Python Creator Guido van Rossum Retires, Heads To Python Conference (zdnet.com) 41
"Guido van Rossum, the creator of the hugely popular Python programming language, is leaving cloud file storage firm Dropbox and heading into retirement," reports ZDNet:
That ends his six and half years with the company, which hired in him in 2013 because so much of its functionality was built on Python. And, after last year stepping down from his leadership role over Python decision making, that means the Python creator is officially retiring....
According to Dropbox, in 2011, when van Rossum first met Dropbox CEO Drew Houston, the Dropbox server and desktop client were written "almost exclusively in Python". Today, Dropbox also relies on Go, TypeScript, and Rust, as well as the open source Mypy static type checker that Dropbox develops to manage Python code at scale. Mypy helps developers overcome the challenge of understanding dynamically typed Python code written by other developers in the past...
Dropbox said van Rossum has had a major impact on its engineering culture. "There was a small number of really smart, really young coders who produced a lot of very clever code that only they could understand," said van Rossum. "That is probably the right attitude to have when you're a really small startup." However, as Dropbox notes, when the company grew, new engineers could not understand the clever but 'short and cryptic' code written by and for earlier developers. Van Rossum called this "cowboy coding culture" and educated the company about the value of maintainable code. "When asked, I would give people my opinion that maintainable code is more important than clever code," he said.... Dropbox also credits van Rossum with sharpening the company's testing processes for its continuous integration program and helping engineers understand why tests were broken.
"Thank you, Guido" is the title of the post on Dropbox's blog announcing the news that van Rossum is now retiring. Sharing that article on Twitter Thursday, van Rossum added "It's bittersweet... I've learned a lot during my time as an engineer here -- e.g. type annotations came from this experience -- and I'll miss working here."
But by Friday he was heading off to the North Bay Python conference in Petaluma, California.
According to Dropbox, in 2011, when van Rossum first met Dropbox CEO Drew Houston, the Dropbox server and desktop client were written "almost exclusively in Python". Today, Dropbox also relies on Go, TypeScript, and Rust, as well as the open source Mypy static type checker that Dropbox develops to manage Python code at scale. Mypy helps developers overcome the challenge of understanding dynamically typed Python code written by other developers in the past...
Dropbox said van Rossum has had a major impact on its engineering culture. "There was a small number of really smart, really young coders who produced a lot of very clever code that only they could understand," said van Rossum. "That is probably the right attitude to have when you're a really small startup." However, as Dropbox notes, when the company grew, new engineers could not understand the clever but 'short and cryptic' code written by and for earlier developers. Van Rossum called this "cowboy coding culture" and educated the company about the value of maintainable code. "When asked, I would give people my opinion that maintainable code is more important than clever code," he said.... Dropbox also credits van Rossum with sharpening the company's testing processes for its continuous integration program and helping engineers understand why tests were broken.
"Thank you, Guido" is the title of the post on Dropbox's blog announcing the news that van Rossum is now retiring. Sharing that article on Twitter Thursday, van Rossum added "It's bittersweet... I've learned a lot during my time as an engineer here -- e.g. type annotations came from this experience -- and I'll miss working here."
But by Friday he was heading off to the North Bay Python conference in Petaluma, California.
See, Python is no better than Perl (Score:1, Insightful)
You can write unmaintainable code in any language and your newfangled language isn't going to change that. And in fact with awful languages like Python you just end up in an even worse position.
Re: (Score:2)
Perl is much better than Python for some things. Python's OOP model is better than Perl's. But it's not as good as Ruby's and Ruby's isn't as good as Smalltalk's (but, then, what is?).
Python is one of those languages (like R) that has no good reason to exist but has found a niche among the inexperienced and plain ignorant.
Translation (Score:5, Insightful)
Mypy helps developers overcome the challenge of understanding dynamically typed Python code written by other developers in the past...
In clear English: Mypy helps competent developers root out obscure bugs from large, unreadable, uncommented piles of poor Python code written by young interns years ago, who couldn't manage their variables' types if their lives depended on it, because the compiler/interpreter doesn't do it for 'em.
Re:Translation (Score:5, Insightful)
That's still preferable to having those interns use a language that would expose them (and their users) to memory safety bugs.
Re: Translation (Score:2)
Re: (Score:2)
There are REASONS why ALGOL has strong static typing.
