« first day (324 days earlier)      last day (1187 days later) » 

8:07 AM
Latest data is now in the data explorer. Time to dust of those SQL queries and get exploring!
 
what do you explore @AndrewStacey?
 
9:12 AM
@JosephWright: With regard to tex.stackexchange.com/q/28972/86 it appears that in the notes, \item is still overlay-aware. Is this by design?
 
9:27 AM
@Raphink Can you see my "saved queries" at this link: data.stackexchange.com/users/6626/andrew-stacey ? At the moment, I haven't explored much, but I have Plans.
@egreg: To risk the war ... (tex.stackexchange.com/q/28836/86), I use the \coloneq(q?) quite a lot but I also back it up with words. When scanning a text, I don't always see every word but the \coloneqqqqq stands out and says "here's where it was defined". Then I read the surrounding text to reach enlightenment. So there's a place for symbols, but I agree that they should not replace the text.
 
@AndrewStacey For some time I had a good balance (unwillingly) between my questions and answers. I always had the same amount. Now I have more answers than questions. I should ask more :-)
 
@Raphink I'm not sure that equal numbers is a "good balance". Obviously, the "more expert" you are, the greater the shift you'll have to answers over questions. So I would expect some sort of curve that started high on the "questions" axis and headed towards the "answers" axis as the user's reputation increased.
 
@AndrewStacey A consistent use of \ceq (for brevity) would require it in many more places than it's actually used by \ceqaddicts. Using a symbol just for emphasis is wrong from a mathematical point of view (but there are worse things around, and this is not a capital sin). Completely different is its use in computer science in syntax rules, for example; or in formal logic as a metalinguistic symbol ("stands for" is what they say when defining a new abbreviation for simplifying formulas.
 
Yes @AndrewStacey. I started with only questions, then I had some long period with a balance, and now I have more answers than questions.
 
9:43 AM
@AndrewStacey Nice queries! It seems you've mastered the fine art of data-explorer-fu. :-)
 
9:54 AM
In tex.stackexchange.com/questions/10558/… I read about pplatex. Until now I wasn't able to make it run under osx. Does anyone have experience with pplatex?
 
@egreg Agree on all points. I'm certainly not setting myself up as an example to be copied, nor saying that my behaviour is exemplary in any form. Just trying to explain what I do. If there's a better way to achieve the same effect, I'd love to know. I do find myself wanting to distinguish visually between "equals by declaration" and "equals by dint of the huge five-page proof we've just slogged through".
 
Lightining 1.0 (stable) for Mozilla Thunderbird will be released in Sept 27th!
 
10:09 AM
@AndrewStacey That's the difference between well and badly written mathematics. I ended studying Latin many years ago, but I can read and understand Gauß's "Disquisitiones Arithmeticae" just because it's well written mathematics (and doesn't use \ceq). :)
 
@egreg That doesn't help! There's a difference between knowing that ones mathematical writing is (to be blunt) rubbish, and being able to do something about it. It's like teaching: no-one ever tells you how to do it, you're just expected to "pick it up" as you go along. (Yes, I do know of texts on writing mathematics, but they only give general principles and I'm very much an examples-person).
 
(Out of curiosity, what does \ceq stand for?)
 
@PauloCereda Shorthand - in this discussion - for \coloneq, aka a symbol meaning "is defined to be equal to".
 
@AndrewStacey Ah thanks.
 
@egreg I can relate a similar story: I stopped speaking French many years ago, but during my PhD had a copy of one of Grothendieck's texts permanently open and could - with one notable exception - understand what he was saying.
@egreg: On other things, thinking about redefining labels and references, on the first such question (tex.stackexchange.com/questions/28894/modify-eqref-command/…) the questioner asked me if there was anywhere else that the equation number would get put, other than the label of the equation and in a reference. I couldn't think of any, can you?
 
