« first day (987 days earlier)      last day (4002 days later) » 

Joel
Joel
7:35 AM
Okay I'm confused.... So I've found both these articles linked to from question answers before basically about how if code is well written it becomes more maintainable than rubbish commented code.
http://butunclebob.com/ArticleS.TimOttinger.ApologizeIncode
http://elegantcode.com/2010/04/18/eliminating-comments-the-road-to-clarity/

But then as replies to this question:
http://programmers.stackexchange.com/questions/198458/working-in-a-company-that-does-not-comment-their-code-at-all

The main feeling seems to be summarized by "There are a lot of reasons for no comments, but I think it usuall…
 
Yannis Rizos
Yannis Rizos
7:47 AM
@Joel Comments are good. Except when they are not. I'm not helping, am I?
I think the key idea is that horrible comments can be utterly catastrophic. And horrible comments are more common than most of us realize. While most of us would prefer a comment here and there, if the choice was between no comments at all and horrible comments, we'd chose no comments at all.
However, "horrible comments vs no comments" is a false dichotomy. Good comments may be hard, but are not impossible.
 
MadKeithV
MadKeithV
@Joel In the real world there are many problems that cannot be solved with a solution that is "obvious" and/or "simple to understand". You better comment those solutions before the new guy that's just read those two articles replaces the solution that has been carefully written and optimized over many hours of work with a solution that is simple, elegant, and wrong ;-)
 
Joel
Joel
8:02 AM
@YannisRizos But can't a good comment turn into a horrible comment if the code base changes is a way that makes it wildly inaccurate?

@MadKeithV Is that really a problem? If you use Unit Tests and Version Control then you can revert and not give the new guy a biscuit.
 
Yannis Rizos
Yannis Rizos
@Joel Sure it can. Which makes the whole situation even more complicated.
 
Joel
Joel
I'm so confused, darned grey areas...

@YannisRizos So when would you comment your code? When you know that is won't be a horrible comment, which is never?
 
Yannis Rizos
Yannis Rizos
@Joel That's the thing, there's no simple answer. I'm working on two projects right now. First one has a couple of very new devs on the team, and we try to comment. Second one, we are all veterans, there isn't a single comment anywhere on the codebase (tests, req specs and rev history is all the documentation we need). I'm not suggesting either approach is "correct", just that team dynamics may also play a role.
Some rules of thumb:

1. Comments rot - http://programmers.stackexchange.com/questions/69538/are-outdated-comments-an-urban-myth
2. Tracking changes is the version control's responsibillity - http://programmers.stackexchange.com/questions/175309/code-maintenance-to-add-comments-in-code-or-to-just-leave-it-to-the-version-con
3. Comments should explain the why, not the how -http://programmers.stackexchange.com/questions/173118/should-comments-say-why-the-program-is-doing-what-it-is-doing-opinion-on-a-dic
 
Joel
Joel
8:22 AM
@YannisRizos That makes sense thanks.
 
Yannis Rizos
Yannis Rizos
@Joel You should also check out our highest voted comments questions. Some good stuff in there.
 
Joel
Joel
 
MadKeithV
MadKeithV
@Joel Breaking changes in the types of solutions I'm thinking of will not be easy to catch by unit tests. They should be caught by tests, perhaps even integration and system tests - but those tests might well be just as hard to write as the solution, if not harder. So in the end you may be spending hours and hours coding up a framework to express something that you could have spent 10 minutes on by writing a decent comment.
 
Yannis Rizos
Yannis Rizos
@Joel Heh. The actual comment also had a couple of four letter words in it.
 
MadKeithV
MadKeithV
@YannisRizos Code is a four-letter word.
2
Continuing the thought from my previous comment - you might end up in the situation where your test code is so convoluted that it needs comments ;-)
 
Joel
Joel
8:45 AM
@MadKeithV What sort of solutions are you thinking of? I mean it seems like there's a lot more factors than I realized.
 
MadKeithV
MadKeithV
:9470059 It's hard to come up with a really good concrete example, but one I instantly thought of is saving a file somewhere. That's simple, right? Except if the current user doesn't have write rights in the location you've picked. Except if the disk is full. Except if the location is a network location that becomes unavailable halfway through. Etc. etc.
 
