Markdown Drama Part III - The Revenge of the CommonMark
-
Ok, so this is a move in the right direction. Still unclear how that's gonna influence Discourse's hybrid markdown implementation, though.
-
Very good question. Especially with the thousands of posts in the dozens (or, if you believe Jeff, 2k+) of forums running Discourse. Changing the markdown behavior now would break the appearance of many posts.
-
And Jeff cares because?
-
TRWTF: They've been working on this spec for like two years. Meaning Discourse is probably already using whatever spec they came up with. Meaning this is probably as good as it gets.
-
I saw him tweet about this earlier, and was laughing for a solid half hour.
-
Why are you following Jeff on Twitter?
-
Because I check Twitter once a year. I was only checking it earlier today to see if Facebook was truly down.
-
You survived the great outage?
(I didn't even notice.)
-
It was only about 20 minutes.
-
I only know it went down because someone posted images of fake t-shirts - "I survived the 9/3/14 outage"
-
I mostly know since I was expecting a phone call shortly after it happened, and was planning on just screwing around until the call.
-
Note that the Markdown Standard calls for a line starting with "10." to be converted to an <OL start="10">.
In other words, fuck you @riking.
-
Soooo...
Does anybody know if the Discourse instance at http://talk.standardmarkdown.com/ uses the standard markdown syntax, or Discourse's own bastardized implementation?
-
It's a Discourse install, what do you think?
-
Use <pre>
10. @blakeyrat
-
It's a Discourse install, what do you think?
In hindsight, I maybe should have added
Filed under: rhetorical question
Filed under: I'll answer your rhetorical question: "no"
-
You can't answer a 'x or y' question with no. It's going to use Discourse's own bastardisation because ain't no way Jeff and co would build a unique version for that site. Or if they did it wouldn't follow that spec anyway.
-
"no" made sense when I wrote that. Dunno why.
...
On the other hand, it does make sense when taken completely out of context:
It's a Discourse install, what do you think?
No.
Or perhaps more like.
(I'm fairly certain that I stopped making any sense at all at this point. Time to get some sleep.)
-
Facebook was down? Excuse me while I badly fake giving a shit :)
-
Note that the Markdown Standard calls for a line starting with "10." to be converted to an <OL start="10">.
Yup, I spotted that too
http://jgm.github.io/stmd/spec.html#example-163
I can't wait for them to fix it!
Filed under: waiting in anticipation
-
Oh look, they even have their own bugtracker:
-
Heresy! THIS is a bug tracker
-
And Jeff cares because?
Oh, he cares:
The blue J avatar is jgm.
No, seriously, this is good, because it gives us something he cares about to get the list behavior fixed.
-
Note that the Markdown Standard calls for a line starting with "10." to be converted to an <OL start="10">.
In other words, fuck you @riking.
https://meta.discourse.org/t/ordered-list-numbering-wrong/19619?u=boomzilla
-
Blockquotes and paragraphs inside list items.... Code tags inside pre tags... Pedantic dickweedery about the number of spaces needed to continue a list...
This is almost front-page material. We just need a racial stereotype and a new hire and the job of HTML-not-invented-here is done.
-
from Jeff's blog:
Markdown is indeed everywhere. And that's a good thing. But it needs to be sane, parseable, and standard.
It's good that it's everywhere while also being insane, unparseable, and unstandardized?
Aren't those really bad qualities to have in something that's ubiquitous?
-
-
And it doesn't matter of you put newlines between the items to try and force it to start a new list
-
If you add one linebreak between just two adjacent items, all the items get an extra space in between them.
If you add two linebreaks, it creates a new list with spacing in between the two and the new list starts at whatever number the item after the double break is numbered....
...uuuuuuuuuggggggggggggggg
EDIT: And try numbering them descending.
-
Blockquotes and paragraphs inside list items.... Code tags inside pre tags... Pedantic dickweedery about the number of spaces needed to continue a list...
This is almost front-page material. We just need a racial stereotype and a new hire and the job of HTML-not-invented-here is done.
That's really the only question, how come we haven't gotten DC there yet?
-
It would be difficult to get such a story past ol' buddy ol' chum Mr Papadimoulis. However, I'd love to hear an audiobook rendition of the epic tale, Jeff vs TDWTF, with 'Make it work' replaced with 'You're doing it wrong'.
-
You write it, I'll read it with my awesome English voice.
-
So the bug report was closed.
WTF does that even mean? No comment, just "STFU". I was looking at it when it got closed, it was like a door slamming in my face.
-
Welcome to the world of Interacting With Discourse Developers. We've all had the door slam at least once.
-
For whatever reason that just pissed me the fuck off this morning. Fuck them and fuck their shitty excuse for customer relations.