10:40 AM
@AndrewStacey I'm a bit afraid about the interaction with hyperref; but \eqref uses \ref and I don't think that hyperref hooks into \eqref as well. I don't see any use of \tagform@other than producing the equation number (or tag) and in \eqref.
 
11:23 AM
Likely to be closed?
0
Q: What is the simpliest way to make a documentation from code

Eugeny89suppose a file with code (*.cpp, *.php, or whatever). There's a lot of symbols like '$', '{', '}' what will definetely be disliked by TeX. I want to make a documentation from a code file. Maybe you heard of some templates for working with code? Thank you in advance!

 
 
2 hours later…
1:18 PM
Speaking of Stallman, now he's talking about Android: guardian.co.uk/technology/2011/sep/19/…
 
@Patrick: I have the beginning of an algorithm to detect rivers (in Lua)
For some reason, it's not detecting the space between "just" and "anotherrr". I wonder why...
 
well, without the code, I can't tell you anything :) I've once crated such an algorithm with a self-written line breaking algorithm and I have made the lines higher and only counted spaces that overlap a certain amount with a helper array. I don't want to look it up though because it was in my diploma thesis and I don't want to read this again.
 
haha
@Patrick I will make the code available
 
thats it for now - you get the point
 
@Raphink, @Patrick: Do you have a precise definition of what a river is?
 
1:32 PM
the images are broken - the black boxes should be aligned within the text
 
@Patrick cool
 
@AndreyVihrov there is none. The best I've found is "Rivers: In a paragraph of text, a series of wordspaces that accidentally align vertically or diagonally, creating an awkward flow of white space within the column of text"
it was from picastudio.com/glossary/glossary.html#rivers - but this link is broken
The difference between images 3 and 4 is that I have removed all glues that are "alone"
 
it's not very precise just yet
it might need to be refined
 
1:38 PM
@Raphink not today (I want more time...)
 
no problem :-)
for now, the code is easier than that of homeoarchy detection ;-)
 
And @Raphink - sorry for reading your mind :)
 
@Patrick: haha ;-)
 
I have to admit that tex.se is much fun these days
Nice challenges from you, @Raphink
 
;)
0
A: Avoiding "rivers" in successive lines of type

RaphinkI have added a first version of an algorithm to detect rivers using Lua to the impnattypo package on github. To use it, simply use the rivers option: \usepackage[draft,rivers]{impnattypo} Here is an example result: Beware that there might still be some bugs ;-)

I think it would be better later on to take in consideration the orientation of the river
you can have diagonal rivers that are problematic, but when it goes left and right alternatively, it's not a big deal in quite a few cases
I'd have to really narrow the definition of problematic rivers
 
1:48 PM
Sure - a good definition would be great.
 
@Raphink: In your package you have multiple occurences of code like (head.char >= 65 and head.char <= 122). Doesn't it only work for ASCII?
 
@AndrewStacey yes, well extended ascii actually. It's a bit ugly I reckon, but I haven't found a better way so far.
that's for the nosingleletter algo
 
2:13 PM
@Raphink you could use unicode.utf8.match(str,"%a") to match for a character.
@Raphink but even this is not correct, as the slots in glyph.char is not in any strict encoding it is font/implementation dependent
 
@Patrick: I only want letters, no punctuation
 
%a should not match punctuation
 
I could try
since the algorithm does modify the line, I'd rather it only touches predictable chars
 
%a is Lu+Ll+Lt+Lm+Lo in unicode.org/Public/4.0-Update1/…
 
I will try that, thanks
so
        if unicode.utf8.match(head.char,"%a") then            -- some kind of letter
?
 
2:18 PM
no, you need to convert the decimal to a string - wait a second
 
ah
 
@Patrick: Ah, so you had edited it already. I hadn't paid attention.
2nd necromancer badge \o/
 
3:27 PM
Quick question: is there a non-etoolbox equivalent of its \appto for patching a simple macro?
Found it: \g@addto@macro
 
@AlanMunn Define 'simple': are there any arguments? Can we use e-TeX?
 
hmmm, the river detection algorithm adds quite a bit of noise so far...
 