Joel
Joel
@MadKeithV Could you not just throw some suitably named exceptions, extract some methods and maybe use some layers of abstraction?
@MadKeithV I mean I guess you could write code that was 100% self-documenting if you were a genius with an unlimited amount of time but I guess it the real world comments are useful sometimes.
Anyway Clean Code is my holiday reading so I guess I'll be less of a noob after that.
 
MadKeithV
MadKeithV
9:02 AM
@Joel You could do an awful lot of things, but it has to be worth the effort. Somewhere there's a cutoff point where your time could be spent more valuably than trying to reorganize the code to the point where you don't need that comment you could have written in 5 minutes.
 
Joel
Joel
@MadKeithV Yay! The real world ruining my code perfection dreams.
Thanks for the help @MadKeithV @YannisRizos
 
 
4 hours later…
user41796
1:31 PM
@gnat - I fully appreciate and understand the motivation behind finding duplicates as the close reason instead of the other 4 options. I have noticed that some suggested duplicates are a bit of a stretch though. I think that those "stretch duplicates" undermine the basis of what you're accomplishing in many of the other closures. Please consider tightening your criteria for what constitutes a duplicate.
 
Joel
Joel
1:50 PM
@GlenH7 Got links to any specific questions?
 
user41796
2:24 PM
@Joel Yes and no. Yes, there a few I could pick out and post, but I'm reasonably certain gnat knows some of the ones I'm referring to. No, I hadn't planned on posting any because I didn't want to create a discussion around my suggestion.
 
user41796
@gnat - if my suggestion doesn't make sense, I'm happy to post some links and / or we can move this into a discussion in a different chat room.
 
user41796
@Joel - There's a longer explanation as to why I'm explicitly avoiding a broader discussion surrounding my suggestion. There are a few folk who aren't as happy with the clean-up process that P.SE is undergoing and that I think is required. I don't want my suggestion used as a starting point to argue against the cleanup. I believe that providing specific examples would generate an in-the-weeds discussion focusing on those questions instead of the broader aspects of cleanup.
 
user41796
Hence a suggestion that gnat is free to happily ignore. :-D
 
Jimmy Hoffa
Jimmy Hoffa
@Joel steve-yegge.blogspot.com/2008/02/portrait-of-n00b.html @YannisRizos you reminded me of this article
@MadKeithV The only reason you like comments so much is because C++ devs are in love with little black boxes of super-optimization here and there which can't be made readable, a lot of us choose the readable solution over the optimal one every single time :P
We live by the mantra "Throw more hardware at it"
 
user41796
2:41 PM
@JimmyHoffa - some systems need the super-optimized routines. Case in point; the backup / archive app I worked on had some sections coded in assembler to eek out the last bit of performance. And yes, it mattered to our customers as well.
 
Jimmy Hoffa
Jimmy Hoffa
@GlenH7 yeah, but what's that? I CAN'T HEAR YOU LALALALA
Just illustrating a large swath of the industry's response to the concept you speak of
I'm not against super optimized little black boxes, I've just never worked anywhere that would even begin to allow such
So I practice brevity and conciseness instead, in reality non-C++/C places rarely do that
which is why I said he only likes comments because he's a C++ dev heh
 
user41796
@JimmyHoffa - no worries; just pointing out one of (many) cases I can cite where the comments were required on top of self-documenting code
 
user41796
I'm secretly a C dev working on C# these days... :-)
 
Jimmy Hoffa
Jimmy Hoffa
@GlenH7 tell me you don't write comments that repeat exactly what the code says every few lines, that's been my experience with C/C++ folks joining C# which in C# it's so unnecessary a lot of the time
@GlenH7 tell me you've overcome the desire for one-exit point in your methods? Good C# should always return immediately.
C/C++ folks drag that single-exit mantra forward to C# forgetting the reasons you do it in C/C++ are no longer there in C#
 
user41796
@JimmyHoffa a) I never subscribed to the one-entry, one-exit camp. It's .... hogwash. Even in my C code, I'd return as needed (and nothing had to be freed). And it's why C# has finally. b) I only comment the counter-intuitive stuff and it's never "how" it's always "why". I still comment in my C# code but only because I work on some mighty obscure stuff. Ergo, the soul behind me needs to see the "scrawls on the prison walls".
 
Jimmy Hoffa
Jimmy Hoffa
2:50 PM
aha! even in C it was hardly necessary! I knew it! Those people who do this drive me nuts and now I know: it's not because they're C devs, it's because they weren't very good C devs to begin with if they felt the need to do this
 
user41796
oh, and I actively delete the comments from the n00b commentators who worked in that code before me when I'm next in the code
 
user41796
@JimmyHoffa .... nice try. ;-)
 
user41796
well, actually, I will agree with that to a point. Comments should be sparse, no doubt. but "no comments" is an equally bad extreme, IMO. If you have to comment the "why" of everything then either you really don't know what you're doing or you're working on some crazy-azz likely-patentable code. I've been lucky enough to be mostly in the latter camp.
 
Jimmy Hoffa
Jimmy Hoffa
@GlenH7 yeah, I could see that, but I disagree with your first point; though perhaps it's because I can read C# as well as english at this point, either way I find standard CRUD apps in C# can easily go commentless.
and are better off for it
and remember, that is like ~90% of .NET development right there
Perhaps your experience is colored by the, well.. colorful .NET code you've worked on
 
gnat
gnat
@GlenH7 here, I rather tend to rely on co-reviewers. I mean if stretch is for real, I would expect other reviewers to cover this. So far it looks like working just fine, there seem to be a fair share of my suggestions to be kept open or re-qualified to more appropriate close reason. Note please that even one reviewer...
...even one reviewer concerned about incorrect closure, could fix this, even if majority vote overrules. One can raise meta question or flag for moderator asking to correct wrong closure. I do that myself quite regularly, works like a charm
 
user41796
2:56 PM
no doubt; it's just my experience, that's all. I certainly don't purport to have worked on "standard" projects by any means.
 
user41796
@gnat by and large agreed. The web services / interface one was one that popped up this AM and I thought there was enough difference to point it out. At most, I'm suggesting a minor course correction is all.
 
user41796
I could also see your point on that with with co-reviews being the check point. Just because I see enough nuance there to indicate them as different doesn't mean others do. And it's potentially worth calling out that difference in the more recent question.
 
Jimmy Hoffa
Jimmy Hoffa
I agree with @MichaelT, I just put the last close vote to a Q and I totally would like to have gotten some fireworks for it
 
MadKeithV
MadKeithV
@JimmyHoffa I live by the mantra "throw more hardware at it" too, but in my case "it" means "management".
 
Jimmy Hoffa
Jimmy Hoffa
@MadKeithV I encourage this behaviour.
 
gnat
gnat
3:05 PM
@GlenH7 ah I see. i would appreciate if you post the links you mentioned, I'd like to take a deeper look at my mistakes (it never hurts to learn, y'know:)
 
user41796
@gnat - will do. I need to take care of a few things first this AM, but will ping you with a few examples and my thoughts.
 
MadKeithV
MadKeithV
Wanted to post some of that black-box code, but it filled a page of chat, bad idea ;-)
 
Jimmy Hoffa
Jimmy Hoffa
lol
It's ok, I actively translate pointer operators into comments whenever I read code anyway
 
MadKeithV
MadKeithV
@JimmyHoffa Real low-level C++ developers read SSE intrinsics as if they are Esperanto.
 
Jimmy Hoffa
Jimmy Hoffa
@MadKeithV my answer to people who think the metal matters is anymore OCaml; because you don't need low level to get C++'s performance, so why bother when pointers are possibly the largest cause of bugs known to man? heh
Now if you genuinely need ASM level twiddling to eek out the perf you want; great, or if you're doing embedded; good on ya, but outside of those two scenarios: There are high level languages that can meet your performance needs, and do so with far less bugs
...and that's my stance on that holy war ;P
hides behind cover
 
user41796
3:13 PM
@JimmyHoffa we'll just flag your inflammatory comment instead. ;-)
 
MadKeithV
MadKeithV
@JimmyHoffa No indeed, you don't need low level to get C++ performance, you can easily make your program that slow in a high-level language too!
 
