Re: [Python-ideas] Consistency in naming [was Re: ...ALL CAPS] (off-list)
James, frankly, it sounds to me like you have found things you don't like about Python, and are frustrated that your ideas here have not been celebrated. That's far from "decline." The process for changing Python is fundamentally conservative, which can be frustrating. I understand that. I myself have experienced that frustration. But claiming that Python is in decline, or that something "is the only real chance of having a successful Python language" is just hysteria that won't win over anyone. Now you say Python isn't growing? Do you mean in features, or usage? Either is obviously false. Perhaps you mean that it isn't growing the way that you want? There are things I would change about Python if I could, but I am not the BDFL, and neither are you. Python is imperfect, because it is made by people, over more than 25 years, and it's got to balance competing demands. It's clear that you are smart and have energy to dedicate to the future of Python. I hope you find a productive way to contribute. (Sorry for top-posting...) --Ned. On 2/3/19 12:34 PM, James Lu wrote:
James, Ned, and everyone else, (I'm intentionally leaving the subject line untouched, but since James' message was sent to the list, I'm replying to the list.) In another thread, you (James) asked for ways to not feel like such a n00b, and you attempted (with mixed success) to start a discussion about improving the quality of communication here. So in the spirit of constructive criticism so that you and others can learn from experience and communicate better: (1) Taking the group discussion off-list should be done rarely, and usually only for personal messages that shouldn't be shared publically, or by mutual agreement for off-topic discussions. I can't see Ned's original comment in either my inbox, or at the archives, so it seems that he took the discussion off-list. I don't know why. It doesn't strike me as either personal or off-topic, so taking it off-list seems to be both unnecessary and a little rude to the rest of the group -- why were the rest of us excluded? But since Ned apparently intended to take it off-list, it is only polite to respect that your reply. (2) The long-standing tradition is to put "OFFLIST" at the *start* of the subject line, not the end where it is easy to overlook. To make it even more clear, we should explicitly state that the message is off-list at the top of the message. (Especially if you intend the message to be confidential.) (3) In the thread about improving communication, I mentioned that the easier it is to make comments, the more likely it is for people to make poor-quality comments. In my experience, posting from a phone is one of those tools that encourages poor-quality comments. It is each person's responsibility to either use a better communication tool (better does not necessarily mean more convenient) or to manage their own use of the tool better. I've seen too many people blame their tools for their own repeated mistakes: "I can't help it, it's my phone, it makes it too hard to do the right thing." Who is the master, them or their phone? We're all only human and therefore make mistakes, but we ought to own them when we do, and learn from them, not fob responsibility off to the tools we choose to use. (4) Speaking of poor-quality comments, you said "Python’s decline is in not growing." Do you have evidence that Python is "not growing" or did you just make that up? On Sun, Feb 03, 2019 at 12:34:02PM -0500, James Lu wrote:
Python’s decline is in not growing.
Sent from my iPhone
-- Steven
On Mon, Feb 4, 2019 at 10:02 AM Steven D'Aprano <steve@pearwood.info> wrote:
Let's see. * You don't usually get much, if any, choice of operating system * Even if you can get a terminal app, it doesn't get true access, especially not root * "Jailbreaking" your phone - that is, getting actual access to it - generally voids your warranty You don't own your phone. It owns you. I don't carry a phone. Terrible UI, locked-down device, the ONLY thing it has going for it is that it fits in your hand/pocket. I'd much rather carry a laptop (running Debian GNU/Linux). True, my mobile connection still comes from a device I don't have control of, but it's JUST a network connection, using standard 802.11 to communicate with the laptop. ChrisA
On 2/3/19 6:01 PM, Steven D'Aprano wrote:
I replied to James privately because I felt that the question of "Python's decline," and James' reasons for believing it to be so, were not on-topic for a list about suggestions for improving Python. It also seemed to me that it could easily devolve into an unproductive discussion for such a large group. But, it was probably also of interest to the group, many of whom were probably wondering the same things I was, so I can see how it could have stayed in the list (where it now is.) It also seemed to me to be a topic which could easily result in James feeling at a disadvantage, being on the other side of a subjective debate from the bulk of the group. I had hoped to discuss it with him in a setting that was less likely to get heated. I didn't mean to be rude to anyone. --Ned.
