« first day (2243 days earlier)      last day (2746 days later) » 

1:15 AM
I guess you might able to ask on Software EngineeringSome Person 11 secs ago
 
 
2 hours later…
2:57 AM
Woot! Duga is spotting our new name :)
 
 
8 hours later…
10:41 AM
@CandiedOrange I did not configure her to do that! WTF!
We noticed that we have some references to your site on our help center on Code Review. Besides the name change, is it still accurate that those kind of questions would be on-topic here?
 
Those two look fine to me. Our topic hasn't actually changed. It's just been polished and rebranded in the vain hope that it'll be understood this time.
 
@CandiedOrange It was not because of the name "Software Engineering" though
 
You talking about Duga?
 
yes. She is not taught yet to use her name
but I added that comment to her training set
in Duga's Neighborhood, 1 min ago, by Duga
@SimonForsberg Result: CLASSIFICATION_ADDED
 
11:46 AM
@SimonForsberg does that mean Duga will keep tracking "programmers"? I would very much prefer this, I somehow feel like worst kind misleading comments will keep using old site name for quite a while
 
12:38 PM
I ran out of time to add some training comments.
I was just swamped this weekend.
@SimonForsberg If you change "Programmers" to "Software Engineering" in your help center, that should be fine. With that change, I don't see anything factually wrong. Maybe update the URLs to point to softwareengineering.stackexchange, too.
 
@gnat Yes it will.
@ThomasOwens Any comments that you can find would be helpful. Of course more data is better data, but any data you have already gathered would be helpful
and I also realized that the comments can probably be included in her existing training set
 
I haven't gone looking for comments. But if I find any, I'll be sure to capture them.
 
1:28 PM
Crowdsourced question: how would you answer "what is perfect CI?"
 
1:56 PM
@enderland Fantasy?
> In software engineering, continuous integration (CI) is the practice of merging all developer working copies to a shared mainline several times a day.
I was not aware that that was a requirement for CI...
 
heh
I'm giving a talk titled "in pursuit of perfect CI" and I am intrigued by the different responses I get to that question
 
perfect CI requires perfect developers
make sure to mention build and test times, the lower latency there is for detecting a broken build the better
 
@ratchetfreak the premise of my talk is an effort I took to reduce our build times by 75%
I'm going to quote both you if that's ok anonymously, @ratchetfreak @SimonForsberg ?
 
@SimonForsberg about Dura, I have seen people putting our name in backticks presumably to avoid the bot. Any way to defeat that?
 
2:16 PM
I think I'm already seeing subtle improvements in the front page after the name change. There are still plenty of people who struggle with asking a good question, and a few who still want codez halp, but there are a substantial number of new questions that are actually interesting and worth answering.
2
 
yeah
I noticed that too, not sure if it's because the mods are just auto deleting stuff or not :(
 
If they are, that would certainly explain a few things.
Very few questions about the actual SDLC, but I'm prepared to accept the idea that most folks are just not all that interested in scrum theory.
The same way you would scale microservices if you weren't using DDD. — Robert Harvey Oct 19 at 23:27
@RobertHarvey thank your for contributing — John 6 mins ago
 
@enderland Sure. I was under the impression that CI just meant "use Jenkins (or any similar system) to make it build after a commit was pushed"
@CandiedOrange Sure. Can clean the string a bit more before. Can you make it a github issue here? github.com/Zomis/Duga
 
I see CI as "everything that needs to happen after my code is checked in to put a build onto the test or production server."
There is plenty that needs to happen in our shop that is done manually, and some of it is error prone.
 
@RobertHarvey Ah, but that's CD. Continuous Deployment
(I think...?)
 
2:27 PM
Ah, I see. Well, CI alone wouldn't help us much.
 
CI/CD gets blurred
 
> The system should build commits to the current working version to verify that they integrate correctly. A common practice is to use Automated Continuous Integration, although this may be done manually. For many[who?], continuous integration is synonymous with using Automated Continuous Integration where a continuous integration server or daemon monitors the revision control system for changes, then automatically runs the build process.
^^ that's me
> Most CI systems allow the running of scripts after a build finishes. In most situations, it is possible to write a script to deploy the application to a live test server that everyone can look at. A further advance in this way of thinking is continuous deployment, which calls for the software to be deployed directly into production, often with additional automation to prevent defects or regressions.[10][11]
 
I would say that few people in practice make the distinction to separate CI/CD if they have a combined system into it's requisite pieces
"CI" becomes "everything automated after I commit"
 
@RobertHarvey I am. I want to try to promote the site more, now that it's easier.
But then again, I'm weird.
I'm pretty sure I've done less deletion of garbage since the name change than before, too.
 
