12:24 AM
@HeikoOberdiek yes I wondered about that but it's documented as allowing utf8 to let simple documents be portable: if you add fontspec it's not portable anyway so if you are having conditional code using fontspec for xetex/luatex then you could just not loatd inputenc in that case which is the current (texlive 2013) situation. I don't know (and I really should be gone:-)

@DavidCarlisle Lots and lots of undocumented features. :) Well I never!:)

4 hours later…
4:00 AM
That moment when you google a Q and come up with your own from a year ago: tex.stackexchange.com/questions/109745/…
2

3 hours later…
6:35 AM
@hugovdberg No problem at all. Figured as much that you were off since I took quite some time to respond myself. Well, in the end I simply copied and pasted the whole .tex file into notepad++ and then saved it again via (via UTF-8 encoding of course). It now works again.
@hugovdberg Would love to know though how such an error happened in the first place.
@SeanAllred That hasn't happened to me once. More like 10 times. :)

6:55 AM
5

This paper by Vi Hart and Henry Segerman has some interesting hand-drawn images that cry out for a treatment with your favourite TeX-friendly package. I like the concept of throwing up one's hands and saying 'forget it; hand-drawn images it is', but it's just as cool to see such interesting figur...

Really?

@JosephWright what do you mean?

@DavidRoberts Well, falls both into the 'too much works for a Q&A' and 'off-topic': clearly the best way to create such figures is in a graphical program
@DavidRoberts Or at least that is how it looks to me

@JosephWright it definitely is a challenge, although a decent 3D aware application/package could probably do it in a breeze, place the cubes in space, connect the correct anchors with a properly styled arrow, color it, and calculate the correct 2D view from a given perspective. Haven't used tikz3d but if it provides anchors and clever connections as tikz in 2D it might be not too hard to do (although I don't have time to try it right now ;-))
@henry windows :-P just kidding, have no clue, just be glad it's fixed ;-) I'd only investigate further if it would happen again and try to remember what you did between then and the last time it worked :-)

7:12 AM
@JosephWright I tend to agree.
While I understand that it's ok to have some "how can I draw this fancy picture" questions, this one doesn't seem to be reasonable at all. The figure is not suitable to be drawn in any way but by hand or in a professional 3D software. However, TeX is not such a software, not even with TikZ or PSTricks. Therefore I think the question is off-topic. — tohecz 1 min ago
@Jos spam:
-5

In this posting Latino traditions intended for children The umbilical Low Concord 11s wire stump Umbilical hernias Submerged fontanel (ca p mollera) Frigid as well as warmth Fright and also anxiety (susto to espanto) Evil attention (mal p ojo)Latino traditions intended for babiesAs a brand new mo...

7:36 AM
@tohecz @JosephWright I thought I'd try. Some people might like a challenge, or contact the authors directly even if this were closed.

2 hours later…
9:54 AM
If the Firefox people want their product to look like Chromium, why don't they just stop developing and tell people to install that, so I can go on using something I'm used to? I just got v29 via Ubuntu and have the feeling I need to puke :-(

You will feel better after you puke, so they say.

@StephanLehmke FF29 hasn't reached Fedora repos yet, but my dad's PC (Windows) and my Mac got updated already. To be honest, since I don't use Chromium at all, I can't spot the difference. :) And to be a little more heretic, I've been waiting for the Australis design to reach the stable branch for a long time. Personally, the problem with FF is not the resemblance with the Google thingy, but the fact they are making major changes on the UI without ever considering the impact on users. :(

@StephanLehmke I can't remember which changes that was (I use FF 31 as my default browser) which bits don't you like?

@DavidCarlisle I think major changes was Australis and the menu hamburger.
Guys, I'm studying temporal logic and got bothered with this: \mathbb{I}nterv. How common is to mix these styles? I'm in an uncharted territory, so I can't say if every material holds.

@PauloCereda oh the --- thing? I've had that for so long:-)

10:06 AM
@DavidCarlisle You live on the edge. :)

@DavidCarlisle The most notable thing for me is no menus any more. But the design of the UI has changed almost everywhere, and honestly my first thought was "why the f*** did FF start chromium after I pressed restart?"