@JosephWright Yes, simple meant no arguments (in fact its the beginning of an environment in an uglily written package.)
 
@AlanMunn \g@addto@macro works globally, while \appto works locally.
 
\def\myappto#1#2{\begingroup\toks@\expandafter{#1}\toks2{#2}\edef#1{\the\toks@\‌​the\toks2}\expandafter\endgroup\expandafter\def\expandafter#1\expandafter{#1}}
 
3:42 PM
@JosephWright Thanks! Since I'm suggesting a fix to someone else (an end user) to fix a problem with package incompatibiliy, and both \appto with etoolbox and \g@addto@macro both seem to solve the problem, I don't think there's a need for my own local equivalent. Are there any unforseen problems with using the global version that might crop up?
 
@AlanMunn Assuming this is not done repeatedly on a local macro, all is well. (Mixed local and global assignments can exhaust the save stack, but this is not normally an issue.)
I'm assuming that it's okay to alter the macro globally, of course
 
@JosephWright Yes, I think so. It's the main environment defined by the package.
 
4:08 PM
@JosephWright Simple is in the sense that it has no arguments; otherwise \apptocmd from etoolbox is necessary (with limitations). Your \myappto should be \long.
\long\def\appto#1#2{\expandafter\def\expandafter#1\expandafter{#1#2}} is simpler (but has no check on #1 being a macro).
 
@egreg It was a quick knock-up (I'd actually started with a pre-appending function, hence the complexity).
 
4:24 PM
@JosephWright Yes, prepending is more complex.
@JosephWright I guess that the kernel does it with `\xdef` in order not to clobber `\toks@`. What about this?

\long\def\preto#1#2{\begingroup
\toks0=\expandafter{#1}\toks2={#2}%
\edef\x{\endgroup\def\noexpand#1{\the\toks2 \the\toks0}}\x}
 
Active character fun:
0
A: Using mathspec’s \setallmainfonts with Russian in polyglossia causes hang during compilation

Andrey VihrovYou observation about shorthands is correct. Loading mathspec redefines \sigma to "σ, makes " active and \lets a certain control sequence to this character. Loading polyglossia with the russian option makes " active again and overwrites the previous definition of ". mathspec also sets up " to be ...

 
4:57 PM
@AndreyVihrov As I wrote in the comment, Polyglossia is at fault because it calls \initiate@active@char also if shorthands are not required. One should modify the code, issuing the call only if the shorthands are asked for. It's possible to enable them globally or locally for each language that supports them, but in this case the \initiate@active@char routine is called.
Supporting mathspec along with shorthands seems quite complicated, but at least one can warn users that shorthands must not be activated if mathspec is loaded.
 
5:15 PM
@egreg Indeed. I will try to report the polyglossia bug. Nonetheless, it's equally a mathspec bug too (not sure where to report on this one, though).
It is ironic that polyglossia, which clams to be A Babel Replacement for XeLaTeX, loads parts of babel code.
 
5:36 PM
@AndreyVihrov Well, not every part of babel must be sent in the waste bin.
2
I'll try to commit some changes.
But, please, send the bug report.
 
5:55 PM
Yay, making progress on the bookmark app!
 
@PauloCereda Hit rep cap again. :(
 
@egreg wow! How many days in a row? :)
 
@egreg Soon you'll be no. 1. Well, at least no. 4. ;-)
 
@lockstep I'm actively chasing you. :) Less than 1000 rep distance.
 
6:10 PM
@egreg you are already #1 filtered by quarter.
 
@egreg try chasing @lockstep regarding voting ;-)
 
@StefanKottwitz Nahh ... try it with regard to editing. ;-)
 
I hope some of the most active voters apply for SE sponsored membership in TUG, because they highlight quality and they ensure that other active members get more access to site features. Reading and validating is a deserving task.
 
6:27 PM
@StefanKottwitz Understood. ;-)
 
@lockstep less than 3000 edits and I got you
@lockstep great ;-)
 
Maybe we should do the reverse nomination for TUG membership. You have to sign the list if you don't want membership. Then SE can go through the list of everyone else to decide who to award it to.
I'd also say that it would be good for us directly if we have lots of people nominating themselves. It's a way of saying "Thank you" to SE and showing that we appreciate them becoming an institutional member of TUG. So don't think of it as putting yourself forward, think of it as saying, "When it comes to this time again next year, we definitely want SE to renew its membership."
3
 
@AndrewStacey That's true!
 
So, for example, anyone who's committed enough to this site to spend time in chat should nominate themselves. Consider it a penance for talking about football all the time. (If it were rugby, now, that would be different.)
 
Now, as some days passed, nobody has to worry that he possibly would take a place somebody else would like to have.
 
6:33 PM
@StefanKottwitz: as you're here, would you like to deal with the top question on the "active" view?
Anyone any views on this question: tex.stackexchange.com/q/26378/86 ? @lockstep: you've a lot of experience with what should be closed (and why), what do you think?
 
@AndrewStacey Thanks, I deleted that SPAM post.
 
Out of curiosity, how much spam do we get? I don't get the sense that there's loads, but then it might all be being dealt with before I ever see it. On occasion I see some deleted answer, but not often enough that I feel we have a problem.
 
tex.stackexchange.com/questions/27588/… 3 hours left for bounty expires... I wonder if @Mico will use it.
 
@AndrewStacey spam is rare now
 
@Raphink: Can your code to highlight repetitions do anything with highlighting? It seems that as you're working at the PDF level, you can find out about the height of the current line - which is one of the main difficulties. See, for example, tex.stackexchange.com/q/24036/86
 
6:39 PM
@AndrewStacey I'd close it as "not a real question". As the questioner made some effort, I might counter an automativ downvote with an upvote.
 
@AndrewStacey what do you mean?
@AndrewStacey well sure I can detect the line height and adjust the boxes.
 
@Raphink What I mean was, if I (in the TeX code) mark a couple of places, can you (via your Lua code and pdfliterals) colour in the background between those places with the correct height?
 
6:58 PM
@AndrewStacey Italy scored nine tries versus Russia. :)
 
@egreg Now you're talking! England, meanwhile, are a complete shambles! Can't wait to see them get completely thrashed by Scotland.
Close as "too localised"? tex.stackexchange.com/q/27948/86
 
@AndrewStacey Argentina is a good team. Let's see what they do with Scotland.
 
@Raphink: Okay, so if I mark a point in the text, what I'd like to do is save the coordinates to the aux file not of that point itself, but of the top and bottom of the current line at that point.
@egreg Yes, they are good to watch. Was it them that got to the semis last time? I remember watching the third-place play-offs. I think it was Argentina against France. Argentina (if it was them) played as if they were in the final. France played as if they weren't there.
 
@AndrewStacey You know the French. Oops, apologies to Raphink.
 
Here's a proposal on area51 some here might find worth looking at: "academia": area51.stackexchange.com/proposals/16617/…
 
7:22 PM
@AndrewStacey I am not Raphink, but this is possible. The green line with the formula would have the height of the formula though and that's not really beautiful.
 
@AndrewStacey you want to complete the highlighting when it's partial like in the case of the math in the question?
@AndrewStacey, @Patrick : why not do that in Lua altogether? What I've done in two packages is that I have two implementations, one in pure TeX and one in Lua, and the switch is done automatically using \ifluatex so the functionalities are nicer when using LuaTeX, but still there somehow when using another rendering engine.
 
@Patrick You mean you aren't one of Raphink's sockpuppets? I agree that it wouldn't be beautiful in that case. However, I'm using highlighting in maths-only places where the highlighter should be wider. I'd also say that this sort of thing should be configurable and shouldn't be totally automatic, but it'd be good to make the configuration as easy as possible.
Here's another that's borderline, and the length of time with no response from the OP tips it over that line for me: tex.stackexchange.com/q/26470/86
 
7:39 PM
@AndrewStacey Well, I am @Raphink. I ask questions and with my other ID I comment my questions that they are really clever questions. :) I am not sure how math is presented in TeX's nodes, so there could be an automatic solution...
 
7:49 PM
@AndrewStacey I tried to understand rugby. But I failed miserably.
@egreg: It seems Brazil has a rugby team, but I never saw them playing. It seems that Argentina is the top south american team by far.
 
@PauloCereda It's not to understand, it's to live.
 
@AndrewStacey That would explain it. :-) Ah, I know what a try is. :-)
The guy at ESPN always explains the rules and how the games work. I don't get some concepts, but it's quite nice and entertaining to watch it. Specially now with the World Cup.
 
@Patrick That would at least explain how you can read my mind...
 
@AndrewStacey I live it indirectly, as my brother's son is quite good. At just 18 he's scrum half in a second division team. And I live in Padova, where there's one of the best Italian teams (at least it was some years ago and it's improving).
@AndrewStacey But I'm more into basketball, having been involved in it for almost 40 years. Ouch!
 
Basketball is older than 40 years? ;-)
 
8:09 PM
@Raphink The first basketball game was played at Springfield, Massachusetts, on December 21, 1891; that's 120 years. :)
The Italian basketball association is 90 years old.
The French association was founded in 1932.
 
@egreg that was a joke :-)
 
@Raphink I know! Your association is younger than ours, but your team is way better. :(
 
we finally qualified for London'12!
 
@egreg And invented by a Canadian, who was the Athletic Director of my alma mater (McGill) (not when I was there...)
 
Do other people think tex.stackexchange.com/questions/29049/… is off-topic?
 
8:21 PM
@AlanMunn Really you weren't there? Somebody here always tells me that I was at Springfield with Naismith: in our local officials' association I'm the one with the least association number.
 
@Raphink Although I noted that, I've actually never been to Graphic Design, so your statement may well be true
 
I see we now have six takers for the TUG membership: looks like a good take-up to me
 
8:31 PM
@JosephWright They are great members. TeX.sx will be surely well represented.
 
@egreg In fact, Naismith graduated from McGill 99 years before me.
 
@Raphink On-topic, though I ponder "Not a real question". I'd say that detection of rivers was the whole of it. Avoiding rivers feels (on the basis of almost nothing) like it's too fine to be left to the computer to do automatically. I'd rather think about the paragraph myself and decide whether or not I like it.
 
@AlanMunn Just a short time, then.
 
@AndrewStacey One day computers will think that the remaining open problems are too fine to be left for people to solve ;-)
 
@egreg In geologic time, certainly.
 
8:45 PM
@AlanMunn Or maybe you were very lazy and graduated after so many years… :)
Your age doesn't appear in your profile. Maybe you've something to hide. :)
 
@egreg Yes, I've discovered the fountain of youth and don't want anyone to know
2
 
9:46 PM
@AlanMunn The fountain of youth? I've been taught that the secret was the broccoli. :-)
 
9:59 PM
@Raphink: Here's a suggested definition of "really problematic" rivers.
 
@lockstep : I saw, very interesting :-)
I'll try to implement that. My idea was also to use the overlap proportion, but your formula seems interesting.
for now, I'm translating the documentation for impnattypo into English since it's beginning to have quite a few features that are not French specific.
 
10:56 PM
@lockstep: no hurries. I'd like to get a nice algorithm that would prove useful for river detection :-)
and there's probably quite a few people here who are better at defining formulas than I am :-)
 
@Raphink And better than I am. ;-)
 
Well good thing brains don't just add up, they complement when necessary :-)
 

« first day (324 days earlier)      last day (1187 days later) »