@enderland I like that definition.
What's SDLC?
 
2:39 PM
@SimonForsberg software development life cycle
@ThomasOwens I've been trying to either up/downvote every question I read on SE^2
 
The systems development life cycle (SDLC), also referred to as the application development life-cycle, is a term used in systems engineering, information systems and software engineering to describe a process for planning, creating, testing, and deploying an information system. The systems development life-cycle concept applies to a range of hardware and software configurations, as a system can be composed of hardware only, software only, or a combination of both. == Overview == A systems development life cycle is composed of a number of clearly defined and distinct work phases which are used by...
 
man. HNQ is interesting with controversial questions/answers, heh
 
This question is probably more suited to programmers.stackexchange.com - unless you have a specific problem using the API? — mabi 56 secs ago
 
@ThomasOwens I wish we could just auto respond with that comment every time P/SESE gets recommended
 
@enderland The one I just wrote?
 
2:50 PM
@enderland Hmm, perfect is a fantasy. So my perfect CI/CD fantasy is a development environment with sufficient tests in place that as soon as I save my code it's tested and, if it passes the tests, is automatically integrated and deployed. All without any disastrous surprises. It's the last part my reality keeps getting hung up on.
 
3:05 PM
@mabi Programmers is now Software Engineering, to better represent the scope of the site. Questions such as this would be considered off-topic (recommending libraries) or primarily opinion based (what is considered production-ready by one team or organization may not be considered production ready by someone else). I can see that you have an account, but no participation on Software Engineering. You should avoid recommending sites that you don't understand - when a question that isn't a good fit is posted and it ends up being down voted, closed, and deleted, that's a poor user experience. — Thomas Owens 20 mins ago
 
@CandiedOrange For me the biggest problem with that is to actually write the tests....
 
I've debated adopting a 100% TDD approach for a while just to see
people who do that claim it increases their dev speed
 
I don't think 100% TDD is realistic.
3
Some tests are going to reveal themselves based on the inner details of the method.
You can't write all of your tests first. You can write some tests, but won't likely be able to figure out all of your boundary cases until you have the actual code written.
 
3:21 PM
I think the bigger problem is when you don't actually know in advance what API you want
 
TDD can help with that, though.
TDD is great for testing API design.
 
right. but that assumes you have a fairly clearly defined API/spec or user story of some sort
 
How so?
You need to have some idea of what you're building no matter what methods you use.
 
If you missed the overwhelmingly discussed meta discussions (15 questions tagged with new site change) over the past eight months and didn't make your objections known at that point or really heavily engage in that process, I guess personally I have minimal sympathy. That being said if you look at any of those questions nearly none of the "Software Engineering as a name" posts had any level of significant downvoting. This is +154 / -4... — enderland 1 min ago
 
@enderland you can write the typical usage code against stub implementation first
then build the test out of the usage code
and then start implementing
(or skip the test and debug as you go)
 
3:46 PM
yeah. just a bit of chicken/egg there
 
but when you write a library you know what you will feed into it and what you expect out of it.
the simplest api is then output = libFoo(input);
 
Apparently, in-browser document preview for multiple formats (PDF, MS Office, Open Document, etc) is hard. Finding a self-hosted solution to product in-browser document previews is harder.
Anyone have any experience in this area?
 
4:19 PM
you know what would be great? if wikipedia provided a copy/pasteable image citation for CC-BY-SA licensed images
 
4:39 PM
I need to give a short presentation on OOP, I remember one of the first OOP's was some scientists modeling boats or water flows or something, anyone got a pointer on it?
 
Tom
I don't, but you might look at object oriented design, the example you want might be talked about there
 
 
Kaitlin Pike on October 24, 2016
Welcome to The Stack Overflow Podcast #92, brought to you by the The Facsimile Association of America. Since the late 1800s, faxes have solved the infuriating problem how to get a message to someone quicker than physical mail, and since the late 1900s, slower than just sending an email. The Facsimile Association of America: When Was the Last Time You Needed to Fax?
 
@gnat BOOM
 
I just gave a final - the students were all thanking me telling me that they "learned a lot," - so I hope they have me back next year.
 
@enderland Painting the bikeshed.
@gnat: What does this mean?
0
Q: Looking for recommendations for external hard drive enclosures

NZKshatriyaI am looking for an external enclosure that meets the following specs: 2.5" drive support is nominal, but 3.5" will work as adapters are readily available. SATA III to USB 3.0. 2 to 3 bay. Price under $100 USD. Has on-board RAID hardware with options for RAID0, JBOD, etc. RAID options are not...