-
Dig deep into the Bugs category and you'll find this done to me like 4 times... hang on I'll look up my favorite example... http://what.thedailywtf.com/t/times-in-user-activity-are-way-off/560
The "resolution" was: "I personally don't use that page, so it's not a bug. CLOSED."
It still appears to be exactly as broken as when I reported it all those months back.
-
At least your bug report / bug follow up is still around. At least, until Jeff decides to 'clean house' and delete old bug reports.
-
I probably wouldn't have closed the topic but to be fair I already answered this twice, even linking to show how the majority of MD implementations work and our current goals for MD standardization + a hint of our long term stuff.
Yes, the behaviour is weird and it's being fixed in the future. When are we going to stmd? Who knows, it's way too early now.
-
"Nobody else does it" isn't a fucking resolution. "We'll maybe do it someday but we don't know when" isn't a resolution either. If you guys had a real fucking bug tracker you could mark it deferred and move on but this bullshit just feels like you're blowing us off altogether.
-
I also have an awesome English voice.
-
Yes, the behaviour is weird and it's being fixed in the future.
How do you know? You just made the topic that was tracking this issue look like it's no longer an issue?
-
I replied as a new topic, because that closing seemed really strange: CLOSED_WILLFIX
RAW POST FOLLOWS:
Continuing the discussion from Ordered list numbering wrong:
This topic is now closed. New replies are no longer allowed.
The linked topic was a bug report (of an issue that causes a lot of grief to users). One developer said that the bug was going to be fixed in the future (timeframe left vague). It's my understanding that the "Bug" category here is the canonical issue tracking system for Discourse, but now an known bug that's planned to be fixed has had the topic that could be used to track it closed, meaning that it looks like something that isn't an open bug any more.
Is there a better way to report this? What's the thinking on tracking issues like this slated for future releases? It seems like there should be a way to do that. Things I've used in other places are tags with specific versions / milestones, but I don't see that anything of the sort has been done here (or that there's a reliable way to do it, in any case).
-
Apparently, the markdown stuff is an external dependency for discourse. It seems like @eviltrout should have been aware of this, especially since he's on the list of contributors for the project.
I don't have a github account, and I bit the bullet WRT meta.d. Who wants to go report this bug?
-
http://www.reddit.com/r/programming/comments/2fdsan/standard_flavored_markdown/
The discussion on reddit about this is a bit more serious, as would you expect. The top-rated comment:
I was ecstatic, but now I'm deeply disappointed: WHERE IS THE FORMAL GRAMMAR?!!111one
To which the answer was, "sorry we can't grammer".
It's impossible to write a EBNF-like grammar for Markdown because of Markdown's requirement that anything should be a valid input.
-
It's impossible to write a EBNF-like grammar for Markdown because of Markdown's requirement that anything should be a valid input.
I am pretty sure that "accept everything" is a context-free language (and hence representable by an EBNF grammar).
-
Not to worry, the pedants at reddit have it covered:
It's impossible to write a EBNF-like grammar for Markdown because of Markdown's requirement that anything should be a valid input.
any_input = . any_input | epsilon
Now that's out of the way…
-
It's impossible to write a EBNF-like grammar for Markdown because of Markdown's requirement that anything should be a valid input and produce indeterministic results.
Also because it's really damn impossible given the actual rules that govern Markdown. Technically, maybe, but it's about as possible as implementing Windows on a long roll of tape, a register, and a magnetized needle.
-
Also because it's really damn impossible given the actual rules that govern Markdown. Technically, maybe, but it's about as possible as implementing Windows on a long roll of tape, a register, and a magnetized needle.
Still sounds better than parsing with regexes.
-
Someone suggested building a state machine specification, like HTML5 has.
That sounds reasonable to me, actually.
-
Well, this was a minor technical / meta-y thing. But it just got way more interesting.
It seems the original author of Markdown threw a hissyfit that things seems to be proceeding without his input. So now Jeff & Co are shutting down standardmarkdown.com and trying to rebrand it into "Common Markdown". You don't get a release more botched than this.
They really should have gotten this guy on board before trying to standardize anything. Instead, now they are trying to appease him by backing out of this whole big "WE ARE CHANGING THE WORLD" launch and trying to rebrand themselves as "just another markdown spec, guys, seriously, not a big fuss". But his royal highness is pissed and ignoring further emails.
Ugh. What a waste of time. This Gruber guy is obviously a whiny idiot. He had a good idea with markdown, but then through his inactivity, he let things deteriorate to the current state of markdown interpretation wild west. IMO they should have told him to get lost and just took over the leadership of the standard. Instead they pussied out and probably lost their shot at fixing things.