11
votes

NOTICE:

Since this site is now in beta, it is time to have these conversations with the folks who are actually building this site. Please bring your comments and discussions to this site's 'meta' support forum so not to split these conversations. Thanks.

Proposal: Emacs

I see that @JoelSpolsky (founder of StackExchange) just committed to the Emacs.se proposal.

A few years ago @JoelSpolsky wrote a blog post called Merging Season in which he specifically argued that proposals for new sites should be rejected if they meet the following conditions:

  1. Almost all X questions are on-topic for site Y
  2. If Y already exists, it already has a tag for X, and nobody is complaining
  3. You’re not creating such a big group that you don’t have enough experts to answer all possible questions
  4. There’s a high probability that users of site Y would enjoy seeing the occasional question about X

emacs questions are on-topic for StackOverflow, StackOverflow already exists and already has a tag for emacs, this is a smaller group than StackOverflow and many of StackOverflow's users are also emacs users and enjoy seeing the emacs questions (even if they aren't expert enough to answer them themselves.)

So the one thing that makes this proposal viable (according to the original blog post) is if the current StackOverflow emacs tag users are complaining. Are they? What are they complaining about? In this case, wouldn't it be better to fix the problem that users are complaining about than start a new site about a niche topic? Or is Joel now in favor of lots of tiny sites? If so, what changed?

locked by Robert Cartaino Sep 30 '14 at 18:25

This question exists because it has historical significance, but it is not considered a good, on-topic question for this site so please do not use it as evidence that you can ask similar questions here. This question and its answers are frozen and cannot be changed. See the help center for guidance on writing a good question.

Read more about locked posts here.

  • 2
    I don't understand the utility of exploring the psychology of JoelSpolsky's decision making process. – Edgar Aroutiounian Sep 15 '14 at 18:41
  • 3
    I didn't ask any psychology questions. The Merging Season blog post is the closest thing that StackExchange has ever given to "official guidance" on what they do and do not want to happen with respect to new site proposals. Joel seems to have just taken action that seems contrary to the spirit of his previous statement. Either something has changed in the policy, or I am misunderstanding the policy. This question asks for clarification on what the current policy is with respect to niche sites. – Wandering Logic Sep 15 '14 at 18:53
  • 1
    seems like only @JoelSpolsky (discuss.area51.stackexchange.com/users/2/joel-spolsky) could answer this one – warren Sep 15 '14 at 19:28
  • The real point of this question has nothing to do with JoelSpolsky, I think. The arguments cited (from JS) are what this question is about. And they are good arguments, IMO. What the question is really asking is why creating this site would help Emacs users. It provides good arguments (from JS apparently - but who cares?) why the proposed site is not needed. Given the lack of any good arguments (so far) showing why it is needed, YAGNI. – Drew Sep 20 '14 at 16:25
20
votes

I can't speak for Joel (I haven't spoken with him about this). But I can tell you what has changed my mind about this proposal.

Last week, I was ready to close this as a duplicate of Stack Overflow. This morning, I woke up to:

Emacs commitment

That's over 150 (currently 184 185) people who committed to a proposal over a single weekend. In answer to your question:

So the one thing that makes this proposal viable (according to the original blog post) is if the current Stack Overflow emacs tag users are complaining). Are they?

I just point to the graph above. And it's not as if these people have never thought to ask on Stack Overflow:

committers active in
50.0% Stack Overflow
15.2% Super User
8.7% TeX - LaTeX
8.2% Ask Ubuntu
8.2% Meta Stack Overflow
7.6% Unix & Linux

The first couple of pages of committers by reputation shows these are folks who are very comfortable with our network, but still want a site dedicated to Emacs. These are people who are voting with their feet. It seems very likely that our existing sites are not meeting their particular needs.

We talked over the prospects of an Emacs site this morning and concluded:

  1. If this had been language split (Ruby, Haskell, Lisp, etc.) we would have shut it down immediately. But Emacs sits in a fuzzy boundary between language, platform, software package, and lifestyle. (I'm only slightly exaggerating on the last point.) As such, Emacs questions (especially those that don't have much to do with programming) are underserved on our network.

  2. If the site mostly consists of Elisp questions or is duplicating content on Stack Overflow, we'll close it down in the private beta.

  3. We aren't interested in starting dedicated sites for each and every text editor out there. Emacs has a long-established community of users who have done some pretty amazing things. (Too bad they can't write a decent editor.) I'd be surprised if we got an Eclipse or TextMate site any time soon.