There are REASONS why PASCAL has strong static typing.
There are REASONS why Ada has strong static typing.
There are REASONS why strong static typing was Stroustrup's first goal for C++. (according to his own writings). OOP was goal #2.
Those ivory tower computer science guys just might actually know what they're talking about.
Re: (Score:2)
So many people see static typing as training wheels, glad you see the light too.
(also like to add Swift with strong static typing and Typescript with weak static typing to your list)
Advantages of static typing:
* Code is more self documenting
* IDEs have a much better time of autocompleting
* Refactoring is so much more robust
* A whole series of bugs that can now be caught by the compiler instead of runtime
* Code typically runs a lot faster depending on compilation environment
With these 2 things, Python would be great... (Score:3, Interesting)
(1) A quick way of getting it installed, just like MS Office or a number of Windows programs...
(2) An official GUI for quickly coding front-end user applications...MS Access & Excel have this capability.
Now do not tell me about all these GUI frameworks that exist. None come close to the drag & drop that those Office applications offer.
For now, the full potential of Python remains unexploited here, sadly.
Re:With these 2 things, Python would be great... (Score:4, Insightful)
If you think Python is hard to install, then you are no person to either criticize or praise any other programming tool.
OTOH, I do wish that the graphics tools were better. tkinter was "almost good enough", but has a problem that it's difficult to track the current text position between mouse clicks. Qt may solve that problem, but is more difficult to handle radio buttons in. Not horrendously more difficult, but I think they want you to use Designer rather than code it yourself, and it just won't handle all the features I need. (Today I'll check whether I can select and copy the "text file line" == "block" that currently has the selection. If I can't I'll need to check out wx somehow, but the documentation isn't great.) (FWIW, the Gtk documentation is unusably bad. And Glade has gotten worse over the years.)
Re: (Score:2)
Enaml [enthought.com] looks interesting.
Re: (Score:2)
That looks to still be a Python2 library. I didn't check much further. When I did check I couldn't even find a textbox or edittext or some such widget.
They've got all number of fancy widgets listed, but they don't seem to have the basics covered.
Re: (Score:1)
Re: (Score:3)
(1) A quick way of getting it installed,
You mean like the silent default installation of it in many OS distributions today?
Re: (Score:2)
I meant getting installed with the respective & working RAD/GUI components...Sorry, my rant wasn't that clear.
Re: (Score:2)
Why not Smalltalk? (Score:3)
If Smalltalk has such good attributes, shouldn't we just be using Smalltalk?
Or is it not being maintained, or there are no up-to-date distributions, or do the distributions require serious coin?
I can understand the hate on Ada because it takes Bondage and Discipline language concepts to an extreme degree, but you can get Ada if you want to use it, either free or paid-for versions.
Both Smalltalk and Ada were introduced at a time when you needed a workstation costing as much as a luxury automobile or m
Re: Why not Smalltalk? (Score:2)
Re:With these 2 things, Python would be great... (Score:4, Informative)
Python is very quick to install, unless you mean it should be installable via the app store.
Python and ActivePython have msi installers.
Anaconda is a feature packed fat distribution.
It does not do well on RAD GUIs though. I feel more productive in Lazarus, based on 90s Delphi, than I do with Python, to put together a decent GUI that talks to a database. At least Python embeds in FreePascal through some nice components, but hardly anyone knows that and uses it that way. Python RAD attempts, like Boa Constructor, never took off. I just wish CPython (not Jython and IronPython) integration with Java and .NET, both of which exist, got first-class attention. That is better than building RAD tooling from the scratch for Python.
Re: (Score:3)
Now do not tell me about all these GUI frameworks that exist. None come close to the drag & drop that those Office applications offer.
For now, the full potential of Python remains unexploited here, sadly.
Well, Python is a language that parses tokens from a flat text file.
I think you're looking for a language which uses buttons, emoticons and animated GIF stickers as its fundamental lexical elements.
Re: (Score:2)
1. Download WinPython: https://winpython.github.io/ [github.io]
2. Extract.
3. Launch Jupyter Notebooks.
What more do you want?
Re: With these 2 things, Python would be great... (Score:2)
Thank God (and Guido, mainly Guido unintentionally I guess) that Python is hard to install. Anything that requires people to have a double digit IQ before they write computer code is good.
