« first day (2158 days earlier)      last day (2758 days later) » 

1:07 AM
Debate is on!
 
 
7 hours later…
8:09 AM
@DavidCarlisle Working on my xparse talk: turns out some of the ideas you/Chris/Frank had were not so bad :)
 
 
2 hours later…
9:44 AM
It was awful.
 
@PauloCereda Trump's going to win
 
@JosephWright :(
 
yo'
10:04 AM
@JosephWright wanted WW3? Here you go... :-/
 
10:15 AM
@JosephWright Really? Was he that good in the debate or was HRC that bad?
 
10:25 AM
@wilx The analysis I've seen says Trump didn't loose badly, which is enough
 
 
1 hour later…
11:38 AM
does anyone use refTeX here?
 
@Werner: when you are online, could you poke me? :)
 
 
3 hours later…
2:24 PM
@PauloCereda How shaky was the bus?
 
@egreg Very shaky. :) And there were no seats in the middle, so I had to sit in the back chairs, which are shakier than the usual shaking. :)
 
@PauloCereda I'm envious, I have to go with this instead of a nice shaky bus
And this is the car I usually seat in
Unbearable. :P
 
@egreg Oh my!
I am envious now. :)
@egreg: does it have wi-fi?
 
2:44 PM
@PauloCereda They charge me 20% more for the monthly fee.
@PauloCereda It has, but it's useless. I go 4G
 
@PauloCereda I'm online...
...getting ready to get on the bus to work...
...there might be slight delays in my response...
 
@Werner Gonzalo is fine. :)
@egreg oh I see.
 
@PauloCereda :)
 
@egreg I never managed to get a 4G connection around. It's only 3G and the area is very restricted. I only use WhatApp, Threema and email on my smartphone.
@egreg I poked him, he poked back. :)
 
yo'
@egreg for 1st class?
 
2:58 PM
Joseph Wright has added an event to this room's schedule.
 
@yo' 20% more than the standard first class fee for the "new" trains
@yo' After the first year as a commuter, I decided that I would only go in first class. It's my 21st, now.
 
@egreg ooh fancy
I never travelled first class. Only coach or in the lugagge thingy.
 
@PauloCereda The monthly fee is about 60% more than second class, but the difference in comfort is not comparable.
 
@egreg oh!
 
@PauloCereda Thanks!
 
3:08 PM
@PauloCereda Just one aspect second class cars have about 90 seats, first class have 54. Not to mention the wine. ;-)
 
@egreg Do they have Coca-Cola too? :)
@Werner My pleasure. :) I can poke everybody. :)
 
@PauloCereda If you really want it.
 
@egreg and Pringles.
 
@PauloCereda ...and have the funds to support it, of course.
 
@Werner Ducks are very good at poking. :)
 
3:11 PM
@PauloCereda No Pringles, I'm afraid. But this month they're offering chips.
@PauloCereda Of course the comfort on modern trains is nothing compared to what we could have in the past
@PauloCereda I travelled on these cars in 1975 and also later; maybe the last time was around 1980, from Brescia to Bergamo
And it was regular service, not historic trains.
 
@egreg Oh my!
 
@PauloCereda Very nice cars, those! There was a door between every pair of seats, so they were familiarly called "cento porte"
 
@egreg It must be very classy!
 
@PauloCereda They actually were for third class; when it was abolished, they were upgraded to second class by changing the number on the doors.
 
@egreg oh my!
 
3:26 PM
@PauloCereda Simple and effective, isn't it?
 
@egreg Indeed!
 
yo'
3:55 PM
@egreg so it's "first first class"? :)
 
@yo' More or less: they're introducing newer trains (actually more than ten year old; the ones we have now are 30/40 year old, refurbished) on the line, so they charge more for them.
 
4:30 PM
@JosephWright just seen Chris at the SPQR evemt (which was a surprisingly good day really)
 
@DavidCarlisle Didn't know he going!
@DavidCarlisle A celebration of a life well-lived, I assume?
@DavidCarlisle How is Chris?
 
@JosephWright range of subjects covered from classical greek to tex to xml to tei to ...
 
@DavidCarlisle :)
 
@JosephWright as ever:-) JF also. (and Phil gave a nice overview) podcasts will be online later
 
@DavidCarlisle I'm not sure I can imagine Chris and JF in the same room!
 
4:33 PM
@JosephWright neither did the organisers:-)
 
@DavidCarlisle I don't think I've ever met Phil Taylor
@DavidCarlisle Ah
 
 
1 hour later…
6:23 PM
@egreg -- i don't think so ... that phil taylor looks about twice as big as the one i know, and the phil i know also has very white hair (and no beard).
 
