TeX, LaTeX and Friends

barbara beeton

00:30

@PauloCereda -- twitter.com/LegoLostAtSea/status/1329139105805049856

barbara beeton

02:34

@Plergux -- For a sheep person, a discussion of spinning and weaving in Iceland and other northern societies, launching a book "The Valkyrie's Loom": youtube.com/… Kind of academic, but you should be able to handle that.

Davislor

04:13

Would appreciate feedback if you have a moment to spare: github.com/latex3/babel/issues/106

Ulrike Fischer

06:06

@Davislor I don't know much about greek apart from that the change is probably my fault as I told the maintainer that it odd that the package changed behaviour depending on if fontspec is loaded before babel or not. But the accent part sounds as if the patterns are wrong and should probably be to the hyph-utf8 maintainer.

Davislor

Chsnging to a different Greek language file did not seem to make a difference. But, yeah, hyphenation should never be allowed before a Unicode combining character.

That said, I don't know if I was doing something with the font map that messed things up.

I haven't yet tried to reproduce the bug without a custom mapping.

Plergux

@barbarabeeton Sounds promising. Though I tend to have a slight aversion to when people start talking about "viking" stuff without it being directly related to the topic. I find it to be sensationalism and unnecessary. That's probably not very "icelandic" of me, but my argument is that if you can't make it interesting without it you're missing something.

Ulrike Fischer

@Davislor oh, didn't you try with direct unicode input? That you should naturally do, I don't remember when xetex does hyphenation.

Davislor

When I wrote that post, I was focused on getting beta-code input to work.

That was just a bug I happened to run into along the way, but found a workaround for.

But, you're right, I should report it to the maintainer. I've got the Unicode input and a little program I wrote to normalize to any of the Unicode forms.

So it won't be hard to test.

Paulo Cereda

08:23

@barbarabeeton ooh

@Plergux ooh

Skillmon

@LaTeXereXeTaL in the brackets which are commented as init options in @DavidCarlisle's code example.

David Carlisle

@barbarabeeton frege :-)

Paulo Cereda

@UlrikeFischer ooh

David Carlisle

@PauloCereda Frühstück

Paulo Cereda

@DavidCarlisle oh no

Davislor

08:37

Tested, and I can definitely reproduce the bug in both XeTeX and LuaTeX and both Babel and Polyglossia.

Ulrike Fischer

@Davislor can you show an example?

Davislor

09:02

@UlrikeFischer tex.stackexchange.com/questions/571643/…

Ulrike Fischer

09:34

@Davislor I knew I had seen this already: tex.stackexchange.com/a/340164/2388

Davislor

That was a similar but different bug,yes.

Paulo Cereda

@JosephWright ^^

Davislor

@UlrikeFischer Also reported on the bug trackers for Babel and Polyglossia.

Paulo Cereda

ooh the L3 team could merge both language package thingies and create either babelglossia or polybabel.

2

Ulrike Fischer

@Davislor you should better write to the hyphenation list, so that hyph-utf8 is adapted.

@Davislor actually there is already an issue: github.com/hyphenation/tex-hyphen/issues/5

@DavidCarlisle many letter classes

David Carlisle

09:53

@UlrikeFischer Germany wins (dinbrief was the only one that worked:-)

@PauloCereda funny you should mention that....

Ulrike Fischer

@DavidCarlisle hm, I thought dinbrief is broken because of the hooks.

Paulo Cereda

@DavidCarlisle ooh l3hyphenation

Ulrike Fischer

@DavidCarlisle but in germany we don't put a space after the name in an address, so it wouldn't matter anyway.

yo'

@PauloCereda i vote for polybabel 100%!

Paulo Cereda

@yo' YES <3

David Carlisle

09:55

@UlrikeFischer oh some classes gave unrelated errors that I ignored (as they required extra data not supplied in that one example) but dinbrief seemed to work just warned I was using 2e (??)