  • 1
    There's a Vi/Vim proposal, too, but that's much more in the Emacs family than in the TextMate family. – hairboat Sep 16 '14 at 0:22
  • 3
    @JonEricson, Just to see if I understand correctly, let me summarize what I think you just said: "We're all programmers who love the same thing, programming, but the tag system on SO never really worked, and certainly failed to scale to 3M users and 7K questions/day so instead of figuring out what went wrong with tags, and fixing it, we're giving up and splintering SO." Did I get that right? – Wandering Logic Sep 16 '14 at 11:21
  • 6
    I was going to weigh in but @JonEricson already said it better than I could have. I love big tent sites like Stack Overflow and emacs questions will still be welcome there but there's a completely different feel to a dedicated site organized around people who love a particular platform, especially a way-of-life platform like emacs... – Joel Spolsky Sep 16 '14 at 14:17
  • Popularity does not imply meaningfulness. – Raphael Sep 16 '14 at 15:48
  • 1
    @Raphael: Agreed. But in this particular case, popularity filled in a missing component: a signal that Stack Overflow, and Super User aren't meeting the need. – Jon Ericson Sep 16 '14 at 15:56
  • @Wandering Logic: That's mostly correct. I'd modify that a bit to say "the tag system on SO never really worked" to create cohesive communities. Tags work exceptionally well for their original purposes: folk taxonomy, filtering, search, etc. Yesterday we talked about various ways that an Emacs community might thrive on SO. All the things we came up with would require months of development. That's not to say we won't try them. Rather, there's no way we could implement anything in time to capture the momentum of this proposal. – Jon Ericson Sep 16 '14 at 16:02
  • 4
    @JonEricson Or that a couple (200, come on...) die-hard fans click on anything that says "emacs". (Beta will show.) – Raphael Sep 16 '14 at 16:06
  • @WanderingLogic This discussion should make you happy: if this is a problem on Stack Overflow as well there's still hope for new features that enable an OS community to thrive in niches across the network. – Raphael Sep 16 '14 at 16:07
  • 1
    @Raphael, I'm only hoping I'm completely wrong that this proposal is the beginning of the end for SO as each beloved platform breaks off into its own "successful" StackExchange site. – Wandering Logic Sep 16 '14 at 16:29
  • 2
    @WanderingLogic FWIW, I tend to enjoy the smaller sites more than the larger, so that scenario does not scare me as much. Stack Overflow will continue to have its place, I'm sure. – Raphael Sep 16 '14 at 17:22
  • 1
    IMHO, SO is too large. Splintering is necessary. If I don't get an answer within 30 minutes, it will never get an answer. Even with a bounty. There's just too much noise, even with filtering, to get good answers to my questions. It's only real use, as a programmer, is when looking up historical answers. – Flimzy Sep 19 '14 at 1:02
  • 1
    StackOverflow is the right site for elisp questions. SuperUser is the right site for Emacs-as-editor questions. Questions about Emacs-as-editor, to which the answer is elisp are unloved on both sites. – SF. Sep 19 '14 at 11:16
  • 1
    "These are people who are voting with their feet... our existing sites are not meeting their particular needs." That's a pretty bold interpretation of 185 people "committing" to a new site for Emacs questions. I almost "committed" myself, before I thought it over and asked myself, "What really is the problem, that this tries to solve?" If I saw any real answer to that, I might still "commit". So far, nothing but hot air. Your little graph is the best argument for the proposal? Showing that 185 people have "voted with their feet"? You are reading far too much into this. – Drew Sep 20 '14 at 16:43
  • @Drew: We are having a private beta precisely because I could be reading too much into this. If at the end of the private beta it's clear that this site does nothing that SO (plus SU and others) already do perfectly well, it will be shut down. If it demonstrates something lacking in our current sites, it will move on to a public beta. I personally committed so that I can find out if this site is warranted or not. (And because I have a few questions that might not get answered any other way.) – Jon Ericson Sep 21 '14 at 3:54
  • 1
    Yes, that all sounds reasonable. Still, I would have expected to see a few arguments supporting the beta itself, including evidence of a problem it is trying to solve. So far, I haven't seen any. Saying that people seem to want this is not really a great argument, IMO. We'll see. On the other hand, giving people a simple way to create a virtual site, which, say, implements a set of filters (e.g. [emacs], [elisp]) under the covers but gives them the impression they are on a dedicated "Emacs site" might be interesting (and useful beyond Emacs). And it would solve any "dispersion" problem. – Drew Sep 21 '14 at 6:27
5
votes

I don't think he renounced anything. I think he committed to participate in a site about a topic he likes. Don't read into it too much.

FWIW, the logic behind letting the Emacs proposal have a shot at life is that condition #1 may not be met. There are many Emacs questions that are off-topic for Stack Overflow, and if that's what this community wants to ask about, they will do very nicely here.

  • +1. Emacs is not only about Elisp, also about usage, and this seems not to belong to SO. – mbork Sep 15 '14 at 21:39
  • 1
    @mbork, I'm trying to understand this. The query "[emacs] -[elisp]" returns over 8800 open questions, only a small fraction of which seem to be about elisp. It looks to me as if usage of emacs (the programming tool) is considered on topic. – Wandering Logic Sep 15 '14 at 21:48
  • 2
    @Wandering Logic: I look at some of the recent questions in that set and some of them probably should have the [elisp] tag added. Many of the rest seem like they aren't really getting great answers if they get any at all. I don't know yet if there is a need for this site, but the time to find out is in private beta. – Jon Ericson Sep 15 '14 at 23:25
  • @JonEricson: Are there in fact lots of Emacs usage questions that are closed on SO because they are not about programming? That's not my impression. Usage questions on Emacs inevitably invite answers of all sorts, including Elisp answers - even when some non-Elisp answers are also helpful. AFAICT, Emacs usage questions are rarely closed on SO, and only a minority get moved to SuperUser. And that is as it should be, based on the nature of Emacs. Using Emacs is hardly separable from programming Elisp, and there certainly is no sense in trying to wall Elisp off. – Drew Sep 20 '14 at 16:34

You must log in to answer this question.

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