Jimmy Hoffa
Jimmy Hoffa
@GlenH7 now I'm actually a little curious; you're a proper engineer which includes the rigor and the formality common to maths and proofs etc; but you're a C guy. Have you never lived the functional approach? I would think it more fitting to a proper real engineer with all the formality and rigor that those languages give you; have you ever looked at a language with dependent types? doesn't get much more formal than defining even your primitive types yourself
dependant typing is all about proving your code does behave exactly as it should, it won't even compile if there are any edge cases you didn't account for
(the classic joke for those languages is the code's never been run, because they only ever compile it)
 
 
2 hours later…
Yannis Rizos
Yannis Rizos
5:38 PM
@gnat Your comments with links to tag wikis aren't particularly helpful. First, nothing authoritative about a tag wiki. Second, the tag wikis you've picked aren't particularly well written. Just a bunch of links and a couple of out of context quotes aren't going to help someone trying to grok the site, especially someone who already failed to read the FAQ.
 
gnat
gnat
@YannisRizos well, I couldn't use the books faq reference because clarification on Q being book request has been added later. - and most importantly, because I felt there is a chance for question to get rid of that part. as for resource requests reference...
...I would appreciate any suggestion of better one
(that would not be limited to book requests)
 
Yannis Rizos
Yannis Rizos
@gnat Check the revision history more carefully, the question was asking for a book from the start.
Title is "JavaFX Book | Learn JavaFX More In Depth", and in the question text:
> I'm having a tough time finding a good book I can curl up with for some good ol' bedtime reading. The Oracle documentation is fine for starters, but I'd like something comparable to the Swing Hacks book or other of the great Swing books.
 
gnat
gnat
@YannisRizos I saw that, per my reading of first rev it wasn't clear enough to qualify as book request "Learn JavaFX More In Depth ...The Oracle documentation is fine for starters..."
that has a chance to slide from book request (had, until rev 2)
not that it would make a good question, but sufficient enough for me to abstain from quoting faq on books
until rev 2, that is
 
Yannis Rizos
Yannis Rizos
6:00 PM
Ok. As for a better comment, you won't find a boilerplate comment that would work on every off topic question. I'd suggest dropping them completely. Either the question is good enough (despite being off topic) to inspire you to spend a couple of minutes writing a personalized comment, or it's crap, in which case downvote, vote to close and move on.
 
user41796
Wall of text warning:
1) http://programmers.stackexchange.com/q/150045/53019
2) http://programmers.stackexchange.com/q/101300/53019
3) http://programmers.stackexchange.com/q/49379/53019
4) http://programmers.stackexchange.com/q/103779/53019
 
user41796
1) For me, the web services aspect is what differentiates this from the duplicate. Yes, services are just classes or methods off of classes, but the manner in which they're interacted with is different. So I see them as related, but different questions.
2) This one is really unit vs. integration testing whereas the dup is "what sorts of testing are there?" And that doesn't go into the level of detail the unit vs. int question could have generated.
3) Another related, but different set here. The question I linked is really focusing on when to use. The duplicate question (to design...) is…
 
Yannis Rizos
Yannis Rizos
As for a better link for resource requests: What kind of questions should I not ask here? (every answer is equally valid)
 
user41796
That said, I had to dig a bit to find those. And I specifically limited myself to ones where I thought the question should remain open. I could have found a few where I voted to close and may have picked a different close reason (NC, NARQ) instead of dup. To better summarize my sentiment, I think it's better to close as NC or NARQ when that's the case instead of finding a "duplicate" that's a bit of a stretch. To paraphrase Shog9's somewhat recent blog post, sometimes it's ok to be an a-hole.
 