This is getting pretty off toipic, but I want to point out hat as a rule, Python has been critisised more for having too MUCH cnage than too little over the last few years: The py2 -> 3 transition Adding "yet another" formatting option (f-strings) Adding := The async stuff Granted, these are bigger deals than changing the spelling of stdlib class, but the point is that people want consistency as much (or more) as new features. And changing something like the spelling of datetime buys us exactly nothing in terms of expressiveness, etc. Finally, what the heck is wrong with datetime.now ? making it a classmethod keeps everything in one namespace, so you can reasonably do: from datetime import datetime, timedelta and be done with it. -CHB On Sun, Feb 3, 2019 at 9:34 AM James Lu <jamtlu@gmail.com> wrote:
-- Christopher Barker, PhD Python Language Consulting - Teaching - Scientific Software Development - Desktop GUI and Web Development - wxPython, numpy, scipy, Cython
On Mon, 4 Feb 2019 at 05:36, Christopher Barker <pythonchb@gmail.com> wrote:
This is getting pretty off toipic, but I want to point out hat as a rule, Python has been critisised more for having too MUCH cnage than too little over the last few years:
[...]
Finally, what the heck is wrong with datetime.now ?
Surely what's wrong is the fact that it adds another item to the list of changes that people will complain is "too much"? The conservatism in Python's development is not a result of the core devs being change averse, or because of a policy of blocking change. Rather it's because *the user community* reacts so strongly when we change things, so that we feel that stability is what our users want, and try to achieve that while still balancing the need for the language to grow and develop. In a list like this, which represents only a tiny fraction of Python's user community, it's easy to misjudge the appetite Python's users have for change, because this is a self-selecting group of people who are interested in change. That's why the pushback on *any* proposal made here is to ask what benefits it gives, and why in the absence of good arguments for a change, "the status quo wins". Paul PS Personally, I'd love to add a lot of the ideas discussed in this list (although I'm "meh" about datetime.now, it's at best a minor annoyance). But with my "python developer" hat on I have to look beyond that and try to help proposers develop a good case for their suggestions (if indeed they do have sufficient merit). Just saying "yay, that sounds cool" doesn't do that, even if it's my personal instinct.
James, frankly, it sounds to me like you have found things you don't like about Python, and are frustrated that your ideas here have not been celebrated. That's far from "decline." The process for changing Python is fundamentally conservative, which can be frustrating. I understand that. I myself have experienced that frustration. But claiming that Python is in decline, or that something "is the only real chance of having a successful Python language" is just hysteria that won't win over anyone. Now you say Python isn't growing? Do you mean in features, or usage? Either is obviously false. Perhaps you mean that it isn't growing the way that you want? There are things I would change about Python if I could, but I am not the BDFL, and neither are you. Python is imperfect, because it is made by people, over more than 25 years, and it's got to balance competing demands. It's clear that you are smart and have energy to dedicate to the future of Python. I hope you find a productive way to contribute. (Sorry for top-posting...) --Ned. On 2/3/19 12:34 PM, James Lu wrote:
James, Ned, and everyone else, (I'm intentionally leaving the subject line untouched, but since James' message was sent to the list, I'm replying to the list.) In another thread, you (James) asked for ways to not feel like such a n00b, and you attempted (with mixed success) to start a discussion about improving the quality of communication here. So in the spirit of constructive criticism so that you and others can learn from experience and communicate better: (1) Taking the group discussion off-list should be done rarely, and usually only for personal messages that shouldn't be shared publically, or by mutual agreement for off-topic discussions. I can't see Ned's original comment in either my inbox, or at the archives, so it seems that he took the discussion off-list. I don't know why. It doesn't strike me as either personal or off-topic, so taking it off-list seems to be both unnecessary and a little rude to the rest of the group -- why were the rest of us excluded? But since Ned apparently intended to take it off-list, it is only polite to respect that your reply. (2) The long-standing tradition is to put "OFFLIST" at the *start* of the subject line, not the end where it is easy to overlook. To make it even more clear, we should explicitly state that the message is off-list at the top of the message. (Especially if you intend the message to be confidential.) (3) In the thread about improving communication, I mentioned that the easier it is to make comments, the more likely it is for people to make poor-quality comments. In my experience, posting from a phone is one of those tools that encourages poor-quality comments. It is each person's responsibility to either use a better communication tool (better does not necessarily mean more convenient) or to manage their own use of the tool better. I've seen too many people blame their tools for their own repeated mistakes: "I can't help it, it's my phone, it makes it too hard to do the right thing." Who is the master, them or their phone? We're all only human and therefore make mistakes, but we ought to own them when we do, and learn from them, not fob responsibility off to the tools we choose to use. (4) Speaking of poor-quality comments, you said "Python’s decline is in not growing." Do you have evidence that Python is "not growing" or did you just make that up? On Sun, Feb 03, 2019 at 12:34:02PM -0500, James Lu wrote:
Python’s decline is in not growing.