\documentclass{dinbrief}



\begin{document}
\begin{letter}{First line\\[1ex]
    Second line\\
    Third line\\
    Fourth line\\[1ex]
    Fifth line\\
    Sixth line\\
    Seventh line}

  \opening{Dear Sirs,}
  Blah Blah Blah\dots
  \closing{Best regards}
\end{letter}
\end{document}

Paulo Cereda

@yo' let's have Twitter fun, I will create a poll. :)

David Carlisle

@UlrikeFischer that'sbecauseyoudon'thaveaspacebaranyway

Ulrike Fischer

@DavidCarlisle ah, it can be that the one error is handled by firstaid, but imho the next one is lurking around the corner.

David Carlisle

@PauloCereda dont

Ulrike Fischer

@PauloCereda better not.

Paulo Cereda

09:57

@DavidCarlisle aw, just for fun... :(

Okay... :(

Ulrike Fischer

@PauloCereda I would love to have only one language package, but ...

Joseph Wright

@UlrikeFischer Yup, we should get back to that

Paulo Cereda

@JosephWright dibs on the name

Davislor

@UlrikeFischer Left a comment, but that issue is from 2016, with no comments and no one assigned to it.

Ulrike Fischer

@Davislor well yes sometimes issues get forgotten, I had forgotten too that I triggered this.

Davislor

10:06

Yep, it sometimes happens. Anyway, it’s been reported.

Joseph Wright

@PauloCereda It's clearly just babel

Davislor

Thanks!

Joseph Wright

@UlrikeFischer One for Arthur?

Ulrike Fischer

@JosephWright he opened the issue after I sent a request to the mailing list. And since then the issue sits there.

Paulo Cereda

@JosephWright poor Arthur :(

Joseph Wright

10:08

@UlrikeFischer Ah, right: he was on the UK-TUG list a couple of days ago, I could prod him

@PauloCereda He and Mojca volunteered for this job :)

@PauloCereda You met Arthur?

Ulrike Fischer

@JosephWright yes at some time, but I guess one would have to meet the relevant people in person to discuss this.

Paulo Cereda

@JosephWright Only by mail

Joseph Wright

@UlrikeFischer We could try to set somethign up, but Javier is not keen on physical/online meetings, mainly email

@PhelypeOleinik I think as I've already got somethign half-done for xparse, I'll probably check something in and ask for your review

Ulrike Fischer

@PauloCereda why? I don't think that he invest much time in polyglossia, imho would have been dead by now without the other new collorabotors

Paulo Cereda

@UlrikeFischer oh sorry

Joseph Wright

10:12

@UlrikeFischer Indeed: I do wish that now babel development is back on, we could convince everyone to shift attention back

Ulrike Fischer

@JosephWright yes, but I never met one of the peoples and I don't know why they push polyglossia and so I don't know how to convince them to shift.

Joseph Wright

@UlrikeFischer Fair point: I guess the best we can do is make sure babel can do the job, then show that

Ulrike Fischer

@JosephWright yes, on other news: I converted tagpdf to l3xref and after correcting the normal amounts of using the wrong expansion it works fine. I only need a better "LastPage" label. And it should change its name again, xref looks odd.

Joseph Wright

@UlrikeFischer Cool: it's still a work-in-progress

Ulrike Fischer

@JosephWright yes, but the basic setup it good, the rest is "adding missing bits". And it loads only one package and not six or seven ...

Phelype Oleinik

11:07

@JosephWright Sounds good

@JosephWright I should be done with \usepackage now. I'll just double check and push the changes. See how Travis behaves

Joseph Wright

@DavidCarlisle Any idea why ltxguide doesn't like a single > in verbatim?

@PhelypeOleinik It's not doing much for my merge

David Carlisle

@JosephWright er no?

Phelype Oleinik

@JosephWright Queued still...

David Carlisle

@PhelypeOleinik this issue "fixed in March" ? traviscistatus.com/incidents/v1vjd2sd2g2v

Phelype Oleinik

@DavidCarlisle :-)

@DavidCarlisle “fixed” is very different from “really fixed”

David Carlisle

11:12

@JosephWright this?

\renewcommand \verbatim@font {%
  \normalfont \ttfamily
  \catcode`\<=\active
  \catcode`\>=\active

@JosephWright @JosephWright documented feature it seems

% Make active <...> produce italics surrounded by angle brackets
% (used in verbatim and \verb).
% << produces a less-than, and >> produces a greater-than.