Yannis Rizos
Yannis Rizos
Better as it's in the FAQ. I don't mind the tag wikis, but 1) not authoritative like the FAQ, 2) no sense of consensus like Meta discussions.
 
user41796
6:02 PM
@gnat - those are some examples based upon our earlier chat
 
Yannis Rizos
Yannis Rizos
@GlenH7 What's that about? Did we just gang up on @gnat?
 
user41796
@YannisRizos - no, no, or at least I certainly hope not
 
user41796
I had put out a suggestion earlier in the day about some suggested duplicates being a stretch
 
user41796
and I had only meant it as a mild course correction type suggestion not a "OMZG ze world is crashing!" comment. :-) @gnat asked for a few specific examples to better understand what I was trying to get at
 
user41796
and I probably could have waited for your discussion regarding that book request to finish up, but that would presume I have manners...
 
user41796
6:05 PM
That said - @gnat if you do feel like I'm ganging up on you then please blow off my suggestion and the examples above. I'd much rather have the work you're doing in cleaning up the site than not.
 
Yannis Rizos
Yannis Rizos
In other news, I got a couple of responses to a comment I posted on another site that if I was a new user would have probably made me quit the site. If I don't see some kind of satisfactory response (to my flags) from the mods there, I'm raising hell on their Meta.
 
user41796
 
user41796
on a related note, it's interesting to see what a well placed comment from a high rep user can do to change the tone of a comment thread. I didn't save the link, but I ran across one where Eric Lippert effectively said "be nice, we all started as beginners" which elicited an apology from the somewhat ranting commenter.
 
user41796
and I'll call myself on the irony of my immediately prior comment and my paraphrasing of Shog9.
 
Yannis Rizos
Yannis Rizos
@gnat (and everyone else) Also, when a question is not horrible (but off topic, nc, narq), it might be a good idea to suggest chat as an alternative.
 
user41796
6:13 PM
@YannisRizos - that's a good point; especially since there have been a few questions from >20 rep users that were off-topic but good Whiteboard fodder.
 
user20683
Any mod or room owner can clear the <20s to talk
 
user20683
Firefox 21.0... Drunken Fox?
 
Yannis Rizos
Yannis Rizos
@WorldEngineer Yeah, but it's a hassle. We certainly can't do it for everyone, and having done it once, I'm not sure I want to do it ever again.
 
user20683
@YannisRizos I didn't say it was a good option, just an option.
 
Yannis Rizos
Yannis Rizos
Also, they have to create a chat parent user by themselves, and that's always tricky. If you are going to post 3-4 comments explaining what they need to do to get into chat, you can simply not do it and post 3-4 comments explaining why their question doesn't work instead. The lazy option is always the better option.
 
user41796
6:19 PM
@YannisRizos +1 just for The lazy option is always the better option.
 
Jimmy Hoffa
Jimmy Hoffa
@gnat that NARQ I flagged for fast-deletion because the guy posted it knowing it's not appropriate
 
user41796
@JimmyHoffa RE: FP No, I haven't delved into or lived functional programming. Shortest explanation is it simply hasn't bubbled up on my priority list in part because I lay claim to more hobbies than perhaps I should. As with many others, most of my technology development path has been pushed by my employer's needs. And until this gig, the opportunity to introduce new languages and technologies was outside of my purview.
 
user41796
Little to nothing I do requires being stamped or certified, so the overhead in switching to FP instead of C# as RAD is an impediment. And to be honest, the hardest part of my problem isn't the engineering it's the requirements elicitation. That's no different than any other domain I've worked.
 
user41796
but I'll admit it's an interesting tech that I wouldn't mind putting some time in with
 
user20683
@JimmyHoffa
 
user20683
6:24 PM
 
Jimmy Hoffa
Jimmy Hoffa
yeah, I just think you'd appreciate it since you have appreciation for formality and provability. In reality half the reason engineering licensure will never truly come to pass though is because as you pointed out; the real problems in our industry aren't ones rigor/formality can solve, they're problems with mismanagement, bad requirements, bad process, etc etc. Only the best company's have problems only with the engineering.
 
gnat
gnat
@YannisRizos no that's what I asked from @GlenH7
3 hours ago, by gnat
@GlenH7 ah I see. i would appreciate if you post the links you mentioned, I'd like to take a deeper look at my mistakes (it never hurts to learn, y'know:)
 
user41796
@JimmyHoffa I agree to an extent. In actuality, what an engineer is doing is saying "this design conforms to our understanding of what reasonable and solid design principles require so the final product will meet or exceed reasonable expectations."
 
