@Werner probably the investigation went like... "oh it's Ok egreg isn't impacted, let's delete".

@DavidCarlisle :D

@DavidCarlisle Darn, neither @DavidCarlisle is.

@DavidCarlisle But I lost 151 on Math.SE for a removed user

@egreg we are terribly sad about that, aren't we, @Werner?

@DavidCarlisle Oh, wait! it was here!

@egreg so it was, that was very late showing up wasn't it?

still I'm alright so all's well.

@DavidCarlisle Yeah.

@egreg although I did lose 15 today:

@DavidCarlisle Great...! Welcome to the elite club.

@Werner :-)

well, the removal of votes hit me too -- down 490 today. i guess i'll never reach 60,000. actual, removel of user took down 580, but i got a few positive votes along the way. (i never manage to get a rep cap ...)

@egreg And @Mico lost 5204.

tex.stackexchange.com/users/5001/mico?tab=reputation

@barbarabeeton I lost 2.2K, even though I had positive votes to offset it.

It is a strange justification: we remove all the votes, because they might be inappropriate. Is this because it is worse to allow people to keep inappropriate votes than it is to strip them of appropriate ones? Otherwise, surely you need a balance of probabilities or something. The reasoning seems fallacious as it stands, so I'd like to know what additional assumptions they are relying on. Poor arguments are a professional irritant.

@Johannes_B Thanks. I've read it before, but didn't have the link.

@Johannes_B I just removed it. The ThesisH.cls I sent back is still awful, it just has slighly-less-awful patches.

@cfr sent back where?

@cfr I lost 2.4k

I lost 2867 in the Papiro deletion. But it's really kind of irrelevant for high rep users.

Seems Papiro didn't ike my answers. I didn't lose any upvotes.

Oh wait, i did. about 1700 points :-)

That is ten percent.

So it is pretty clear who the serial upvoter was in my case. The message "Voting reversed" has been changed to ^User removed" and that's why I did not expear another loss of rep.

@Werner Thanks for following this up

@cfr Once there is some evidence of voting patterns by a user, the Powers are keen on pretty 'swift and strong' action (the mod team have in the past had some interesting comments from the Powers in this regard). Once you are suspicious of the actions of a user then all of their behaviours are tarnished.

@Johannes_B Having many rep captures will prevent a bigger loss

@Johannes_B Possibly shows some form of pattern, which of course could be to do with the type of questions you answer

Did the number of unanswered questions just jump up by a few hundred? (In other words, was there a lot of answers where Papiro's vote was the only one to any of the answers?)

@Werner this will make your day:

@TorbjørnT. Unupvoted... Answered, but no upvote. I´ve got a bunch of them ;-) So Papiro was the one who cast the single vote, yes

@DavidCarlisle This makes for a good start of my day! :-D

@PauloCereda see how mean @egreg is ^^

@DavidCarlisle oh no

@DavidCarlisle: I also got stabbed

@PauloCereda -620 is nothing I got -2814

@PauloCereda With 50000 votes, most high-rep users will have been hit (I lost 1550).

@DavidCarlisle oh

@TorbjørnT. indeed

@DavidCarlisle If you consider percentages, then I think @PauloCereda was worse off.

@TorbjørnT. yes but some people like @egreg just lost probably fake points they didn't really deserve anyway but I lost 2814 well deserved points

@DavidCarlisle Certainly.

@TorbjørnT. percentages? too technical for me.

@DavidCarlisle And on top of it I stole your tick -- I'm really sorry :-(((.

@UlrikeFischer I'm sure you didn't steal it but was awarded it on merit (unlike some other people who I could mention)

@VincentVerheyen I don't know how scr* sets up page styles but in general that sounds like you should have a new page style for the modified footer and then \thispagestyle{...} on the final page

@VincentVerheyen sorry no idea about scrheadings, ask a q on site:-)

@VincentVerheyen Is this what you want?

