Meeting minutes
Symbols
[discussion of timeline and reqs for TPAC]
matatk: If we have a couple of fonts ready for TPAC (Bliss and maybe ARASAAC) then I can take out 95% of the code for the extension and authoring tool I wrote, and we can make it a 'real' demo. (Except for using the PUA.)
Russell: Hannes and I had the same thought last week!
Russell: Knowing that we want to use SVG glyphs, I looked at Global Symbols, Mulberry popped out to me (SVG, CC licenced).
Russell: The mapping needs to be done.
Russell: I had a quick look; they have 3,200 to 3,400 symbols. Some duplicates. I did some matching and curating, and came up with 498 symbols.
Russell: Hannes is as we speak working on those two fonts. I think he understands the need for the ligatures. We'll work together on it.
Russell: I'm sure we can get this done by October.
Russell: Note though that this is a very large number of ligatures, my research indicated this ought to be feasible, but we'll see in practice.
matatk: I can start stripping away code.
janina: We should be aware that surprises will happen. Any project without surprises would be unheard of.
janina: We need to declare our intent to meet at TPAC by tomorrow.
matatk: If it's not practical, then we can demonstrate _that_
Russell: In theory it's doable, but maybe assumptions in systems' implementation mean that it isn't in practice - it's reasonable that it should work, though, and we ought to be able to do it by then.
janina: Eager that we not lose momentum. This sounds really good.
matatk: +1! Thanks Russell and Hannes
matatk: Asks whether we could have a placeholder font soonest in order to practice mappings
matatk: Is it relatively easy -- a glyph per word or such?
matatk: Suggests variety of representatioh for testing -- double line below, one red, one green, etc
Russell: there are triditional ligatures that have both separate codepoints and the individual glyph
Russell: how it's accessed in a sequence of codepoints -- code is always checking it's codepoints for a sequence of text
Russell: So, we can do that, but it won't work that way once developed
matatk: We need to work out the steps so the AAC devs can follow the guide to create the fonts
Russell: not by tpac
matatk: of course!
matatk: The kernel of how it goes would be good to have though
Russell: agree
Russell: still wondergin whether ligatures work in privade Unicode space
Russell: any font should work for testing font interchange
Russell: obviously for a demo you want AAC, but for dev work any font should do
matatk: Also want to demo authoring
matatk: Important we document anything that doesn't work as expected
matatk: Would like to understand better
matatk: Now we're mapping from a codepoint to a glyph
matatk: We need the list ...
Russell: I have it. Got it from Michael's public draft document.
matatk: We may not want to pub Registry with private use references; but we'll need it for our work
matatk: Trying to see the data struct of the Registry
matatk: we need all these things to match up
matatk: all based on valid codepoints that make up the Registry
matatk: Wondering how to help? How to marry up with glosses?
Russell: I map BCI IDs, have it in spread sheet, not codepoints but easily matched
matatk: we can use internally though?
Russell: On an aside -- Doug Crawford, inventor of json, very into Bliss
Russell: he came up with his own version and was doing things with itmaSo, canonical needn't be json, but it's useful to us at this point to use json
matatk: our goal is interoperable font, but we're not quite there. However, we want to dev efficiently so that we don't unnecessarily repeat processes as we work this out
Russell: bliss string would be the key
Russell: gloss, id, codepoint string
Russell: I basically have this, as does Hannes; we can put it together as is useful
matatk: I propose to create a schema we can test against some json; perhaps for next week?
Russell: I'm currently mostly spread sheet, Libreoffice Calc
Russell: Google SS useful for scrolling to see the symbols
matatk: it's imp[ortant we can roundtrip among these various formats
matatk: want to be ready when the font is ready
Russell: So from me a small set using codepoints?
matatk: maybe not yet
matatk: a pointer to the source of truth for the bcp
Russell: on BCI web
Russell: will send link and one to the Google Sheets version
matatk: sounds great
matatk: json so it has everything we build everything else from