Jimmy Hoffa
Jimmy Hoffa
@WorldEngineer can you mods see the edits which aren't marked because they're before the 5 minute window?
 
user20683
@JimmyHoffa probably, I've not checked
 
Jimmy Hoffa
Jimmy Hoffa
6:28 PM
-4
Q: Has any theory about software development and the human brain similar to mine been postulated before?

JubbatThis question will probably be closed, but I'm really keen on sharing a theory of my own on the relatively small number of outstanding developers as I'm genuinely curious to know if someone has suggested something remotely similar before. I had the realisation when I read this piece of news some...

development-process abstraction thinking
 
user41796
there's a nuance within that statement that says "it's as good as X so it will meet your service expectations." Software expectations are .... difficult to generate consensus around
 
Jimmy Hoffa
Jimmy Hoffa
I swear when I first looked at this the first sentence didn't have 'probably' in it
 
Yannis Rizos
Yannis Rizos
@JimmyHoffa If it was edited in the first 5 minutes, there wouldn't be a revision history.
 
gnat
gnat
@YannisRizos thanks I'll give it a try. Here's canned comment I drafted using it, feel free to trash it... :)
resource requests (every answer is equally valid) are not quite welcome at Programmers. As far as I understand, one would rather present an underlying problem instead - a problem that was intended to be solved with particular resource requested
 
Yannis Rizos
Yannis Rizos
@gnat I'm not going to trash it, I'm going to say you should try it out. Your other comment wasn't working, I spotted 2-3 askers responding negatively to it. Not saying that's your problem, but I'm assuming the whole point of posting such comments in the first place is that you want to help newer users understand what the problem is. If they repeatedly fail to understand it, try a different comment.
Remember, the target audience for such comments is people who in all likelihood have already failed to read the FAQ. You'll have to take that into consideration if you want your comments to make it through.
 
gnat
gnat
6:36 PM
@YannisRizos thanks, i'll give it a try
 
Yannis Rizos
Yannis Rizos
And for brand new users, failing to read the FAQ is not that bad (don't you dare quote me on that ;).
 
user41796
@YannisRizos ding! done.
 
thorsten müller
thorsten müller
Even if they read it, they often just fail to understand the finer details.
 
user41796
@gnat - how about "Instead of asking for a resource, please ask a question about the underlying problem that your resource request is trying to solve." after the "not welcome" sentence
 
thorsten müller
thorsten müller
maybe we could find a way to directly migrate towards reddit
 
user41796
6:41 PM
I'd vote for /dev/null instead. /dev/null's bandwidth is simply amazing
 
thorsten müller
thorsten müller
cross post by two different users:
3
Q: Re-architecting a classic inheritance design

bruiseruserI have the opportunity to rewrite a core piece of a project (C#) that is inheritance-heavy and feels increasingly restrictive in how it is designed. The scenario is pretty simple, imagine an application that is built to handle various small tasks. Every task is unique in it's functionality so it ...

c# design-patterns architecture
1
Q: Re-architecting a classic inheritance design

snappymcsnapI have the opportunity to rewrite a core piece of a project that is inheritance-heavy and feels increasingly restrictive in how it is designed. The scenario is pretty simple, imagine an application that is built to handle various small tasks. Every task is unique in it's functionality so it imple...

c# design-patterns architecture
and re-migration from SO
 
Yannis Rizos
Yannis Rizos
@thorstenmüller Same user, different login credentials.
 
thorsten müller
thorsten müller
6:57 PM
Does anybody know a Linux Terminal that accepts simple Ctrl-C/Ctrl-V instead of requiring Ctrl-Shift-C?
 
Yannis Rizos
Yannis Rizos
You can change the key bindings in gnome's default terminal.
24
Q: Making Ctrl+C copy text in gnome-terminal?

