« first day (2128 days earlier)      last day (1573 days later) » 

1:12 AM
any Python guys in the building?
2:04 AM
Me in a bit
2:31 AM
@daOnlyBG did you need something?
3:05 AM
Well if you did, you missed the Python window. I'm tired and going to bed now.
5 hours later…
7:53 AM
Q: Call for volunteers: close voters

Robert HarveyProgrammers could benefit from a few more regular visitors who close questions on the front page and the Unanswered page that are clearly off-topic. At the moment, many such questions get three or four close votes, and then languish on the Unanswered page, sometimes for several days. Ideally, t...

8:31 AM
^^^ star that if you support the call
8:51 AM
It's more of a question for programmers.stackexchange.combaudsp 15 secs ago
The help on programmers.se explicitly states that career advice is off topic there as well. — Gerald Schneider 49 secs ago
9:23 AM
^^^ looking for dupe votes at meta to close old discussion in favor of newer one. @RobertHarvey - since you participated in both, please check if it makes sense
4 hours later…
1:54 PM
Eh, they seem like two different discussions.
2:37 PM
@gnat What's the rep minimum?
Hey, I'm giving a talk framed as an architecture talk about Python, any things you guys want to know about Python (so I can include the material in my talk)? (seriously! :) )
ok, not there yet
It should be lower on the smaller sites.
Probably 2000.
3000 on Programmers
Gimme Q's to A!
I mean, it should be lower on the smaller sites.
2:43 PM
12 mins ago, by Aaron Hall
Hey, I'm giving a talk framed as an architecture talk about Python, any things you guys want to know about Python (so I can include the material in my talk)? (seriously! :) )
Or maybe anything you already know but think I should include?
I would very much like to know some of the principles behind the "pythonic" philosophy. For example, who decided it was a good idea to use exceptions for flow control and why?
AFAIK python is the only language that takes that position.
Also, how do pythonistas work around the fact that python is such a poorly-performing language?
Thanks, I've gotta be ready for these kinds of comments.
Oh, and the double-underscores. For the love of God.
Did you find that delete link yet? :) — Robert Harvey 10 secs ago
@RobertHarvey You know, I don't know that it really does, I think where you see Python programmers using exceptions to avoid race-conditions is where non-Python programmers would do the same.
But I think it makes sense, checking for something to be there and then doing the action that fails if it isn't isn't atomic in any language, but in Python, it's much less atomic.
2:59 PM
@AaronHall No, exceptions for flow control is common in Python.
Well yes, and, I'm just thinking from my experience with a code-base with over 5000 contributors - I've seen copious exceptions where we've decided it's wrong. But where we have decided it's right, I mean, it's not as common as condition checking. Oops, gotta get on a call.
@AaronHall rep minimum is 125, don't believe what @RobertHarvey says. :) Downvotes work too, they 1) help indicate close-worthy questions, 2) prevent sympathy upvotes from hiding these and 3) increase chance of question becoming eligible for insta-deletes by 20K users. OTOH one can say that rep minimum is indeed 15. Upvotes to unfairly downvoted questions pull them out of the attention of close voters, also a useful thing
@gnat Are you referring to flagging? I think you still need 3000 rep for that to register as a close vote.
3:19 PM
@RobertHarvey no I refer to regular voting. Bad question voted down from 0 to -1 helps 3Kers quicker find it. Good question voted up from -1 to 0 helps 3Kers avoid wasting time on a false alert. Flagging to close is available since 15 rep, except for dupe which takes 50 (to generate comment on behalf of user) - but in our case flagging is useless because nobody looks into CV review queue
3:36 PM
I think maybe Python uses exceptions as much as other languages, it's just that Python is at such a higher level, exceptions stick out a lot more. Well this is something for me to look into more.
Maybe the relative cost isn't a big, since Python stack frames are probably more expensive than functions optimized by a compiler into machine code.