@UlrikeFischer would be proud of me for reading the comments not the code

@JosephWright @PhelypeOleinik have any of us signed up for the beta of "sync .org to .com" ?:

Phelype Oleinik

@DavidCarlisle Not that I know of

David Carlisle

I am not sure how that works for organisational repositories, do we have to all do it if it is synced, so I have left it so far

Joseph Wright

@DavidCarlisle I had some mail I ignored

@DavidCarlisle Oh. Great (!)

David Carlisle

Is this telling us that it is running flat out at 400 linux jobs and has 6000 in a queue, that seems... less than ideal traviscistatus.com/#day

Joseph Wright

11:31

@DavidCarlisle Ah, so that's the issue then: either pay up or sit in a queue

Phelype Oleinik

@JosephWright We can manually add ✔️ at the end of every commit message, then we won't need Travis

2

Joseph Wright

@PhelypeOleinik :)

@DavidCarlisle @PhelypeOleinik Looking at this, we are likely to have to move I think

David Carlisle

@JosephWright just sent mail to team list (on a thread started by Frank)

@JosephWright this is supposed to be temporary though:

Q. Why are some of my queued jobs taking longer than usual to build? #

A. At peak usage times, you may see your build times are longer than they previously have been - we need to make sure all users have equal access to the .org platform until the end of the year as we move our infrastructure across to .com. You may want to consider migrating across to .com sooner rather than later, or consider scheduling builds at a quieter time of the day if remaining on .org for a while longer.

Joseph Wright

@DavidCarlisle Seen it

@DavidCarlisle Hmm, that's all great but why not just migrate people now!

@DavidCarlisle Forcing a migration now

Paulo Cereda

12:53

@DavidCarlisle And the hammers batter down the door, YOU BETTER RUN

yo'

Rmano

@PauloCereda Helicopter noise in background? vvvvvv

Paulo Cereda

@Rmano YES

yo'

14:04

@PauloCereda your last two starred messages nicely complement each other (see above) :)

Paulo Cereda

@yo' LOL

yo'

@PauloCereda how're you, pal?

Paulo Cereda

@yo' recovering, so to speak... :) And you?

yo'

@PauloCereda quite fine. Just busy, it seems that Overleaf users always need something :-)

Paulo Cereda

@yo' :)

Paulo Cereda

14:44

barbara beeton

15:35

@DavidCarlisle -- This will royally screw up TUGboat if it becomes the default. For TUGboat, there's a definition \<#1> and I've forgotten what is defined for the escape character. That way, only the escape character is removed from its usual meaning in verbatim. Please rethink.

Ulrike Fischer

@barbarabeeton I don't think that David suggested any change. This code is already in ltxguide.cls.

Joseph Wright

@barbarabeeton It's not a proposed change: it's just something I was having 'fun' with

barbara beeton

@UlrikeFischer -- ltxguide,cls isn't the kernel. (I'm not sure I've ever used this class for anything.) But if verbatim goes into the kernel, it really would spell disaster for TUGboat.

@JosephWright -- Okay, thanks. But please remember my plea if anyone suggests moving verbatim into the kernel.

Joseph Wright

15:53

@barbarabeeton WE only every make change when it's necessary and meets a safety 'level': the only change I could imaging to verbatim would be to cover all of the 8-bit range for pdfTEX/all of Unicode for LuaTeX (using catcode tables)

Paulo Cereda

@JosephWright ^^ taking the day off?

Paulo Cereda

16:13

Mac people, this seems to be my last waltz:

Warning: You are using macOS 10.13.
We (and Apple) do not provide support for this old version.
You will encounter build failures with some formulae.
Please create pull requests instead of asking for help on Homebrew's GitHub,
Twitter or any other official channels. You are responsible for resolving
any issues you experience while you are running this
old version.

LaTeXereXeTaL

17:02

I've a feeling I'm not yet picking up on a lot of inside jokes here.

Paulo Cereda

Good news: arara runs on Apple M1 chips!

@DavidCarlisle > Why Wear a Mask When You Can Wear an Inflatable Dinosaur Costume for Just $38? amazon.com/gp/product/B00TO6E0T8

2

Paulo Cereda

17:29

Plergux

17:41

@PauloCereda LOL

Paulo Cereda

@Plergux <3

barbara beeton

17:58

@PauloCereda -- except that it doesn't tell you how to convert from kilometers to miles, or vice versa.

Paulo Cereda

@barbarabeeton the tomato has failed. :)

barbara beeton

@PauloCereda -- does that mean it's a rotten tomato? (And "drunken lobster" sounds like something I've seen on a menu soewhere ...)

Paulo Cereda

@barbarabeeton ooh

dedded

18:35

@PauloCereda Except miles aren't the problem. Bushels are the problem. You might look up "bushel" and see that it's 4 pecks and think you're set. But that's not how bushels are used commercially. Commercially, bushels are units of mass, not capacity. And the definition depends on the quantity being measured (rice, wheat, turnips, what-have-you).

And those definitions vary by state, so different in Massachusetts vs. Illinois vs. Missippippi, etc. "Drunken lobster" doesn't even begin to describe it.

See also "ton", and "yard".

Paulo Cereda

@dedded oh no

/quacks in confusion

dedded

Yes, we all quack in confusion :-)

Paulo Cereda

ooh :)

@dedded this reminds me of this video:

A Guide to Imperial Measurements with Matt Parker | Earth Lab

►

dedded

18:54

@PauloCereda :-) It gets worse though. It's a mistake to think that the US uses the imperial system, which wasn't devised till the 1820s (well after our independence). We use older units. Length works out to be the same (by treaty), but capacity is wildly different. The imperial gallon is about 20% more than the US liquid gallon (different, of course, from our dry gallon).

Paulo Cereda

@dedded oh my

dedded

British visitors are always disappointed to enter a US bar, order a pint of beer, and get the (much smaller) US pint.

Rmano

19:07

@dedded ...this is serious!

2

^^^^ I will save this on my phone ;-P

Paulo Cereda

Birra media?!

Medium beer?!

LOVE IT

Rmano

@PauloCereda yes --- in Italy is piccola, media and grande

Paulo Cereda

@Rmano got it!

Voglio una birra piccola, per piacere.

Rmano

@PauloCereda you nailed it! (media is better ;-P)

Paulo Cereda

@Rmano :)

Rmano

19:12

Although normally you say "una piccola bionda". I normally ask for "una media rossa"

Paulo Cereda

@Rmano ooh :)

Davislor

What, you don't order grande, venti and trenta? :)

(a trenta is, of course, thirty-one liquid ounces.)

Paulo Cereda

Es ist nicht meine Schuld

Rmano

@Davislor that's fun. (I think that in Italy the three main beer size are more or less standardized by law for draft beer. In Spain it's much more variable).

Ulrike Fischer

@PauloCereda in english "blame Ulrike"?

Paulo Cereda

19:17

@UlrikeFischer <3

Ulrike ist mein Freund <3

Davislor

Ulrike has an account on expertsexchange?

dedded

In the US you hear "tall" (20 oz) and "short" (16 oz).

cfr

19:31

@DavidCarlisle Thanks. Less confused now. (I confused me, too, misreading something years ago.) Also, the 3rd bullet on page 294 of source2e.pdf seems to contradict the code on page 295. 294 says no entries for weight+width requests, giving bx as an example, but 295 includes several lines with bx as second argument. (The code looks sensible, if I've understood the new stuff, but the bullet seems wrong.)

Paulo Cereda

@cfr cwac!

cfr

@DavidCarlisle source2e.pdf says 9 weights, whereas the Font Installation Guide listed 10. Guessing the discrepancy concerns FIG's listing for 'medium', which gets a mb designation. What do you recommend doing about fonts/packages which use mb? ('medium' here is distinct from 'regular'/'book'). I'm trying to update nfssext-cfr.sty which is tricky as I didn't write the nfssext bits and I'm not sure what purpose they always serve. But one of the Venturis fonts uses mb (& m &&).