MehrdadIs it possible for me to make Ctrl+C perform a copy command if there is highlighted text in the terminal? Otherwise, it should retain its normal behavior. (If there is a terminal that can do this other than gnome-terminal, that's probably fine too.)

command-line gnome-terminal clipboard
 
thorsten müller
thorsten müller
I use KDE, but works just the same. thanks
 
Yannis Rizos
Yannis Rizos
@thorstenmüller KDE? I should suspend you for that.
 
thorsten müller
thorsten müller
Hey, Gnome sucks so much. And I think last time he was ranting Linus mentioned that he uses it now. (Because of the wobbly windows or something)
 
Yannis Rizos
Yannis Rizos
Yeah, Gnome sucks too, and it seems to be getting worse over time. I haven't used either for quite some time to be honest, all the linux boxes I use are either remote servers (hence, terminal) or virtual (xfce is the simpler choice there).
 
thorsten müller
thorsten müller
7:12 PM
I used xfce when it was very new and (then) was missing a lot. I went through quite some variations over many years. For some time I was playing around with the Qt libs, so using KDE was a quite natural choice. And ssh sessions on the command line on our servers anyway.
But for everyday work I just like to have a graphical web browser, pdf reader and stuff, so command line isn't really an option.
 
Yannis Rizos
Yannis Rizos
It's still missing a lot, but it's significantly lighter. I've been meaning to try LXDE, which is also supposed to be extra lightweight, but never found the time.
 
thorsten müller
thorsten müller
I do not really care about lightweight anymore. My smallest computer has 8G memory and its cores rarely go beyond 5%, so why worry how much resources my windows manager uses? the moment you start chrome or any other large tool (especially anything running in JVM) that's just history.
 
Yannis Rizos
Yannis Rizos
12 (or so) virtual boxes, and I kinda need at least 3-4 of them to run in parallel when testing new stuff. A stupid setup I inherited and I don't care much to change.
 
thorsten müller
thorsten müller
ok, vm is a good argument.
 
Yannis Rizos
Yannis Rizos
7:55 PM
Links to Meta and chat have gone awol (on all sites)
60
Q: Missing: "Meta" from top navigation. Reward offered

MattThe "meta" link has disappeared from the top navigation bar of Stack Overflow all sites in the network... I'm feeling quite lost at the moment. Additionally, I could create separate "Missing"/ "Wanted" posts for the FAQ, and Chat links as well, but I think you get the drift...

bug status-reproduced top-bar
 
 
1 hour later…
gnat
gnat
8:57 PM
@GlenH7 thanks, this is certainly worth trying. I recall similar change a while ago to one of my canned comments, it worked very well. Regarding slippery dupes, I bookmarked it for further study, will ping you after done

to study: slippery dupe votes

6 hours ago, 2 hours 56 minutes total – 31 messages, 5 users, 6 stars

Bookmarked 2 hours ago by gnat

 
psr
psr
9:37 PM
@JimmyHoffa I don't agree with Steve Yegge on one sort of important point. Metadata isn't documentation for the next developer, it's data for the next meta-program you want to write. I do agree that what's useful for such a meta-program often doesn't have a heck of a lot to do with the type system in whatever language you are using.
 
Jimmy Hoffa
Jimmy Hoffa
9:49 PM
> Haskell is specifically designed for researching Haskell
Jon Harrop is anti-everything-non-F#, but that still made me laugh
 
Jimmy Hoffa
Jimmy Hoffa
10:02 PM
@psr Keep playing with haskell, there's a reason hoogle let's you search for type signature: it's so expressive you really can intuit what the function is going to do just by the signature relatively frequently, what would you think this might do: Ord a => [a] -> [a]
@psr also in the case of languages that compile to IL, the resultant assembly is dripping with meta data just from the static type system themselves which you can write applications agains.. Actually I'm not sure I understand how you disagree with Yegge's point.. I read that he says comments are metadata, not that all metadata are documentation... unless I misread
it's accurate to say comments are metadata; there's programs that will compile all of your codes comments out of your source code into a referenceable form, treating them like meta data..
 
psr
psr
I think he said people often make sure things have a statically inferable type purely as documentation, when this adds noise. His examples are C# and Java, not an algebraic type system.
I would call comments documentation. I would only call it metadata if it's reasonably machine readable.
 
