« first day (1615 days earlier)      last day (1707 days later) » 

9:09 PM
@egreg For a very brief moment, I'd thought you'd actually asked a question.
 
@SeanAllred Huh?
 
@JosephWright As letters -- 'bop' and 'pop' sound very similar (at least with an American accent)
 
@SeanAllred Ah
 
Thanks I'll try this. Been looking for something more automated than flowfram for a long time.
 
@SeanAllred I'm wiser than that. :P
 
9:16 PM
@egreg A very brief moment -- too short a time to actually read any of the question-widget in chat :)
 
@SeanAllred :)
 
yo'
my mobile 4G connection doesn't work :-(
 
@1010011010 it is sound but hard - TeX is not made for multiple columns really ... the LaTeX3 Or is supposed to resolve that one day and as @DavidCarlisle pointed out one could build something on top of multicol based on the ideas outlined in that answer but it is really requiring a different approach compared to the balancing ideas that multicol persues
 
9:36 PM
@FrankMittelbach I'd really like to help out but I'm not savvy enough with the functionality to produce something as well-coded as what's currently in place. If I can do or mean anything in helping this idea become something tangible, I'll be happy to do it. Judging from the people I spoke to there's a reasonable interest in this kind of functionality without the intrusive changes in workflow from e.g. flowfram
 
@1010011010 I playing with this since 2000 (and earlier) .. so ... have a look at latex-project.org/papers/xo-pfloat.pdf basically in full generality (ie automatic placement as well as balancing etc) the problem is simply unsolved and by now that is starting to get recognized in the DocEngineering research community finally as well.
 
yo'
@FrankMittelbach the problem is that it is unsolvable.
 
@yo' yes and no ... you can build "impossible documents" for any algorithm but you can do that even with \pagenumbering{roman} and a single column latex document with suitable cross references
but you can design algorithms that would work on a large class of documents
 
yo'
@FrankMittelbach yeah. But still: my opinion is that the reliable way would be to place all floats in code at the beginning of a "chapter block" (basically anything separated by \cleardoublepage) and say precisely where they belong.
 
9:53 PM
@yo' that's what the xor algorithm was offering as an alternative (it does write a placement file or take the one you provide so you can have both ways ... there are many cases where you want such level of control and then specifications such as flowfram or anything else is really what you need/want. On the other hand there are good reasons to want automatically produced documents as well (or automatic as far as possible)
anyway no here nor there as it doesn't exist yet
at least not in a relibale productionready form
 
yo'
@FrankMittelbach yes. As you say, you need both ways. As a journal typesetter/copy ed, I need to be able to position the floats as easily as possible. I get the final version of the paper, so I can do this easily by hand. No further edits that would need a float position modification should happen.
 
@yo' that was precisely the idea with xor ... you get a file that gives oyu the calculated layout and you essentially change that slightly (or drastically) and it would obey.
 
yo'
@FrankMittelbach any recent progress on that? :-)
 
@yo' no not looked at it for quite some time (for a number of reasons) but that will change soon
 
yo'
@FrankMittelbach I'm happy to hear that :)
(btw, I'm currently trying to get some money for the Darmstadt journey paid by my university -- that would be a success. I hope to meet you all there)
 
9:59 PM
@yo' seems that Germany it becoming a popular location :-)
night folks ... one minute to midnight ... need to go in search of some nice book
 
@yo' Cool
 
yo'
@FrankMittelbach good night to you too! I still need to get my ass out of my parents' place.
@JosephWright we've got some money left on the student grant. I just need to convince my supervisor that it makes sense. The grant treasurer has already agreed.
 
Thanks for the information. I'll look into the document you sent me Frank. Good night all.
 
10:23 PM
No arara in TeXworks 0.4.6. :(
 

« first day (1615 days earlier)      last day (1707 days later) »