Talk:Python (programming language)/Archive 8

Archive 5

Archive 6

Archive 7

Archive 8

Archive 9

Archive 10

Archive 11

Indentation not significant?

Does anyone besides Strombrg think that Python's use of indentation isn't worth mentioning in the lede? I strongly disagree as it's one of the more unique, useful, and visually striking attributes of the language. --Cybercobra (talk) 05:44, 15 March 2010 (UTC)

I don't consider it's use of indentation it's core/defining feature, so I also don't consider it significant enough for the lead. Section 4.1 on indentation could make more of a deal about it in comparison to other languages. peterl (talk) 05:55, 15 March 2010 (UTC)

strombrg here: When Python is attacked, it's normally attacked for its whitespace for blocks. It's now one of the things I now love most about the language, but when I first saw it, I thought it was awful - as is the reaction of many others new to the language. Putting something this likely to discourage newcomers in the first paragraph is just asking for trouble. Yes, unusual != bad, but it's important to keep in mind that this perspective is "truer" in technical discussion, not marketing discussion, because it's very common for humans to mistake the descriptive for the normative. —Preceding unsigned comment added by Strombrg (talk • contribs) 00:41, 19 March 2010 (UTC)

I believe, with Cybercobra, and with the longstanding consensus of article editors, that mentioning Python's indentation block style should be mentioned in the lead. This feature is one of the most widely mentioned aspects of Python in other introductions and discussion of Python. While I entirely agree that the significance of this feature is often exaggerated–both pro and con–the fact it is so frequently mentioned in 3rd party sources makes it notable to readers of the article. LotLE×talk 03:01, 19 March 2010 (UTC)

strombrg: No one's saying it shouldn't be mentioned, and in fact it is mentioned (again, as if once isn't enough) later in the article. I'm just saying it shouldn't be in a "you only get one first impression" spot where newcomers are more likely to be driven away by it - this in combination with the previous "claims to be powerful" stuff read like it was written by someone who hated the language. Please see http://www.se-radio.net/podcast/2006-02/episode-4-scripting-languages for a typical example of how whitespace for block scoping is misperceived outside the python community. —Preceding unsigned comment added by Strombrg (talk • contribs) 18:36, 20 March 2010 (UTC)

I think the indentation is important enough to be mentioned in the lead. Ulner (talk) 21:37, 20 March 2010 (UTC)

I think indentation makes Python "special". We could add the GvR's argument: you already use space/tabs to indent your code, why not make it part of the language (for better and for worst)? --0x6adb015 (talk) 12:28, 22 March 2010 (UTC)

No "mere" in scripting

I think the word mere should be dropped when describing scripting in the heading as it smacks of being (wrongly) ashamed of your parents. There is nothing "mere" about scripting. It is the essential glue that binds many a system and central to how things get done on Unix-like systems. Yes, it may be thought by some to be beneath other programming but Python practitioners should know better.

Drop the word! --Paddy (talk) 06:44, 15 March 2010 (UTC)

Example?

Wouldn't it be nice to have an example code? --JokerXtreme (talk) 11:35, 26 March 2010 (UTC)

Yes, it would. Perhaps a classic 'Hello world'. 95.147.185.198 (talk) 19:20, 25 April 2010 (UTC)

I've added a couple of simple examples that may help. We need to be careful not to add too much - Wikipedia is not a 'How To'. peterl (talk) 10:38, 27 April 2010 (UTC)

Improve "Mathematics' section

NumPy and SciPy definitely should be mentioned in the "Mathematics" section (currently they are mentioned in the "Usage" section only). For now there are some garbage about %, /, rounding floats to integers etc.

Dmitrey (talk) 12:37, 26 September 2010 (UTC)

Introduction

It could be valuable from a novice's point of view if the article mentioned that the software is free to download. 74.105.52.232 (talk) 18:56, 23 May 2010 (UTC)

Criticism

Can someone put a criticism section in? Python can't be totally free from criticism. --Volumebass112 (talk) 12:29, 23 June 2010 (UTC)

The problem with such things is sourcing. When they've been done in the past, even highly regarded industry commentators have been removed as "not being WP:RS", because they were recognised as experts in Java, Ruby or Perl etc., but not Python. To be WP:RS for Python itself, one's assumed to have to be drinking the Kool-aid and thus unlikely to criticise it.

