Meeting minutes
trackbot, prepare teleconference
Agenda Review
Action Items
r12a: I suggest we appoint a scribe.
<addison> https://
<addison> action-996?
<trackbot> action-996: Addison Phillips to Follow up with mark davis on liaison status and google participation -- due 2021-02-25 -- OPEN
addison: I pinged Mark again. He is hopefully going to take up our case-folding issue soon.
… He wants to continue being our liaison.
… He made some suggestions for dealing with Google and this issue.
… So keep the action open.
<addison> action-1006?
<trackbot> action-1006: Richard Ishida to Reply to coga issue with a proposed note addition to the text -- due 2021-03-18 -- OPEN
<addison> close action-1006
<trackbot> Closed action-1006.
r12a: That is done.
<addison> reopen action-1006
<trackbot> Re-opened action-1006.
<xfq> https://
<xfq> https://
Discussion about who the action was for. Was it for Felix? So is it not done?
xfq: Those two ^^ are related.
<xfq> https://
r12a: I replied to a similar one, I think, similar to 162
… I can add a comment to 162.
<xfq> https://
xfq: r12a also replied to 231 ^^
r12a: They all seem to be saying the same thing.
xfq: Yes, different issues but about same things.
r12a: Was I supposed to write the text? It wasn't clear from the minutes.
… I asked them to write the text, rather than write the text for them.
addison: You could extra credit by writing the text. :-)
<addison> close action-1006
<trackbot> Closed action-1006.
<addison> action-1011?
<trackbot> action-1011: Addison Phillips to Send a note to google acrep asking for potential i18n support -- due 2021-04-01 -- OPEN
r12a: Let's close the action-1006. I'll keep the web page open and will add a comment. (But wondering if it should be Felix.)
<addison> close action-1011
<trackbot> Closed action-1011.
<addison> action-1013?
<trackbot> action-1013: Addison Phillips to Comment on geolocation 50 about on going discussion of i18n on this topic -- due 2021-04-08 -- OPEN
<addison> close action-1013
<trackbot> Closed action-1013.
addison: About error msg localisation, it's on the agenda. Marcos replied, but I haven't read it in depth.
r12a: Marcos is really trying to help. We can let him off the hook.
addison: I don't expect they will change the spec.
r12a: They say they won't make substantive changes, aren't they?
addison: Sometimes we have to ask for changes, though.
r12a: At this stage they are just making editorial changes.
Info Share
addison: Anybody any info to share?
Nobody has.
RADAR Review
<addison> https://
addison: Cascading & Inheritance 4 & 5, anything left to do?
atsushi: Yes.
addison: That's our agenda item 5.
… Move it to completed?
atsushi: Yes, should be OK to move.
CSS: Use of logical properties in examples
<addison> https://
atsushi: (I'm looking through my email for the issues.)
… Fantasai wrote that level 3 won't change.
… But we may be able to delay it to level 4.
addison: It seems our request is reasonable. We would need to make a new issue agains level 4 or 5.
r12a: Looking... Maybe possible to use the same issue and just change the label.
addison: r12a, do you change the labels?
r12a: Yes, I can do that.
Action: richard reopen csswg 5570 against cascade-4/5
<trackbot> Created ACTION-1014 - Reopen csswg 5570 against cascade-4/5 [on Richard Ishida - due 2021-04-15].
geolocation 50 comment
<addison> https://
addison: There is not a lot here, but I call attention to my long note and Marcos's reply.
<xfq> https://
addison: Should we write something into specdev for this?
xfq: We already have a section about error messages, see ^^
… Maybe the section title can be made clearer.
r12a: Sound like a good idea.
xfq: The last few guidances in that section are relevant.
addison: Should we extract anything as explanation?
r12a: Yes, discuss it in specdev.
Action: addison: harvest useful text from geolocation 50 for specdev
<trackbot> Created ACTION-1015 - Harvest useful text from geolocation 50 for specdev [on Addison Phillips - due 2021-04-15].
addison: We were about to discuss it in the contet of string-meta.
String-meta "unnatural language text"
r12a: We can write it in one and link from the other.
<addison> https://
addison: I did some more editing on syntactic content.
… Based on our conversation last time.
… It has forward reference to the base example.
Bert: Text is clear, heading was a bit hard too parse, to many nouns in a row.
addison: r12a suggested mentionoing the words data and values.
r12a: I'm not sure, in the 4th para, if all of those strings are syntactic.
… We might be shooting ourselves in the foot a bit.
addison: You might indeed want to bidi-isolate them, but they are still identifiers.
r12a: Isolation might not be enough.
addison: And set direction to ltr, because there may be a strong directional letter.
