4

Instead of creating an external blog to replace the Stack Exchange blog, why don't we just leverage the normal Q&A to do this?

Here's how it would work:

  1. Write your blog post in any ordinary editor.
  2. Split your post into a question & answer pair or title/abstract/body combination (Abstract goes into question, body goes into answer).
  3. Answer your own question.
  4. Flag the post and ask a moderator to add a "Wiki Lock" to it (not Community Wiki). The Wiki Lock allows anyone to collaboratively edit the post (if they have editing privileges), but prevents other answers from being posted.

I know what you're thinking: "This won't work, because reasons." Here's why I think it will work, and why it will work better than a blog:

  1. The community owns it (with first-authorship preserved).
  2. There is already some precedent: it's called Canonical Q&A. We've already done that here at least once, with great effect.
  3. Canonical posts benefit from all of the facilities that are already provided by the Q&A machinery, such as full visibility, voting, collaborative editing and the ability to comment.

But what about closing?

In order to prevent closure, your blog post would have to be high enough quality to overcome objections like "Too Broad." Not everyone is going to rise to this level of quality, but that's why it's a Canonical post and not ordinary Q&A.

The usual "on-topic" rules would have to be relaxed a little bit. Obviously they would still have to pertain in some way to Software Engineering. But let's be honest: most questions are not closed because they are off-topic; they're closed because they're just not all that interesting, and "off-topic" provides a good excuse.

On Stack Overflow, canonical questions have allowed people on many occasions to post useful, blog-like material that extends well beyond the usual parade of code troubleshooting questions. Similarly, we could use this mechanism to provide those folks who have an itch with an outlet to create better content than the usual parade of pattern matching, name that thing and "am I doing it right" questions and answers that frequent our front page.

Further Reading
List of Canonical Questions on Stack Overflow
Encyclopedia Stack Exchange

  • if moderators are technically capable to migrate to main from meta this could possibly work even better by utilising meta as a sandbox for such collaborative Q&As. Wrt little relaxing of the rules for such questions, this seems to be in line with the way how collaborative lock was intended to work: " small handful of questions that are not exactly appropriate for Stack Overflow, are popular and have useful answers where the primary value arises from the community's regular efforts to keep them updated via edits..." – gnat Jan 12 '17 at 6:26
  • 2
    I'm not convinced that the on-topic rules would have to be relaxed. Maybe it's because it's early, but I'm having trouble coming up with a question that I would wiki lock that isn't on-topic. I'm also worried about the size limit in an answer of 30000 characters (~500 words), so we may need to summarize and defer details to third-parties. Note that I think that's characters in the text field and not visible characters in the answer, so markdown and URLs eat characters. I'd be willing to help out in this, and I think @gnat is onto something with using Meta to draft these questions. – Thomas Owens Jan 12 '17 at 10:21
  • @ThomasOwens do you by chance know if it's technically possible to have a collaborative lock with more than one answer? If yes, this can help address 30K char limit (may be a bit tricky to fit into Q&A format but still). All examples I could find so far had only one answer – gnat Jan 12 '17 at 10:29
  • 1
    @gnat I'm not sure. If it is, though, you would need to post the answer before it is wiki locked. And then you fall into the 0, 1, or infinity problem - if we allow 2 answers to exist, why not 3? 4? 10? And then we don't avoid the problem that it appears we are seeking to avoid. – Thomas Owens Jan 12 '17 at 10:31
  • 1
    with drafting at meta making as many answers as needed prior to locking will probably be easy @ThomasOwens. As for your note wrt problems related to allowing more than one answer I fully agree, that's exactly what I meant when I said "tricky to fit into Q&A format". Drafting at meta, again, would make these problems less painful compared to doing so at main site but nevertheless it is probably going to be difficult – gnat Jan 12 '17 at 10:37
  • 1
    I would agree - if we do this, I would hope they would be drafted on Meta. Using chat would be helpful, too. That should alleviate some of the pain and make sure it's close to canonical. – Thomas Owens Jan 12 '17 at 10:39
  • speaking of 30K char limit, I just figured that we can estimate chances for it to be a problem @ThomasOwens by simply checking the sizes of prior blog posts, does that make sense? – gnat Jan 12 '17 at 11:11
  • @gnat: It's a small problem. MichaelT's blog doesn't have any posts that are over 30000 characters AFAIK. – Robert Harvey Jan 12 '17 at 15:31
  • that's really great to hear. Taking into account that part of the blog post should go to question (so that we actually have more than 30K) it looks reasonably safe to assume that limit won't be an issue – gnat Jan 12 '17 at 15:35
  • 1
    This isn't normal Stack Exchange, but then again, Puzzling is also weird. – haykam Jan 16 '17 at 14:43
9

Absolutely not. This is neither good for the site, nor are Stack Exchange questions a very suitable medium for a blog. I would prefer us to use an external blog and to find other mechanisms to integrate it with this site.

The primary reason why I think blog-answers are not a good fit here is that on-topic questions could just be posted as ordinary questions, no blog mechanism needed. But a blog would be great for stuff that doesn't fit the question-and-answer format: opinions, essays, and the like.