@PauloCereda Cwac!

Paulo Cereda

@cfr How are you? How is dad?

cfr

@PauloCereda Grumpy, so he must be doing OK. Just wish I could escape. Feels like I've been in jail since March!

@PauloCereda How are you? Cats?

Paulo Cereda

@cfr ouch, good to know. :) And Lucy?

Ulrike Fischer

19:38

@cfr stick to the list in fntguide.pdf

Paulo Cereda

@cfr Trying to survive. :) Fubá says hi. :)

David Carlisle

@LaTeXereXeTaL you are making the fundamental error of assuming that anyone has any idea what @PauloCereda is talking about most of the time.

5

Paulo Cereda

@DavidCarlisle ooh

cfr

@UlrikeFischer Lucy's OK. Went to vet today for B12 & she's lost a bit of weight, as she does. Been very sicky lately. Taking her to vet next week. (Only saw nurse today for the B12.)

Paulo Cereda

@cfr You actually replied to Ulrike. :) I wish her a speedy recovery. :)

cfr

19:41

@PauloCereda Oops. Sorry.

@UlrikeFischer Sorry.

Paulo Cereda

@cfr No worries. <3 Fubá is now 18 years old!

@cfr ^^ from yesterday :)

He has a cow pattern blanket.

David Carlisle

@cfr er not sure since m-width and weight-m mean something ow, mb seems a bit problematic, I'd probably ask Frank who holds more of the high level design in his head, or you could ask him

cfr

@UlrikeFischer I understand that, but I'm not sure how best to do that without breaking existing documents. I don't think, for example, that I should switch stuff to using scit because that will break lots and lots of things.

Ulrike Fischer

@MarcelKrüger regarding the math node question. Does that mean that when one wanders through the shipout box one could miss math parts? (I think the space code tries to detect math, and now I wonder if it fails sometimes, but it to long ago I last looked at it).

@cfr difficult to say without a concrete example. But you can always detect if a current format is used and then branch to the older setup or use the new one.

David Carlisle

@cfr are people really using \fontweight{whatever}\selectfont in the document? aren't these (most of he time) just internal codes? using changing bx to b doesn't break a docuemnt if the document just uses \bfseries

cfr

19:48

@DavidCarlisle Where can I find Frank? I think nfssext-cfr will have to switch scit to si.

Davislor

@cfr Why is that?

David Carlisle

@cfr latex-team at latex-project.org would reach all of us.

cfr

@DavidCarlisle Hopefully, they're doing \mbwidth, where it needn't matter. I guess I was really wondering what it made sense to use in the .fd files. The real worry is that I don't know if nfssest-cfr.sty is used by anybody with fonts I didn't package.

@Davislor Because I can't figure out a way to switch to scit without potentially breaking a whole bunch of stuff. The new stuff lets me partially do away with some of the code in nfssext-cfr, but there are a couple of tricky cases. One is mb which the Font Installation Guide used as a weight. Another is si which it used for italic small-caps. The third is swash, where the approach is quite different.

@UlrikeFischer Yes, I'm working on a 3 file setup where the main package just loads one of two other files, depending on whether the new init macro is defined. (May not be the best test to use, but I'm really just trying stuff right now to see what might work.)

Skillmon

20:15

Regarding the following question, wasn't there a special \relax TeX only inserts during some error recovery that can't be altered and isn't the same as a normal \relax (I don't recall when it's inserted).

0

Q: Are there any commands that cannot be redefined?

i.e. Can I redefine \def, or \renewcommand? Is there a definitive list?

David Carlisle

@Skillmon well yes and no, its called frozen relax but it's like the end of paragraph thing that is usually but not always \par it can't be referenced by a csname at all it's just an internal behaviour so you can't redefine it but I don't think it counts.

Skillmon

@DavidCarlisle "frozen" that was the name :) Thanks.

Phelype Oleinik

@Skillmon \edef\x{\ifnum1=1\else\fi}

Skillmon

@PhelypeOleinik stop mocking me :)

Phelype Oleinik

@Skillmon And indeed, redefining it is not allowed: \expandafter\def\x{frozen relax}

@Skillmon NEVER! (Why mocking?)