I probably need to have more conversations about this with you guys so I don't sound like a complete idiot.
4:07 PM
I guess I need to talk about the Global Interpreter Lock
For a non-Python book, Clean Code sure does have a lot of exception handling.
4:22 PM
@AaronHall I've never actually read Clean Code.
4:41 PM
@AaronHall I don't personally know python but it might be a good idea to cover exception handling best practices, I've been digging into that subject for a while and it seems like for both C# and C++ the conclusion is "Good exception handling is hard"
If exception swallowing is allowed in python it might be worth bringing up cases where that doesn't work, and how you can design without doing that
don't write functions that can "fail"
5:03 PM
From now on, every question that asks for the "best" way without explaining what "best" means in objectively satisfiable criteria gets a "Primarily Opinion-Based" close vote from me."
Same goes for "most common" or "most popular."
@ratchetfreak "File not found."
5:30 PM
@jrh There's no way to guarantee that you won't have a race condition. You can check for the presence or absence of a file, and some milliseconds later someone else could have added, deleted or renamed said file, and you'll still throw an exception when you try to perform your file operation.
@RobertHarvey oh, that wasn't a reply to your post. You're right that file not found is unavoidable
I just posted that case here because it's a rather well received answer with exceptions more or less being used as flow control.
"Computing effective permissions is not for the faint hearted and your code to attempt creating a file and catching the security exception thrown is probably the path of least resistance. "
the comments on this answer stackoverflow.com/a/1281638/4975230 show what additional problems come up if you try to use the ACLs
6:03 PM
According to Clean Code, C++ doesn't have exception handling in spite of efforts to insert it, but C# does.
6:30 PM
Single Responsibility Principle - "an object should only have one reason to change" is this about mutability of the object?
7:00 PM
@AaronHall C++ does have exceptions cplusplus.com/doc/tutorial/exceptions
C does not have exceptions, though
IIRC (I don't use C++ very often) C#'s exceptions have more/different features than C++'s exceptions, but C++ still has try/catch so there's some similarities
also Aaron take a look at this for SRP: en.wikipedia.org/wiki/Single_responsibility_principle
"The single responsibility principle states that every module or class should have responsibility over a single part of the functionality provided by the software, and that responsibility should be entirely encapsulated by the class. All its services should be narrowly aligned with that responsibility."
if you decide to talk about SRP I'd recommend finding some practical examples of where this has eased the development process
e.g., if you find a repository that used SRP extensively (to good effect) that would be a great thing to talk about; try to find a blog of one of the devs to see if they talked about it, etc
if you find something like that please let me know, by the way. I've seen a few okayish C# references that looked good on the surface but they were actually anemic domains
7:16 PM
I'm sorry, the distinction was "Checked Exceptions"
C# doesn't have them, C++ doesn't "despite valiant attempts"
ah, okay
yeah, C++ has checked exceptions but it doesn't seem like they're as strictly enforced as they are in Java
> Checked exceptions can sometimes be useful if you are writing a critical library: You must catch them. But in general application development the dependency costs outweigh the benefits.
@Ana: This post applies only to questions that are clearly and unambiguously off-topic, always have been, and still will be after the scope change (if it happens at all), such as coding help questions and product recommendations. I'm sorry if I didn't make that clear. Is it your assertion that we suspend all close voting activity until SE gets around to evaluating the scope change? How is this any different from the various close cabals that already pepper the Stack Exchange network? — Robert Harvey 2 mins ago
7:37 PM
1 hour ago, by Aaron Hall
Single Responsibility Principle - "an object should only have one reason to change" is this about mutability of the object?
If I recall, it's not, it's about having to change it based on stakeholders changing requirements.
Dependency inversion is about depending on an abstraction/interface instead of the idiosyncracies of specific types.
inversion is the wrong word.
"abstraction?" The Dependency Abstraction Principle.
No, it's not about mutability. Sheesh. If I had a nickel for every time someone got confused over SOLID, I wouldn't have to work anymore.