As one of Python's obvious targets for criticism would be dynamic typing itself, we'd first need a WP article that managed to get an accurate description of dynamic typing in Python, something we don't currently have. Andy Dingley (talk) 12:36, 23 June 2010 (UTC)

To assume that one is favorably biased just because one has been using a product or service for a while is in and of itself being biased, and is no indicator whatsoever that the source is unreliable. There are products that I've been using for years with which I have a bone to pick, often because I've seen better 10 or 20 years ago and all that's available now is product "X" with numerous flaws and quirks. I'd be happy to accept a critical review any time under such conditions, especially if criticisms can be backed up with evidence. All too often I've found that experts on product "X" are blissfully unaware of shortcomings, because they've been working with "X" exclusively and don't realize that products "Y" and "Z" are much better. In such cases the criticisms of product "X" from experts on products "Y" and "Z" carry more weight.

The very style and content of the article as it now stands illustrates my point. I visited it several months ago to find out what Python was all about and came away just as ignorant as when I started. The article gets bogged down in minutiae from the very first sentence in the lead section and didn't even tell me if Python was run on an interpreter or a compiler. It just got worse from there. That's what you get when something is written by "recognised experts" or copied from the writings of such "experts". (I added a few words in the lead section earlier today in an attempt to alleviate this criticism.)—Quicksilver^{T @} 05:02, 29 October 2010 (UTC)

Python rounding

There have been a few attempts to insert phrasing such that Python follows "mathematical laws" or "floor and ceiling mathematical rules" on rounding. Most seem designed to simply promote Python over over languages, and suffer from some NPOV issues. However, there are substantially larger problems with these statements, which are:

• The claims are not properly cited.
• Rounding to the nearest smaller integer is floor semantics, not "floor and ceiling".
• Floor and ceiling are simply utility functions, not any sort of "mathematical law". Both of them -- along with the "nearest to zero" rounding performed by some other languages are equally valid mathematically (and to be strictly precise, none of them follow the generally accepted guidelines for mathematical rounding).

Fell Gleaming^talk 11:13, 25 August 2010 (UTC)

Hi Fell. I can see your point, but I'm not happy with the current wording. It seems to imply that Python is out of step with other languages (which it is), but that there is no basis with the decision for Python to floor towards -infinity (which is wrong - it's a clear decision, as your ref clearly shows).

I agree with your NPOV statement, but I feel the current wording is not neutral either. How can we word it so that Python is shown to be different but considered, and relevant based on mathematics?

PS: Excellent ref you found - well done.

peterl (talk) 10:13, 26 August 2010 (UTC)

Maybe could state something like 'The result of integer division rounds towards negative infinity. This maintains the validity of of the equation b*( a/b)+a%b = a for all integers a, and positive integers b, as Python's modulus operation a % b results in a number in the range [0,b), following usual mathematical convention.' That's pretty verbose, though, and may not draw enough attention to the a < 0 case. --Aflafla1 (talk) 18:46, 30 August 2010 (UTC)

Language or software?

I don't think this article (nor indeed Python's official website) makes enough of a distinction between a language called Python and piece of software called Python. Balfa (talk) 13:57, 6 October 2010 (UTC)

I think the problem lies with the Python user community in making too many assumptions about the non-cognoscenti's knowledge of the surreal British comedy series of the early 1970s, Monty Python's Flying Circus. Until I edited it today, this article didn't even tell me if Python was a compiled language or an interpreted language. In fact, Python is the language, and the interpreter on which it runs is called IDLE, probably a reference to Eric Idle, one of the members of the Monty Python comedy troupe. The whole subject matter is rife with "inside jokes" and is bogged down in too much detail for a Wikipedia article. It needs a thorough enema.—Quicksilver^{T @} 05:16, 29 October 2010 (UTC)

Oh no it doesn't --Paddy (talk) 10:49, 29 October 2010 (UTC)

Hydrargyrum, you talk as if there is a clear distinction between interpreted and compiled. Pascal was called compiled when, for some time, an important "compiler" generated p-code that was then "interpreted". --Paddy (talk) 10:55, 29 October 2010 (UTC)

Oh yes it does, and yes there is. Pretend you've never heard of Python before and then read the article. See if you can make heads or tails of it by the time you reach the bottom. For an even better experience, read the version before I added material about its interpreted nature. See the Python Tutorial at http://docs.python.org/release/2.5.2/tut/node1.html. The authors of that page clearly state just a few sentences into the "Front Matter" that Python has an "interpreted nature". Now, why do you suppose they did that?

I'm not stupid; I've been programming in a number of different high-level and low-level computer languages since about 1970 and I fully realize that many computer languages can be run either in an interpreted or compiled environment. A classic example is Dartmouth BASIC: It was always a compiler in the beginning while under development at Dartmouth College, and when resource-poor minicomputers and microcomputers appeared, interpreter versions were created, such as PC-BASIC for the IBM PC. Anyone familiar with both flavors knows there are differences in the way the source code can be written and big differences in performance. When I encounter a new computer language, that's one of the crucial things I want to know right from the start.