Jimmy Hoffa
Jimmy Hoffa
Ah yes, well I already know Yegge prefers dynamic languages, I don't fault him for it but I disagree
@psr compilers read your comments and appropriately forget them, so they're machine readable :P
He says haskell folks are insane
 
Jimmy Hoffa
Jimmy Hoffa
10:25 PM
According to Yegge:

Assembly language: Batshit liberal.

Perl, Ruby, PHP, shell-script: Extremist liberal.

JavaScript, Visual Basic, Lua: Hardcore liberal.

Python, Common Lisp, Smalltalk/Squeak: Liberal.

C, Objective-C, Scheme: Moderate-liberal.

C++, Java, C#, D, Go: Moderate-conservative.

Clojure, Erlang, Pascal: Conservative.

Scala, Ada, OCaml, Eiffel: Hardcore conservative.

Haskell, SML: Extremist conservative.
he's largely speaking in reference to type system, and he thinks the right side of the puzzle to be is Ruby, I mean he's heading a team coding in Ruby at google and thinks it's the best thing since sliced pajamas...
Other than that one thing; I largely agree with him on most all of his writing. His attempts at language research and blogging on it show something I think far more engineers should be doing
> In the conservative/liberal framework, language designers can make more accurate, less bait-and-switchy claims; for instance: "Haskell is the best choice for every radical extremist conservative programmer!"
 
user20683
Yegge says lots of things
 
user20683
many of them good, many of them also overblown in my view
 
Jimmy Hoffa
Jimmy Hoffa
True enough I suppose, if nothing else his writing makes ya think, and nobody can call him an idiot
 
user20683
@JimmyHoffa I'm just glad he's not Stallman
 
user20683
we only need one of him
 
user20683
10:35 PM
 
user20683
This made me laugh
 
Jimmy Hoffa
Jimmy Hoffa
That is awesome
I love that it's addressed from Palo Alto
 
user20683
That's where they are
 
Jimmy Hoffa
Jimmy Hoffa
I get that
 
user20683
I've been by their headquarters on a number of occasions
 
user20683
10:37 PM
just the gates
 
user20683
 
user20683
This one is <3
 
Jimmy Hoffa
Jimmy Hoffa
The concept of people in Palo Alto not knowing anything of computers is like bizarro universe
 
user20683
@JimmyHoffa There are a surprising number of biotech people there
 
user20683
or were 15 years ago anyway
 
user20683
10:39 PM
that might not be the case anymore
 
Jimmy Hoffa
Jimmy Hoffa
I can only suspect there's more biotech abound these days than previously
 
user20683
right but biotech is also more computerized now
 
user20683
less wetwork more bioinformatics
 
psr
psr
@JimmyHoffa Don't really fall much into either camp - on specific issues I might go either way.
 
Jimmy Hoffa
Jimmy Hoffa
@psr yeah, he over simplifies it a lot for sure. I strongly disagree with here he says type inference is a tool of static typing lovers, people who are nuts about static typing want every detail spelled out and nothing inferred
That's why a lot of .NET devs hate the var keyword
 
user20683
11:02 PM
@JimmyHoffa I have Dragon's Milk :)
 
Jimmy Hoffa
Jimmy Hoffa
Thought you didn't like beer? Unless I have no idea what you're talking about..
 
user41796
11:28 PM
@JimmyHoffa re: yegge post on statics & meta blah blah blah. I think he had some good points in there, but he also let the hyperbole carry things away. His evolution of thought regarding what level of the team to write code for is telling regarding his progress as a coder. My not-so-dirty secret about some of my comments are that I put them there to remind me what the heck I was thinking when I wrote that chunk cause I invested a fair degree of brain power to get there.
 
user41796
It's just like when I'm working in my wood shop - I make notes all the time along with my story stick so I can remember what the heck I meant to do. It's not useless "meta diarrhea" (okay, so now I'm hyperbolizing), it's valuable information for the next time I come back to it. I have no intent of keeping those decisions in active memory. The comment is there to queue the long-term storage recall and explain what was going on.
 
psr
psr
If you just used Haskell the 76 page type signature would document it automatically (to the point where a reasonably competent person could infer what it did and your DNA sequence).
 

« first day (987 days earlier)      last day (4002 days later) »