Given the following question:
Is programming using QISKit for IBM Q experience (QX) on-topic?
Considering the fact that the questions about Q# are on topic as per Are questions about Q# on topic?
$\begingroup$
$\endgroup$
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
18
Programming and using quantum computers should definitely be considered 'on topic'. It's what the field is all about.
It might mean that we get lots of questions regarding technical issues for certain quantum SDKs, which might be not so interesting for some people. But those technical issues concern actually running programs on quantum computers. There needs to be a stack exchange for them. Is this not it?
-
2$\begingroup$ There is a stack exchange for them. Stack overflow. It's about a simulation library, only tangentially related to QC $\endgroup$– ItamarG3Commented Mar 21, 2018 at 15:45
-
1$\begingroup$ Btw. Drupal.SE allows question which asks PHP programming within Drupal framework. Same with programming in APEX on Salesforce.SE. $\endgroup$– kenorbCommented Mar 21, 2018 at 15:47
-
1$\begingroup$ @ItamarG3 The question that inspired this meta question is about runs on real IBM devices. Where is the reference to a simulator? $\endgroup$ Commented Mar 21, 2018 at 15:55
-
$\begingroup$ @kenorb But both working with Drupal and in PHP can be considered tasks for a programmer. I think allowing such questions means you ask programming questions to scientists, while you should ask them to programmers instead $\endgroup$ Commented Mar 21, 2018 at 16:21
-
$\begingroup$ Some more remarks: "Programming" (...) "It's what the field is all about." Not nessecarily. I don't like to accuse you, but quite often, programmers are rather ignorant about what CS really is. CS isn't about programming. It is about the science that allows you to program, among others. These are very fundamentally different. (Although some ancient people say 'programming' is CS, 'programming' has a very different usage here than the colloquial one) $\endgroup$ Commented Mar 21, 2018 at 16:30
-
3$\begingroup$ There's one exception here: 'programming' a quantum computer is such a highly non-trivial task that writing a non-trivial 'quantum program' can be considered a task fitting for a scientist. $\endgroup$ Commented Mar 21, 2018 at 16:30
-
4$\begingroup$ @Discretelizard That's true. But is this to be specifically a quantum computer science SE? Hasn't there already been a decision that quantum programming is on topic (regarding Q#)? I would have thought that precedent would apply equally to QISKit, especially since QISKit has actual quantum hardware. $\endgroup$ Commented Mar 21, 2018 at 16:37
-
1$\begingroup$ @ItamarG3 Your first comment here is essentially saying that the linked question would be on topic on SO. This is true. However, by itself, that does not mean that it isn't on topic here $\endgroup$– Mithrandir24601 ModCommented Mar 21, 2018 at 22:16
-
$\begingroup$ @JamesWootton That is a good question to ask. This example is a good example to let the community decide what it wants. I'm afraid the programmers that are amateurs in the field of QC outnumber the experts currently, so popular vote might not be ideal, but it is the best we've got. I can tell you that I very much would appreciate it if all 'pure programming, devoid of science' would be limited to a few well-constructed tags, so that I and other uninterested ones can avoid it. Personally, I'm here for the science. $\endgroup$ Commented Mar 21, 2018 at 22:29
-
$\begingroup$ But well, community should vote and all. I'm willing to accept consensus if needed $\endgroup$ Commented Mar 21, 2018 at 22:29
-
$\begingroup$ I do think that this hasn't been properly answered at the Q# question, as this case it is clear to me that it has nothing to do with quantum computing but a hint on how to use a service to do quantum things. (so, while turning on my computer is needed to actually program, I wouldn't ask on SO how to turn on my computer). The Q# question considers actual 'quantum programming', which I think makes it valuable to remain. Additionally, I fear that leading experts will shun this site if it gets flooded by questions that are basically programming. But well, this is a democracy, I guess. $\endgroup$ Commented Mar 21, 2018 at 22:32
-
$\begingroup$ I do hope people know what is at stake here and think well before voting. That is all I ask. $\endgroup$ Commented Mar 21, 2018 at 22:33
-
$\begingroup$ @Discretelizard I would say to trust Cartano's advices on this one. Let's not be too strict at this stage and see what happens. There is always time to change policy later if needed. We don't have that much traffic to deal with for now anyway. $\endgroup$– glS ModCommented Mar 21, 2018 at 22:46
-
1$\begingroup$ I agree with this answer. If the question relates to a quantum computing topic/frameworks, I'd rather error on the side of inclusiveness rather than force StackOverflow. The added context helps. $\endgroup$ Commented Mar 22, 2018 at 16:54
-
4$\begingroup$ I completely agree with this answer: questions about programming and using QISKit for IBM Q experience should on-topic. And I'm really puzzled why there is even such a debate about this under this answer. It would be very-very unnatural to make this off-topic.... $\endgroup$ Commented Mar 23, 2018 at 0:27
$\begingroup$
$\endgroup$
9
No, this particular question asks why a library for python that just happens to be an interface for a QC service (i.e., 'compute on my QC as a service') gives some error. This is a question that is purely about programming, not about QC. Perhaps the community disagrees with me here, but I do think we have to be strict here, as else we might get flooded by all sorts of questions that should be on stack overflow
The reason why having Q# on-topic, but this question not can be consistent (although perhaps initially confusing!) is that 'programming' a quantum computer is very much on-topic. The latter is often quite non-trivial and requires extensive knowledge of the inner workings of quantum computers!
Note that if you asked instead how to construct some quantum algorithm using your library, this would be on-topic, I think.
Forgive me for sounding harsh, but I think the best actions would be to
- Delete the question. Migration is impossible, so just delete here and ask on SO.
- Option 1: kill the tag python and burn it very hard.
- Option 2: keep the tag python, but mention in the tag usage that if you're using this tag, you are probably asking an off-topic question (although not necessarily. Perhaps someone would ask whether certain language features of python can be used on quantum computers or something like that)
Add a tag for the specific library you have used in python, see if you can ask a question related to the current that is about creating a quantum 'program'. (This has already been done here, but might be useful in the future)
So, in summary, I think that the attitude of this site should be the same as CS.SE to regular 'programming questions', but be accepting towards clearly quantum programming questions (or quantum algorithm questions)
A bit of motivation on why we wouldn't want this. Look around you. Do you see questions about programming? Are this the important experts of the site? I don't think so. Most experts here are scientists of some sort, and only programmers by accident. So, to help you with programming problems, I think you should ask programmers, on StackOverflow.
-
1$\begingroup$ I am pretty sure the question was for runs on real devices, rather than simulators. $\endgroup$ Commented Mar 21, 2018 at 15:59
-
$\begingroup$ Ah okay, this is basically Quantum computer as a service. Will update. $\endgroup$ Commented Mar 21, 2018 at 16:17
-
1$\begingroup$ The problem I see in this reasoning is that basic "quantum programming" questions will likely almost always be just about the programming, not the physics behind it. I don't think we can consistently enforce a distinction between "quantum programming" questions asking about the physics behind the code and those just asking about the code. The way I see it, this question is on-topic if and only if other Q# questions are to be on-topic $\endgroup$– glS ModCommented Mar 21, 2018 at 16:49
-
1$\begingroup$ The original question needed to be answered by someone with experience of running jobs on QISKit, rather than specific python knowledge. That would be someone here or on the QISKit Slack, but I don't thing SO would be able to help. Regarding your 'look around you' comment, I and many other scientists use QISKit. I am not someone who answers questions on SO. In fact, I only learned Python to use QISKit and Project Q to run jobs on quantum devices. Nevertheless, I would have been happy to answer the question if the OP hadn't answered it already. $\endgroup$ Commented Mar 21, 2018 at 16:56
-
$\begingroup$ @glS I disagree with your general statement, but I agree with your sentiment that such questions should be treated with caution. Do make a post of your own if you feel you have another position. $\endgroup$ Commented Mar 21, 2018 at 22:40
-
$\begingroup$ @JamesWootton Yes, but not by someone with a PhD in Quantum Computing. I do agree that some (perhaps many) scientists like to use these tools but I nevertheless believe that the discussion of this isn't the place of this site, it is on SO and any discussion here is meaningless clutter at best. The lack of other resources does not excuse this. Nor do the motivations for learning to program. Note that I am not saying you shouldn't have asked this question, simply that you shouldn't have done so here. $\endgroup$ Commented Mar 21, 2018 at 22:43
-
$\begingroup$ Perhaps my 'look around you' comment was out of place, in hindsight, I should have said look at the questions around you, do they look like this? Why could that be? Do these questions really fit here, or do they stick out like a sore thumb? $\endgroup$ Commented Mar 21, 2018 at 22:45
-
$\begingroup$ @Discretelizard I’m afraid I don’t quite understand one of your comments. To what does ‘not by someone with a PhD in Quantum Computing‘ refer? And what question do you not think I should have asked here? $\endgroup$ Commented Mar 21, 2018 at 22:55
-
$\begingroup$ @JamesWootton I'm not referring to any question made by you, only the one mentioned at the OP here, sorry for the confusion. I mean that the question mentioned in the OP here can be answered by any programmer using the given tools, so 'no need for PhD in quantum' to answer the question. $\endgroup$ Commented Mar 22, 2018 at 8:08