12:02 AM
@PeterGrill tex.ch says
@!save_size=600; {space for saving values outside of current group; must be
at most |max_halfword|}
so you are probably bumping into max_halfword

@DavidCarlisle The problem only occurrs on a fairly large doc, that output might be EXTREMELY large, but let me try..
what is max_halfword?

@Alan: Did you see PS4 will cost R\$ 4.000,00? :)

@PeterGrill half the size of a word (it says:-) (I don't think it's changeable without recompiling, and probably code changes) but I'm out of my depth (and it's late at night:-)

@DavidCarlisle Ok, show how do I turn on and off the tracing? I probably need to control so that the tracing is only turned on where the problem is.

12:10 AM
tex.web does confirm s is save size odd that it chnages b:
  wlog_ln(' ',max_in_stack:1,'i,',max_nest_stack:1,'n,',@|
max_param_stack:1,'p,',@|
max_buf_stack+1:1,'b,',@|
max_save_stack+6:1,'s stack positions out of ',@|
stack_size:1,'i,',
nest_size:1,'n,',
param_size:1,'p,',
buf_size:1,'b,',
save_size:1,'s');

@DavidCarlisle Not sure what you are saying here.

@PeterGrill They are integer parameters: just set \tracingstats2 anywheer and you get information about menroy usage every page shipout rather than at the end (which is \tracingstats1)

@DavidCarlisle ok, sounds good... let me try...

@PeterGrill that is tex.web (ie the original source of tex the program) where it prints out that stack positions line and the number it puts next to s is its save_size variable

Something is definitely wrong when a chicken is better at snooker than you are. :P
ooh @egreg is the 2nd top voter! :)

12:21 AM
@PauloCereda Especially since that's not snooker. :) (It's pool.) And where do you find this stuff?

@AlanMunn Oh my! :)

@AlanMunn I was going to mention that, but you can't expect a chicken to know the difference

@AlanMunn Exactly! @PauloCereda Where do you find this stuff?!!?!

@kan Hey! <3

9 hours later…
9:51 AM
@egreg I am using the both of those commands, any idea why that might still be happening?. Even more strange, when i remove the \usepackage[T1]{fontenc}, and just keep \uspackage{lmodern}it seems to work fine

10:31 AM
@Eric No idea. Minimal example?

10:51 AM
I wonder why I noticed many do not think large and deeply nested TOC is a good idea. For me, it is a very good thing. Because it allows one to see, in one place, everything in the book. Better than having to go to each section and see a small TOC there, then find out what they want is not there, go back and go to next section and so on. What Am I missing?
I like the TOC to go all the way from part to subsubsubsection.

11:09 AM
@Nasser Do you also like to read telephone directories? ;-)

@egreg Well, not really. But this is not the same. is it? Having a small TOC will give one the 5,000 miles high view of the book, but what if one wants to see more details? I like when I open a book to read its TOC to find what is inside. A more detailed TOC tells me more information. I think Latex needs a new \maketableofcontents option. So one can make a mini TOC (for the managers) and detailed TOC for the technical people to use.

@Nasser This is already possible; check the shorttoc package.
Gotta go now.

@egreg thanks! will check it out.

11:38 AM
The awkward moment when you see a question that reads Can we make love to LaTeX? and the next instance when you read it properly hence getting confused whether you are dirty or the question title...

1 hour later…
12:39 PM
Just discovered lineno, great for line numbers for proofreaders/editors :)

3 hours later…
3:34 PM
Was texwelt.de separately built from stackexchange network because its proposal (to create tex.sx in German language under stackexchange network) has no enough users who make a commitment as active members?

4:22 PM
@egreg so now you get people to pop up asking ooalign questions for you to answer.

4:40 PM
@DavidCarlisle I made a reputation of being expert in \ooalign.

@egreg and I voted because of your expertness:-)

@DavidCarlisle And, while the question was about breqn, I didn't add “don't use it”. ;-)

4:55 PM
@egreg LOL

Hello!
Don't you guys think that this edit should have been a comment under my answer (without the code part)?

1 hour later…
6:23 PM
@Alenanno It's not quite clear what Sverre is after. It seems that he's not content with something in the solution, but it's quite too long for a comment. Any ideas, @JosephWright ?

6:51 PM
@tohecz I know what it is, he asked for something where you can manually set the position of the mother node. But that was not in his original question. He added that later. I told him this in a comment as well.

7:03 PM
@Alenanno well, I'm certainly not going to revert it and I don't know if anything can be done with it. Sometimes it happens that the idea of the OP is different than the idea of the answer he recieves...
anyways, I'm off home, it's quite late here

@tohecz Still here?

@AlanMunn yep, for a while

@tohecz Are you familiar with Andrew's tikzmark library?

@AlanMunn not at all

@tohecz Oh well. I'm trying to use it but can't get it to do what I want, and the documentation is a bit on the thin side.

7:18 PM
@AlanMunn well, good luck. I'm off now really

@tohecz Ok. Good night.

8:09 PM
@tohecz Well, I didn't ask you to revert it. :P And yes, I agree with the rest you said.

@JosephWright I've noticed a number of journals put the DOI in the .bib file as "http://dx.doi.org/<DOI>". While wrong this means I have to hand edit the bib file. Is there a chance you could add code to detect this and correct for it by linking to the URL they give instead of blindly linking to "http://dx.doi.org/dx.doi.org/<DOI>";?

