« first day (249 days earlier)      last day (4686 days later) » 

7:42 AM
@PauloCereda: I just used both answers together and they worked out pretty good. TeX Font, TTF, OpenFont all there (the first 255 symbols).
One thing is that they are all top-aligned and not sitting on the baseline.
 
 
3 hours later…
10:45 AM
@MartinScharrer: Cool! Khaled posted an update on how to align all the glyphs on the baseline.
This would be a nice entry for the TeX.SX blog. =P
 
 
1 hour later…
11:50 AM
I wonder if the output routine could be adjusted easily enough to do tex.stackexchange.com/questions/22511/…
 
 
2 hours later…
2:04 PM
I will write a package called donothing. In the preamble, put \usepackage{donothing}. In the document body, you may use the following commands: "Lorem ipsum dolor sit \donothing amet, consectetur \donothingagain adipiscing elit. \alsodonothing Pellentesque \redonothing egestas, risus \pleasedonothing eu consectetur \yaydonothing posuere, risus urna placerat \donothingparttwo nisl, non rutrum dolor tellus non leo. \lastdonothing".
 
2:18 PM
We have a tag for pdf(la)tex but we don't have one for the DVI latex, do we?
 
@MartinScharrer we have a , I guess.
 
2:43 PM
@PauloCereda: Yeah, I saw that one. It's probably the best way to go with. A tag looks so ugly.
 
@MartinScharrer What if becomes a synonym to ?
 
@PauloCereda: Mmmm, might be an idea. Maybe I should open a post on TeX - LaTeX Meta.
(BTW: [meta] is now turned into TeX - LaTeX Meta)
0
Q: How to tag questions specifically about the classic DVI-producing latex compiler?

Martin ScharrerWe have a pdftex tag to refer about pdf(La)TeX, but no actual way to tag a question specifically for the classic latex or tex compilers. For example I don't know how to tag questions ala "How to use PNG images with latex". In contrast, a question like "How to use EPS images with pdflatex" can be ...

 
@MartinScharrer: Cool!
 
3:29 PM
@MartinScharrer: We already have a use case for the nounproject package. =)
 
Apparently one of my accepted & +10 up-voted answers I got an Enlighted badge for got un-accepted (or down-voted below +10). In this case you don't loose the badge again, but don't get one for the next accepted +10 answer. However, when you have then again such an answer you get the badge for the last one! That's quite confusing!
 
@MartinScharrer: Oh. My brain hurts.
 
@PauloCereda: It's a pity that they don't have a nice symbol for that. My suggestion is OK, but I would like to have a better one. I tried to search for 'Web' or 'Internet' and they don't have anything.
 
@PauloCereda I explain it more clearly: You answer a question which gets +10 in votes and accepted. Then you get an Enlightened badge for it. Let's call that answer "answer (A)".
 
3:34 PM
Hm OK.
 
If this question (A) gets unaccepted you keep the badge.
But if you have another accepted +10 answer (B) you don't get another badge, because it now kind of count for this answer instead.
 
Wow.
 
Now if you have a third accepted +10 answer (C), you get a new badge, but for (B) not for (C)!
Then if you would have a new accepted +10 answer (D) you get a badge for (C)!
etc. etc.
 
Oh my.
 
It's confusing if you suddenly get a badge for an older question which didn't changed at all.
 
3:38 PM
@MartinScharrer: You found an epic bug in the StackOverflow framework!
 
@PauloCereda: that's might be actually status-by-design
 
Programmers. =P
 
4:13 PM
I just had an idea to code an \shiftbox macro, i.e. an horizontal version of \raisebox.
Basically \hbox{\rlap{<content>}\hspace{<..>}} or \hbox{\hspace{<..>}\llap{<content>}}
no wait: \hbox to \dimexpr\width-\shift{<content>\hss}
 
Cool!
 
Do we need a \includegraphics[shift=<..>]{..}?
 
@MartinScharrer: I can't think of a case.
 
4:28 PM
@PauloCereda: I like to support it for adjustbox which re-uses/shares the same options as \includegraphics ...
 
@MartinScharrer: It'll be one of the additional features, so it may be a nice idea. Someone will need it.
 
 
2 hours later…
6:53 PM
@PauloCereda: Maybe I call it \lapbox instead ...
 
@MartinScharrer: Nice! =)
I fail at giving names to commands, files and softwares.
 
7:39 PM
Friends, should I finish the TeXPrinter app?
 
8:18 PM
@PauloCereda: i.imgur.com/Z0Y6l.png
2
 
@MartinScharrer: Wow, impressive!
You are my hero!
 
Thanks goes to Khaled Hosny for his script!
 
That Python script is pure gold.
 
@MartinScharrer nice
 
We just have to see how to handle the 256 characters per TeX font limitation
It's a pity that font glyphs only support one color! But understandable of course.
 
8:38 PM
The black diamond ♦ gave such a headache to figure out what was wrong in my parser. =P
 
8:49 PM
Yay, I finally got TeXPrinter generate a tex output!
 
@Martin Are you spying on us when we write our answers?
 
o.O - Spooky.
 
9:06 PM
@MartinScharrer A while ago I remember you asking about a script that would speed up the CTAN submission process. Did you ever find/write one?
 
@Alan: I coded something by myself.
 
@MartinScharrer Are you willing to share?
 
@egreg: No, moderator rights don't go that far :-)
@AlanMunn: If I find it again .. :-)
 
@MartinScharrer I see. It must have been very useful... :-)
 
@AlanMunn: I didn't uploaded anything in the last month because my thesis keeps me busy.
 
9:17 PM
@MartinScharrer As a prof, I can't argue with that!
 
It's actually part of a Makefile which simply starts the CTAN upload page using firefox with the form values set
firefox 'http://dante.ctan.org/upload.html?contribution=${CONTRIBUTION}&version=${VERSIO‌​N}&name=${NAME}&email=${EMAIL}&summary=${SUMMARY}&directory=${DIRECTORY}&DoNotAnn‌​ounce=${DONOTANNOUNCE}&announce=${ANNOUNCEMENT}&notes=${NOTES}&license=${LICENCE}&‌​freeversion=${FREEVERSION}' &
Then you only need to select the filename for upload and press sent.
I tried to set the filename as well but the browser doesn't allow that for security reasons.
Which makes sense.
 
@MartinScharrer Cool. Thanks. I'm trying to convince someone to put his stuff on CTAN but they think it's too much work; maybe this will help.
 
@AlanMunn: Oh, yes, the Makefile is also on my SVN repository which is online: latex.scharrer-online.de/general/browser/filemod/dev/Makefile
 
9:33 PM
@AlanMunn: I found it more difficult to write a DTX file and documentation for my first package, than to upload it to CTAN. See also my sty2dtx Perl script (on CTAN as well), which is very helpful.
 
hello
I have some simple question, that I believe is not worth separate question on TeX.SX
 
@przemoc: hi there. =)
 
I'm using zlevel solution brought by Martin in tex.stackexchange.com/questions/20425/z-level-in-tikz
but I'm going a bit further
I have: % current pgf layer
\newcounter{cpl}
\setcounter{cpl}{1}

\tikzset{%
zlevel/.style={%
execute at begin scope={\setcounter{cpl}{#1}\pgfonlayer{\alph{cpl}}},
execute at end scope={\endpgfonlayer}
},
}
which works fine
but I would like to setcounter and return alph at once in pgfonlayer
 
@MartinScharrer Yes, I find .dtx a bit intimidating. I haven't made that plunge yet. It seems a kind of cumbersome way to work, although I can see the advantages.
 
\def\cpl#1{%
{\setcounter{cpl}{#1}%
\alph{cpl}}%
}
is not working, \pgfonlayer{\cpl{#1}} spits error
Missing \endcsname inserted, the question is how I should fix \cpl to make it work
 
9:52 PM
@przemoc: Do you need the cpl counter for anything else?
 
@MartinScharrer: no
but I wanted to make \cpl macro more advanced after fixing this issue
 
Use \pgfonlayer{\expandafter\@alph\numexpr#1\relax}
 
but it requires \makeatletter, that I would like to avoid if possible
I'll explain then my idea
 
Or simply \pgfonlayer{\@alph{#1}}
Or \pgfonlayer{\csname @alph\endcsname{#1}}
 
\cpl was meant to do later something like: main for 0, a,b,c for -1,-2,-3,... and A,B,C for 1,2,3,...
the problem is I don't know how to fix this \cpl at this moment
it's some basic issue I'll gladly understand with your help
 
9:56 PM
See \alph uses \@alph, which uses \ifcase #1\or a\or b\or c\or d\or e\or f\or g\or h\or i\or j\or k\or l\or m\or n\or o\or p\or q\or r\or s\or t\or u\or v\or w\or x\or y\or z\else \@ctrerr \fi
@egreg: Say, is there any way to have an expandable version of \char?
Like \edef\mycmd{\echar`A} which is basically defining \mycmd to A.
I tried once to use \futurelet to look ahead and turn the token back into a letter (if it is a letter), but there I could go further.
 
@MartinScharrer: yeah, I can go with \ifcase... but apart from it, can you explain why my current \cpl definition is wrong and how is the easiest solution to fix it?
 
@przemoc: I assume \pgfonlayer{..} uses \csname .. .. \endcsname internally, i.e. the input must be fully expandable.
@przemoc: the setcounter command makes an assignment and is therefore not expandable
 
I'll hide now while you guys talk TeXese. =P
 
@PauloCereda: Before you go
 
but cannot \setcounter be somehow hidden for expanding?
it would be extremely convenient
as it does not return anything, as opposed to \alph
 
10:04 PM
@PauloCereda: There you can upload t-shirts and sell them with commission (which can be 0). We could upload a TeX.SX Shirt and sell it for the cheapest price available, i.e. 13.90 euro + 4 euro shipment for a basic shirt, extra color makes it more expenside)
 
@MartinScharrer: Great idea!
 
@przemoc: You don't need a counter for this. Simply use \@alph instead of \alph.
 
Besides, I need to learn German. =P
I'm gonna make a template for a TeX.SX shirt. Should we ask permission to use the .SX name?
 
@PauloCereda: They are located in Germany apparently, but sell international
@PauloCereda: Maybe we should open a general meta post about it first.
 
Good idea!
 
10:07 PM
@MartinScharrer: I'm asking to learn something, I got already that the counter is not needed here. so it is not possible to "hide" \setcounter for \csname?
 
I'll be back in an hour or so. See ya guys!
 
@przemoc: No, you can't. See, the expansion is before anything gets executed/assigned.
Everything inside a \csname .. \endcsname must expand to text only. That \setcounter doesn't produce any output in the PDF doesn't matter here.
 
so if I use \ifthenelse, there will be problem too?
 
There is however the \numexpr command which is expandable if you place a \the in front of it, e.g. \the\numexpr 1+1\relax will expand to 2.
@przemoc: Note sure, if it is not fully expandable then yes. Try to do it numerically and use \ifnum or \ifcase instead
 
I see, thanks for your help @MartinScharrer. Maybe later I'll add another answer in mentioned z-level question showing ready to use boilerplate
 
10:13 PM
@przemoc: Why exactly do you need letters instead of numbers?
 
because \pgfdeclarelayer does not support numbers
so going with letters is IMO nice workaround
 
@przemoc: Not numbers at all or just not numbers by themselves?
 
I haven't investigated it, seen error when number was used, so skipped it
 
@przemoc: \pgfdeclarelayer seems to simply define two boxes:
\expandafter \pgf@newbox \csname pgf@layerbox@#1\endcsname \expandafter \pgf@newbox \csname pgf@layerboxsaved@#1\endcsname
 
ok, it was \pgfonlayer giving an error
sorry for misinformation
 
10:20 PM
No problem, I thought so.
You could simply prefix the layer name, like zlevel#1 where #1 is the number
 
damn, I had messed something before, because numbers are working for me now...
 
In my quick tests numbers work fine, too.
Also you should alias the normal default layer to 0.
I can show you how to do this.
You need to set up all layers beforehand, before the picture I mean, do you?
\expandafter\let\csname pgf@layerbox@0\endcsname\pgf@layerbox@main
@przemoc: Wait a minute, there is a second box which must be \let as well.
 
I see, I was trying \pgfonlayer{\ifnum#1=0\relax main\else #1\fi} which doesn't work
but some of my previous tests showed that drawing on main doesn't work, I'll repeat it now
 
Try it without the \relax but make sure to keep one space behind the 0:
\pgfonlayer{\ifnum#1=0 main\else #1\fi}
 
10:36 PM
yeah, works without \relax (I should use it only if there were other parameter like #2, right?). but my repeated test using main also directly reveals that main is supposed to be used without defining it, at least this is my hunch
 
@przemoc: I'm just looking at the PGF code. In basiclayer/pgfcorelayers.code.tex there is some code which checks if the argument is exactly 'main' without expanding it.
Yes main is special.
 
i.e. whatever is not within pgfonlayer, it is in the main, and it's fine to me, one scope less ;)
 
exactly
I'm just seeing that the code I mentioned just now is only for the \pgflayerlist, so having a \ifnum in \pgfonlayer should be fine.
 
yes, it works, only not for special main, as layer explicitly using main is lost :)
overall I am quite impressed by tikz, but it's a shame it is embedded within inaccurate TeX, i.e. engine not supporting high precision or big numbers
 
@przemoc: So \pgfonlayer{main} doesn't work?
 
10:44 PM
no
 
@przemoc Yes the numeric accuracy of TeX is bad :-(
 
Creating TikZ/PGF solution outside of TeX would be nice
but it requires definitely a lot of effort
 
@przemoc: I think main doesn't work as extra layer, because it is already used around all layers.
Layers seem to be simple TeX boxes and everything inside the tikzpicture is stored inside the main box, the layer environments open a new box.
 
@MartinScharrer: Ok, as I said there would be no gain in explicitly defining main, maybe only in aesthetics terms
 
So you store a box while you already storing in the same box. The outer operation simply overwrites all changes from the inner.
It's like \savebox\mybox{ stuff \savebox\mybox{ inner stuff } more stuff }
Then \mybox will hold { stuff more stuff }, the inner assignment was overwritten
Anway, I'm going to bed. Good night.
 
10:53 PM
Good night and thanks again, @MartinScharrer!
 
@przemoc: You could make zlevel=0 check if it is used inside another zlevel scope (e.g. by setting a macro to true when any other zlevel option is used) and then simply close and later reopen the \pgfonlayer environment.
I.e. escaping to main
 
11:30 PM
@MartinScharrer It's easy: \def\mycmd{A}. :) You probably want to give a number and get a character in return. Not expandably, I'm afraid. \def\assign#1#2{\begingroup\catcode?=\catcode#2\relax\lccode`?=#2\relax\lowerc‌​ase{\scantokens{\endgroup\def#1{?}}}}` Now \assign\mycmd{64} is the same as \def\mycmd{A}. A bit dangerous if one says \assign\mycmd{`\%}
 

« first day (249 days earlier)      last day (4686 days later) »