@JosephWright Is there an easy way to maintain backward compatibility with old versions of LaTeX3? I've just had a query regarding the hobby package. That package needed updating when some functions in L3 were changed, but there are lots of old distributions out there that don't yet have the update. Any "best practices"?
@AndrewStacey This is a tricky one at least in part as we don't (yet) want to find ourselves tied down by an ever-growing set of macros that cannot change (the issue for the LaTeX2e kernel).
@AndrewStacey My position with siunitx is that the version have to match: the first test I do is for version of expl3, and I bail-out if it's wrong. My thinking there is that most people get packages 'together', so either (1) have an old siunitx and matching expl3 via say Ubuntu,TL2011, ... or (2) have an up-to-date TL or MiKTeX and so can update the support easily.
@AndrewStacey We've wondered about taking the position that we don't remove anything else, but only deprecate, but I don't feel that we are in a position at present that this is workable.
@JosephWright I certainly wouldn't want L3 to get locked in so it's more about at the package level. I guess the problem is that since hobby is new, people download it directly from CTAN (especially as the latest version added quite a bit more functionality).
The difficulties arise mainly with functions being renamed (though there's a bit with the fp library). So I could test for the version of L3 and put a few \lets in place.
But maybe I don't want to go down that route either! The alternative would be to test for the version of L3 and if it is wrong then only load the Quick Hobby algorithm (since that doesn't use L3).
@JosephWright So with siunitx, do you update the package every time that L3 gets updated? Even if it is just to change the version number?
@AndrewStacey I update siunitx where there is a 'breaking' change in a release of l3kernel. That doesn't happen every time there is an l3kernel release, as it depends on the exact changes.
I do worry pretty hard about this sort of thing, hence the rather dull deprecation e-mails that get sent to the LaTeX-L list
@AndrewStacey We've not had anything dropped for a while: there was something of a 'clean up' about 6 months ago. We have a pretty small number of functions that now might change: one obvious pair are \tl_to_lowercase:n and \tl_to_uppercase:n, which are badly wrong (the pattern of names means these should be expandable, which they are not).
@AndrewStacey It's hard for the L3 distribution to provide compatibility with every previous version but typically a package such as hobby only uses a fraction of that and so there you could (if you want) just define the commands that you need. Depending what changed this can be as simple as renaming things or just stealing the definitions from the new sources.
@JosephWright sorry not me The wifi had a little fit and kept telling me it could not connect so chat kept saying unknown error: retry. Then I got logged out:-)
Hey guys! Quick question about TeXMaker: is there any way for me to see a quick output of some minor changes I do to the document? Instead of doing the whole compilation every time I just add a comma and want to see what happens?
@ach Optional arguments in [] are used for various purposes. In the case of "moving arguments", such as captions and sectional titles, the optional argument is "what moves", that is, is used for lists, headers and similar things that require writing to the auxiliary file (this is what makes "moving arguments" problematic with some commands, that are therefore called fragile).
I would really love to have a software or tool which provided real time compilation of my LaTeX document (to be able to see the final pdf or dvi document as I type the latex code), especially for when I am drawing pictures using tikZ. So far, I know of two things, but none of them work for me. Th...
I have exer environment which takes two arguments, one is a reference to the question number for which the solution is being given and the other is difficulty level. I want to sort these exers based on the order of questions, irrespective of the order in which they appear in the source code?
@ach [] doesn't change the functionality of the main argument really but it is writing to the toc that is the moving argument and the default value of the optional argument is the main one, so fragile commands don't work in the optional argument, but if you don't supply the option then the main argument is used (so fragile commands dont)
Of course, David Carlisle has already mentioned in a comment the correct solution, which is to use the fixltx2e package. Among other goodies, this package will ensure that figures appear in the same order in the output as you specified them in the input.
If, however, you like how LaTeX does it ...
