TeX, LaTeX and Friends

Yossi Farjoun

4:14 PM

hey! we have a chat room!! OMG!!

2

Andrew Stacey

5:36 PM

Must admit, I'm highly amused by the description.

Yossi Farjoun

8:08 PM

yes, but what happens when someone posts a \end{chat}?

I'm quite amused by the synergy between english.SE and tex.SE....although no-one over there seems to care about that question anymore....

Seamus

The "double spacing at end of sentence" question is pushing up our "answers per question" ratio quite alot...

Yossi Farjoun

@seamus: LOL

I'm actually torn about whether answers/question is a good metric....

when I visit questions with 5 pages of answers.....well, lets just say that I definitely do not read every answer....not even those with non-negative score.

Seamus

Yeah I agree. If a question only gets one answer, but that answer solves the problem, and does it well, then that's valuable.

Yossi Farjoun

which is why that metric doesn't really measure anything interesting.

Martin Tapankov

@YossiFarjoun uhh, probably the world will end..

Seamus

8:21 PM

But it is at least an objective measure that is easy to implement...

Yossi Farjoun

@Martin: its that easy? hmmm. don't tell anyone.

Seamus

Do you think "average upvotes for accepted answer" would be a better metric?

Yossi Farjoun

@Seamus: yeeeesssss....not the most positive praise I've heard about something...

TH.

Seamus: not really. That's more a function of the number of users than anything else.

Martin Tapankov

@Seamus that would only promote vote inflation..

Yossi Farjoun

8:23 PM

why? we get presents when we have high metrics?

Martin Tapankov

We get to receive a shiny new design.. and that's about it.

Yossi Farjoun

well, for that no-one is going to change any metrics...and I don't think that there is any doubt TeX will launch...is there?

TH.

shrug

Martin Tapankov

@YossiFarjoun I think not -- we do tend to have an exceptional high quality questions, very rarely something unsuitable slips in.. But the real question is when it will launch officially, with the new policy it could linger in beta longer than GMail..

TH.

heh

Yossi Farjoun

8:26 PM

so what? do we have a candidate design waiting to be launched? the site is functional (as you said) and people seem to be using it as it is. what real difference does it make if it launches in a week or 3 months?

TH.

We get new mods?

Martin Tapankov

Well if it does linger seemingly forever, SE guys might decide that it's stagnant, and get rid of it,

But we're kinda in the middle of the pack, so I think we should be safe.

TH.

It doesn't really seem like something worth worrying about.

Seamus

Even if the site does get shut down, that's not a massive problem: we'd all just migrate back to StackOverflow...

We'd lose all our juicy points, but oh well...

Yossi Farjoun

oh no! I was planing on buying a trip around the world with rep points!

TH.

8:32 PM

This place is kinda fun, but I'm not sure it's really the best place to get help. The lag between question and answer is too variable.

At least for the people who need immediate help.

Yossi Farjoun

this place? you mean chat or the parent site?

TH.

The parent site.

Yossi Farjoun

actually, that makes me think of a much more important metric...I wonder if we can use the spiffy SQL engine to make a plot of that: average time until first answer.

or average time until first upvote on an answer.

TH.

That could be interesting.

But again, probably just a function of active users.

Yossi Farjoun

well...that's also important...

how do you measure active users?

Martin Tapankov

8:34 PM

Hmm, probably doesn't take so much time for somebody to upvote a sane question, but I've seen ones that have been waiting for answers for half a day

Yossi Farjoun

how active are they?

TH.

Especially the voting. I vote a lot (I might have had the most votes), but some people will seemingly vote up every question. Even the truly terrible ones.

Seamus

Yeah I asked a question about biblatex that's gone days without an answer...

Yossi Farjoun

ok...so here is a more complicated metric: average time until the answer that eventually gets accepted, is posted.

TH.

heh

Yossi Farjoun

8:35 PM

I want to see the plots!!!

TH.

I think you can design increasingly complicated metrics, but it's not clear exactly what they'll tell you.

Yossi Farjoun

I have no idea how to do it though...

Martin Tapankov

On a related note.. Can we get some sort of statistic about the geographical distribution of users? I'd guesstimate that most of us are from Europe, but still could be interesting to see.

Yossi Farjoun

true...

TH.

Why Europe?

Yossi Farjoun

8:36 PM

by the activity, I woudl actually say that most are from the usa...I always feel that in my afternoon everyone wakes up....

anyhow...going out for a beer. good to chat with you guys.

TH.

later

Martin Tapankov

@TH dunno, just an impression

TH.

Right. I'm just curious what that impression is based on.

Martin Tapankov

Well, from checking out user's profiles, and also on names (last is not a good metric, but usually gives a good idea)

But of course I could be wrong.

TH.

Ah, okay.

I'm a bit confused why (at least the first few) answers to the double space question keep guessing at how much space it is. Just look at the \sfcode for a period and find out!

Martin Tapankov

8:43 PM

TBH, I am actually surprised that this turned into such a debate -- for me, it's like an argument what's the right size of a full-stop in cmr10.

Seamus

And given that the space is variable if the text is justified, it's pretty meaningless...

TH.

Not meaningless. The space expands as a function of the sfcode.

The bigger the sfcode, the bigger the space relative to other spaces in the line.

Well, as a function of the spacefactor, really.

Seamus

Ah I see. So there is still a question of whether it should be relatively bigger...

TH.

Yeah.

TH.

9:21 PM

Well, that was really too much text to have written on the subject.

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends