9

I am pretty sure I am stating a risky question here, and maybe it is the hot weather that makes me see things in a different way, but what is the main goal of what we are doing on AU?

The reason for asking is that -I imagine increasingly- AU is flooded with programming- related questions with a seemingly realistic value, but actually can be solved in one or two lines in any language. Nevertheless they generate a huge voting activity and a lot of answers, as if they were life saving issues.

Although I, as well as many others, benefit of it every now and then, I have the feeling these questions, even if it is bash, grep, awk or whatever language, should be asked elsewhere, unless they have at least some real practical value and are related to an existing problem of OP. It is a bit weird that maybe (just a guess) 50% of my reputation I earned in less than 5% of my spent time here, and mostly on answering questions, that, if we are honest, are about completely nothing.

  • 4
    I also often see programming questions in Ubuntu-related languages, but not-Ubuntu-related situations. As I also see this as a problem (although a minor one), I agree with you that we should at least close them more often and tell people to ask this on Stack Overflow. – Byte Commander Jul 2 '15 at 13:16
  • 2
    Your question relates to another one, meta.askubuntu.com/q/13807/295286 . Basically, my opinion is that all these languages are integral part of any linux distro, including Ubuntu, hence questions are on topic here. By the way, can you explain this part : real practical value and are related to an existing problem of OP ? For example, how would you alter an existing question to fit "practical and related to existing problem"? – Sergiy Kolodyazhnyy Jul 2 '15 at 13:47
  • @Serg I have to run, but of course they are on topic, as long as they are not meant as only solving or thinking over programming issues, without a (real) link to an existing problem. – Jacob Vlijm Jul 2 '15 at 13:52
  • 1
    @Serg what distinguishes on- or off topic is the fact that their goal is to solve a problem, other (or more) than a programmatic one. In the example I mentioned, there is no problem to OP whatsoever, and if there were one, OP wouldn' t need us to solve it for him, given his history. I am not pointing fingers to anyone, this happens a lot; questions with the only goal to explore a number of programmatic options to approach an imaginary problem. Don't get me wrong, my "personal happiness" is not moved in any direction if the majority on AU decides tomorrow to only post theoretical problems. – Jacob Vlijm Jul 3 '15 at 6:38
  • @Serg My opinion however is that the main goal on AU should not be to do what is done so much better elsewhere. Of course there is no strict fence, but what we're good at here is solve Ubuntu related questions. With the help of coding or not. Again: when questions about non-existent problems like this (while being a minority in number) generate a substantial amount of activity on this site, far bigger than many real problem, it seems a bit ridiculous to me. BTW, as said below kos' answer: – Jacob Vlijm Jul 3 '15 at 7:06
  • I want to make clear that it is not a personal thing to anyone: all of the people involved in the question are constructive members of the band with great skills. – Jacob Vlijm Jul 3 '15 at 7:07
  • It would be much easier to understand what you're talking about with a couple of examples. What are the questions you consider pointless? Anything involving core *nix tools like grep? Anything involving scripting languages like awk? Anything about the main Ubuntu shell bash? Surely you realize these tools are at the core of Linux, how could questions about them not be on topic? What are these questions with no "realistic value"? – terdon Jul 6 '15 at 8:06
  • 1
    @terdon The occasion was this question: askubuntu.com/questions/643520 I am pretty sure it was a made up problem, not to solve an existing issue, but just with the intention to create a broad collection of answers. That makes it a purely programmatic question, apart from the fact that this kind of questions is already taken care of extensively on AU. – Jacob Vlijm Jul 6 '15 at 9:18
  • 1
    @terdon Not so rarely, we even see initially "real" and simple questions explode into endless (programmatic) discussions, which seem irrelevant here if the problem can be solved either way in one or two lines. At the same time, questions in which the shown skills can be applied in a comprehensive way receive hardly any attention. As mentioned below Fabby's answer, usually this is just an observation, not something that bugs me. Probably time for a holiday. – Jacob Vlijm Jul 6 '15 at 9:18
  • I don't understand. That's a perfectly good Q&A. It showcases various approaches and useful tools using a simple problem that can very easily be modified to fit different needs. That is not a programming related question at all! It is asking about file manipulation, specifically renaming, on a Linux system. What programming? The closest you come to "programming" in any of the answers is a shell loop which, of course, is completely on topic since it is one of the most basic tools of a *nix admin. I really don't understand what the problem is. – terdon Jul 6 '15 at 9:24
  • @terdon Obviously my understanding of "programming" is different. I do think a showcase of options on a fictional situation is actually a programming question, as said, apart from the fact that the subject is kind of exhausted here. The question was meant however to get an impression on the general idea on AU. I have an opinion, but I have no problem with whatever that general idea may be. – Jacob Vlijm Jul 6 '15 at 10:15
  • I completely agree. I have not been around here for a very long time, haven't asked a single question, and I have still been able to provide quite a few relevent and accepted answers even though I'm not using Ubuntu! This means that the answers to these questions could be found through a quick Google search or other Linux knowledge/man pages! Thus meaning these questions shouldn't be asked at all or should be asked on another site. – nixpower Jul 8 '15 at 12:44
  • Have you got an example of a question you think is a programming related question with a seemingly realistic value? I'm struggling to imagine what you mean currently – matt freake Jul 15 '15 at 12:14
7

I'm afraid your question is too vague for me to understand what class of questions you are referring to. I will therefore base this answer on the information you gave in your comment where you pointed out this Q&A as the one that prompted this discussion. You considered it:

[...] a made up problem, not to solve an existing issue, but just with the intention to create a broad collection of answers. That makes it a purely programmatic question [...]

Well, anything involving computers could be considered a programming question since computers are useless with no software running on them. However, basic system administration tasks like file renaming or other manipulation are right at the core of this site's topic. How could they not be?

I think you are seeing this as programming simply because you tend to use python for everything so, of course, you end up writing programs. There's nothing wrong with that but because you're so good at python, I get the feeling you're not really aware of the general purpose POSIX tools that all Linux flavors have by default.

Tools like grep (which is not a language, by the way) and rename do not require programming knowledge to run them. The fact that you would approach an issue with a python script does not mean that the question was about programming.

The question you highlight is a case in point. It is a broad question asking about way of removing an extra character from many file names. That is a problem that i) could indeed be shared by future users; ii) can be used to highlight the abilities of the various tools making the answers easily adaptable to other problems.

I think that's where your pythonista view is clouding the issue. A python (or Perl, for that matter) script written to solve a particular problem is almost always "giving a man a fish". A solution using the core utilities available on any *nix system, one that highlights what these utilities can do, is usually "teaching a man to fish".

In other words, an answer showing a python or perl script that can rename files can be fun but it is unlikely to help someone who doesn't know how to hack in those languages already. An answer using grep and regular expressions opens a whole world of possibilities that require no programming knowledge whatsoever except, perhaps, a working knowledge of regex.

More to the point, where would we be without these questions? Personally, I have no interest participating in a site whose questions are 90% "no wirlesses helps plase" and "black screen no X server". I find them uninteresting, repetitive and boring. They don't challenge me to find an answer. if I can give one, I either know it already or just need to spend the 5 minutes googling that the OP was too lazy to dedicate.

As for actual problems, yes, that's part of the site's mandate but not its entirety. The SE sites also aim to become a "repository of knowledge". For example, my post on U&L about replacing strings has received 168930 views at the time of writing and more than a hundred votes. There was no problem to solve there, however it serves to highlight certain commonly available tools and the things you can do with them, it serves as an answer to many similar questions. It's a reference. That is a Good Thing©.

So, in conclusion, no, the site is not only for "real" problems. It is also for imagined ones which, however, bear a close resemblance to everyday problems users might face. If you don't like these questions, just ignore them. As you can see, many of us like them and welcome them. Work on others if you don't. There's room for all of us! :)

  • Thank you for this extensive answer. It is a lot of information, and a well argumented one, which doesn't mean I agree. What I do agree on is the fact that only answering "this or that doesn't work, help me" -type of questions would be boring on a live- threatning level. I also looked into my own history to see what was my own "behaviour" on this. This one: askubuntu.com/questions/576604 made me think for a while I was doing exactly the same as OP in the mentioned occasion, but I don't think it is the case. – Jacob Vlijm Jul 6 '15 at 16:33
  • My question was motivated by a problem; a problem I ran into repeatedly when answering questions on window placement etc. For me, that is exactly the key. At the same time, I cannot guarantee I never broke the rule in answering questions myself. Another thing I do agree on is the fact that I might produce too many "ready to use" answers, without explaining what actually happens. Especially when writing in a "compressed" style, I can imagine there is not much to understand if you are not into python. – Jacob Vlijm Jul 6 '15 at 16:34
  • Funny, by the way, is that I started to minimize the code since my activities here. At the same time, an explanation on answers like this: askubuntu.com/a/631467/72216 would be too extended to post on AU. As far as awk, grep, sed and so, I agree completely those are powerful tools and valuable to have knowledge on. The fact they I have poor knowledge on those actually bugs me. – Jacob Vlijm Jul 6 '15 at 16:35
  • We've had this discussion before on other occasions, but while these tools are extremely powerfull, and fast in many situations, I really disagree to the statement that you do not need any (programming) knowledge to use them. Well, strictly maybe no programming knowledge, but the linked question proves their use is not too easy. In python (sorry) this would take me less then ten seconds to write a solution, test it, maybe fix a typo and it works under all circumstances. – Jacob Vlijm Jul 6 '15 at 16:35
  • Look at the answers on the linked question and their corresponding discussions; while I really highly value the posters and their skils, it seems to have a lot of pitfalls. I think your own level of experience on *nix tools, and your talent on patterns and comprehensive structures might be giving you a false impression on the difficulty and readablity of those tools. No need to say that I appreciate the fact that you have taken the time to write this answer – Jacob Vlijm Jul 6 '15 at 16:35
  • @JacobVlijm yes, it would take you seconds. It takes me seconds to do the same thing with grep or sed or bash. You seem to think that python is somehow simple and easy to understand but that's because you know it. I'm sure I do the same with the other tools, yes, but basic grep or cut or any other coreutil is very simple. They are the very blocks from which Linux is built. Try man cut or fold or paste. They tend to be tiny, simple manuals. They are not programming or even programming related. They have no variables, no structures, no loops. They're simple C tools. – terdon Jul 6 '15 at 16:41
  • @JacobVlijm As I've told you before, I very rarely vote on your answers because I usually have no idea how or if they work. It's all import something from somewhere and significant whitespace :P. Those are programming answers but, as you know, I don't usually see any need to bring in external scripting languages when there are tools specifically designed for a task. So yes, those things are (or should be) extremely welcome here since they are the very backbone of *nix systems. – terdon Jul 6 '15 at 16:44
  • If you want to provide Python solutions for them, that's great, but it doesn't make them programming questions. Just because I can use a car to drive a nail into the wall does not make the car a hammer, if you see what I mean. – terdon Jul 6 '15 at 16:44
  • IMO the fact that the (no doubt lower level) tools, discussed in the linked questions are more familiar around here does not prevent the fact that the difference, between those tools and a language like python, is not an absolute one. Both need a syntax and an understanding how to use., both need to be interpreted. The fact that (a language like) python is not very commonly used here and people do not easily understand the code, is a property of (e.g.) you and me, not so much of the discussed solution(s). – Jacob Vlijm Jul 6 '15 at 18:06
  • I am perfectly fine with that by the way, as I mentioned on earlier occasions. A fact is however that python is highly integrated in Ubuntu, but that is kind of off-topic. – Jacob Vlijm Jul 6 '15 at 18:07
  • @JacobVlijm sorry, but no. They don't both need to be interpreted (none of the coreutils tools do) and they mostly work with flags. The only exception is Ubuntu's default rename which takes PCREs. And yes, python is integrated but so is C and so is Perl. That's not really relevant. The question is what the objective is. File manipulation is a basic task of any OS so it must be on topic here. – terdon Jul 6 '15 at 18:08
  • That surprises me: en.wikipedia.org/wiki/AWK – Jacob Vlijm Jul 6 '15 at 18:09
  • @JacobVlijm note that I didn't mention awk :) It is i) a scripting language, just like python and ii) not part of coreutils. Run apt-cache show coreutils to see what that package provides. They are all compiled C programs. – terdon Jul 6 '15 at 18:11
  • @JacobVlijm again, I specifically avoided mentioning either awk or sed since they are indeed scripting languages. My point was more about the coreutils tools and that questions about file management need to be on topic. It would be absurd to make them otherwise. – terdon Jul 6 '15 at 18:13
8

What's your objection to them? That they get too much attention or that they seem too abstract? Who decides what "practical value" means to a user?