\documentclass{scrartcl}
\usepackage{scrlayer-scrpage}
\usepackage{lastpage,refcount}

\usepackage{lipsum}

\pagestyle{scrheadings}
\cofoot{\ifnum\getpagerefnumber{LastPage}=\thepage\relax\thepage\fi}

\begin{document}

\lipsum[1-20]

\end{document}

@VincentVerheyen You didn't ask for that.

:(

@samcarter "only" 935 for me

@VincentVerheyen Here's the idea

\documentclass{scrartcl}
\usepackage{scrlayer-scrpage}
\usepackage{lastpage,refcount}

\usepackage{lipsum}

\pagestyle{scrheadings}
\cofoot{%
  \ifnum\getpagerefnumber{LastPage}=\thepage\relax
    This is the last page of \pageref{LastPage}%
  \else
    Page \thepage\ of \pageref{LastPage}%
  \fi
}


\begin{document}

\lipsum[1-20]

\end{document}

Hiiiiiiiiiii to everybody. Good morning

@egreg my warmest greetings to you

Best regards to everybody users to TeX.SE

@VincentVerheyen Get as far from those templates as you can.

@VincentVerheyen \thispagestyle{...} in the sense: define a new page style as you like it to be and in the first page issue the choice.

@VincentVerheyen what have you got against \thispagestyle which is defined in the format, for defining a different style on specific pages?

Can we align a tikz text node on the baseline of a multi line text (text width=10cm)? I'd reather not have to wrap the text in an additional minipage.

@egreg @VincentVerheyen Neither \getpagerefnumber{LastPage} nor \thepage needs to give really a number. So this test is not very reliable. see e.g.

\documentclass{scrartcl}
\usepackage{scrlayer-scrpage}
\usepackage{lastpage}
\usepackage{refcount}

\begin{document}
\pagenumbering{roman}
\ifnum \getpagerefnumber{LastPage}=\thepage blub \fi
\end{document}

You should better use zref.

I wonder what would happen if I delete my account...

@UlrikeFischer Yes, of course, if the document is not supposed to have arabic page numbers.

@cfr: Cwac!

@PauloCereda Cwac! @JosephWright One of my students just complained that my/your/TUG's materials on macro creation are 'a bit dull in places'. What do they expect? Fireworks?

@JosephWright I just think the reasoning is dubious.

@Johannes_B To the student who (eventually) sent it to me so I could try to give them an approach which would avoid the errors you get if you try to compile a document with neither a class nor a document environment.

@cfr The UK-TUG notes stuff? I'm not really sure what to say!

@JosephWright Mostly, yes. They might have only found the bits I wrote dull, of course :(.

@JosephWright Maybe I should do expl3 syntax instead? Less dull, do you think?

@cfr Takes you into 'real' programming: well outside a beginners course

@cfr One of the issues is trying to cater for those users as well as the bulk who find 'you have code and an output' difficult

@cfr -- it's hard enough writing documentation that is accurate and comprehensible. asking for it to be exciting as well is a bit much ...

@JosephWright Yes, I was joking.

@JosephWright Yes, though at least one student said that the workshop was good precisely because it was structured to accommodate users at different levels. (My version of the handouts includes a couple of appendices extending the coverage, so that students can move on if they are bored during the exercises or can build on what they've learnt later.)

@barbarabeeton I wonder how exciting you can make an introduction to \newcommand, \newenvironment and \ProvidesPackage. Maybe I just lack sufficient pedagogical imagination or something.

@barbarabeeton Maybe I should have changed \newcommand*\authorname{Joseph Wright} to \newcommand*\authorname{<something in Elvish>}?

@cfr ooh Elvish Presleysh

@PauloCereda ;)

@cfr -- i like the idea of elvish -- do it! (as for @PauloCereda's suggestion, i had the misfortune to have the same birthday as elvis. my sister was more fortunate -- she got mozart and lewis carroll. sigh.)