7:58 PM
Robert Harvey Needs a Posse
You got that right.
Sometimes I get the feeling that all of this talk about site scope is for naught, because our illustrious leader thinks that "how do I move the turtle in Logo" is a valid question.
it is a valid question, for Stack Overflow and also for 2009.
Deputize me.
Progse in 2016 is a different story
Also, I'm really worried about this name change. It's gotta have a handy 6-letter abbreviation.
That's what I said.
8:03 PM
Why is that green car parked outside of my house?
Is it still there?
Oh, wait. Not six letters.
almost looks like Sofa Engineering
  puts 'What is your name?'
  name = gets
  puts 'What is your quest?'
  quest = gets
Rewrite in Python ^^^^^
8:07 PM
Has a nice ring to it. So fing.
python "What is your name?"
name = python
python "What is your quest?"
quest = python
8:33 PM
So, I got a question about the name-change wiki y'all have been working on
The answer is, Yes.
feels... Not that much different from the current definition
Which is to say, the actual definition of what is on topic still kinda hints that programming questions are OK
...but then, of course, there's like a page of text explaining how they're not.
So my question is, where's the line being drawn here?
Presumably most of this is aimed at folks coming from Stack Overflow; at least, that's how it reads to me
So assuming I'm someone who's here because at least part of what I'm doing is writing code, where should I be seeing the split between SO and SOFENG?
It's not clear to me where the point of confusion lies, can you quote the specific parts you're referring to?
ok... Well, let's start with the proposed On Topic page intro:
> Software Engineering is a Q&A site for practitioners, researchers and students of software engineering and related fields. Questions asked here are ideally those which can be answered as if one was standing in front of a whiteboard with no access to a computer, reference books, or documentation.
"software engineering and related fields" - yeah, ok, I the SO user definitely fall in there somewhere. Good, another site for me!
so "and related fields" is too spongey?
8:41 PM
"questions can be answered standing in front of a whiteboard with no references" - um, ok. So, no questions about API declarations or studies need apply.
> If you have a question about...

- software development methods and practices
can be answered without != won't be answered with
yeah, that's definitely me still.
Ok, so this is good. My question about why my self-devised encryption method is throwing a NPE is still on topic. Hooray!
...then there's a big-ass interstitial that tells me it isn't.
So the short intro is too vague?
@AaronHall what does Software Engineering mean to you?
same question for @Robert & @Rachel, since y'all wrote a lot of this ^
Software engineering is the application of engineering to the design, development, implementation and maintenance of software in a systematic method. Typical formal definitions of software engineering are: "research, design, develop, and test operating systems-level software, compilers, and network distribution software for medical, industrial, military, communications, aerospace, business, scientific, and general computing applications." "the systematic application of scientific and technological knowledge, methods, and experience to the design, implementation, testing, and documentation of software...
>Software engineering is the application of engineering to the design, development, implementation and maintenance of software in a systematic method.[1][2][3]

Typical formal definitions of software engineering are:
1. "research, design, develop, and test operating systems-level software, compilers, and network distribution software for medical, industrial, military, communications, aerospace, business, scientific, and general computing applications."[4]
2. "the systematic application of scientific and technological knowledge, methods, and experience to the design, implementation, testing,
8:49 PM
application of engineering to the design, development, implementation and maintenance of software; isn't Programmers supposed to be design, but not development and implementation (SO) or did I misunderstand the help center?
I like definition 2 & 3 best.
or is that meant to be read "how to design and implement" and not "what is designed and implemented"?
@AaronHall didn't ask what Wikipedia said.
Let's try something else; you wake up in a strange house; as you wander through the rooms, you see no one. On a table next to a davenport, you find a dusty magazine with only the words, 'Software Engineering" on the cover. What topics would you expect to find in the list of articles upon opening it?