@StephanLehmke LOL pretty much it. :)

@StephanLehmke well yes there is a tendency for everything to move towards chrome (although I restored the file/edit/history/bookmarks/tools/help menu toolbar at the top the toolbox is still there you just need to check a box to show it in view/toolbars

@DavidCarlisle But first I have to find view/toolbars in this mess :-(

@StephanLehmke alt v :-) then check "menu bar"

10:11 AM
@DavidCarlisle No luck. Probably a different shortcut in the German interface.

@DavidCarlisle Guys in HCI publish stuff like water just because they change the position of things according to some sort of "study". :) Meanwhile, in the theory alley... "hey buddy, could you spare me a dime? Wanna hear about Solomonoff's theory of inductive inference?"

@StephanLehmke v alt :-) dann "Menüleiste"
@StephanLehmke (I have absolute faith in google translate)
3

@DavidCarlisle LOL
Found it, though in a place with absolutely no connection (translation-wise) to anything I could associate with "view/toolbars". Now I have two menus, what an improvement :-(

@StephanLehmke oh no! Blame David. :)

Well since I saw the "menu area" of the current word, I have to think maybe I'm too mentally inflexible from old age to adapt to modern software tools any more...
First, the aspect ratio of computer displays was adapted to resemble movie theater screens. Now, software UIs are adapted to fit an aspect ratio which has relatively much less space vertically. So, we are getting insane user interfaces because screens have sizes more fit for watching movies than actual computer work. That's really monty python-style humour on a global scale ;-)
3

10:45 AM
@StephanLehmke LOL
Hm Monty Python was mentioned. Where's @Joseph? :)

11:10 AM
@StephanLehmke I'm not looking forward to that, as I hear you. After Gnome3, it seems that FF29 is another step in the f***ing "destroy all useful stuff in UI" direction.

1 hour later…
12:27 PM
@PauloCereda Travelling today

@JosephWright Is github.com/plk/biber the correct site for raising issues for Biber?

@JosephWright Oh. Have a nice trip.

@egreg I thought sourceforge.net/projects/biblatex-biber, but as I'd just e-mail @plk in any case ...

@egreg: I'm studying temporal logic and found something along the lines of \mathbb{I}nterv. I believe you also think it's awful. :)

12:43 PM
@PauloCereda I'm regularly forced to write \mathcal{A}ut :-/

@tohecz Oh no!

@PauloCereda well, I at least write \mathcal{A}\mkern<minus-something>\mathit{ut} but it's not much better.

@JosephWright There's a manual near me right now, so shoot. :)

@PauloCereda What's the 'correct' way of handling functions now that we are not supposed to use module? All the examples assume you can dolocal mod = require 'mod' with a file mod.lua, but there is nothing explaining how you handle this for a case where everything is 'in line'.
  local l3kernel = { }
function l3kernel.strcmp (A, B)
if A == B then
tex.write("0")
elseif A < B then
tex.write("-1")
else
tex.write("1")
end
end
return l3kernel
l3kernel.strcmp
(%
"\noexpand\luaescapestring{#1}",%
"\noexpand\luaescapestring{#2}"%
)%
}%
E.g. the following fails
Ideally I want the two pieces of code in different \directlua calls
I can put the l3kernel.strcmp stuff in a separate file and it works if I then use require

12:51 PM
@JosephWright Hm I believe that's probably a requirement from the module concept. If you have a return statement where it is, nothing below it will be reachable.

@PauloCereda All rather tricky!

@PauloCereda I'm also working on it (not currently driving so ...)
@PauloCereda I think I have it :-)

@JosephWright Yep, Lua modules are, roughly speaking, a table of functions, so after creating them, you return the table so functions are enclosed in that structure. Since you have a return in the last line, require'ing to a local scope will ensure the namespace. :)
From the manual:
> A well-behaved module also arranges for require to return this table.