Skillmon

20:21

@PhelypeOleinik (just because my memory is like a sieve doesn't mean you all have to proof that you remember :P)

Phelype Oleinik

@Skillmon My memory is crap too :-) I just happened to have stumbled on that one not so long ago

@Skillmon Or maybe I was mocking you. Who knows ;-)

Skillmon

@DavidCarlisle I think it counts :)

David Carlisle

@Skillmon yes I was just going to suggest to @PhelypeOleinik that he posts that, I was thinking it unfroze if you accessed it that way bit it doesnt:

\edef\x{\ifnum1=1\else\fi}
\expandafter\show\x
\expandafter\def\x{frozen relax}

Skillmon

@DavidCarlisle I knew it doesn't (I remember playing with it), I just didn't remember the name nor how to create it.

Phelype Oleinik

@DavidCarlisle If it doesn't unfreeze, maybe it counts as an answer...

Skillmon

20:25

@PhelypeOleinik I think it definitely counts.

@PhelypeOleinik wanna have the internet points? :)

Phelype Oleinik

@Skillmon Nah, go for it. I have a LaTeX to fix :)

Ulrike Fischer

@PhelypeOleinik did you break it?

Phelype Oleinik

@UlrikeFischer That one issue with paths

@UlrikeFischer But I can blame you if you want :)

Ulrike Fischer

@PhelypeOleinik I was quite impressed yesterday about the ../../whatever path.

Phelype Oleinik

@UlrikeFischer :-) I tried implementing something like that before (in fortran though). Eventually I gave up :)

Ulrike Fischer

20:31

@PhelypeOleinik Imho at some point you have the right to say "are you mad"?

David Carlisle

@PhelypeOleinik I still think the original 2e position has merit: it's user error to put a path in a package or class name. we are too kind to people now making it work:-) It's definitely a feature creep not a bug fix:-)

Phelype Oleinik

@UlrikeFischer :-)

@UlrikeFischer It is not too hard (and too absurd) to support path/class and ./path/class, but it's much easier and robust to just look at the name

Ulrike Fischer

I'm just trying to convince someone that it is not a good idea to call external applications with write18 on windows and try to pass arbitrary unicode values as arguments, but they don't want to hear this ...

Phelype Oleinik

@DavidCarlisle Users seems to have been using that feature creep (though definitely not as many as I had thought)

@UlrikeFischer Users like to learn the hard way

Marcel Krüger

@UlrikeFischer If you try to detect math by going though the shipout box and tracking math on/off nodes then yes, you can miss some math segments. But that approach also has other issues, e.g. when math mode material is broken across lines or even pages. Therefore such code should always run before linebreaking. (The space code already does that, so it should be fine)

Skillmon

20:40

@DavidCarlisle it unfreezes if you \let something to it, so \edef\x{\ifnum1=1\fi}\expandafter\let\expandafter\x\x will let \x to the normal \relax.

David Carlisle

@PhelypeOleinik the ProvidesPackage warning about /foo/bar/longtable not matching longtable was more or less explicitly added to stop people using paths in \usepackage :-)

@Skillmon that sounds familiar

Phelype Oleinik

@DavidCarlisle Oh...

@DavidCarlisle Should I keep that?

David Carlisle

@PhelypeOleinik yes even if you make it work you should warn people not to do it

Phelype Oleinik

@DavidCarlisle Okay, that's easy enough to keep

@DavidCarlisle I'll make a list of the main changes so others can review

David Carlisle

@PhelypeOleinik at the time we did 2e variant path syntax was far more important, VMS had [abc;xyz]file (if I recall correctly) some OS used . separator IBM mainframes exposed a non hierarchical filesystem etc, the ltdirchk tests at format time meant you could handle this in controlled ways but if people started using literal OS paths explicitly in the document things got too hard and portability was impossible.