@jrh this is a good - and I dare say, common - question. And I'm not really seeing it answered in either the current or proposed help center.
To attempt to relate this to another field, SE is like a civil engineering firm and SO is like skilled trades? I'm not sure
I've already been eaten by a grue. um... what's a davenport? uh... Topics on practices, principles, and processes to develop software?
8:55 PM
Ideally, if we're going to all the effort of changing the name, we don't pick another one that sets up false impressions of what the site is about, only to spend the next few years trying in vain to correct those misconceptions via walls of text elsewhere.
@AaronHall such as...?
Methodologies (Agile/Scrum/Extreme)? Conceptual frameworks (SOLID)? OOP and functional programming?
The mechanical analogy sort of breaks down when you get into more complex subjects though, SESE might say "in order to make it so that this gasket does not leak, you must put 200 pounds of force on this surface" but then is it design work again to say "do that using a hydraulic pump" or is that SO's realm?
which is almost exactly where I personally start getting confused as to where SO ends and SESE would begin (I'm not sure it this is such a good example anymore)
Code answers given in pseudocode (perhaps executable pseudocode - i.e. Python)
@RobertHarvey you're missing out! :)
@Shog9 was I right in what I said? Programmers is ""how to design and implement" and not "what is designed and implemented"?"
@jrh beats me
hence my question
9:01 PM
something simple like "no real code" might help, that sort of rules out the "is this <OOP Standard> best practice" questions (not sure if those are good questions for this site?)
Yeah, I think we've said that before, it's about how, not specifically what.
Any specifics in answers should be mere examples to illustrate a greater principle.
"Clean Code" has lots of specific examples, but I don't think you would pick it up to learn Java.
please anybody guide me from developing CAN Library for STM32F4 discovery board ... i am begging .. please .. if u have source code please give me... i am begging u .. pleasr
CAN = Controller Area Network, the car protocol?
yes sir
i need to develop CAN LIBRARY for stm32f disocvery board without using CMSIS Library .... please help me.. please guide me..... if u have example code give,,,, it will be great help
@user143252 I have no idea how to help you, but I think you need to ask this kind of question on StackOverflow, not in the programmers chat.
Maybe frame it as, "How do I hello world with CAN on ..."
9:11 PM
whenever i ask question there, they delete my question ... here i get some answer.. thats why i ask here....
@jrh can u help me
you might not be able to get around the CMSIS library, I've never used it but browsing around this arm.com/products/processors/cortex-m/… makes me less than optimistic that what you're trying to do is going to work
why don't you guys take it to another chat room, jrh, maybe you can invite him to follow you.
I can't create rooms, maybe I can steal one but I'm not sure if that's frowned upon
sure, use a preexisting one, what's the harm.
can u give me email id ... i will not spam .. swear
9:14 PM
my apologies in advance to Robert and Kai if this makes the older logs harder to read
I wouldn't worry about it.
go hit it.
@jrh ... so sir what to do .. i search over github ... i saw example code of CAN.. it uses CMSIS library ... any help is appreicated ... my english not good apology...
ok, maybe a mod can wipe the above noise? "Clean Code" was the last relevant message.
9:38 PM
Yeah - that works, I guess.
@RobertHarvey "implementation"... "maintenance"...
welcome to Stack Overflow - this is a programmers forum; for general computer problems (hard- and software), please ask on Super Userumläute 21 secs ago
@Shog9 We would specifically disallow Stack Overflow's "how to fix my broken code" questions.
And "How to use my programming tools" questions.
Which is about 90% of what gets asked on Stack Overflow these days. Stack Overflow already adequately addresses those questions, and we don't want them.
Let's keep in mind that this is for the established community. New users that want to split hairs over it will get the boot even faster than the ones who just refuse to read it.
9:53 PM
@Shog9 I have no problem with "How To" coding questions, so long as they're not Too Broad.
You mean, "Too Specific"?
No, I mean "Write my program for me" questions.