Update the l3kernel and l3experimental packages. You will find in the latter bundle a package meant for sorting, namely l3sort. Here is some code that defines a sortbibliography environment, which (1) captures everything until \end{sortbibliography} (using the environ package, which you probabl...
@DavidCarlisle @egreg ah, that makes sense. It's confusingly worded in the Companion however ".. in fact, LateX allows multi-paragraph expressions only if the short-text argument is present."
and then it goes on to suggest the error that you get - but I got a different one
@egreg Do you remember the \findrootlanguage patch you provided for glossaries? It seems the \@@input babel.sty bit is causing an error, but I don't know why. There's a thread on the tex-live group about it. I thought the category codes were those in use when the command was defined. (Page 36 of glossaries.pdf)
Could someone please help me with this? (I commented last in the thread) http://tex.stackexchange.com/questions/70158/syncing-feature-between-editor-and-pdf-viewer-not-working
@NicolaTalbot Adding \makeatletter after \begingroup in the definition of \findrootlanguage allows to go past the \spacefactor error. But this is bound to fail miserably: the patch was tailored on the form of babel.sty 3.8 that started with a list of \DeclareOption statements; now it's completely different. :(
Cbs228 writes "During last week's Google I/O conference, the company announced a replacement for its aging Talk instant messenger: Google Hangouts. Hangouts, which is only available for Android, iOS, and Chrome, offers closer integration with Google+. Unfortunately, the new product drops support for the XMPP instant messaging protocol, which has been an integral part of Talk for over ten years. XMPP delivers instant messages to desktop clients, like Pidgin, and enables communication between users on different instant messaging networks. Hangouts users attempting to communicate with contacts…
@egreg Oh drat. Thanks. I'll have to have a look at the new version of babel. It would be useful if it stored a list of the defined languages and the language root.
@NicolaTalbot IIRC, the problem was that \languagename used to store the language name as a string of catcode 12 characters, so a comparison test couldn't work. However, \bbl@main@language seems usable for the purpose.
@NicolaTalbot Maybe it's something that Javier can do better. The big problem was that Babel was in a suspended state, so it was impossible to get new features. I guess it shouldn't be too difficult to add something like \bbl@main@rootlanguage that refers to the ldf file corresponding to \bbl@main@language. Other packages might benefit from it.
@egreg Yes, definitely. I sent an email to the babel address last December asking about the possibility to test which languages had been loaded, but knowing the root language would be really useful.
Psmith, the TeX bot, in fixed font mode: Let's take a look at the last cricket results:
- Leicestershire v Glamorgan 240/5 *
- Netherlands v Northamptonshire
- Chennai Super Kings 192/1 v Mumbai Indians 13/1 *
Our cricket expert David might explain these results later on.
walterbyrd writes "In 2012, IBM started retiring the Lotus brand. Now 1-2-3, the core product that brought Lotus its fame, takes its turn on the chopping block. IBM stated, 'Effective on the dates listed below, [June 11, 2013] IBM will withdraw from marketing part numbers from the following product release(s) licensed under the IBM International Program License Agreement:' IBM Lotus 123 Millennium Edition V9.x, IBM Lotus SmartSuite 9.x V9.8.0, and Organizer V6.1.0. Further, IBM stated, 'Customers will no longer be able to receive support for these offerings after September 30, 2014. No serv…
Good old times of Lotus 123. :)
En general todos los artículos pueden enviarse en WORD, LATEX o PDF. En caso de que su artículo haya sido seleccionada para publicar en el Libro [conference name here] y se encuentre en formato LATEX o PDF, deberá reenviar una versión en formato WORD.
^^ This is sad. :(
In general, all papers can be submitted in Word, LaTeX or PDF. If your paper has been selected to be published in the conference book and happens to be in either LaTeX or PDF formats, you must submit a Word version.
@PauloCereda There is a mischievous solution to this demand: convert your Latex to Word for Unix (which is a fork of the v2.0 "Spaceman Spiff" version of MS Word from 1991) and subit that. Modern versions of Word don't even try to read Spaceman Spiff .doc files or versions older than v8.0.