They generate so much activity because of the type of problem they are. You can't debug a kernel module or resize partitions nineteen different ways, but I don't see that as a fault of the problem, it's just programmers showing off as they are wont to do.

And that's a reason why having this sort of question around is probably a good thing. It brings people in. I've seen people learn things in one question and apply them to help another user. The site's value and ability to attract users is based on our wide range of topics and fervour to answer them.

I'll agree that the lasting value of some of the text-processing questions "convert this format to that format" can seem pretty small but again, who are we to say whether or not it's a real problem or that it won't help anybody in the future? Every problem was original once.

Ultimately I don't think these questions are a problem. I know this changes, but only one of the current newest fifty posts are one of "these" questions. If that becomes one-in-ten or higher, we might have a scope issue.

  • 1
    @JacobVlijm But you have to understand, not everyone is a developer. Everyone starts off somewhere. In some cases even, system administrators usually have no idea how to code (especially ones coming from a Windows background - I'm not bashing Windows or anything, but unfortunately this was the case with quite a few sysadmins I've met) . What I'm trying to say, what seems to be an easy/dumb issue for you, can be a hardcore issue for others. In the case of replacing the dots, I can understand how someone can have issues with such a case, specially since . in regex means any character. – Dan Jul 2 '15 at 16:10
  • 2
    @Dan to be honest, I totally have no problem with problems of inexperienced people. In cases like this, often there is no problem, and OP here obviously needs no help if it were. It is simply an invitation to throw all possible (programming) options into answers. That is done much better elsewhere. – Jacob Vlijm Jul 2 '15 at 21:16
4

I think I can answer this one, because I'm an extensive user of the questions you're talking about. Point by point:

[...] AU is flooded with programming- related questions with a seemingly realistic value, but actually can be solved in one or two lines in any language.

I.e.:

What's the value of such questions?

  • For the asker: the most an answer can give him, since it solves its problem;
  • For the answerer: the most the answerer can learn from answering it; personally, I've learned a lot just in trying to answer those kind of questions;
  • For the others: usually nothing, aside from what they are willing (if they are willing) to learn from a given answer, unless they have the exact same problem, unless they somehow manage to adapt an existing answer to their specific problem or unless the problem is wide-spread enough. So, overall, I totally agree with the fact that such questions are usually so narrowed and specific they won't help the majority of the users experiencing a similiar problem;

Nevertheless they generate a huge voting activity and a lot of answers, as if it were live saving issues.

I agree with this. Being honest I have to say I've earned the most of my reputation from those, and some were really excessively upvoted, while I gave a couple of way more interesting answers, one here and one on Unix & Linux which, at least IMO, would have deserved it way more. But shrug, people upvote the answers to questions they are interested in that they can deem as helpful, not the one they are not sure or have no idea about (this as a general rule). In any case, going to the point, questions should be upvoted only when well written and useful. Answers should be upvoted regardless, because this is necessary to identify the best answers and that's how the system works. Personally I agree with the fact that being their scope so limited, unless really covering frequent problems of text-processing such as, say, removing duplicate lines from a file, such questions should not be upvoted.

TL;DR

  • In 99% of the cases, such questions are useful to the posters, to the answerers and to a very low number of other users;
  • They indeed generate lots of reputation to their askers and to the answerers; the problem is that these kind of questions shouldn't be upvoted if not really of interest; good answers, on the other hand, should be upvoted regardless, because this is necessary to identify the best answers; I agree that the amount of reputation generated is excessive, however this is due to how the system works and due to the high number of users partecipating in such questions.

So, in a nutshell: I think those questions should be kept, mainly because they are useful to the askers; I won't deny they somehow introduce a reputation problem. However I also think that this is not solvable. People are interested in them and vote on them (me included), what are we going to do? Setting those questions as off-topic would indeed solve the reputation problem but mainly would decrease, in my opinion inacceptably, the value of the community for the askers.

