(I know that there are approximately 10000 posts here with essentially the same topic. I wanted to explain my position and to put forward some ideas about changing the rules. If nothing happens as a result, it doesn't really matter.)
It is well-known that this website, Retrocomputing StackExchange, has a broader scope than its name might imply. For example, questions about 1940s computer hardware are perfectly on topic.
As such, this site has become not just a place for questions about (say) old game consoles or personal computers, but also a place for questions about computer history in general. Posts on the meta show that the majority of users are OK with this; after all, no other StackExchange site really fits for history questions.
I recently saw this question, asking about the reason for Markdown's wide adoption. It was closed a few hours after posting for being off topic.
As it stands, the question is perhaps not answerable objectively. I argue, however, that it should be considered on topic.
Obviously I can't contradict the fact that there is a fixed, well-defined boundary for what is on topic, based on the age of the post's subject. Going only by the rules, the close votes can't be contested. My claim is that the rules are at fault.
StackExchange has no site, other than Retrocomputing, oriented specifically towards computing history. The information for Stack Overflow's [history] tag says
DO NOT USE THIS TAG. Questions about the history of programming are off topic.
Questions about history are scattered about other sites, such as Unix and Software Engineering, but they are in the minority.
The question about Markdown is a inarguably a question about computing history. One user, in response to the suggestion to close it, commented
… Retrocomputing is de-facto also a computing history site, and as far as its subject goes, this question would be a fine history question. But I don’t think it is answerable with objective evidence.
And a moderator responded by saying
Yesterday is history. This is a site about ancient history – long enough ago that we think we need only two digits to specify the year.
First of all, that's a hilarious way to put it. But more importantly, the statement contradicts the idea that computing history questions are on topic. (At least that's what I'm claiming.) Once again, I agree that the question would be off topic for a site exclusively about “retro computing”, but that isn't what this site is.
Here we have a question clearly about history, on a forum that is widely agreed to be the de facto Stack Exchange site where computing history questions can be asked. The question is, however, considered off topic. This seems illogical. If a question about history doesn't belong here, then where else should it go?
Therefore I suggest that the rules should change somewhat, in order that all questions about computing history can be allowed. The possibilities I have thought of are the following.
Restrict the time frame of “pragmatic” questions only, imposing no such bounds for history questions. A pragmatic question is one that might belong on Stack Overflow, usually having the form “How can I do this?”. On the other hand, a history question is one that has the history tag (applied correctly). I can't find any questions that seem to belong to both categories at the moment, but they doubtless exist. When the pragmatic part of an ambiguous question is off topic, that part should be factored out, leaving only the historical part.
History questions could be split into two categories, by creating a new tag recent-history. This tag is applied to questions about the history of things that wouldn't be considered retro by the site's definition. The downside of this approach, which makes it practically intractable, is that the definition of “retro” changes often, so that some recent-history questions become valid history questions.
As a possibly better alternative to the previous idea, add post-2000-history, dividing history questions based on whether the subject precedes the year 2000. The time frame requirements for retrocomputing do not apply to post-2000-history questions. It would be possible to use any cutoff point if 2000 isn't deemed good idea, as long as the cutoff point remains constant.
(Impossible; I know this can't happen.) Split this site into two. One remains Retrocomputing, and the other becomes Computing History or Computer History. The time frame requirements stick around on Retrocomputing, but go away for Computing History.
A less extreme, though still impossible, alternative to the previous is simply to change the name of this site to Computing History. The problem is that such a change would cause most of the pragmatic questions to become off topic, and it would make references to Retrocomputing become hopelessly confusing for those who are unaware of the change. (I'm not saying that the name change itself is impossible, but that it would never happen. StackExchange site names have changed many times before.)
By far the most reasonable choice is the first one, since it doesn't invalidate existing tags.
It is ironic that, even if any of these changes were made, the Markdown question would probably still not be acceptable, for the same reasons why questions like What cause the downfall of Pascal? get closed. People might answer “Because Markdown is convenient and user-friendly”, or “Because it's easier to implement Markdown than to implement a more sophisticated markup language”, or “Because [some popular website] was using Markdown”.
To fix it, the question could be changed, for example, into
When did Markdown start becoming the typical “simple” markup language?
Even though the answer might be hard to pin down, at least the kind of answer expected has been explicitly stated.