That's too specific.
yeah, but... The most compelling argument for a name-change was confusion caused by the name. So if you still have to explain this...
9:54 PM
It can't be done with a two-word name. It's a good start, but you still have to explain something.
Any thoughts on the proposed interstitial?
@RobertHarvey so do that in the first paragraph / tagline
We have:
> We don't do code troubleshooting here, so if your question is about how to fix your broken code or how to use your programming tools, ask it on Stack Overflow, making sure that you provide a minimal, complete, and verifiable example.
I propose that it's 6:00 PM (Eastern) on the Friday before a long weekend and that we adjourn and pick this back up on Tuesday...
I still sense a disconnect between the Progs community and SE corporate.
eh; I just want this to do what y'all hope it will.
10:00 PM
The interstitial is a big part of it. We hope that "Software Engineering" will hint at "this isn't coderz-r-us", but you can't possibly hope to put your whole message into the site title, or even a subtitle.
We worked very hard to get "Who we are" into as short a space as possible, but it can't be done in a subtitle.
Doesn't even sound right.
I agree that the title change will be a big help, and then mods can better focus on shutting down the desperate pleas for code from those who leak from SO.
Don't expect too much from just a title change. The title change is very important, but it's not enough.
You still have to communicate who you are.
It will help shut down all the people who try to be helpful and send bad askers this way from SO.
To be fair, it did seem to bring a big positive change to Electronics.SE.
Yeah, that's like, "hey, I like to play with batteries and soldering guns, I can ask there!" vs. "electrical engineering? that's high level stuff."
10:06 PM
It's hard to imagine anyone saying "Well, this didn't work. Let's go back to 'Programmers.'"
And it's the best site title anyone can come up with.
Yeah, no one is going to say that.
How about "Software Design"?
I don't hear anyone complaining. I was the loudest objector, and my thoughts were I didn't think SE would do it regardleess, so it was a moot point.
@jrh No. See the meta discussions on the name.
@jrh We kinda had that discussion already. "Software Design" doesn't encompass the entire SDLC, and we have several members who still want to see questions about things like CM.
10:09 PM
Configuration Management.
Continuous Integration, things like that.
ok, makes sense
The line in the sand for implementation questions is very clear. Just look at 9 out of 10 questions asked on SO about "how to fix my broken code." These are they questions SO wants, and the ones we don't.
I see those questions on SO and I do not want to respond to them. I don't know why I have that objection, but I do. Anyhoo, I gotta head home or I'll be in trouble. Good night!
11:02 PM
@RobertHarvey I think it does, but you gotta reinforce that right away. And right now, I'm not seeing that.
I went through a few questions flagged as OT on the site, and you got quite a spectrum. There's straight-up, "my code doesn't work, why?" questions, but there's also "how do I do X?" questions and "How can I make X do Y better?" questions.
And I stopped, because... I don't feel like I know where y'all want that line right now.
Is it, "design and strategy, but not low-level execution?"
Or is it, "The entire SDLC, but no troubleshooting"
I'm kind of curious about that too, what's too narrow for Programmers? A single function? A single assembly? A single application?
is it OT to ask how to best design something high level and complex, e.g., "some question about this <high level> tree searching algorithm for 3d collision"?
if the answer to the question would be something like "visit each node and check to see if its nearest neighbor is on the correct side of a plane", and not code, would that be OT?
or to put it another way, are Fabien's source code reviews similar to an on topic answer? (very little code, but lots of analysis) fabiensanglard.net/doomIphone/doomClassicRenderer.php
11:28 PM
@Shog9 "how do I do X?" and "How can I make X do Y better?" are often gray area, my understanding is call to action was originally not about these. Note how it focused on questions that people just don't want answering because these clearly don't belong. The edit referring close queue sort of bent the original intent
> many such questions get four close votes, and then languish on the Unanswered page, sometimes for several days

« first day (2128 days earlier)      last day (1573 days later) »