Sent from my iPhone
-- Steven
On Mon, Feb 4, 2019 at 10:02 AM Steven D'Aprano <steve@pearwood.info> wrote:
Let's see. * You don't usually get much, if any, choice of operating system * Even if you can get a terminal app, it doesn't get true access, especially not root * "Jailbreaking" your phone - that is, getting actual access to it - generally voids your warranty You don't own your phone. It owns you. I don't carry a phone. Terrible UI, locked-down device, the ONLY thing it has going for it is that it fits in your hand/pocket. I'd much rather carry a laptop (running Debian GNU/Linux). True, my mobile connection still comes from a device I don't have control of, but it's JUST a network connection, using standard 802.11 to communicate with the laptop. ChrisA
On 2/3/19 6:01 PM, Steven D'Aprano wrote:
I replied to James privately because I felt that the question of "Python's decline," and James' reasons for believing it to be so, were not on-topic for a list about suggestions for improving Python. It also seemed to me that it could easily devolve into an unproductive discussion for such a large group. But, it was probably also of interest to the group, many of whom were probably wondering the same things I was, so I can see how it could have stayed in the list (where it now is.) It also seemed to me to be a topic which could easily result in James feeling at a disadvantage, being on the other side of a subjective debate from the bulk of the group. I had hoped to discuss it with him in a setting that was less likely to get heated. I didn't mean to be rude to anyone. --Ned.
This is getting pretty off toipic, but I want to point out hat as a rule, Python has been critisised more for having too MUCH cnage than too little over the last few years: The py2 -> 3 transition Adding "yet another" formatting option (f-strings) Adding := The async stuff Granted, these are bigger deals than changing the spelling of stdlib class, but the point is that people want consistency as much (or more) as new features. And changing something like the spelling of datetime buys us exactly nothing in terms of expressiveness, etc. Finally, what the heck is wrong with datetime.now ? making it a classmethod keeps everything in one namespace, so you can reasonably do: from datetime import datetime, timedelta and be done with it. -CHB On Sun, Feb 3, 2019 at 9:34 AM James Lu <jamtlu@gmail.com> wrote:
-- Christopher Barker, PhD Python Language Consulting - Teaching - Scientific Software Development - Desktop GUI and Web Development - wxPython, numpy, scipy, Cython
On Mon, 4 Feb 2019 at 05:36, Christopher Barker <pythonchb@gmail.com> wrote:
This is getting pretty off toipic, but I want to point out hat as a rule, Python has been critisised more for having too MUCH cnage than too little over the last few years:
[...]
Finally, what the heck is wrong with datetime.now ?
Surely what's wrong is the fact that it adds another item to the list of changes that people will complain is "too much"? The conservatism in Python's development is not a result of the core devs being change averse, or because of a policy of blocking change. Rather it's because *the user community* reacts so strongly when we change things, so that we feel that stability is what our users want, and try to achieve that while still balancing the need for the language to grow and develop. In a list like this, which represents only a tiny fraction of Python's user community, it's easy to misjudge the appetite Python's users have for change, because this is a self-selecting group of people who are interested in change. That's why the pushback on *any* proposal made here is to ask what benefits it gives, and why in the absence of good arguments for a change, "the status quo wins". Paul PS Personally, I'd love to add a lot of the ideas discussed in this list (although I'm "meh" about datetime.now, it's at best a minor annoyance). But with my "python developer" hat on I have to look beyond that and try to help proposers develop a good case for their suggestions (if indeed they do have sufficient merit). Just saying "yay, that sounds cool" doesn't do that, even if it's my personal instinct.
participants (7)
-
Carl Smith -
Chris Angelico -
Christopher Barker -
James Lu -
Ned Batchelder -
Paul Moore -
Steven D'Aprano