« first day (327 days earlier)      last day (4596 days later) » 

7:01 AM
@BrunoLeFloch No, I would have been 7 or 8 :-)
 
 
3 hours later…
9:41 AM
@JosephWright Just the best age to begin learning TeX! I was 3 (more or less) :-P
 
9:54 AM
Is anyone interested in testing a bash completion script for texdoc?
 
@Raphink What should it do?
 
@egreg complete the command with the known manuals. Just drop it in /etc/bash_completion.d (if you're on a GNU/Linux OS) and source /etc/bash_completion. Then when you press tab, it should complete, like texdoc l2tab completes to l2tabu l2tabuen l2tabues l2tabufr-heavy l2tabufr-light l2tabuit
 
@Raphink I'll try. My usual shell, you won't believe it, is tcsh. ;-)
 
@Raphink It's a great idea!
 
10:09 AM
@egreg I used tcsh for a long time. It's very good. =)
It reminds me of a Futurama episode:
- These are the dark matter engines I invented. They allow my starship to travel between galaxies in mere hours.
- That's impossible. You can't go faster than the speed of light.
- Of course not. That's why scientists increased the speed of light in 2208.
 
@PauloCereda At that time the choice was rather restricted. I found out tcsh and managed to compile it on the HP/UX workstation I had access to, where the default shell was csh. The first versions of Mac OS X used tcsh as default shell, so I continued.
 
10:25 AM
@egreg Cool! I used it during my Slackware times. It was perfect for my use, I didn't use a GUI so I had to rely on a good shell.
 
@PauloCereda That was the time when we had to compile the TeX system programs "by hand", but kpathsea was already available. Only later on the first version of teTeX appeared. I think I contributed the HP/UX binaries, at least once.
 
@egreg Was the compilation time very long?
 
@PauloCereda The longest part had been before! HP/UX shipped with only a very basic C compiler (one had to pay for the big version). So we had first to compile gcc (v. 2, I believe) which was done in three steps: the first one with the basic compiler, then the second gcc compiler was built with the first and then the full compiler with the second (later we could build gcc3 with gcc2). Compilation of TeX required half an hour or so, IIRC.
And I still know nothing about C. :)
 
10:44 AM
BTW, it's the usual procedure for bootstrapping and compiling GCC for a non-cross-platform use:

* GCC stage1 is compiled using the host C compiler
* stage2 is compiled using stage1
* stage3 is compiled using stage2
* stage2 and stage3 are compared to ensure that the produced compiler works
 
@AndreyVihrov Cool! :-)
 
@AndreyVihrov I haven't compiled gcc for years. I thought that it was the usual procedure.
 
@egreg neither do I. :-P I remember when I was first introduced to parallel and concurrent programming with C, I didn't sleep for a whole week.
 
11:24 AM
@PauloCereda All I did was ./configure; make all or something like that, trying to figure out how to solve problems when they showed up.
 
12:18 PM
Actually my only programming language is TeX. :)
 
12:48 PM
@egreg never been interested in learning another one?
 
12:58 PM
@Raphink TeX is Turing-complete. :) I can do some basic shell programming; I have to learn Lua, some day.
 
Sure egreg, I know quite a few Turing-complete programming languages. Yet, I wouldn't use the same programming language depending on what I want to achieve. It's not about which language is best or which I prefer, it's just that some are good for some things, others for other things. I wouldn't make a website in TeX or try to hook Perl in TeX, because PHP or Python are much better for website programming, and Lua is perfect to hook in TeX.
 
@egreg Lua is very nice. I started learning it, but sadly I had millions of things to do and could not dedicate myself. There are great resources on Lua, specially Programming in Lua, by Roberto Ierusalimschy. Ah, Lua is a brazilian project!
 
1:35 PM
@Raphink Lua is known to be good as a scripting language embedded into bigger software, thus no wonder this applies to TeX, too
 
2:04 PM
@AndreyVihrov sure. My point is just that some languages are good for some tasks.
 
3:04 PM
@MartinScharrer Can you look at tex.stackexchange.com/review/suggested-edits where there is a proposed edit to a closed question?
 
@egreg Mmm, I wonder what to do.
 
@MartinScharrer The provided example doesn't present any unexpected behavior.
OTOH, rejecting an edit by a freshly registered user is not a nice policy.
 
@egreg I just added a comment pointing out the possibility of unwanted spaces. This is often the case.
 
Such a shame the slides weren't made using LaTeX! webcast.web.cern.ch/webcast
 
@egreg: Could you up-vote hat comment so that it is directly visible. Thanks.
 
3:22 PM
@GonzaloMedina Not even Keynote. I suspect PowerPoint (2007+). :-(
 
I'm slowly getting somewhere...
 
0
Q: Tikz figure in Beamer shifting way to the right of the frame

Dave KincaidI'm trying to include a tikz drawing into a Beamer slide. When I do it like this: \begin{frame} \frametitle{Overview} \input{smartservice-diagram.tex} \end{frame} the image is way too big. I tried this: \begin{frame} \frametitle{Overview} \resizebox{6in}{!}{\input{smartservice-diagram...

@MartinScharrer I flagged this because I don't see the point of editing a closed question. But someone approved the edit, so either the editor (who isn't the OP) should be asked to ask a separate question or the question should be reopened, no?
 
@AlanMunn Indeed. I tend to reopen it.
 
3:39 PM
@MartinScharrer I approved the edit, for the reason I stated some messages above: not rejecting a freshly registered user. Asking him/her to provide a better minimal example would be useful. But I won't object if the mods decide to roll back the question.
 
@egreg Sorry, I didn't notice the previous discussion on the question. If it's possible to get an answer, then we should vote to reopen it. The new edit isn't a full MWE either, unfortunately. w.r.t. the spaces issue, is it possible to construct an example that shows how spaces could cause this and make that the answer?
 