<addison> HEBREW HERE YOUR BOOK IS <span dir=auto>978-1234666666</span> <!-- all rtl -->
r12a: E.g., a publication date might be in Japanese and need an appropriate font.
addison: The date is in ISO format.
… People use JSON-LD and other strings for data. They don't need lang or dir there.
<addison> Content-Type: text/html
r12a: Yes, because those things are part of the syntax, but is a date also part of the syntax?
<addison> pubDate: "2021-07-15T14:00:00Z"
addison: Look at the example. It is a value in some data structure.
r12a: Yes, but you don't say that in the text. It might be that the ISO date is just one example of a date and people could have use a different representation, rather than a syntactic requirement to use an ISO date.
… I suggest we clarify that.
… And ISBN?
addison: It is a identifier, and of couese you do need some care in presenting it. Take a MAC address, e.g.
r12a: So how do you prevent pieces from being swapped?
addison: Probably need to set dir=ltr, "auto" might not work, when there are no strongs.
addison: IP addresses and phone numbers are other examples.
… You need dir when displaying them.
r12a: I'm not clear whether we have looked into all the ins and outs.
addison: I think it is important we collectively have an idea what to recommenda and how to recomments.
… If I haven't expressed it well enough yet, that is not good.
… Are there other things we should say in this text?
r12a: I don't know.
addison: Obviously I need to clarify that para.
… Want to think about it for a week?
r12a: Yes.
epub
<xfq> https://
xfq: I have 3 pending issues
… Encoding of CSS files (1353).
… Why is UTF-16 mentioned (in addition to UTF-8) there?
addison: Is that the sole mention of utf-16 in CSS for epub?
xfq: Yes, this is the only sentence.
<xfq> > It MUST be encoded in UTF-8 or UTF-16 [Unicode].
r12a: Maybe they can refer to the encoding standard instead of Unicode. And take out utf-16.
addison: The question is why they mention utf-16. Are there books encoded in it?
… I second r12a's suggestion of referring to Encoding standard.
r12a: We can ask the question and see what they say. And depending on the answer we can suggest referring to Encoding.
addison: Do that in a second issue?
atsushi: There is a legacy encoding section.
r12a: The Encoding spec says explicitly to not use anything else than utf-8.
<xfq> https://
addison: It defines the other encodings as well, but indeed says to not use them.
addison: So objections to forwarding this issue to them?
xfq: With label needs-resolution?
<atsushi> +1
addison: Absolutely.
<Bert> +1
<xfq> https://
xfq: 2nd issue is 1354, limit of file name to 255 bytes.
… They also say file names must be unique after case normalization, but not clear what normalization they mean.
addison: That seems to call for a separate issue.
<addison> https://
addison: They have *two* restrictions on the file name. And they say the file name is utf-8, but they don't say anything about path name.
JcK: [missed]
addison: They also exclude the emoji flags. They try to make it file system safe.
JcK: I understand that, but nthey're not getting it right.
r12a: Typo U+FDDO instead of U+FDD0 (Oh instead of zero)
… The U+FDD0 … U+FDEF is indeed unassigned, but why not other unassigned characters in other blocks?
addison: And unassigned now may become assigned later.
addison: So it is a whole raft of issues...
<xfq> atsushi: U+1FFFFF might be a typo
<r12a> also U+1FFFFF is out of range
<xfq> https://
xfq: In the reading system interface there is an object for the name of the reading system. I want to ask if the name is meant for human reading.
addison: Indeed not clear if it is an identifer or text.
… You should split the file name issue into several issues and raise them all.
… case-normalization should refer to charmod.
r12a: The two typos can be together in one issue.
xfq: And the language meta data issue?
addison: Yes, we should engage in that discussion.
<r12a> https://
r12a: Different topic: Shadow DOM and inheritance of directionality. See ^^
addison: Direction also depends on base direction. Cannot answer from the fragment alone. It is an example I use.
r12a: And with a hyphen instead of a space? It can do different things, depending on the character before it.
Addison shows an example on his screen.
"123 456" vs "123-456" when they follow a Hebrew character or not.
The 456 may end up to the left of the 123. Same with "123@456" and "123~456".
"123-456" (with a hyphen-minus) keeps the 123 to the left of 456, while "123 456" (with space or @ or ~ or some others) puts 456 on the left.
addison: Should we make this issue an agenda item next week?
… We should read the issue thread. Are they reinventing bidi?
<addison> (for next time)