3
Q: Is there a LaTeX/ConTeXt Rosetta stone out there (i.e. LaTeX and ConTeXt files producing the same output)?

JohnWCompared to LaTeX (as a set of macros, a "language", not as the latex implementation), there are far fewer ConTeXt documents available. Yet ConTeXt seems really interesting and I want to explore it a bit more. What I would be very interested in is a kind of Rosetta stone for LaTeX and ConTeXt. I...

^^^ This is definitely »too broad«.
6
 
6:51 PM
@egreg don't be silly, he's not american:
 
7:02 PM
@HenriMenke Yup
@DavidCarlisle > After some ten years of development and testing
@DavidCarlisle I like the last bit!
 
@JosephWright the "We want to thank David and Joseph" bit?
@JosephWright building now....
 
@DavidCarlisle :)
@DavidCarlisle Guess once Akira has a build made I'll look at l3build
 
@JosephWright still haven't got biber running (after multiple emails with Ken brown (who does cygwin texlive) and Philip Kime) strace shows it is looping in dll loading ken found a similar looking issue on a cygwin forum but that was resolved by turning off virus checker and i can't do that (have admin access but somehow or other not for that and anyway can't really do that on a works machine) previous texlive i got it to work by using it as a perl script (not compiled) but now ..
@JosephWright... it needs a bleeding edge perl and i'd have to build perl from source as cygwin is .2 of a version behind :(
@JosephWright bibtex is a lot easier to install:-)
 
@DavidCarlisle :)
@DavidCarlisle Before Biber was in TL I used to compile my own, but not with the Cygwin 'fun'
 
7:21 PM
@JosephWright I don't think it used to need perl so close to the development release. cygwin has 5.22 but biber refuses to run on less than 5.24
 
@DavidCarlisle That's the version in Strawberry, so would have been fine
 
@JosephWright cygwin will probably catch up soon but that was the situation yesterday:-)
luatex still building....
 
@DavidCarlisle ^^^
 
7:42 PM
@HenriMenke New bug search starting off
 
$ lualatex
This is LuaTeX, Version 1.0.0 (TeX Live 2017/dev)
 restricted system commands enabled.
**\documentclass{article}
LaTeX2e <2016/03/31> patch level 3
Babel <3.9r> and hyphenation patterns for 1 language(s) loaded.

*\begin{document}
(/usr/local/texlive/2016/texmf-dist/tex/latex/base/article.cls
Document Class: article 2014/09/29 v1.4h Standard LaTeX document class
(/usr/local/texlive/2016/texmf-dist/tex/latex/base/size10.clo))
No file texput.aux.

*Hello World

*\end{document}
[1{/usr/local/texlive/2016/texmf-var/fonts/map/pdftex/updmap/pdftex.map}]
@JosephWright ^^
 
@egreg This release date seems to be completely arbitrary. It's just like in the Linux kernel where they just slap a major version number on the same code which was before 3.9.14.26r245 or whatever.
@egreg If they follow Google Chrome's lead we will have LuaTeX 20.0.0 by next year.
 
  $ texlua build.lua check
^^^^ what could go wrong....
 
8:03 PM
@DavidCarlisle Ask Mr Murphy
 
8:21 PM
$ ls build/test/*diff | wc -l
84
@JosephWright ^^
 
@DavidCarlisle is this good?
 
@PauloCereda well 84 failures wouldn't be great except....:
@JosephWright just ran over the diffs by eye looks like apart from one supercalifragilistic test they are all instances of the new discretionary logging that came up at 0.98
build/test/tl2e3.luatex.diff
::::::::::::::
*** ../build/test/tl2e3.luatex.tlg      2016-09-27 20:48:00.460379200 +0100
--- ../build/test/tl2e3.luatex.log      2016-09-27 20:48:00.472404100 +0100
***************
*** 95,100 ****
--- 95,102 ----
  ....\OT1/cmss/m/sl/10 h
  ....\OT1/cmss/m/sl/10 a
  ....\OT1/cmss/m/sl/10 p
+ ....\discretionary (penalty 50)
+ .....< \OT1/cmss/m/sl/10 -
  ....\OT1/cmss/m/sl/10 p
  ....\kern0.27779
  ....\OT1/cmss/m/sl/10 e
so I suppose it's just a matter of deciding whether to normalise that out in l3build or just to check in the .luatex.tlg versions,
just build/test/tlb0435.luatex.diff to check ..
 
