Re:Python 3 makes it easier to develop high-quality software
by mangeletti on Wednesday November 18, 2015 @03:51PM
(#10590648)
I could not agree more.
To add to this; if it included even a 2x speed up / memory reduction, I think 90% of everyone would have switched by now, and you know what else? Go probably wouldn't exist.
Re:Python 3 makes it easier to develop high-quality software
by thrownaway2424 on Wednesday November 18, 2015 @05:40PM
(#10591288)
This is a good point. Go is a direct response to Google's experience with Python.
Re:Python 3 makes it easier to develop high-quality software
by k_sze on Wednesday November 18, 2015 @07:22PM
(#10591805)
I would argue that removing the GIL would scare the shit out of developers and slow adoption instead of hasten it. A lot of accidentally thread-safe code would suddenly become thread-unsafe.
Re:Python 3 makes it easier to develop high-quality software
by rdtsc on Wednesday November 18, 2015 @10:47PM
(#10592608)
It was just a hypothetical example of the kinds of changes that would have flocked everyone to Python 3.
Re:Python 3 makes it easier to develop high-quality software
by jrumbut on Wednesday November 18, 2015 @04:48PM
(#10591012)
I think the energy spent supporting Python 2 and the inability to really unify and focus the community on Python 3 is preventing or delaying that kind of dramatic improvement.
If people would rather move to another ecosystem rather than update their code that's fine. The Python community doesn't get paid a nickel every time the interpreter runs, (at least slightly) engaged users are what move Python forward.
Re:Python 3 makes it easier to develop high-quality software
by rdtsc on Wednesday November 18, 2015 @10:52PM
(#10592626)
The difference is if we look at it from the point of view of the Python community (and they are all about Python 3). But another view is that of developers who see themselves as developers of product X or company Y and they value their bread on their table or that X and Y wins more than Python is popular or Python succeeds.
They are not skeptical. Nobody thinks "yeah if I just wait long enough, Python 3 will go away and 2.8 will come out".
They just don't see an incentive to update. Simple as that. It is basic economics of time and money.
Python 2 wasn't terrible and Python 3 isn't dramatically better. Unicode and other stuff in Python 3 is nice. Is it nice enough to start digging into stable, working, money making code just to say "Yay, Python 3"? I think the development community has spoken by how it was handled. It's gonna happen but it will happen at a very slow pace.
Imagine this alternative scenario -- "Python 3 brings 3x speed improvement + removed GIL". Ok, I bet you, there would have been a much faster adoption because now there is a greater incentive to justify it. People would think "Yeah, I'll destabilize my code a bit but I gain performance, I can do that".
Someone in a sibling comment suggested "They should have just stopped updating security patches for 2.7, that would have forced them to use 3.0". That's ok if Python holds a monopoly on development languages. But it doesn't. As soon as that happens developers will not just look at updating to Python 3, they'll start looking at other ecosystems: Go, Elixir, Java, Clojure, Scala, NodeJS etc.