@barbarabeeton oh my!

@barbarabeeton I only share my birthday with Helen Hunt and Neil Patrick Harris. :)

@PauloCereda @barbarabeeton I read this line first first and totally mis-parsed it as "Only Helen Hunt and Neil Patrick Harris know when Paulo's birthday is." :D Context matters!

@AlanMunn ooh :)

@barbarabeeton Sadly, I doubt that OverLeaf has the fonts installed. Though I could use it as an example .,..

@cfr It's a classic 'fruit of the poisoned tree' argument though. It's certainly widely used in e.g. American criminal law. And in this specific case it doesn't surprise me that the voting pattern was suspect, since he was also a reknowned robo-reviewer.

@DavidCarlisle you may find this interesting:

github.com/tatzetwerk/luatex-harfbuzz

it is new Harfbuzz shaper for LuaTeX, it can use several methods for accessing the library, some of them may work even on Windows.

@cfr Good idea: any chance I could have a copy to add into our version?

@michal.h21 oooh

@michal.h21 I'll be looking at that at the weekend, certainly

@DavidCarlisle My apologies for my late reply! I didn't see the notification. Yes, I was pretty happy with the response. The reason why I didn't post a full document, is because my questions are usually related to the latex code I need for the mathematics on the Mathematics forum, so I didn't work with a full document to begin with!

@ShaVuklia David is mean to me, beware. :)

@JosephWright Sure. Do you want it right away? I'm going to put LaTeX 2 on OverLeaf when I've finished it, but it is split into 4 so I've only done part of it at the moment.

@AlanMunn I know. But the reasoning is still dubious. That is, the details of a case might justify general suspicion in that case. But the move from simply this token is untrustworthy to this type is untrustworthy is fallacious. This is how Descartes gets from 'my senses have sometimes deceived me' to 'my senses may be deceiving me in all cases'. Maybe, but not without further argument.

@cfr But if I understand the process correctly, in the case of large vote counts they do undergo further scrutiny, and if there's evidence of irregularity (independent of the reason for deletion of the account) then all votes are invalidated, which seems to be what happened in this case. (I'm assuming here that your use of 'type' and 'token' applies to tokens of voting patterns by a particular user, not tokens of individual votes on questions.)

@cfr I am confused :-)

@cfr No panic!

@cfr All true, but we can't ultimately know the reason for any one vote. I do notice in this context that @Johannes_B has lost a lot more rep than many people with a higher total number of votes, which may of course suggest coverage of particular topics but may also point to targetting.

@cfr The Powers have two options when someone decided to remove their account: keep the votes or drop them. When the user has had questions about inappropriate vote patterns its not that surprising that they take the latter path.

@JosephWright relative, not absolute figures.

@Johannes_B Yes, but if voting is done purely on merit of answers one might broadly assume that the relative impact of a deletion would be similar across most higher-rep users

@JosephWright Would be interesting to know how many of those votes turn my answers back by the community user. I answer abandoned questions once in a while, maybe i did the upvotes just to get the list of unanswered cleared (which i suspected anyway, but that was the reason for asking). I just never considered that it could be the same guy over and over again.

Good evening to everybody. All users to TeX.SE.

@Sebastiano Evening

@JosephWright egreg and DPC answer questions from today, seldomly older question. I seldomly answer new questions.

@Sebastiano Hi.

@Johannes_B Yes, sure: as I said, there could be a range of reasons for being an 'outlier' without targetted voting being involved

@UlrikeFischer Hi to you. You are an excellent users of LaTeX. Always I vote your posts.

Hi to everybody.

@ChristianHupfer Look on the bright side, Christian. You will have the unique distinction, I am guessing, of being the only TeX.SE user to break the 100K barrier twice!

@StevenB.Segletes You were at risk, too! –2760

@egreg Alas... -2760. I was expecting the red mark on the "reputation" graph to be more alarming, but they truncated it.