Phelype Oleinik

20:57

@DavidCarlisle I did notice there is code in graphics that inserts a variable \path@separator and wondered why that was for (having never seen an OS with that type of file system :)

Ulrike Fischer

@MarcelKrüger did it? ;-) it is so long ago. But it is something I have to keep in mind, and in case it matters to set attributes in an earlier callback.

David Carlisle

@PhelypeOleinik I guess you have never used an acorn Archimedes either, they had a fun filesystem no extensions allowed so had to use a directory level so article.cls was in cls/article made an interesting test of the 2e filename parsing code. (that machine had in interesting novel chip design that went on to have a life of its own as well:-)

Ulrike Fischer

@PhelypeOleinik users don't like to learn ;-(.

Phelype Oleinik

@DavidCarlisle No, never even heard of it. In the decade I was born, acorns were already edible :-)

@DavidCarlisle So portability was a fun part of the work :)

David Carlisle

@PhelypeOleinik The Arm chip that was designed for it you may have heard of though....

Phelype Oleinik

21:11

@DavidCarlisle Oh, ARM yes

@DavidCarlisle The first computer I've used was a PC already, with a pile of floppy disks with win 98 (I didn't install it though. The best I could do on a computer back then was MS paint :)

David Carlisle

@PhelypeOleinik now it's just a name but it was Acorn Risc Machine in the beginning, weird British micro computer that essentially no one ever bought

Marcel Krüger

@DavidCarlisle There was a talk about that machine at the last CCC.

Phelype Oleinik

@MarcelKrüger “No eureka jokes” :)

David Carlisle

@MarcelKrüger we had some friends who had one, I used it a bit

Plergux

@PhelypeOleinik I bought my first Linux in '98. SuSe 6.0. :p It was a book with one floppy for installation. :p Our first computer didn't even have an "OS" it was command line all the way. and the floppy discs were literally floppy. And now I feel stupidly old XD

Phelype Oleinik

21:28

@Plergux Oh, the big 8-inch ones. I've seen some of those (though never got to actually use one). Dad had boxes and boxes of the 3 1/2-inch discs with bank stuff

Skillmon

@PhelypeOleinik yay, same here. I used to draw circles and fill them on my father's PC.

Phelype Oleinik

@Plergux I remember, as the average windows user, being terrified of a command line and completely wrecking the computer with it :)

David Carlisle

@PhelypeOleinik how old is your father?

Phelype Oleinik

@DavidCarlisle 52

David Carlisle

@PhelypeOleinik figures. Younger than me:(

Skillmon

21:31

@DavidCarlisle don't worry, there is still @egreg around.

Phelype Oleinik

@DavidCarlisle :)

@Skillmon I drew the Simpsons once, copying from a magazine or something like that. I had discovered the magic Ctrl+Z, so if I messed a curved line I could just try again... Apparently I had nothing better to do with my time

Skillmon

@PhelypeOleinik well, drawing like 50+ circles and filling each overlap with a different colour doesn't sound like the most interesting thing to do to me as well... Or using two colours and trying to always have all the diagonally adjacent areas with the same colour.

Phelype Oleinik

@Skillmon You were into abstract art then :)

Skillmon

@PhelypeOleinik I'm pretty sure I've saved quite a few of my "art" pieces back then, but those are long lost on some old drive.

Phelype Oleinik

@Skillmon I did too. That computer had a 500 MB hard drive (half of it taken by the OS), which I tried my best to preserve after the computer died, but eventually it just stopped working and then I learned about backups

egreg

21:49

@PhelypeOleinik Wow! My first hard drive for the Macintosh Plus had 100MB. A big leap from the one at the university that had 20MB.

@Skillmon :-P

Skillmon

@PhelypeOleinik I don't remember the size anymore, maybe I never knew (I was quite young back then, and at the time not into computers, only cared for playing football and stuff). And I first learned about backups when our family's Windows 2000 PC was killed by lightning.

barbara beeton

22:05

@Plergux -- Punch cards and paper tape.

egreg

22:22

@Plergux Once I assisted somebody trying to install DOS 6 on a new machine; the keyboard had the Italian layout and the first thing one had to do was to tell the installer about the layout. Of course the command to input had characters in different position than on the available keyboard…

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends