Python, of course), and it doesn’t even take much time, thanks to the “2to3” tool which automates a significant part of the conversion. It’s well worth the effort anyway, because Python 3 is so much cleaner, more consistent and easier to work with.
 
It’s well worth the effort anyway, because Python 3 is so much cleaner, more consistent and easier to work with.
The migration strategy for Python 3 is disruptive, so your arguments are false. We have spend a lot of time learning Python 2 and building what we want. Now we have to redo everything again to get a so called cleaner, more consistent en easier code? Do you have any idea how much that costs? Money that could have spend improving our Python 2 projects. We do not experience the value added by Python 3. We only experience the expenses. Python 2 is easy, clean and consistent enough. Python 2 is secure. It is not broken. Do not fix things that are not broken. Keeping Python 2 alive is not a waste of time. It is worth the effort because it is cheaper. They killed Python 2 because the market didn't move to Python 3 out of choice. It is aggressively forced upon the community. Python 3 is a way to beat competitors for big tech.
 
We have spend a lot of time learning Python 2 and building what we want.
I don't know how long ago you started but the impending end of python 2 was known well in advance. The EoL shouldn't come as a surprise.
 
What I never understood is, why they kept Python 2 in parallel to Python 3 for more than 11 years. Only this made the inevitable hard break much harder.
 
Not so long ago, I took a book with some snippets of python. Maybe it's time to learn this very used language, I've told to myself.

At this time, there were both python 2.7 and python 3.x on my FreeBSD box (as dependencies of some packages). I started to run the scripts with python 2.7. All was fine. Then, I tried python 3.x: several errors came.

Ok, no problem, the book was old and the scripts were very simple. I managed to convert one of them to python 3.x. The funny thing is that, despite the conversion, the script also worked for python 2.7.

I found this strange. Then, I realised the python developers have invented a new type of compatibility: not backward as most language are but forward...

It instaneously clicked for me: if I learn python 3, I probably should get rid of it when python 4 will come.

I closed the book. I will stick to C/C++.
 
We have spend a lot of time learning Python 2 and building what we want.
If you’re really so good at Python 2, it shouldn’t take you much effort to adjust to Python 3. It’s not a completely new language. The changes are sensible and necessary, because sometimes it’s better to cut backwards compatibility in order to progress and make a language better. There are certain languages that try to maintain backwards-compatibility at all cost, which means that they can never get rid of their design mistakes.

I’m afraid I have to say, if you are unable to get used to changes like that, then you shouldn’t be in the IT business.
 
What I never understood is, why they kept Python 2 in parallel to Python 3 for more than 11 years. Only this made the inevitable hard break much harder.
It was only hard because people decided to make it that way and not take the hint. Now that an obnoxious amount of time has passed, the Python team is done wasting their time.
 
Now we have to redo everything again to get a so called cleaner, more consistent en easier code? Do you have any idea how much that costs? Money that could have spend improving our Python 2 projects. We do not experience the value added by Python 3.
Funny, that's extremely similar to the argument the Python team makes for not supporting Python2 anymore.

We only experience the expenses
Hey, guess what, you paid nothing to write in Python, and now you're done getting the value of the nothing you paid into the support contract that extended the EOL six years! If you've been using Python for more than (checks watch) 12 years, that's 12 years of paying nothing where it would have been an irresponsible business decision to not save whatever money you were making off your Python2 programs to invest the bare minimum into updating your scripts for Python 3. Seems like you could have scrimped a little here and there over that timeframe. In the business world, that's simply the cost of doing business.

Surely, you haven't been intentionally writing 2.7 code in a 3 incompatible way for that long, so you should be pretty good at doing it the correct way by now, making transitioning whatever legacy codebase easier than if you ignored the warning signs to your business for twelve years.

I think most proprietary development tools stop working after at least decade, and you pay for those. Those people have to buy new dev tools every 5 years or so to keep current. Given that Python is free, seems like you'd still have a leg up over them. Considering they have to go through sunsetting of their programming tools and changing paradigms (MFC, Silverlight, XAML, WPF) more often than Python2 to Python3, I think you're still better off.

Of course, if you started developing 2.7 code after 2008 (or really, 2006 when it became obvious that Python3 was the upgrade route) without a 3.x plan, that just seems like poor planning and a bad business decision.
 
  • Like
Reactions: a6h
Non of the argument for the EOL of Python 2 are valid. It is about money. A way to get payed for development because Python is FOS. We are now forced to pay developers for upgrading instead for enhancing our Python 2 projects. Money can only spend once. To make money to pay developers, we need enhancements and (more) hosting of our Python 2 projects instead of upgrading tot Python 3. The EOL of Python 2 is a form of unfair competition. It is a way of killing projects that can compete with big tech. Upgrading to Python 3 does not give us more sales. Enhancing our Python 2 project does.

Never seen something like this in the more than 20 years we use FOS.
 
Guido van Rossum is even threatening with lawyers if you want to maintain your own python 2 stack
No, that's not what he's saying. The legal talk is about the name, not the code. They can't call it Python or some variation thereof.
 
Guido van Rossum is even threatening with lawyers if you want to maintain your own python 2 stack, making it impossible to keep Python 2 alive https://github.com/naftaliharris/tauthon/issues/47 How unfair is that? It is FOS!
They locked that thread because of you on January 21. Now you're here complaining about the same thing 9 months later? Are you trying to resurrect your argument here just to keep it alive somewhere?
 
It is about money.
What's wrong with that?
We are now forced to pay developers for upgrading instead for enhancing our Python 2 projects.
Developers should charge more. Good for them.
Upgrading to Python 3 does not give us more sales.
You probably don't have competitive advantages, thus trying to lay the blame on python.
Guido van Rossum is even threatening with lawyers if you want to maintain your own python 2 stack, making it impossible to keep Python 2 alive
I'm not sure that's true, but he's the creator of python. More power to him.
How unfair is that?
Q. How to tell when someone is losing an argument?
A. They eventually use the plug-word: "unfair".
 
They locked that thread because of you on January 21.
And it's about to happen here too, for exactly the same reason.

Juronski if you have a problem with the EoL of Python 2 you can take that up with the Python developers. This is not the place for that discussion. Python 2 will be EoL'd upstream and as a consequence it's going to be EoL on FreeBSD too. No amount of discussion is going to change that.
 
Non of the argument for the EOL of Python 2 are valid
Hey, glad my post was too much for you to read and point out exactly what was wrong.

We are now forced to pay developers for upgrading instead for enhancing our Python 2 projects. Money can only spend once. To make money to pay developers, we need enhancements and (more) hosting of our Python 2 projects instead of upgrading tot Python 3.
I think I already indicated earlier that you had almost a decade to budget in costs for Python 3 upgrades (or simply changed your coding style to be prepared for them, which costs just about nothing) when charging your customers. If you didn't do this, that's your problem, not Python's.

It's hard to imagine that a poor decision has caused a problem that is now out of your control. You're right, the money can only be spent once, and it should have been spent on preparing to update to Python 3 so that your business could continue to function in the future. It may not have seemed "essential" at the time, but as you realize now, you can not force someone to do work they do not want to do such as continue to work on Python2, just like the Python developers could not force you to update your code.

Your lived reality is that you have a problem now, but if you won't acknowledge that your refusal to prepare for the future contributed to the problem, you're not going to get much sympathy here. Trying to cite "business" as an emphasis actually undermines any point you try to make, as business is a "do or die" environment with no consideration to the welfare of companies that go under. What you should have "done" is update to Python3, because now you're facing the prospect of the company "dying."

The EOL of Python 2 is a form of unfair competition
EoLing products, particularly when the EoL date is set in advance so you can make plans, is not a form of unfair competition.

Let us look upon the Python 2 license and despair:

* Provided on an "as is" basis
* No warranty
* No merchantability
* No fitness for any particular purpose
* No partnership
* No permission to use the trademark (aka "Python")

Python2 had no obligation to you. You chose Python 2 at your own risk. Your only exit hatch is that you can do the work of keeping Python2 alive yourself.
Never seen something like this in the more than 20 years we use FOS.
20 years, you say? Let's see... You must be used to this kind of disappointment already, when you realized that FreeBSD stopped providing security fixes, packages and ports that would work on version 4.0.
 
What's astonishing is your out of hand dismissals.

I'm sorry your business is built on Python2 and the consensus is that nobody wants to support the environment you built your software on.

That's just how business goes, however.

What you're mistaken about, is that berating people on the internet is going to create the consensus necessary to solve your business problems.
 
Unbelievable... Software goes EOL all the time... It's amazing what some people expect for free...
 
I've had enough of this. Thread closed.

Juronski If you start a new thread or take any other thread offtopic with this same discussion you're going to be banned.
 
Status
Not open for further replies.
Back
Top