@egreg It also makes for a very diverse "weekly" reputation list of users.

@egreg, @DavidCarlisle Thoughts on tex.stackexchange.com/a/351777? Should I add an \ifhmode test to the \ifinner one?

@StevenB.Segletes I feel deeply (!!!) honored to have this chance twice ;-)

@JosephWright I'd go with \ifhmode: from your description, \ifinner is the wrong test to do.

@StevenB.Segletes Indeed. Seems there are about 500 pages of users with a negative score for the week. Bernard is at the "top" of the list, -5030.

@StevenB.Segletes :) I'm still at top, anyway. :-D

@Sebastiano Thanks.

@egreg We wouldn't have it any other way.

@egreg I suspect a simple \iffmode is wrong as one could have blah blah.\blockquote... akin to blah blah.\begin{figure}

Do I need to add any package to use \foreach in TikZ?

@OskarTegby No, tikz loads pgffor which makes \foreach available.

I've added \usepackage{pgffor} but I still get the same errors.

\foreach \i in {0,...,3}
{
   \draw (\i * pi/2 r:1) --
              ({(\i + 1) * pi/2} r:1);
}

@OskarTegby What errors?

"Missing number, treated as zero", "Illegal unit of measure (pt inserted)", and "Overfull hbox (23,92719pt too wide) in paragraph". The missing number stuff I get four times, which is the number of iterations.

@OskarTegby Nothing to do with the \foreach: you get the same if you hard-code some values

Okay.

I got this from a TikZ manual. Maybe they didn't check the code.

@OskarTegby Which page?

Page 216: tug.org/TUGboat/tb30-2/tb95mertz.pdf

I said a not the manual. ;)

I wanted something easy to get going.

@OskarTegby I noticed, so I deleted my comment. But:

\foreach \i in {0,...,3}
{
   \draw ({deg(\i * pi/2)}:1) -- ({deg((\i + 1) * pi/2)}:1);
}

Yeah, sorry!

So, what was the issue?

@OskarTegby Don't actually know, but what I posted above works.

Yeah. It's nothing of importance for me; I'm just learning the basics, it's not part of any course or project.

Thanks anyway!

@OskarTegby The TUGboat article you refer to was written for TikZ version 2.x. There were quite a few changes to version 3, so probably something changed in the parsing of polar coordinates.

Okay. Yeah, I noticed that. Should I use the manual instead?

@OskarTegby Most of that article still applies I think. The only other thing that I noticed right away that is deprecated is the intersection of syntax shown on page 217. See the manual, section 13.3.2 about the intersections library for info about the new syntax.

@michal.h21 ooh thanks, I'll look later

@JosephWright ifinner test doesn't seem that useful why do you want different quote display behaviour in inner vmode compared to the main vertical list? (or was the intention just really to test for inner h mode?)

@DavidCarlisle Well PL didn't really go in for comments so it's hard to be sure

\long\def\csq@bquote#1#2#3#4#5#6{%
  \csq@ifnested
    {\csq@err@nbquote}
    {\csq@getpunct{%
       \ifbool{inner}
         {\csq@bquote@ii}
	 {\ifbool{csdisplay}
	    {\csq@bquote@i}
	    {\iftoggle{csq@parbox}\csq@bquote@ii\csq@bquote@i}}%
	 {#1}{#2}{#3}{#4}{#5}{#6}}}}

@TorbjørnT. Okay. Thanks! :)

@DavidCarlisle I suspect it was to test for inner hmode

@JosephWright guess so

@DavidCarlisle Something doesn't stack up though as \parbox is mentioned in the manual

> Note that csquotes will force inline quotations in footnotes, par- boxes, minipages, and floats by default.

@JosephWright oh so arguably by design, but seems a bit odd

@DavidCarlisle Well the funny thing is the test as set up doesn't work: see the display part above which isn't actually working ...