Ease of Pythoning (Score:2)
(1) A quick way of getting it installed,
Waht do you mean "quick" ? It comes pre-installed on your distribution!
How coud it more qui....
just like MS Office or a number of Windows programs...
Ah. I see where you're coming from ! :-D
None come close to the drag & drop that those Office applications offer.
I come from a different language background (mostly C/C++ with Perl and Bash for the duck-taping) so I can't speak with extensive python experience.
But to me it looks like Qt at least it's drag and drop UI creation (Qt Designer for Qt Widgets) and even more so the modern Qt Quick and QML makes it easy to make drap-and-drop complex UIs, and Qt has python binding.
So for making
Good. I hear the Python death knell.. (Score:1, Troll)
In my experience, Python has allowed totally incompetent developers to produce barely functional shit more rapidly than any language since Visual Basic.
The abuses allowed, even encouraged, by Python have been a constant reminder of why strongly typed, fully compiled languages should be mandatory for anything other than simple scripting.
And don't get me started on the dumpster fire that is Python version incompatibility.
Re: (Score:3)
Yep, we have one dev in our team (or actually not in our team, as all teams write Java, and he doesn't "like" Java).
Got highly offended when we rewrote his Python turd into Java when it became clear this shit had to go into production use and would need to be supported for the next few years.
Re: (Score:3)
> Python is not the be all, end all and C/C++/Java won't go away, just because you wish for that
Indeed. From the summary:
-
According to Dropbox, in 2011, when van Rossum first met Dropbox CEO Drew Houston, the Dropbox server and desktop client were written "almost exclusively in Python". Today, Dropbox also relies on Go, TypeScript, and Rust
-
The company that had the creator of Python on staff moved away from using Python for everything, instead using languages appropriate for the project.
Python is a grea
Re: (Score:3)
C++ code is so hard to maintain and Python isn't
I wouldn't phrase it that way but there is an element of truth in there.
So they came to the conclusion that everything must be rewritten into Python
I agree: those consultants were being childish and unprofessional.
So grow up and become a real programmer, you know, one that can produce code in whatever language is best for the product to be delivered.
Sometimes the best language really is Python though. I wrote and maintain two services that are important at th
Hallway conversation on Machine Learning (Score:2)
Overheard one Electrical Engineering professor telling another "I need to learn me some Python!"
Stepping into that discussion, it appears that Machine Learning or whatever-the-heck-you-call-it is the hot thing in EE and ECE departments nationwide. The concern about Python was "how do we teach it so our students can get machine learning jobs in industry" (Comp Sci is already teaching Python in some sections of the Intro-to-Programming course that EE requires.)
When I described intrinsic shortcomings of
Re: (Score:2)
Python is almost always being used as glue logic for C or C++ code in those libraries. Sometimes FORTRAN, sometimes assembly, sometimes JIT compilation.
Re: (Score:3)
Python gets used for the configuration and specification of the computational structures and graphs as well as their many of diverse ancillary parameters. Things that are computed once upon problem setup, or once per some large outer loop (like change a learning rate per epoch) that have negligible computational impact but need to be transparent and easy to change by the scientist.
Various algorithms and variants newly invented have varying new definitions of free parameters, less convenient to represent in
Re: (Score:3)
What does this say about the seriousness of Machine Learning as a research area?
That the intelligence is going into Machine Learning research and not into learning more complex programming languages. That the time of researchers is being spent in acquiring and understanding data sets instead of programming. That the network effect of the first mover using python has led everybody else to follow suit.
Machine learning researcher: I've spent four weeks acquiring data and devising the algorithms needed. Do I spent a day writing this shit in python and run it in three days, or spend two day
Re: (Score:1)
Not a troll, the cold hard truth.
Gee, 95% of the time is spent ... (Score:3)
... READING code while only ~5% is spent writing it.
Who knew that write-only code wasn't meant for the long term! /s
Re: (Score:2)
Possibly the United States Department of Defense, when they started the Ada project?
Possibly Jean Ichbiah, who lead the team that designed the language that became Ada?
Almost that exact statement appears in the document that describes their rationale for why they did things the way they did.
FYI he's 63 years old (Score:4, Informative)
There's fear and trembling in Petaluma tonight (Score:1)
Never been to a Python convention. Is that some kind of wrist wrestling competition ?
Dynamically-typed languages are best for scripting (Score:1)