@DavidCarlisle oh those are failures
 
@PauloCereda yes diff between what we get and the checked in expected log
 
@DavidCarlisle got it
 
8:39 PM
        FATAL ERROR
        Luaotfload requires a Luatex version >= 0.95.0.
        Please update your TeX distribution!
oops...
 
@DavidCarlisle Day three of reporting bugs
 
@DavidCarlisle 0<95, as everybody knows.
 
@JosephWright so I think we see how much testing against latex went in to this release....
 
@DavidCarlisle Sure: I'll look at it once I can get 1.0
 
@DavidCarlisle awesome!
 
@PauloCereda podcasts of all the talks will be appearing
 
9:21 PM
A bit dated, but I was suprised to see that webkit (safari) is actually still improving their mathml rendering:
https://webkit.org/blog/6803/improvements-in-mathml-rendering/
I heard also rumours that google is open to getting mathml enabled if harfbuzz would get opentype math table support, which I also saw commits on.
Could this MathML web dream come true at last?
 
@Kasper have faith:-)
 
I saw that now in Safari Technology preview, I can add eventlisteners to mathml nodes (like mouseover, click etc). I was quite suprised about that. In firefox I can not even set a style to a mathml node it seems.
 
@Kasper a long story:-)
 
What is a long story :) ?
 
@Kasper mathml but in particular the state of the dom api to event listeners and style, and whether it applies to mathml
 
9:36 PM
ah okay, well I really hope it will get there one day. I work on elearning company (making interactive math (applied math) books for university). Having native mathml would make many things easier for us (and give the students a better experience).
not sure why I did this overkill of parenthesis :P
 
@Kasper what can I say, I've been working towards that for 15 years, if it isn't there yet It's not my fault :-)
 
Oh wow :) That's great
Just checking your website :)
Do you happen to know Arjeh Cohen?
 
@Kasper yes
 
Okay, I work for him :)
 
@Kasper or check the mathml spec (I'm the editor)
 
9:40 PM
Ah, what cool
 
@Kasper say hello (not seen him for ages since we worked on the OpenMath projects)
 
I'll do! I'm now writing an interactive calculus course with him. The WYSIWYG editor of the project is actually based on OpenMath still:
http://mathdox.org/formulaeditor/
 
@Kasper yes, thought it might be:-) I hope we'll still be able to do EU projects in the future:(:(
 
Not sure if I understand your eu project remark, I guess Arjeh hasn't told me everything about that yet :)
 
@Kasper oh sorry I meant because of the UK leaving EU (the openmath projects were funded by EU for many years)
 
9:52 PM
ah I see
We are currently considering switching from mathdox editor to mathquil though. Mainly because we want to get copy pasting working. Also trying to make copying subexpressions from mathjax possible. But it's not that easy.
It seems like it would be more straight forward if native mathml was used.
 
10:13 PM
It seems like Fred Wang is doing commits for OpenType Math tables for harfbuzz today:
https://github.com/behdad/harfbuzz/commits/master
 
@Kasper oohhh
@Kasper yes we have a web based annotation system for our documents at work and with firefox you can highlight/annotate math just the same as text but in every other browser I provide a button that removes the mathjax then you highlight the "raw" text from unrendered mathml then turn mathjax back on with (if you are lucky) highlights in the math, but it is so much more natural in firefox
 
@DavidCarlisle ah interesting approach
you may understand better than me if this is significant for getting mathml in blink, I saw fred writing this:
> In order to add MathML support in Blink, it is very easy to import the OpenType MATH parsing code from WebKit. However, after discussions with some Google developers, it seems that the best option is to directly add support for this table in HarfBuzz. Since this library is used by Gecko, by WebKit (at least the GTK port) and by many other applications such as Servo, XeTeX or LibreOffice it make senses to share the implementation to improve math rendering everywhere.
So basically, it seems that getting it right in HarfBuzz would make Firefox, Safari and Chrome have kind of the same math rendering? Using mathml?
 
10:29 PM
@Kasper yes it might be (google are hard to guess:-) As you may know they had mathml for a while but basically took it out as they stripped out lots of code when they forked from webkit
@Kasper that's Fred's long term plan and I hope it works!
@Kasper then we need to get harfbuzz into luatex....
 
@DavidCarlisle harfbuzz all the things!
@DavidCarlisle It seems like some real hope for mathml, I kind of lost some hope that mathml would ever get there since peter from mathjax lost faith in it.
 
when will apple invent a fingerprint sensor that works when you've been eating peanut butter
 
10:50 PM
@PauloCereda They don't want we to eat peanut butter, it is bad for health.
 
@Kasper peter got impatient but all good things come to those who wait...
 

« first day (2158 days earlier)      last day (2758 days later) »