@AlanMunn @egreg: I reopened it and posted an answer based on my comment. In the case of the MWE it seems to be really just spurious spaces.
 
3:55 PM
@MartinScharrer I too added some other remarks. You can vote for it without worrying about adding to my rep. :) But I already voted for Martin's answer, so the question is not any more among the unanswered ones.
 
 
1 hour later…
5:24 PM
Step 1: detecting all rivers (even not bad ones): done :-)
It occured to me that spaces don't have to overlap. If I allow diagonal rivers up to 45°, then the overlap is - or + the line height
 
@AlanMunn: About the CTAN upload automation again:
Jul 7 at 21:06, by Alan Munn
@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?
As you might have seen I published a ctanupload script in the meantime.
I also wrote an extended Makefile which uses it: bitbucket.org/martin_scharrer/ltxpkg/src/default/ydoc/Makefile
 
@MartinScharrer: Funny I wrote to Robin Fairbairns about that just a few days ago. I suggested an automated system based on GnuPG signatures, as done for Debian/Ubuntu packages.
 
It also includes the old browser based upload as an alternative.
 
5:39 PM
+ tds.zip standard
so, a tds+zip + a changes file signed by a GnuPG key (and containing the tds.zip md5sum, along with the update message / changelog)
which could be authenticated using the GnuPG key and processed automatically
I'm used to automating this kind of things with Debian/Ubuntu packages
 
@Raphink Yeah, GnuPG signatures would not be bad, but I don't think the CTAN admins are ready for that.
 
well it's not very hard to manage, it's just a keyring with a list of keys allowed to upload for a list of packages
if the interface is well made, it's quite easy
and it's a very secure authentication method. You can be 99.99% sure that the guy who uploaded IS who he claims to be, and so you can be OK to automate the process.
With the current system, I could pretend to be anyone and update whatever.
You can't automate updates unless you have a strong authentication method.
 
@Raphink The only automation I implemented is for the upload. It saves you from filling static information into the webform over and over again. It is still handled manual or semi-automatic on the CTAN side.
 
Yes, and I use your script @MartinScharrer. I still think an automated system on the CTAN side would be great. It could be on top of the existing system. All it requires is an ftp server, GnuPG keys and a bunch of scripts running in cron.
Automating the client side is nice, but it still leaves a lot of work for the CTAN maintainers :S
 
I like the idea of a signed package.
 
5:47 PM
On the servers I manage at work, I have GnuPG authentication, and there's something like 200 uploads a day. If it wasn't automated, I'd just go crazy ;-)
@PauloCereda: yes, the idea is very simple, too. You upload two files: a .tds.zip which contains the actual data, and a .changes, which is a file containing your name, a changelog (the change message for the announcement), and the checksum for the other file (size + md5 + sha1 + sha256 for example).
This last file is a text file and is self-signed by your GnuPG key. This way, when you upload the files, the .changes is checked against a keyring, authenticating you, and the .tds.zip file is checked against the checksums (which have been authenticated by your key). The upload is fully authenticated, the .tds.zip can be automatically included in the CTAN and the announcement made. Any mistake is full authenticated and you know who to blame.
 
@Raphink Sound interesting. I wonder what the CTAN adims say about it.
Anyway, I have to run. See you guys tomorrow.
 
You can also allow several keys on one package, allowing multiple uploaders.
Bye @MartinScharrer
 
See ya, @MartinScharrer!
 
6:47 PM
@egreg Are you planning to post a sparkline answer to tex.stackexchange.com/questions/29293/… ? I was curious so I started to play with it and got an example ready
 
7:42 PM
12
Q: Did Luke accomplish anything by going to Cloud City?

PeterDCIn Star Wars - Episode V: The Empire Strikes Back, when Luke Skywalker goes to Cloud City, he pretty much Falls (almost literally) into the trap set for him Loses a duel against Darth Vader Loses his (father's) lightsaber Gets his hand crudely amputated And Is barely saved by those he came ...

I like one of the comments: He got some good Chinese food there.
 
7:55 PM
@NN I only remembered its existence. Go forth.
 
8:39 PM
Posted by Jeff Atwood on September 23rd, 2011

It’s been a little over a year since our last improvement to the bounty system. Question bounties have been working well enough that we’re comfortable encouraging even more use of the bounty system.

We used to limit people to one question bounty at a time, but now you can have up to three simultaneous question bounties. We also show the history and number of bounties you’ve started or participated in on your user page, on the bounties tab.

Upon further reflection, we realized that it can be difficult to tell exactly what a question bounty is for. That is, what is the bounty own …

Now bounties might have descriptive texts to inform why they are offered.
 
Cool
 
@StefanKottwitz (or any other moderator): Could you "un-CW" this answer of @Raphink?
 
@lockstep: the thing is I've edited it more than 10 times and I keep editing it, so it gets to be CW automatically
 
@Raphink I don't think so -- the mods have the power to remove CW from a post that has been edited at least 10 times. (Joseph Wright already did it with an answer from Audrey.)
 
8:49 PM
ah ok :)
thanks for caring then @lockstep
 
9:03 PM
@lockstep Thank you for reminding me to post meta.tex.stackexchange.com/questions/1774/… :)
 
@lockstep I changed the wiki status
Contest:
0
Q: Who designs the best tag wiki?

Stefan KottwitzTags are a map of our site. Tag description pages can be like small home pages, including information, internal links and links to helpful external resources. And they can be written with markdown syntax. Many users alread helped in writing wiki entries for the most used tags. What is the most i...

 

« first day (327 days earlier)      last day (4596 days later) »