I'm voting to close this question as off-topic because drive bays do not fit in the definition of hardware that we can recommend (electronic computer hardware, essentially). — ArtOfCode ♦ 1 hour ago
Hmm...
 
5:16 PM
@RobertHarvey this has been proven to work surprisingly well on folks from Math.SE, in the sense that they tend to delete their inappropriate questions when hinted that way (2 of 3 examples listed in comments to that post were self removed). Can't read their mind but my theory is these folks are mostly smart and reasonable and main reason they post like that is they are used to this way at Math.SE, where it is normal and acceptable and when pointed to cultural difference they often get the hint
 
You seem to have better luck than I do at this sort of thing, then. It took me awhile to realize that people don't click my links, 95 percent of the time.
So I no longer bother. I didn't click your link either; on its face, it doesn't seem to have much relevance.
 
@RobertHarvey I guess Math.SE folks react to magic word "hint"
 
Ah, I understand now. Absent that explanation, I doubt that the rest of the Programmers Software Engineering community has the slightest idea what you're doing.
How do you know they're coming from Math.SE?
Oh. Linked accounts.
 
@RobertHarvey Does a fellow get a chance to paint a bikeshed every day?
 
@AaronHall Your students will want to return to your classes because you have long hair, man.
(Just an observation)
 
5:27 PM
well unless they let me teach "Advanced Python" I doubt they'll get me again. They are supposed to take an SQL course in the spring, but I doubt they're going to tap me for it.
I'd love to be qualified though.
 
Do you know SELECT, UPDATE and JOIN? That's all you'll really be able to teach in the first semester anyway.
Well, that and GROUP BY, WHERE, HAVING and aggregate functions.
And maybe some simple DDL.
Pretty sure you know all that.
It would have been really nice to have had someone point the way when I was first learning SQL. It would have taken days or weeks instead of months.
 
That's the prerequisite handout for the PyCon sqlalchemy tutorial
 
If there's one thing that new-ish people really need, it's some objective, non-dogmatic hand-holding in basic data-application architecture. That, and fundamental tool skills like basic IDE usage. Most courses teach you how to write a for-loop. booring.
 
yeah, we did for-loops fast.
 
Interesting read:
27
Q: What to do when part of a community lacks trust in ♦ moderators on small stacks

SmurftonI would like to emphasize that I personally do not lack trust in this ♦ mod, nor do I wish for his removal. As such, I will not be mentioning the mod's username. While it is possible to work out who this mod is from this question, please don't. I have been noticing that several members1 of Role ...

Fair warning: this one is like TVTropes. You might get sucked in for awhile.
In particular, this (linked):
39
A: What can I do against moderators with double standards?

Shog9Well, there are a few popular strategies here, and considerable debate as to which is the "best"... I'll lay them out and leave the decision up to you. The Archivist This strategy involves carefully monitoring a moderator's activity over time, potentially many months or even years. Notes are ta...

Also, just in case someone here hasn't seen this yet:
309
A: Let's Plan The Second Iteration Of The Stack Exchange Quality Project!

Robert HarveyNew users need better awareness of what is expected of them when they ask their first question. This is the single most important thing that needs to happen on the Stack Exchange network, in my opinion. If the new Question Ask page will facilitate that, then I am enthusiastically supportive. H...

Apparently SE has been thinking about this for awhile, but didn't show their hand during the name change.
> We plan to test a new, 'guided' version of the ask question page soon. This page would essentially break down all of the elements that make a great question, and give the user plenty of guidance as to why it's to their advantage to understand what's needed in each of them, and provide it.
 
5:45 PM
@RobertHarvey sweeeeeeet!
 
@RobertHarvey I hope that it's customizable. Guidance for a good SO question isn't the same as a good SE question or a good Travel question.
 
Wait, they're going to take my advice?
Because I've been whacking that decomposing pony for what seems like years.
I want to see how I'm ranked by impact on SO. There's no easy way to do it. It's like I'll have to work or something to get it...
 
6:04 PM
@AaronHall Who's taking your advice?
 
6:16 PM
33 mins ago, by Robert Harvey
> We plan to test a new, 'guided' version of the ask question page soon. This page would essentially break down all of the elements that make a great question, and give the user plenty of guidance as to why it's to their advantage to understand what's needed in each of them, and provide it.
 
