5

If someone comes along with a set of so specific requirements that there is currently no software fulfilling that needs, is it acceptable to answer

Currently no software fulfills that needs. If you can afford it you could hire someone developing that for you.

| |
8

No, it's not an answer IMO. We want each answer to recommend a product. If there isn't a solution for a problem, the question should stay unanswered until one is found.

| |
  • @juergend Heh, yeah. Although I'm not sure... – Undo Feb 5 '14 at 1:20
  • 2
    “There is no product” has to be a possible answer. It may be correct or not, but “we simply need to search further” is not a workable policy: sometimes there is nothing further. – Gilles 'SO- stop being evil' Feb 5 '14 at 1:25
  • @Gilles Yes, you're right (and I am wrong). I'll leave the answer here in favor of having two options. – Undo Feb 5 '14 at 1:27
  • An answer that says (authoritatively) "there is no such beast" seems reasonable to me, although I don't know how the answerer can be sure in most cases, or whether such an answer is stable. "Let someone develop it for you" is a pointless response; it an implicit answer to every question, so there is no point in saying it. – Ira Baxter Feb 5 '14 at 3:24
  • 1
    Let's nix the "belongs in a comment" suggestion. That simply is not what comments are for, and I would likely flag/remove such activity. – Robert Cartaino Feb 5 '14 at 14:37
  • @RobertCartaino Nixed. Wrote that late at night and wasn't thinking properly ;) – Undo Feb 5 '14 at 14:38
  • While I'm tending to follow the "No", I see a little contradiction in your justification: "we simply need to search further" ist not really acceptable – unless we see ourselves as "personalized search engines". I thought we should give recommendations. Can we truly recommend things we just googled? IMHO we should at least have some backing of knowledge on the "product" we recommend. But if we do not know any match, the question should stay unanswered to indicate that, and to encourage others to put a "real" answer (see Angelo's answer). – Izzy Feb 6 '14 at 0:28
  • @Izzy First, note that I've changed my mind on this already, and don't agree with the answer (it's just here for the sake of voters having two options). Second, does the edit look better? – Undo Feb 6 '14 at 0:31
  • Yepp (though my comment now looks a bit stupid, as the relation is lost ;) I'm not sure, though, whether to favor your approach or Angelo's. I tend to the latter: If there's no "perfect match" (yet), at least let's recommend the "closest match" if it makes sense (it makes no sense e.g. if it just covers 1 of 3 nice-to-haves, but none of the what-I-really-needs – you get my point. In that case, I'm 100% with you :) – Izzy Feb 6 '14 at 0:37
8

Sure. Quoting the ground rules:

You must also be prepared for answers that simply inform you that no single thing meets your needs, but goes on to recommend things that cover what you've identified to be the most important.

And if there is nothing that meets your needs, “write it” or “get someone to write it for you” may be the best answer.

Obviously, such answers need to be justified, just like positive answers.

Bad:

Use FooBarSoft, it's the bee's knees.

Good:

You can use FooBarSoft. Requirement #1 is met in that …. Requirement #2 is met in that …. To meet requirement #3, configure it in this way. Nice-to-have #4 is partially met: blah blah. Additionally, to do what you're trying to do, many people find this feature useful. A limitation of FooBarSoft is that it doesn't work in this rare circumstance.

Bad:

No. You'll have to write it yourself.

Good:

The de facto standard for requirements #1 and #2 is FooBarSoft, nothing else comes close. Requirement #3 has been requested (link to FooBarSoft's issue tracker); the author decided not to implement it because he had other priorities. He estimated the workload at 1 man-week. So you could adapt FooBarSoft yourself.

| |
  • 2
    -1 because a simple "does not exist answer" is just as likely to be wrong because the answerer didn't know about some obscure software; AND because a pure "doesn't exist" answer without backing it up is as awful as rightfully-much-maligned "here's a link to software without explaining how it fits". I will reverse the DV if you actually meant "'No' answer is OK but must be severely backed up to shy WHY you answred 'no'" and just omitted that for brevity; and add the clarification. – DVK Feb 6 '14 at 0:37
  • @DVK Of course a “does not exist” answer has to be backed up, just like a “this exists” answer has to be backed up. P.S. Dialog with you would be easier if you didn't start right off the bat with hostility — you give the impression that getting a downvote from you is the worst thing that can happen to anybody and that if I bow down to you you might condescend to grant me the utmost favor of deigning to pay attention to me. – Gilles 'SO- stop being evil' Feb 6 '14 at 10:20
  • short answer: blog.stackoverflow.com/2012/08/…. Quoting Shog9 (who quotes Help pages I think): "Above all, be honest. If you see misinformation, vote it down. Add comments indicating what, specifically, is wrong....". – DVK Feb 6 '14 at 11:56
  • ... As far as I'm concerned, my comment was the Most constructive of criticism, and a lot LESS hostile than a random unexplained downvote which means that I either think the post isn't worth improving or just don't care enough to help you improve it. BTW, you saw enough of my discussion on the topic on SFF to know that I consider that to be a desirable behavior for ALL users, so I am flabbergasted at how you can view it as an ego trip as opposed to trying to be constructive and giving your post the benefit of the doubt. Silent downvotes are hostile. Explaining what can be fixed is not. – DVK Feb 6 '14 at 11:59
  • However, if it's your preference, in the future I will omit the "how you can fix" from my criticism (if any) of your posts if you feel it will be more constructive to you. – DVK Feb 6 '14 at 12:03
  • @DVK On the contrary, what I ask is that you comment with the criticism. Leave out mentions of downvotes. – Gilles 'SO- stop being evil' Feb 6 '14 at 12:14
  • downvotes or upvotes? – DVK Feb 6 '14 at 12:23
  • @DVK Upvotes too. Mentioning votes in comment is very rarely productive. – Gilles 'SO- stop being evil' Feb 6 '14 at 12:26
  • @DVK, what's wrong with wrong answers? Voting Can Fix That! Imagine an answer like "I work in related area X and I've long been looking for something like what your question is requesting. A is the closest thing to a recommendation but blah blah blah. Currently there is nothing that meets all of your requirements." followed by a comment or answer "B, just release, now does all of the wonderful things you require!". Bam! The-dynamic-nature-of-an-SE-site plus voting-to-reinforce-the-introduction-of-new-information to the rescue! – Kenny Evitt Feb 12 '14 at 17:50
5

Very much related to this, where I suggest a very simple guideline - don't answer unless you can point to something that exists prior to your answer being written.

That said .. if you know something is open source, or can be modified with some kind of plug-in, and a combinations of plugins or known modifications could suit the criteria laid out by the question author, and the question author indicates that they're open to recommendations that entail a little develop-y stuff - fine. But we're getting more into granular case-by-case criteria there.

There's nothing wrong with saying:

I'm confident that what you want doesn't exist exactly how you laid it out, but [xyz] matches most of your criteria, and if used in conjunction with [abc], most of your needs would be met.

That's a far cry from you probably need to consult a developer to have something written to suit your need though.

Just make sure that answers point to something immediately actionable by the question author, and I think we're good.

| |
3

No. It could be that it exists and you don't know it. It could be that it does not exist yet and the answer will be different in the future.

If a user asks a question and despite some views doesn't get an answer (maybe even after a bounty) he may just conclude that if nobody here knows a tool, there maybe isn't one.

Answers in the form of "I don't know a specific tool that does all of what you need but this tool here can do most of it and it can be improved to do all of it with a developer" should be good. Even more so if they are the only answer.

| |
  • Other than the "No, go and have it developed" (which I cannot accept in this context), this is something I can share: Finding the closest match, even if it's not perfect. – Izzy Feb 6 '14 at 0:33
3

Is it acceptable for the answer to be summarized by "no, no such software"?

Yes.


Is it acceptable to answer literally with Currently no software fulfills that needs. If you can afford it you could hire someone developing that for you with nothing else in the answer?

NO!!!.

That answer is impossible to evaluate for correctness. You didn't provide any EVIDENCE that there's no such software. May be you don't know what you're talking about and randomly guessed. May be you just didn't hear about some useful software due to limited experience. May be you tried every single relevant software and found 100% wanting. By reading that threadbare answer, we can't tell which of those cases you represent.

As noted here, any good answer should be backed up.

A good "No such thing" answer should include:

  • List your qualifications for why you are familiar with this type of software. Maybe you have same needs as the OP and seeked similar software. Maybe you are a developer of similar software and competitor market research is your professional responsibility. Maybe you read trade publication dedicated to such software.

  • List software that you personally researched and evaluated that is in this area; and which features of each disqualify it from OP's needs.

| |
  • +1 for: Its a good answer if it is throughly backed up – Angelo Fuchs Feb 6 '14 at 9:49

You must log in to answer this question.

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