R. A. Dehi @radehi@qoto.org

@allenholub In a Bayesian sense there's no such thing as a completely unknown unknown because there's no uniform probability distribution over the whole real line or the whole positive real line. You probably don't know the volume of Lake Superior (to take an example from earlier today) but probably know is more than a liter and less than a googol liters. An unknown is always just a question of the probability distribution.

I agree that narrowing the scope of stories to the minimal possible scope is very valuable.

@allenholub I agree that is more extreme than most cases in practice, but is an extreme version of a universal problem in programming: we never *really* know what's hard and what isn't until we're done. Our Bayesian belief about completion time has a lognormal distribution with usually enormous variance, sometimes until the very moment of completion.

Often, though, a spike provides much information to reduce that variance, and is often valuable.

Thanks to some experiments with @ambihelical, we've characterized the particular way #Qoto's Markdown support is broken with respect to code blocks.

Its Markdown mode is exactly backwards with respect to newlines: are ignored inside code blocks, but inside normal paragraphs newlines are treated as `<br>`. But another bug in Markdown support partly cancels this out if the code block is at the beginning of the post, because then, treats the code block as a normal paragraph.

This suggests a straightforward, if inconvenient, workaround: instead of using a code block, use a normal paragraph, but wrap each line within the paragraph in `` ` ``. Unfortunately, due to a third bug in Markdown support, this doesn't actually work; the first line is fine, but the typewriter text on subsequent lines incorrectly fails to be recognized:

`def main():`
` print('hello')`
` print('world')`

A fourth bug means that newlines after the first one are ignored, perhaps because are incorrectly parsed as being part of another code block; I think in a case like this you don't have that problem:

def main():
print('hello')
print('world')

@jeffjarvis Can probably get by with less defensive measures than that, just because the background level of aggression here is so much lower than on Twitter. Except among recent arrivals from Twitter, but hopefully they'll acculturate soon enough when they realize their aggression isn't a winning strategy here.

@ambihelical @freemo See, there was a newline after the colon, but got eaten. Inspecting the page DOM shows that there's a `<pre>` and a somewhat unnecessary but harmless `<code>` inside it, but the actual text node in the DOM has no newline in it.

@ambihelical @freemo By "plain text mode" mean "not Markdown mode"? I don't have a problem with prepending four spaces to each line (that's standard Markdown, fenced code blocks are a common extension) but that didn't keep the buggy implementation from eating my newlines.

Let's try four spaces:

def main():
pass

@allenholub We always do at least subconscious estimates because there are always more likely improvements than we have the time to make.

Consider [this now-obsolete extreme example from xkcd](https://xkcd.com/1425/). Wouldn't have been reasonable to equally prioritize working on both of these possible features; an afternoon or a week spent on the bird-recognition feature would have been only wasted.

In less extreme cases, an hour or two of estimation work can tell us whether a feature is more likely to take two days or a month. Many features that are worth trading off two days of other features for are not worth trading off a month of other features for.

Unfortunately, almost no companies have incentives set up to reward doing that estimation work, so formal estimates add little more value than subconscious ones.

@allenholub Can extract the principal components of the user/like matrix so that after a few clicks you can project a new user into the low-dimensional likability subspace and re-sort by statistically predicted likability *to you*

@AccordionGuy It showed up in a typewriter font, though? In theory, this version of Mastodon supports Markdown, but it's pretty broken.

@mmasnick Actually on Qoto (my instance) and on Mathstodon we can use LaTeX with Mathjax: \[\mathbb R, \mathcal R\], and maybe even \[\mathscr R, \mathrsfs R\]. But on #Qoto Mathjax is kind of broken, and when the post gets syndicated to other instances it just shows up as raw LaTeX code unless they install Mathjax too.

Everyone ought to, because if people are discussing things without being able to use equations, there's really no hope of coming to agreement on basic facts, and Mathjax makes it enormously easier to use equations.

(Edited: I screwed up the Mathjax delimiters! Possibly because of Markdown actually.)

@AccordionGuy the formatting in that post looks broken to you too, right? It's not only broken on my instance but in the actual post? Haven't yet figured out how to view the JSON-LD source.

@mmasnick Yes, although I imagine screenreaders will have exactly the same amount of trouble. But if your instance supports it (am using Mastodon v3.2.1) you can also format in Markdown: *italics*, **bold**, _**bold italics**_, `typewriter`, [links](https://www.techdirt.com/), _**[`bold italic typewriter links`](https://www.techdirt.com/)**_. Pleroma has a similar feature. For Fraktur, blackboard bold, and script you still need to resort to Unicode.

------

Actually am not sure how much of that is going to work because the "Markdown" implementation in Mastodon is super janky and has a lot of problems. Let's see!

@enoch @mmasnick I just wish they'd put all the Twitter crap behind a birdsite CW so we can see the interesting things they say without having to wade through all their complicated feelings about their abusive ex

@shriramk @rmerriam Which languages would you think mainstream programmers (JavaScript I guess) would learn most from?

@science_quotes@c.im I think this one is particularly informative, BTW; try doing a quantitative estimate, with 95% confidence bounds, in terms of units you can visualize, then look up the answer.

@freemo why is the formatting in this post broken? The pseudo-Markdown implementation seems to be eating all the newlines. How does that happen and how can we fix it?