People come to Wikipedia to get a quick overview of a subject with which they're unfamiliar; they don't come here to get an advanced degree. The articles should be written accordingly. Anything more should be left to "Further Reading" and references in the footnotes, or Wikibooks and Wikiversity.—Quicksilver^{T @} 23:07, 29 October 2010 (UTC)

Python is a language; there's a specification. However, there is also one primary implementation, also known as CPython. CPython is a bytecode-compiled implementation of the Python language. The two are closely connected, more so than with some languages (like C, Forth, or Common Lisp) but less so than with some others (like Perl). Most of the time, when someone says "Python" they could reasonably mean both the language and the primary implementation. If someone wants to refer to Jython or IronPython or CL-Python, they say so. This is all discussed in the specification itself --FOo (talk) 15:36, 30 October 2010 (UTC)

IDLE is not an interpreter, it is an IDE. InverseHypercube (talk) 08:21, 2 March 2011 (UTC)

And there are interpreters for the 'compiled' languages, such as the 'classic' compiled language, C. See Ch_interpreter#C.2FC.2B.2B_interpreter. There's a difference between the language (which has nothing to do with how it is implemented), and the implemenation (which could be compiled, interpreted, p-code, byte code or one of many other options.) How do you classify a language running under a JIT?

peterl (talk) 08:17, 3 March 2011 (UTC)

Further reading

I think we need some guidance as to what should go into this section. It's starting to become a self-promtion list of any books about Python. What do others think? peterl (talk) 05:06, 30 November 2010 (UTC)

How about a 'List of books on Python (programming language)' page? —Preceding unsigned comment added by Paddy3118 (talk • contribs) 08:46, 30 November 2010 (UTC)

That would be unusual - I don't think any other programming language has that. I don't think there's anything that wikipedia would be adding that isn't already covered elsewhere, such as [1]. peterl (talk) 09:10, 30 November 2010 (UTC)

I've decided to prune severely under Wikipedia:Layout#Further_reading, particularly "any book included should have received more than one good review in RS - newspapers and scholarly journals being the norm, and the clear balance of RS reviews should be positive." peterl (talk) 05:10, 30 November 2010 (UTC)

Agreed, and thanks. We should not even list all books meeting the above criteria. Johnuniq (talk) 07:50, 30 November 2010 (UTC)

Least someone be left with an incorrect impression, there was no Wikipedia:COI#Self-promotion in my listing Hetland's Python book. Peterl, I am assuming you were concerned that the list was getting out of control and did not mean anything personal. --Javaweb (talk) 03:37, 6 December 2010 (UTC)Javaweb

I am new to the editing of articles, so maybe I am missing something obvious to others. The Further Reading section needed improvement, so I added a good book and organized the list. Later the book list was severely pruned, but I do not understand what qualifies the books that remain. Selected Wikipedia policy was duplicated in the Further Reading section, but it appears that the policy is not being followed. Could someone please explain what I am missing here, or show an example of the proper way to add a book. Thanks. Lance Albin (talk) 03:37, 2 January 2011 (UTC)

Performance

I've heard that Python's performance (unless optimized) isn't that great, however, I know nothing about this myself. Can someone put something about Python's performance compared to other high-level languages? —Preceding unsigned comment added by 212.68.15.66 (talk) 06:25, 3 February 2011 (UTC)

Not really. It depends on too much. You might want to check this page for why plus have some fun. P.S. I am assuming that you meant run-time performance. Ease/accuracy of development is another measure. --Paddy (talk) 12:22, 3 February 2011 (UTC)

I totally agree with what you say about development/runtime performance (especially when it comes to Java, people are way too aggressive about its runtime performance in my opinion). Thanks for the link! 212.68.15.66 (talk) 06:58, 4 February 2011 (UTC)

Python and Java *can* perform well. But they rarely *do* in my experience. That's because high-level programming languages tend to discourage the kind of optimizations that are second-nature when doing lower-level programming. An exception to this is PERL, because it allows (and frequently uses) inline and XS compiled code.... where high level and low-level code can be freely mixed. —Preceding unsigned comment added by 66.194.114.178 (talk) 16:07, 18 May 2011 (UTC)

Python 2 vs Python 3

Are we assuming that nobody cares about Python 2 now? The table of built-in types was "simplified" by having it refer only to Python 3. I'm OK with that, as long as we also add a section describing the differences between 2 and 3. (I'm not sufficiently fluent in both to create such a section myself.) Joule36e5 (talk) 02:18, 10 June 2011 (UTC)

Python 2 will be in use for probably another 5-9 years. Most production is running under Python 2 now since programs need their library packages converted to Python 3 before they move over. --Javaweb (talk) 03:04, 10 June 2011 (UTC)Javaweb