When they ask why you've submitted in such an old version of Word, you can say that it is the old version of Word MS have released that works with Linux (a claim that is more or less true). It will do your standing with the program committee no good at all :-)
Regardless of the matter, please do not anonymize certain users if you are going to address them by the implications of the questions. If that's Papiro you have in mind say openly that you have Papiro in mind.
This is not a good way to discuss matters, since you are observing one user's behavio...
Do we have anything like a policy about edits that just delete thanks?
I've certainly indicated that I think that, while an edit that removes greetings and other fluff before the problem statement is good for the site, just removing thanks is not, and I'm aware that many agree with me. But I don't think we really have a policy to that end, at least not that I recall seeing.
I am drawing trees in LATEX. The qtree library seems to be what I need but the parent is always placed in the center of its children like this:
parent
/ \
child_1 child_2
(I'm a new user and does not have enough reputation to post an image.)
What I want is a right-growing tree lik...
@AlanMunn Is it just the screenshot or are the vertical rules in the left image (down at the bottom) of different widths? I mean, the top of each vertical rule seems to be thicker than the bottom.
I am also confused how child1 is displayed in the first tree (having no children it should IMO be lower than GChild1)
@dıʞsdoʇ It actually is the screen rendering it seems (I thought it was because the lines were drawn one on top of each other). When you show the page at 200% they disappear, and they look fine when printed. Thanks for asking, because now I can update the answer a bit.
@PauloCereda Just finished debugging a machine code which drives a 10 M€ worth machinery. I can safely say that some matlab employees are idiots or secretly hate mathworks.
@PauloCereda I for one am beginning to doubt whether I can entrust anything important to Google in the long term. Dropping Reader was, for me, a catastrophe; I've abandoned the memory-bloated Chernobyl that is Chrome; and dropping support for XMPP suggests a new Google, less interested in open standards and other signs of the absence of evil.
@CharlesStewart I think they are bad and have consistently rejected them if they came past me in the review queue (but they always got accepted by enough people in the end anyway)
@PauloCereda I'm not that worried about G for search (until the day you can't search without logging in) - I'm more worried about mail, chat, and calendar. Ferinstance, what happens if G stops supporting SMTP for email?
@Brent.Longborough Yes, I have the same concerns. And things are pointing towards that direction. Did you know Google is creating a new rendering engine and giving up on WebKit?
@PeterGrill Why? After \def\MacroToUse{\FormatText*}, when you say \MacroToUse{x}, TeX will expand \MacroToUse and then see \FormatText*{x}, which has the correct shape.
I'm having trouble aligning some of the elements.
I would like align Check Box and Cross Box to be same size so that they look nicer below is what i got so far.
Also it would be great if they can change in size with respect to environment with (some ratio 60% of text or so)
[Optional] have ba...
Of course, David Carlisle has already mentioned in a comment the correct solution, which is to use the fixltx2e package. Among other goodies, this package will ensure that figures appear in the same order in the output as you specified them in the input.
If, however, you like how LaTeX does it ...
Partial solution: the font is loaded with the package mathpazo.
Regardless of the matter, please do not anonymize certain users if you are going to address them by the implications of the questions. If that's Papiro you have in mind say openly that you have Papiro in mind.
This is not a good way to discuss matters, since you are observing one user's behavio...
I am drawing trees in LATEX. The qtree library seems to be what I need but the parent is always placed in the center of its children like this:
parent
/ \
child_1 child_2
(I'm a new user and does not have enough reputation to post an image.)
What I want is a right-growing tree lik...
I'm having trouble aligning some of the elements.
I would like align Check Box and Cross Box to be same size so that they look nicer below is what i got so far.
Also it would be great if they can change in size with respect to environment with (some ratio 60% of text or so)
[Optional] have ba...