  • Hi Kos, thank you for your answer, and I want to clear out one thing, my second comment on Oli's answer was not very nicely formulated, and I definitely do not want to point to anyone personally. On the contrary, all of the answerers there, I consider very constructive members of the band, and a good chance I would have posted an answer in reflex if an original one would have popped up in my mind. Looking from a distance however, especially in this case, it is not about a real problem (OP seems not the one who needs any help on that), but just an invitation to explore the options. – Jacob Vlijm Jul 2 '15 at 20:56
  • ...And there ar much better places to do that, it makes AU a pale shade of what is done on a much higher level elsewhere. – Jacob Vlijm Jul 2 '15 at 20:57
  • Again: 22 votes and 7 answers on how to replace two dots by one? It must be my question, generating the traffic on the question :) – Jacob Vlijm Jul 2 '15 at 21:18
  • @JacobVlijm No worries, I understood since from the start that you weren't addressing anyone personally, but that you were bringing up what you think in your opinion are question only with (quoting you) "a seemingly realistic value"; now on the question you're addressing explicitly I agree also on the fact that if it doesn't correspond to a real problem OP had than it's (almost?) plain useless. That's the kind of thing that shouldn't happen, altough I'll be honest, I didn't even notice this before answering. But that doesn't seem (at least to me) the general case [cont] – kos Jul 2 '15 at 21:25
  • 2
    [cont] I've seen lots of questions from the same users who clearly still don't (or don't mind to) have the right tools to answer their question themselves, but I didn't notice a particular pattern of question-answers on that particular topic. If that's the case, then surely I disagree with this particular behavior. In all the other cases, I'd be prone to accept them, again, aware of the fact that they generate ridiculous spikes of reputations for both the askers and answerers. – kos Jul 2 '15 at 21:31
  • @JacobVlijm However again, more explicitly: in general, if OP doesn't really have the problem subject of the question than I agree with you and the question should better vanish per its unusefulness. If he does, than to me that's fine. I wanted to make this clear in a stand-alone comment for future readers. – kos Jul 2 '15 at 21:57
1

The main goal of Free/Libre Open Source Software is to advance the human race without any intellectual encumbrants. AU is to allow you to contribute to the advancement of FLOSS, therefore you're just helping to advance the human race... :-)

Whether it's getting a server back on-line, finally convincing someone to actually take back-ups or convert a text file meaningless to you but used to sort data on retro-viruses finding the ultimate cure to cancer, doesn't really matter, does it?

Tip: When you hit 20K and you're an established user, take a small break from answering questions, go lie on a beach somewhere, read a good book and stay away from the site for a week or so. Even top-level guys like you need a break every now and then and it seems you need a break! ;-) :P

  • 1
    Hi Fabby, thank you for taking the effort to write an answer. Be sure there isn't much here that I think is meaningless. On the contrary, I still remember the days that I read about $PATH a lot, but was too embarrassed to ask what the peanut butter it meant :). Simple, but well asked questions are the ones I love (too). The objection I had was on a question that was purely aiming at programming, with no "external" goal whatsoever. We see quite a lot of those lately, and I still think this is not the place for that. – Jacob Vlijm Jul 5 '15 at 21:30
  • 1
    But..... usually I don't care at all, I am just having fun in my own formula here, and looking back on my comments (which I removed), I wasn't too nice, while I really like the people involved. Time for a holiday? probably :). The good news is that my holiday just arrived. Thanks, have a great summer, and see you around! – Jacob Vlijm Jul 5 '15 at 21:45
  • 1
    Enjoy! Don't log on! And when you're back, drop by the chat room! ;-) – Fabby Jul 5 '15 at 21:49
  • I will! Don't log on at all I can't promise, but it will be much less :) – Jacob Vlijm Jul 5 '15 at 21:53
  • 1
    +1 for the break after 20K ;) – Sylvain Pineau Jul 9 '15 at 11:46
1

Well, this forum is for asking questions and getting them solved. As long as they are related to Ubuntu, then asking those questions is ok. because AU is about learning and getting to know Ubuntu much better with the help of other people from around the world.

0

Ask Ubuntu is about solving any and all Ubuntu-related problems, rather, Ask Ubuntu is about generating documentation for Ubuntu in Q&A format collaboratively.

What's the purpose of documentation? Obviously, it's only there to help users solve their problems, but there are some problems that documentation is simply not the solution to.

(Programming can benefit from documentation, for example, but something like vim vs emacs battle can't be resolved through documentation.)

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .