Can I ask the pros and cons between 2 programs on SR? Example: I have used Open Office for the longest time for my free word processor. I am interested in trying LibreOffice. Why should I change? Why should I not change?
No, this is not a valid question.
Why you should switch or not can only you decide. And if you seek a recommendation you need to be very specific what you need from that really big software solution and why you consider switching in the first place.
Do you have requirements your current software does not fulfill? Name them. Be specific.
The problem with this type of question is commensurate on the number of possible differences. If program 1 is only marginally different than program 2, then the question isn't by definition too broad. If one could spend all day enumerating the differences between the two, then you do venture into the 'too broad' territory quickly when it comes to our format.
When questions are very narrowly scoped, this is less of a problem. Let's pretend that you're looking at two different word processing programs, and you really care about:
- How grammar and spelling auto correction behave
- Keyboard shortcuts
- Storing and exploring previous versions of documents
Then you could ask a question not unlike any other on the site and say what you ultimately hope to get as far as experience using the software, identify the two things that you're considering and indicate that you'd be eager to hear from folks that have perhaps used both in recommendations. But, be open to possibilities that you hadn't considered as well.
After all, you never know when some clever person at Microsoft is going to dust off edlin and see that it has a comeback. (I kid, I kid, but you get the point).
After seeing this trial question: no.
The principle of this site is to have recommendation questions. There are two important parts to a recommendation: a user story (what task you want to accomplish), and a requirement list (what needs to be there, what should preferably be there).
X vs Y leads to answers that each tack on their user story (if they're well-written) and proceed to analyze what requirements can be met by each of X and Y. So the result (if the answers are good) is a collection of user stories, with no particular answer being more or less appropriate for the question.
If the question specifies a user story — as IntelliJ IDEA vs Eclipse vs NetBeans for "pure Java" development does, albeit in a very weak form, then it becomes closer to our core business. But I think that the preselection of two answers detracts from making a strong question. The user story is very weak and the feature list is by definition non-existent. This makes for a question that is too vague.
I think this could be turned into a good question by strengthening the user story and listing at least some requirements. Then, mention that X and Y both meet the core requirements, and that:
- You have some additional requirements and wonder which of X or Y comes closest to meeting them all.
- You're open to alternatives if someone thinks that Z is a better match than X and Y.
Given that, I'm unconvinced whether X and Y need to be mentioned at all. It looks like the question already contains more of the answer than it should.
Depends on whether we want to cover "Software Comparisons" as well. Sure, most of our questions contain that as parts ("I need a software for X. Already tried a and b, but because [...] I need something different", or "Alternative.TO", both naturally compare things). But that shouldn't be all it is about.
IMHO there should be a focus on the task behind it. And the options on how to do it. Taking the task out of the equation doesn't make much sense to me. Looks a bit like "too localized" (for those who remember it).
I don't agree with the answer by juergen d
People don't know what they want until they know what they can choose.
Look at this question: C++ IDE with rich features; code completion, refactoring, etc. In an amicable way, every question about IDE should start with this list of features: autocompletion, refactorings, etc. Answerers mention their own favourite tools and it's score mainly depend on popularity of the mentioned tools.
But such question is not very useful for someone really choosing C++ IDE. OK, didn't you know that all mainstream IDEs support syntax highlighing, VCS, refactorings? Furthermore, most people aren't ready to choose anything except 2-3 most popular options. For example, Visual Studio and Qt Qreator, if we are talking about C++ IDEs. Or LibreOffice and MS Office, if we are talking about desktop office suite for Windows.
Considering all of this, I think it would be great to approve questions like "Visual Studio vs Qt Qreator" or "MS Office vs LibreOffice vs OpenOffice in Windows". Answerers would list key differences between examined tools. Examples:
VS supports integration with MSSQL
Qt Qreator is better for cross-platform GUI programming, because ...
[answer 2 years later] Now Qt Qreator has brand new really cool feature X, but VS does not
It would be never too late to answer such question.
Most interesting or comprehensive answers are upvoted.
This question becomes a brilliant resource for those who google "visual studio vs qt qreator".
(On @NicolasRaoul 's request, I'm posting my comment as an answer:)
I believe that "tool X vs tool Y" questions should be accepted, even though right now they aren't. My reasoning:
- When you want to ask "What tool should I use for task ABC?", you often do enough preliminary research to narrow it down to a few options, and a comparison is what you need answered in order to complete an answer to a valid question. It would be artificial to force you to ask the question you no longer have.
- More "meat" for the website, in terms of attracting users capable of answering such questions, who might also answer other, straight-up-recommendation questions.
- No other place on SE to ask this, nor a good alternative venue outside SE (AFAIK anyway).
- Answers to such questions can always refer to a third tool which settles the dilemma differently - further blurring the difference from straight-up-recommendation questions.
- Other SE sites tend to exhibit some flexibility regarding the margins of their area of scope for questions.
- There can be guidelines in place for these questions, e.g. "focus on the task, not just on the tools", like @Izzy suggests - instead of just banning them outright.
Yes, this is a valid question, as long as it allows people to describe the benefits and, possibly, drawbacks of each solution (or just one) in an answer, as long as the final decision lies with the questioner (who should be told that they have to judge by themselves and that the answers are just guidelines and point out features or things like that).