@PauloCereda For my purposes, I don't actually need to worry about the local business, so
\directlua{
lua.bytecode[1] = function ()
l3kernel = { }
function l3kernel.strcmp (A, B)
if A == B then
tex.write("0")
elseif A < B then
tex.write("-1")
else
tex.write("1")
end
end
end
lua.bytecode[1]()
}
\long\def\pdfstrcmp#1#2{%
\directlua{
l3kernel.strcmp("[[#1]]","[[#2]]")
}%
}
\pdfstrcmp{a}{a}
\bye
@PauloCereda For my purposes, I don't actually need to worry about the local business, so
\directlua{
lua.bytecode[1] = function ()
l3kernel = { }
function l3kernel.strcmp (A, B)
if A == B then
tex.write("0")
elseif A < B then
tex.write("-1")
else
tex.write("1")
end
end
end
lua.bytecode[1]()
}
\long\def\pdfstrcmp#1#2{%
\directlua{
l3kernel.strcmp("[[#1]]","[[#2]]")
}%
}
\pdfstrcmp{a}{a}
\bye
works
@PauloCereda For my purposes, I don't actually need to worry about the local business, so
works

1:01 PM
@JosephWright: ah, you are just considering l3kernel to be a function repository. In that case, it works. :)
So don't go with the module approach.

1:26 PM
@JosephWright l3kernel.strcmp("[[#1]]","[[#2]]") looks very dangerous to me, #1 and #2 can contain anything, thus arbitrary Lua code can be executed. The arguments should be "sanitzed", e.g. \luaescapestring (without \noexpand, because the Lua code should not see \luaescapestring, but its result.

@HeikoOberdiek I was fiddling around trying to get the 'Lua native' escape business to work: can't seem to make it happen, so for the moment will stick with \luaescapestring
@PauloCereda Module = namespace prefix, as I understand it
BTW, the bytecode business is there for format mode, where we want the Lua to be 'baked in' and which LuaTeX doesn't currently do automatically

@JosephWright I do not know, how the bytcode register numbers are managed: "Who says which number can be used by whom?" An alternative could be adding the initialization code via the token register \everyjob. \dumped code looses the Lua code of \directlua, thus the definition of l3kernel.strcmp can then done in \everyjob.

@HeikoOberdiek Finding docs on this is tricky: I have the code for this from Taco

@JosephWright Yes. :)

@HeikoOberdiek As you say, we could do the entire Lua part in \everyjob (currently we just reread the bytecode)

1:37 PM
@JosephWright \pdfstrcmp must work with arbitrary strings. Consider if #2 is ]]"); nowRunningEvalLuaCode() --.

Am still fiddling about a bit

@HeikoOberdiek TeX injection. I like it. :)

@HeikoOberdiek I guess you can understand why I'm asking these questions :-)

@JosephWright As you see, the function is very trivial. But how the function gets defined or how it is safely called, is complete different matter and non-trivial at all. Additionally the primitves \directlua must be checked, whether it is available; version check of LuaTeX; \luaescapestring does not exists at the start of iniLuaTeX and needs to be enabled, ...

1:54 PM
0

In many fonts I come across, I see that the glyph uppercase i - I is same as lowercase l. Why is it so? Why don't font designers add a differentiating factor between the two glyphs? This specially creates a problem with the word 'Ill'. Whether to read it roman 3 or short for 'I will'. Extra ...

2:04 PM
@PauloCereda No, it's not awful. It's appallingly abominable.

2:20 PM
@egreg :)

@PauloCereda maybe @egreg would be happier if you did \mathrm{I\!Interv}

2:38 PM
@DavidCarlisle Wow!

@AwalGarg blame those Romans:-)

Ahoy.

@Pops Quack.

How big of a deal is access to the TUG members area online?

@Pops You mean in terms of exclusive content?

In terms of how useful you would find it, I guess? I know absolutely nothing about it.

3:00 PM
@Pops Ah got it. :)

Apparently access to that is a new perk of institutional membership, but they grant access based on IP, and trying to set up and maintain a whitelist for every TeX SE user would be madness.
So I'm trying to figure out whether and how to pursue this.

@Pops When I got the institutional membership, TUG contacted me and gave me a login/password.

Apparently they're widening their net a bit.

3:46 PM
@PauloCereda: Duck mode registered.

@IamwhoIsayIam LOL

@PauloCereda :-)

1 hour later…
5:09 PM
Why cannot I see the images in pstricks.net/pstricks/examples/all?

@HeikoOberdiek Don't worry we have some tests (for things we support)

5:25 PM

@JosephWright I see. I think I am the only person cannot see the images.

@JosephWright I really need to work on the test framework. :)

@PauloCereda Not that sort of test!
@PauloCereda But yes ...
@IamwhoIsayIam No pictures here either

@JosephWright It could be. :)

@JosephWright OK. Thanks.

5:48 PM
Fortunately, there was a regional train, which is slower than my usual one. But, you know…

6:02 PM
:) ^^^^^

6:43 PM
@PauloCereda tried latex2e 2014 yet?

@DavidCarlisle I'm scared. :)

@PauloCereda well no need to be sacred, any faults we'll blame some bloke who promised to write us a test framework...

@DavidCarlisle Yay! Oh wait.

@PauloCereda You can try the .tds.zip files from latex-tds, pre-release for testing is available at dl.dropboxusercontent.com/u/71873949/tmp/latex-tds/base.tds.zip and same path for tools.tds.zip, graphics.tds.zip and cyrillic.tds.zip. You can install them in a separate TEXMF tree for testing.

@HeikoOberdiek Oh separate TEXMF. Good idea. :)
@Nicola: I think I have a solution for the file with spaces.aux problem in arara. I will fix it in v6, and there will be a lot of goodies included. :)
@Heiko: speaking of which, I think I might ask for your insights (as well of David's [if he doesn't charge], egreg's and other blokes) when writing the upcoming predictive interface for arara`.
@DavidCarlisle: hey speaking of test framework, if you didn't break my code with your bullet-proof logic, the tool would be ready by now. :)

7:03 PM

@HeikoOberdiek Nah don't worry. :) I'm in a hurry in the last few months. PhD stuff is really draining my energy. I hope I can come back to coding soon. :)
@DavidCarlisle: ^^^ is this a valid excuse for not working? :)

@PauloCereda No! the tex community relies on phd students procrastinating and avoiding writing their thesis. If people start saying they are too busy working on what they're supposed to be doing to work on TeX, then the whole system breaks down.
13

@DavidCarlisle Sir, now I want my thesis to have at least twenty chapters, so I can add twenty epigraphs from you. :)

1 hour later…
8:24 PM
If anybody has thoughts on this, I'm scheduled to print tomorrow at noon :(
0

I'm making a poster with beamer. The department's printer has a specific printable area that I must stay within, but I'd like to have a gradual fade-out of my background pattern. How can I fade out to pure white by the time I hit this printable area? (Let's assume a .5in long-edge margin and a 1...

@SeanAllred I thought you probably needed to know about tikz, so moved on to look at another question:-)

@DavidCarlisle Haha! You know I did think about solving it with tikz; setting something like a transparent rectangle node and then fading out to pure white along the margins

1 hour later…
9:31 PM
@DavidCarlisle around?

@FrankMittelbach yep

9:49 PM
@DavidCarlisle seen my mail?

@FrankMittelbach yep replied just now (twice) but had replied on latex-l first testing...

10:06 PM
@FrankMittelbach damn my connectivity is coming and going, but it's not impossible that I'm wrong:-)

10:17 PM
@DavidCarlisle could be .. but I doubt it ... nevertheless I like to sleep over it

@FrankMittelbach good plan (I'm going to mail a plan b in a bit to the team list, but no need to wait up, it's late and I need to test a few things:-)

@DavidCarlisle moving to the wine cellar stage then

1 hour later…
11:47 PM
@DavidCarlisle Plan B = rollback. :)

@PauloCereda that's plan c

@DavidCarlisle ooh! :) Gotta love plans! :)