8:43 PM
@Marienplatz texwelt.de was founded because we like the Stack Exchange way very much, and sadly a German SE version wasn't accepted. SE is commercial and thinks in business. I don't think so, so I just pay the server and loose the money, that's fine. :-)
@Marienplatz The way it grows, is fine: TeX has staying power and a long breath. No need for high traffic as here: duplicates will come soon enough ;-)

@AlanMunn I tried to use it for an answer the other day, I marked two spots but I couldn't get tikz to draw a line between them, the syntax appears not to be learn-able by osmosis.

@Marienplatz It just feels good to read and write in native language. And to not have a burden of rules and restrictions, which originate from a crowded site such as SO.

@DavidCarlisle Yes, I agree. That's why I prefer my version, which although simpler, is a bit more intuitive. I just adapted a version similar to what I had previously used with pstricks and didn't know about Andrew's version until later.

@AlanMunn I meant the syntax of tikz rather than that of tikzmark:-)

@DavidCarlisle Oh. The syntax of TikZ isn't so bad. Have you ever used R?

8:56 PM
@AlanMunn briefly (we produced some R interfaces to the NAG library, but R had a manual, I've only tried tracing tikz...

9:16 PM
@StefanKottwitz: It is interesting! Thank you. Who is the developer of this site? Is it PHP based or ASP.NET MVC based?

@Marienplatz It is Python based, build with the Django framework, which I also use with texample.net
and it runs on Linux

@StefanKottwitz Very good. I still don't understand why German version of TeX.SE was not accepted. Have you made a proposal before?

@StefanKottwitz, looks like OSQA right?

@ForkrulAssail Yes, it's OSQA!

@StefanKottwitz, using that for subject Q\&A for my students.

9:25 PM
@Marienplatz No, it was somebody else. TeX is a niche, non-english TeX even more. There's more success in Gaming, Cooking, Travelling, ...
@ForkrulAssail Nice to hear! Did you customize it?

@StefanKottwitz, we're just doing an initial study to see if it helps with student engagement, might customize it with more gamification type features when time allows.

@StefanKottwitz OK. Thanks. I will investigate whether or not there was such a proposal in Area 51.

@Marienplatz There was one! I'll have a look...
@Marienplatz At least for German StackOverflow, and StackOverflow as such would be much bigger, and even that failed

9:44 PM
@StefanKottwitz: There are less than 100 members in texwelt.de so it is still not enough to create a new proposal for German TeX.SE immediately.

@DavidCarlisle: That's interesting... that unboxing provides a different result, even if it contains the some stuff...

10:05 PM
Texshop (Tex Live 2013) does not update from version 3.23 on Mavericks OS... "Update Error! An error ocurred in retrieving update information. Please try again later."
Any hints? Tex Live (2013) Utility is ok.
Also, problems with fink... "fink selfupdate" returns "We don't support upgrade from 10.8 to 10.9, Check the 'Clean Upgrade' section... of INSTALL for information about how to proceed."

@Papiro On the TeXShop issue, just download a new version from the TeXShop web page. This was a known issue with the automatic update and Mavericks. As for fink, it seems like you need also to reinstall a new version of it.

10:23 PM
@AlanMunn Thanks!

@Werner not the unboxing, the boxing. It's a bad premature optimisation in tex-the-program to save a few bytes that mattered in 1979: in inner horizontal mode tex doesn't add the penalty and language nodes used in linebreaking/hyphenation as hboxes don't break, but of course they can be unboxed...

10:50 PM
@DavidCarlisle Has anyone tried to rewrite tex to avoid things like that? I know it would horribly break compatibility, but it might be worth it.

@DavidCarlisle I believe that penalties are preserved. Not language nodes; and luatex doesn't preserve them either. :(

11:19 PM
@egreg penalties are preserved if they are added, but they are not added (\discretionary node not added to be more exact) compare:
\makeatletter

\def\Sbox#1#2{%
\setbox\z@\vbox{\hsize\maxdimen%
#2\par
\global\setbox#1\lastbox
\global\setbox#1\hbox{\unhbox#1\unskip\unskip\unpenalty}}}

\showoutput
\sbox0{abc-xyz}\showbox0

\Sbox1{abc-xyz}\showbox1
> \box0=
\hbox(8.33331+2.33331)x39.49265
.\OT1/cmr/m/n/12 a
.\OT1/cmr/m/n/12 b
.\kern0.32639
.\OT1/cmr/m/n/12 c
.\OT1/cmr/m/n/12 -
.\OT1/cmr/m/n/12 x
.\OT1/cmr/m/n/12 y
.\OT1/cmr/m/n/12 z

! OK.
l.37 \sbox0{abc-xyz}\showbox0

?
> \box1=
\hbox(8.33331+2.33331)x57.11746
.\hbox(0.0+0.0)x17.62482
.\OT1/cmr/m/n/12 a
.\OT1/cmr/m/n/12 b
.\kern0.32639
.\OT1/cmr/m/n/12 c
.\OT1/cmr/m/n/12 -
.\discretionary
.\OT1/cmr/m/n/12 x
.\OT1/cmr/m/n/12 y
.\OT1/cmr/m/n/12 z

! OK.
l.39 \Sbox1{abc-xyz}\showbox1

@DavidCarlisle I wonder if Taco thought about this. I think it could be easy to correct this behavior in LuaTeX