On porting to Python 3 as the answer
This really upset me:
On Sun, Mar 23, 2014 at 3:17 AM,
I think asking developers to make significant modifications to their code is besides the point of the PEP. However, if they are willing to make changes, I'd still recommend that they port their code to Python 3, as that is the better long-term investment.
This is a completely unrealistic form of wishful thinking, and repeating it won't make it more true. At Dropbox I work with a large group of very capable developers on several large code bases that are currently in 2.7. We are constantly changing our code to make it more secure (there are several teams specifically in charge of that). And yet porting to Python 3 is completely out of scope, for a variety of reasons. Please stop your wishful thinking. (TBH, I expect that none of the changes to Python 2.7 under consideration would make any difference for the security of Dropbox. But neither would switching to Python 3.) -- --Guido van Rossum (python.org/~guido)
Am 23.03.14 17:22, schrieb Guido van Rossum:
At Dropbox I work with a large group of very capable developers on several large code bases that are currently in 2.7. We are constantly changing our code to make it more secure (there are several teams specifically in charge of that). And yet porting to Python 3 is completely out of scope, for a variety of reasons.
Please stop your wishful thinking.
I can stop expressing it; I don't think I can stop wishing it :-) If it's really unrealistic that Dropbox will ever port the code to Python 3, would you then think that Python 3 is a doomed project, since it won't ever see significant usage? Regards, Martin
On Mar 23, 2014, at 12:33 PM, Martin v. Löwis
Am 23.03.14 17:22, schrieb Guido van Rossum:
At Dropbox I work with a large group of very capable developers on several large code bases that are currently in 2.7. We are constantly changing our code to make it more secure (there are several teams specifically in charge of that). And yet porting to Python 3 is completely out of scope, for a variety of reasons.
Please stop your wishful thinking.
I can stop expressing it; I don't think I can stop wishing it :-)
If it's really unrealistic that Dropbox will ever port the code to Python 3, would you then think that Python 3 is a doomed project, since it won't ever see significant usage?
I think expecting every production instance of Python 2.7 to port is unrealistic. New projects will start to get written in Python 3, some existing projects will get ported over time. But I doubt there is ever going to be some mass exodus where everyone suddenly starts moving to Python 3. For people/companies/projects where Python 2 is currently working fine and porting would be a significant effort they have to decide between adding more value to their project through bug fixes, new features, etc or essentially sinking cost into Porting to Python 3 which their end users will not benefit from in a very large way. It’s hard to make that business case for existing code, especially given that for a lot of the truly hard parts of porting to Python 3 there’s no way to port module by module so you have to basically do an entire code base at once. This adds extra churn, increases the risk for new bugs, and is quite an annoying process. That doesn’t mean Python 3 is bad or is inherently a doomed project, but it’s essentially a new, but very similar, language.
Regards, Martin
_______________________________________________ Python-Dev mailing list Python-Dev@python.org https://mail.python.org/mailman/listinfo/python-dev Unsubscribe: https://mail.python.org/mailman/options/python-dev/donald%40stufft.io
----------------- Donald Stufft PGP: 0x6E3CBCE93372DCFA // 7C6B 7C5D 5E2B 6356 A926 F04F 6E3C BCE9 3372 DCFA
On Sun, Mar 23, 2014 at 9:33 AM, "Martin v. Löwis"
Am 23.03.14 17:22, schrieb Guido van Rossum:
At Dropbox I work with a large group of very capable developers on several large code bases that are currently in 2.7. We are constantly changing our code to make it more secure (there are several teams specifically in charge of that). And yet porting to Python 3 is completely out of scope, for a variety of reasons.
Please stop your wishful thinking.
I can stop expressing it; I don't think I can stop wishing it :-)
If it's really unrealistic that Dropbox will ever port the code to Python 3, would you then think that Python 3 is a doomed project, since it won't ever see significant usage?
No, it's just that the timescale is drastically different. -- --Guido van Rossum (python.org/~guido)
On 24 Mar 2014 03:48, "Guido van Rossum"
On Sun, Mar 23, 2014 at 9:33 AM, "Martin v. Löwis"
wrote:
Am 23.03.14 17:22, schrieb Guido van Rossum:
At Dropbox I work with a large group of very capable developers on several large code bases that are currently in 2.7. We are constantly changing our code to make it more secure (there are several teams specifically in charge of that). And yet porting to Python 3 is completely out of scope, for a variety of reasons.
Please stop your wishful thinking.
I can stop expressing it; I don't think I can stop wishing it :-)
If it's really unrealistic that Dropbox will ever port the code to Python 3, would you then think that Python 3 is a doomed project, since it won't ever see significant usage?
No, it's just that the timescale is drastically different.
Yep, and Paul was ultimately right in guessing that part of my motivation here was seeing trouble on the horizon in terms of RH's ability to affordably sustain our own long term support commitments for the Python 2 series. That said, I didn't actually realise the full implications for us until *after* writing, publishing and receiving feedback on the PEP, though - otherwise I would likely have done things in a different order. Isn't open source fun? :) So I'm now going to switch my focus on this topic to the Fedora community for a while: I think the second draft of the PEP is in a pretty reasonable state, and from my perspective, anything we come up with is going to have to at least pass muster with the Fedora Engineering Steering Committee for it to actually solve the upcoming supportability problem on that side of things. It would be good if folks engaged with other downstream redistributors could also get them to take a look and provide feedback. A plan we like, but key redistributors choose not to go along with, would be rather missing the point... Cheers, Nick.
-- --Guido van Rossum (python.org/~guido)
_______________________________________________ Python-Dev mailing list Python-Dev@python.org https://mail.python.org/mailman/listinfo/python-dev Unsubscribe:
https://mail.python.org/mailman/options/python-dev/ncoghlan%40gmail.com
participants (4)
-
"Martin v. Löwis" -
Donald Stufft -
Guido van Rossum -
Nick Coghlan