6:46 PM
Stick to reddit for these types or try programmers.se — self 1 min ago
@self There is no more Programmers Stack Exchange - it has been rebranded as Software Engineering Stack Exchange to better indicate what is on-topic there. Questions like this are not considered good questions - we have the same too broad and primarily opinion based close reasons as Stack Overflow. A question like this would be quickly down voted and closed, leaving the asker with a very poor user experience. — Thomas Owens 12 secs ago
 
Very well written comment, @ThomasOwens
 
Thanks.
 
History of C++ (pdf): stroustrup.com/hopl2.pdf
 
I met him once.
 
Me too! He said I could do lunch with him.
I have no idea what we would talk about though.
 
6:59 PM
He likes hoppy beers.
If you do lunch with him, bring him the hoppiest IPA you can get your hands on.
 
He likes bitter, eh?
For the past year, I've been avoiding IPAs like the plague, preferring anything dark, fruity, or otherwise.
Cherry-wheat, Octoberfest, Stouts, Lagers, etc...
IPAs have been my last choice. But it seems like it's all I can find in my hosts' fridges. (I go to tons of meetups.)
 
7:15 PM
I like stouts and porters in the late fall and winter, like around now.
So nice when it's cold outside.
 
Oh hey, I pushed an update to my new website yesterday. now it's light-grey text on black instead of the harsh black on white. Feedback is welcome - aaronchall.github.io
 
The three click thing is confusing.
 
three click? on the headings?
 
Yeah.
It doesn't even look like it's going to expand, but then it does. And you would expect a second click to minimize, but it doesn't
 
Yeah, it works kinda like orgmode. I didn't program it. I wrote as little configuration as possible.
 
7:19 PM
It's not user friendly, that's for sure.
 
That's my new mantra for emacs - as little configuration as possible.
 
I also prefer light backgrounds. I'm one of the few people at work who don't use a dark theme.
 
Yeah, I wish it could just be optional on like all sites: light or dark, click a button. Like clicking a button for the language.
 
7:35 PM
0
A: Are external hard drive enclosures on topic?

Robert HarveyThe Programmers.SE site suffered through a long bout of "byzantine scope rules." We finally got religion and threw out whole categories of questions that touched on scope rules that we came up with in good faith, but which nobody could understand. Intuitively, it feels like hard drive enclosure...

 
8:18 PM
Hi, which stackexchange can I use to get an answer about specific API's I can use for a project?
 
@develop1 If you're looking for a recommendation for which API to use, none I'm afraid.
Google is better suited for that kind of research anyway.
 
ok, I tried google but haven't had any luck finding what I need.
 
What sort of API are you looking for?
 
Real time stock information
I tried google and yahoo both are unsupported anymore. The API has to be HTTPS because I am running my web server through Heroku.
 
what language
 
8:33 PM
You'll probably have to pay a data provider for that information.
 
@enderland Javascript, JSON
@AaronHall Yeah, but this is a school project for one semester so I am trying to avoid it
 
Real time is kinda silly for a school project, you could scrape Yahoo Finance and update a small portfolio every few minutes.
 
I have considered it. Both Yahoo and Google are unsupported anymore, I either get a Cross-Origin Request Blocked for Google or server can't be reached error from Yahoo.
@AaronHall you are right that real time is not completely necessery. Even a couple hour delay would still work.
 
8:52 PM
Then there you go. I just got done teaching a class how to do this with Python, but since you're using Javascript, I can't help you further.
 
@develop1 I think api/data sets recommendation questions are on topic on opendata.stackexchange.com. However, I'm not a regular there, I could be completely wrong. Do check their help center and search for similar questions before you ask your own.
 
@Yannis Thanks I will look into that
 
@develop1 actually, you should look for a paid data provider and see if they'll give you complimentary access. I'd bet you can arrange it.
 
that's not a bad idea if you can do it, but I would also make sure you don't spend 1/2 the semester trying to chase down an API if you can find a workable older one
 
 
2 hours later…
10:47 PM
Questions including words like "better" are not answerable without more specifics. What do you mean by "better?" Note: don't answer with a tautology like "best practice" or "most popular." State your specific acceptance criteria.Robert Harvey 9 mins ago
 
jrh
11:03 PM
Just wondering, does anybody know of a Journal that publishes papers on programming paradigms? E.g., something that would have articles on functional programming, OOP, etc., I think that's something I'd like to read
I see a lot of blogs but I'm sort of looking for something more formal and academic
something that would go into the problems that some new programming style was meant to solve, and how it did it
I'd like to do some research that compares the goals/theory behind several different programming styles (OOP, functional, structured, etc.)
 
11:19 PM
@jrh Something that you should be aware of: many of the new tools and techniques that are emerging in the existing language spaces are actually old ideas. Functional programming enhancements being added to languages like C# actually derive from McCarthy's Lambda Calculus in the 1950's.
 
jrh
Yeah, I was surprised when I was reading the wikipedia page on functional programming, it really is quite old
 
wikipedia in my experience is pretty weak when it comes to meaningful CS related articles
 
There's no dearth of technical papers. Read Okasaki's Thesis on Functional Data Structures, or get his book.
 
jrh
hmm, that's a pity that there aren't many papers. I've learned a lot reading papers on algorithms, I was hoping I'd be able to see something in the format of a new OOP concept to solve <whatever problem> like I've seen for algorithms.
 
This feels a lot more like a Stack Overflow question than a SoftwareEngineering question to me. Are you asking how to implement this? Or for design guidance? — enderland 5 secs ago
 
jrh
11:24 PM
I will check those out, thanks
 
@jrh While OOP is quite useful, it's not all that remarkable from an academic perspective. At the end of the day, it's just a collection of code organizing principles.
If you're really interested in academic OOP, study Smalltalk.
That's the real OOP.
 
jrh
I'm kind of interested to know what the original design decision behind OOP was, i.e., something that explains what brought about the idea of taking the first parameter into a function and making the function a member of the type of that parameter
 
depending on how deep into the rabbit hole you want to go, you can always look at the citations in those papers Robert linked
and see what papers cite them - they are probably related to the subjects
 
jrh
I think I should probably try smalltalk, it might answer some questions, for whatever reason C# style guides/etc. don't quite give me a clear enough answer for what I'm looking for
I'll dig around, thanks guys
 
11:30 PM
@jrh Are you referring to dot notation? f(x,y) --> f.x(y)
 
jrh
@enderland yeah, most of the books focus on the how, not the why, I'm mostly interested in the why
well, yeah f(x,y) -> x.f(y)
 
Dot notation is just dereferencing a member of a class. It's pretty much baked into class design. But you see its implications in, for example, method chaining in C#.
Where this is returned from the function, allowing you to dereference again.
 
jrh
Sure. e.g., structs in C use it to access data. Though I'm kind of curious about at what point they made the leap from "here's a group of functions that operate on structs A and B", to "these functions are now grouped with A, and they take in B as an input".
 
That's just the nature of classes and the Kingdom of Nouns. noun.Verb()
As opposed to Verb(noun).
 
jrh
Right, that's pretty much what I'd like to know, why noun.Verb instead of Verb(noun)?
 
11:35 PM
@jrh there does need to be a syntactical distinction
 
Because a class is a noun. And you have to refer to the class first, unless you're working in an OO language that allows free functions like C++.
I think the latest version of C# finally allows such free functions. So you can now say Verb(noun) instead of StaticClass.Verb(noun) or noun.Verb()
 
jrh
Were these nouns introduced primarily to improve the problem of namespace pollution, though?
I guess I can follow that, it's better than GTK_DoStuff, though you might not need classes to fix that
 
All of the class ceremony in languages like Java serves mostly as scaffolding, documentation and encapsulation.
You can do the same thing in functional languages with much less (and arguably more expressive) code. But there are also far fewer programmers in the marketplace that will understand it.
You can find a Java programmer on every street corner. Lisp programmers, not so much.
 
jrh
Smalltalk and Lisp seem to be the ancestors of what I'm trying to research, I think I'll look into those next when I get some time.
 
Happy hunting.
 
jrh
11:44 PM
I bought this book amzn.com/0201543303 but I was kind of disappointed that there wasn't more about the why of OOP; it makes sense since Smalltalk may have already made that case; though it was a pretty cool book that tells some interesting things about starting a new language.
 
@jrh "unknown binding" $700. better be bound in gold leaf
 
jrh
I bought the $2 version
 
@jrh Incidentally, Alan Kay (the inventor of Smalltalk) says: "Object-orientation doesn't mean what you think it means." lists.squeakfoundation.org/pipermail/squeak-dev/1998-October/…
Also, Object-Orientation in C++ is partly a response to the problem of scalability (in terms of writing code) in C. While you can certainly write large programs in C, C++ is specifically designed to facilitate such programs.
 
jrh
IIRC Bjarne mentioned that C++ was also intended to replace macros in C, and to improve static type safety.
which is naturally related to the scalability of the program
anyway, thanks again guys
 

« first day (2243 days earlier)      last day (2746 days later) »