Your suggested mechanism for Q&A-blog posts muddies our site scope. We have historically had some difficulty communicating the scope of the site. Adding an escape hatch to our site scope like “everything is OK if you call it a blog post and get at least 5 upvotes” will be confusing.

  • It sets a precedent for content on the site that we don't want to see as ordinary questions. On a fairly regular basis new users ask “Why was my question closed? There are similar questions like this!” – which usually refers to a locked or closed question from the Not-Programming-Related era. Fresh blog posts would worsen this, especially as they would appear on the front page.

  • Please tell me wiki-locked questions don't appear in the hot network questions list. If they did, this would damage perception of our site scope considerably.

Your suggested mechanism also adds work for moderators. In particular, applying the lock requires prompt action by mods to prevent additional answers from being posted. That's probably manageable when coordinated in a chatroom.

However, this also puts the final decision what should and shouldn't be blog-protected into the hands of our moderation team. How should they decide? Should they only accept submissions by a select few? That won't go over well. Or should they apply their own judgement? When submissions are declined, how long do you guess we'll have to wait for the first “Mod Abuse!” meta questions?

As mentioned above, I think SE isn't a good blogging engine, any community considerations aside.

  • No full-text RSS feeds of blog posts/answers. The next best thing is a feed for questions of a tag.

  • Fewer and less visible sharing widgets (although that's good from a privacy perspective).

  • SE isn't very accessible. While I don't use assistive technologies, I do sometimes use text-extraction or read-later services like Pocket or Instapaper. On SE, they currently interpret only the question as main content; any comments and answers hidden as clutter (and the SE pages have a lot of navigation clutter). For a Q&A that's OK since participation like voting and answering are integral parts, for a lengthy blog post it is not.

  • Fairly restricted markdown only. No tables, definition lists, or figures.

  • Embedding content (tweets, slides, videos, pdfs, gists) isn't possible (although that's good from a privacy perspective).

  • Extremely restricted commenting. 500 chars isn't sufficient to tell someone on the internet they are wrong. Well, it's enough to tell them that they are wrong but not for an in-depth explanation.

Even though you may not consider these points an issue, I do not see how any of this makes SE a good choice for blog posts when compared with something like GitHub pages or a simple Wordpress site.

  • Your suggested mechanism for Q&A-blog posts muddies our site scope -- The scope is already muddied, so long as we have wiki-locked questions here. People already point at those questions all the time as evidence they can ask their crappy shopping questions here. – Robert Harvey Jan 12 '17 at 15:28
  • 2
    wiki locked questions do appear in HNQ but (and this is a very very big but) 1) amount of answers in these is limited (typically to one) which makes most of the negative effects of HNQ do away (in particular it won't stick for too long) and 2) if we first sandbox these on meta and delay migrating to main for a few days this could essentially guarantee that these won't even enter HNQ – gnat Jan 12 '17 at 15:40
  • I re-checked the points you make, look like solid justification against preparing this kind of posts at main site. How about giving it a try at meta? (assuming further migration to main if we find that it works okay) Also I wonder if the drawback of restricted commenting can be somehow balanced by ease of collaborative edits and an option to create chat rooms linked to question – gnat Jan 12 '17 at 17:31
  • @gnat: Feedback is not a problem. Since when are 600 character limits in comments an unreasonable restraint? – Robert Harvey Jan 12 '17 at 17:53
  • 1
    @RobertHarvey you think so because you are too used to the way how things work at SE. :) I am also too used to it but more thorough re-checking of the points made in this answer made me recall that in typical blog engines comments feature is indeed much richer than here (tree views, higher char limits, collapsing etc etc) – gnat Jan 12 '17 at 18:05
  • 2
    @gnat Good to know, so HNQ isn't a factor here. I'm still not a fan of this approach, and am surprised you of all people seem so enthusiastic about it :) I'd have no problem if the posts are created and stay published on Meta – after all, Meta is for all the discussiony stuff that has no place on the main site. That negates the possible scope-muddying risks (which I care about) but still keeps the “SE is a crappy blogging engine” problems (which I care a lot less about). But I'd like to hear what Morons has to say about this, since they seem to have the most interest in blogging here. – amon Jan 12 '17 at 18:15
  • 1
    I am not that much enthusiastic about hosting it here as desperate about finding the appropriate blogging platform after reading all the discussions related to this here at meta. With all the drawbacks you mention hosting here offers a benefit of reliably keeping content (as a 10Ker I would be able to get it even if deleted) – gnat Jan 12 '17 at 18:25
  • 2
    @gnat: Regardless of the platform chosen, they will all require hosting, and will all essentially have no visibility, save for a link from the chat room and this meta site. – Robert Harvey Jan 14 '17 at 16:51
  • 1
    @RobertHarvey We can look at other sites for inspiration – Worldbuilding uses events in the community bulletin to advertise blog posts via a persistent meta question, I think SF&F does something similar. There seems to have been a lot of existing discussion about this, Monica Cellio explains some results on MSE. I don't think getting visibility for an external blog will be a problem once its going steadily for some time. The difficult part is getting started but I see no sufficient initiative and am not interested myself. – amon Jan 21 '17 at 7:48

You must log in to answer this question.

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