Respond to session chat (separate from Q&A) in revision #19
Description
Aug 6, 2021 1:34 PM
Vincent Lizzi
What is the origin of the quote Michael Kay referenced "If you don't believe me then you're not using your imagination well enough", or is this original?
David J. Birnbaum
https://obdurodon.org/misc/gin.xhtml
http://dev.obdurodon.org/djb/Balisage2021.pdf
Norm Tovey-Walsh
Bombay Saphire East?
It's a nice list of gins.
Bethan Tovey-Walsh
Is David crackly for everyone or just me?
David J. Birnbaum
I'm crackly for everyone. It's my way.
Tommie Usdin
David's voice is crackly for all of us. David may also be crackly, but that's a different thing.
Bethan Tovey-Walsh
Let me try that again - is David's audio crackly for everyone or just me?
No, it's the audio, not the voice, and not the David.
David J. Birnbaum
David is muted, and will try to stay that way.
12:00 = 2:00
Evan Lenz
For some reason, this session is not showing up in the mobile app
David J. Birnbaum
@evan: Does https://whova.com/portal/webapp/balis_202107/Agenda/1677911 work?
C M Sperberg-McQueen
/me waits for a shoe labeled kerning to drop.
blam
Hugh Cayless
Boom
Bethan Tovey-Walsh
Yep, I was just wondering about that!
Fixed width would fix that, presumably?
Tommie Usdin
"Fixed" in some meaning of the word. Kerning is good!
Bethan Tovey-Walsh
Though then you also wouldn't need to get the widths of the individual chars, I guess.
Gerrit Imsieke
Aren't the kerning pairs also encoded in the font metrics?
Tommie Usdin
It is POSSIBLE to have variable width characters without kerning.
Bethan Tovey-Walsh
That is very cool.
John Lumley
Impressive for an undergraduate - we had one who was good enough to let them present a paper (on Gbit networking) at an ACM conference.... Congratulations Charlie!
Tommie Usdin
John: I completely agree. The work is impressive and so is the presentation.
Bethan Tovey-Walsh
Impressive full stop, I think.
C M Sperberg-McQueen
Do fonts ever use negative kerning? Or is it a safe assumption that the sum of character widths will be a strict upper bound for the width as laid out?
Bethan Tovey-Walsh
I think that depends massively, Michael.
I've seen negative kerning, e.g. for combining diacritics.
Liam Quin
yes, -ve kerning is possible as well as glyph substitution
for example "rn" often has a kern pair to space them further
(SVG can contain script elements of course)
Joel Kalvesmaki
Fonttools will convert your fonts to XML: https://pypi.org/project/fonttools/
See also https://groups.google.com/g/fonttools
From there the candy store is WIDE open
C M Sperberg-McQueen
I think he must have meant alternating turtles and elephants ?
all the way down
Aug 6, 2021 2:35 PM
Gerrit Imsieke
s and
C M Sperberg-McQueen
@gerrit, right
Syd Bauman
The other answer to Charlie’s question “how do I get the main axis label to be somewhere presentable with respect to the individual column labels” is to put the main label closer to the axis than the column labels. :-)
Liam Quin
there was a graphics diagramming language that used dependencies
called Ideal, i think by Lorinda Cherry (Cherygh?)
John Lumley
There was some work by Cameron McCormack, which coupled SVG to a constraint resolution system
C M Sperberg-McQueen
@Syd, your solution leaves open the question 'how wide do i have to make this thing in order not to have too much whitespace on the extreme R or bottom?
John Lumley
I used it significantly
Steven Pemberton
Yes, redesign SVG to use constraints! The current SVG is far to dependent on being told where to place elements
John Lumley
I have developed a constraint-decorated SVG
C M Sperberg-McQueen
@john, pointer?
John Lumley
e.g. 'all my children are in a Y-flow
Something in DocEng c 2013- will have to check
C M Sperberg-McQueen
I was just thinking that a pre-processor or wrapper would be the best way to do what SP suggested without having to get into the task of persuading the SVG WG to change things
Evan Lenz
Or a JavaScript-based framework for enhanced SVG that does it for you
Bethan Tovey-Walsh
@john - this one? https://dl.acm.org/doi/10.1145/2494266.2494274
John Lumley
Yes - I no longer have ACM DL access but that is the one
Evan Lenz
A preprocessor couldn't produce perfect results without the browser
short of reimplementing an SVG formatter in XSLT :-)
John Lumley
I have the PDF and original source (somewhere)
Bethan Tovey-Walsh
Devanagari would be a good challenge... I kinda want to try now!
Evan Lenz
Maybe the constraints-enhanced SVG could be implemented using Saxon-JS 😲
John Lumley
I could, but it would involve a lot of new coding without reference to stuff done while I was employed
Bethan Tovey-Walsh
https://github.com/harfbuzz/harfbuzz
^^ that's the text-shaping engine Liam just mentioned
Gerrit Imsieke
(starting to think that the occasional non-update of the chat message list is due to the fact that the last viewed messages stays in focus while more messages are getting added at the bottom, but it won't scroll down by itself -- you need to pick the scrollbar and move it down manually)
Bethan Tovey-Walsh
@gerrit - nope, sometimes I'm missing whole chunks of chat, higher up than the current message.
Gerrit Imsieke
I wonder why they didn't fix it then in the past year if it apparently affects many users and should be somehow reproducible.
Ari Nordström
I've given up on the chat, mostly. I'm experiencing the same as @bethan, and it's quite frustrating.
Marion Riley
Great job Charlie! Super interesting and clever ideas:)
Evan Lenz
Re: PDF, I once did a project for paginated flow charts. I ended up using XSL-FO for each box, getting the dimensions from AH Formatter's area tree XML, generating the SVG from those, and embedding those in XSL-FO. To summarize: FO within SVG within SVG. It worked great! :D
Oops, I mean FO within SVG within FO
Tony Graham
That was in the requirements for XSL 2.0.