[{"author": "Michael Prorock", "text": "

this is always one of the best WGs

", "time": "2024-03-22T05:00:14Z"}, {"author": "Francesca Palombini", "text": "

yes! :D

", "time": "2024-03-22T05:00:25Z"}, {"author": "Chris Lemmons", "text": "

We're not small, we're just packed efficiently.

", "time": "2024-03-22T05:00:31Z"}, {"author": "Christian Ams\u00fcss", "text": "

:-D

", "time": "2024-03-22T05:00:37Z"}, {"author": "Maria Mat\u011bjka", "text": "

shall i tag myself then?

", "time": "2024-03-22T05:00:48Z"}, {"author": "Eduard V", "text": "

24 people

", "time": "2024-03-22T05:01:23Z"}, {"author": "Chris Lemmons", "text": "

I\u00a0vote yes, because I\u00a0personally would benefit and because it's recorded.... which he just said. :D

", "time": "2024-03-22T05:01:56Z"}, {"author": "Michael Prorock", "text": "

i resist the urge to heckle Orie

", "time": "2024-03-22T05:02:56Z"}, {"author": "Francesca Palombini", "text": "

Thank you guys, it was a pleasure to be your AD! <3 I will still be around :)

", "time": "2024-03-22T05:03:33Z"}, {"author": "Michael Richardson", "text": "

My name is Michael. It's been 7 weeks since I registered a CBOR Tag.

", "time": "2024-03-22T05:03:50Z"}, {"author": "Christian Ams\u00fcss", "text": "

:rolling_on_the_floor_laughing:

", "time": "2024-03-22T05:04:05Z"}, {"author": "Chris Lemmons", "text": "

My name is Chris. It's been -x weeks since I\u00a0registered a CBOR Tag because my first one is still in review. :D

", "time": "2024-03-22T05:04:46Z"}, {"author": "Michael Richardson", "text": "

(ps: I have a RUST macro that consumes cbor-pretty as binary)

", "time": "2024-03-22T05:05:13Z"}, {"author": "A.J. Stein", "text": "

Chris Lemmons said:

\n
\n

My name is Chris. It's been -x weeks since I\u00a0registered a CBOR Tag because my first one is still in review. :D

\n
\n

Is this a confession or the beginning or a support group declaration?

", "time": "2024-03-22T05:05:26Z"}, {"author": "Michael Richardson", "text": "

support group.

", "time": "2024-03-22T05:05:50Z"}, {"author": "Christian Ams\u00fcss", "text": "

pretty macro is sure useful. I'm about to pack cbor-diag up in a way that you can write cbor!('[0, b\"hello\"]') and get a serialized constant.

", "time": "2024-03-22T05:06:24Z"}, {"author": "A.J. Stein", "text": "

Would still be interested in this CSV draft, interesting!

", "time": "2024-03-22T05:06:45Z"}, {"author": "Michael Richardson", "text": "

Hmm. That would be way more interesting and descriptive in my unit test cases.

", "time": "2024-03-22T05:07:01Z"}, {"author": "Chris Lemmons", "text": "

Floating point numbers are silly anyways.

", "time": "2024-03-22T05:07:23Z"}, {"author": "Michael Richardson", "text": "

Chris Lemmons said:

\n
\n

Floating point numbers are silly anyways.

\n
\n

27.314% of the time.

", "time": "2024-03-22T05:07:46Z"}, {"author": "Orie Steele", "text": "

Don't want to interrupt the speaker

", "time": "2024-03-22T05:09:55Z"}, {"author": "A.J. Stein", "text": "

Michael Richardson said:

\n
\n

Chris Lemmons said:

\n
\n

Floating point numbers are silly anyways.

\n
\n

27.314% of the time.

\n
\n

\"0.27314\" of the time can be a string, I saved us pain in CBOR, JSON, and many other places where I don't plan on doing math with any precision. But I know that won't make 100% of people, so I give up.

", "time": "2024-03-22T05:10:01Z"}, {"author": "Orie Steele", "text": "

are text strings always required to be quoted?

", "time": "2024-03-22T05:10:09Z"}, {"author": "Orie Steele", "text": "

in EDN ?

", "time": "2024-03-22T05:10:18Z"}, {"author": "Chris Lemmons", "text": "

Someone working on a claim in a spec I'm working on elsewhere wanted to use a floating point value and after running into all the corner cases, just decided to use a pair of integers to form a fraction instead. It wound up being more straightforward.

", "time": "2024-03-22T05:11:32Z"}, {"author": "A.J. Stein", "text": "

Orie Steele said:

\n
\n

are text strings always required to be quoted?

\n
\n

Unsure. Which spec is \"the spec\" I don't know, but RFC8610:

\n

o Text strings are enclosed by double quotation '\"' characters.
\n They follow the conventions for strings as defined in Section 7 of
\n [RFC8259]. (ABNF users may want to note that there is no support
\n in CDDL for the concept of case insensitivity in text strings; if
\n necessary, regular expressions can be used (Section 3.8.3).)

\n

(I assumed there may be a dedicated RFC for EDN but I was wrong and Christian definitely knows more than me. :laughing: )

", "time": "2024-03-22T05:13:17Z"}, {"author": "Christian Ams\u00fcss", "text": "

text strings are always quoted; no bare words

", "time": "2024-03-22T05:13:20Z"}, {"author": "Orie Steele", "text": "

resisting the urge to make leap second jokes

", "time": "2024-03-22T05:13:25Z"}, {"author": "Orie Steele", "text": "

... carsten is too faster for my snark

", "time": "2024-03-22T05:13:38Z"}, {"author": "Orie Steele", "text": "

Thanks Christian, I think we've got some bugs in our SCITT examples, and it my EDN tooling.

", "time": "2024-03-22T05:14:10Z"}, {"author": "Christian Ams\u00fcss", "text": "

\"Everybody knows\", I wish. (I don't.)

", "time": "2024-03-22T05:15:15Z"}, {"author": "A.J. Stein", "text": "

There are CI systems that don't tell people everything they did wrong is immediately most of the time? I didn't know I was missing out.

", "time": "2024-03-22T05:17:18Z"}, {"author": "A.J. Stein", "text": "

And next, we shall have best practices on spaces versus tabs in the CDDL ...

", "time": "2024-03-22T05:18:59Z"}, {"author": "Orie Steele", "text": "

Sounds like a linter is in our future.

", "time": "2024-03-22T05:19:41Z"}, {"author": "Michael Richardson", "text": "

Chris Lemmons said:

\n
\n

Someone working on a claim in a spec I'm working on elsewhere wanted to use a floating point value and after running into all the corner cases, just decided to use a pair of integers to form a fraction instead. It wound up being more straightforward.

\n
\n

Yeah, in ROLL, after going for mattissa/exponent in 12 bits, we dropped the mantissa, because it was just measuring scale.

", "time": "2024-03-22T05:20:01Z"}, {"author": "Christian Ams\u00fcss", "text": "

or just a reformatter.

", "time": "2024-03-22T05:20:03Z"}, {"author": "Orie Steele", "text": "

but the \"_\" vs \"-\" would be a lint rule I think.

", "time": "2024-03-22T05:20:32Z"}, {"author": "Orie Steele", "text": "

this e' syntax is much more readable

", "time": "2024-03-22T05:21:21Z"}, {"author": "Christian Ams\u00fcss", "text": "

no, the reformatter would just rewrite and do
\nimport COSE_KEY as cose-key :-D`

", "time": "2024-03-22T05:21:24Z"}, {"author": "Orie Steele", "text": "

ahh...

", "time": "2024-03-22T05:21:34Z"}, {"author": "A.J. Stein", "text": "

I was being flippant and saying I am all for style requirements but I don't expect people will make CDDL require a style so uniform it's just syntax as that point.

", "time": "2024-03-22T05:21:39Z"}, {"author": "Michael Richardson", "text": "

https://archive.org/details/anatomyoflisp0000alle/page/n5/mode/2up I highly recommend this book, speaking of 1958.

", "time": "2024-03-22T05:22:49Z"}, {"author": "A.J. Stein", "text": "

In other data-formats vars and file names Orie would hate it: I will group things with dashes and then pseudo-suffixes with underscores so the file names or vars can be both. Now, more than one downstream consumer complained about this, and I inherited it this way, I came to enjoy it, but still others strongly disliked it. I got the dislike, believe me.

", "time": "2024-03-22T05:23:02Z"}, {"author": "Christian Ams\u00fcss", "text": "

The example is a prime speciment of \"Tell me you are German without telling me you are German\"

", "time": "2024-03-22T05:27:46Z"}, {"author": "Orie Steele", "text": "

We'll need a document shepherd for this soon

", "time": "2024-03-22T05:29:25Z"}, {"author": "Orie Steele", "text": "

and reviews... https://datatracker.ietf.org/doc/draft-ietf-cbor-cddl-more-control/

", "time": "2024-03-22T05:29:52Z"}, {"author": "Christian Ams\u00fcss", "text": "

Yes; added to the current-documents section coming up after the tutorial. (Let's try for a non-chair shepherd.)

", "time": "2024-03-22T05:33:02Z"}, {"author": "Orie Steele", "text": "

+1 for non chair shepherd, feel free to ask the list for volunteers

", "time": "2024-03-22T05:35:11Z"}, {"author": "A.J. Stein", "text": "

Oh boy we got to namespaces, I didn't know this is on the horizon for CBOR. (I am actually intrigued by this, not snark.)

", "time": "2024-03-22T05:35:19Z"}, {"author": "Brendan Moran", "text": "

we desperately needed import semantics. That made namespaces inevitable

", "time": "2024-03-22T05:37:37Z"}, {"author": "Orie Steele", "text": "

^ that

", "time": "2024-03-22T05:39:56Z"}, {"author": "A.J. Stein", "text": "

So I am not sure I go to mic: I spend a good deal of time at work with XSLT3 which has a similar import/include semantics (please don't throw rocks at me). As a developer this was confusing and I guess I should have known better but this is different from many commonplace module mechanics in programming and templating and other languages. so ... yeah.

", "time": "2024-03-22T05:40:38Z"}, {"author": "Orie Steele", "text": "

you want me to ask that at the mic?

", "time": "2024-03-22T05:41:10Z"}, {"author": "Christian Ams\u00fcss", "text": "

Please go to the mic

", "time": "2024-03-22T05:41:13Z"}, {"author": "Brendan Moran", "text": "

I also prefer pythonic syntax for this kind of thing (note: I'm an author on this draft)

", "time": "2024-03-22T05:41:16Z"}, {"author": "Michael Richardson", "text": "

A.J. Stein said:

\n
\n

So I am not sure I go to mic: I spend a good deal of time at work with XSLT3 which has a similar import/include semantics (please don't throw rocks at me). As a developer this was confusing and I guess I should have known better but this is different from many commonplace module mechanics in programming and templating and other languages. so ... yeah.

\n
\n

so, the \"empty\" gets you the set of stuff with the current module, so... it's /usr/include.

", "time": "2024-03-22T05:41:25Z"}, {"author": "A.J. Stein", "text": "

I mean I can but this is just like an opinion. Might seem dumb, I am also more excited by this module targets issue.

", "time": "2024-03-22T05:41:32Z"}, {"author": "Brendan Moran", "text": "

No, please share your opinion

", "time": "2024-03-22T05:41:47Z"}, {"author": "Michael Richardson", "text": "

Brendan Moran said:

\n
\n

I also prefer pythonic syntax for this kind of thing (note: I'm an author on this draft)

\n
\n

What exatly do you mean here?

", "time": "2024-03-22T05:41:53Z"}, {"author": "Orie Steele", "text": "

I like import from URL, and from RFC, and from relative file, and from absolute file, etc....

", "time": "2024-03-22T05:41:53Z"}, {"author": "Christian Ams\u00fcss", "text": "

the import/include is confusing to me too; main motivation AIU is that it's geared to even work with linters after conversion to CDDL1.

", "time": "2024-03-22T05:41:55Z"}, {"author": "Brendan Moran", "text": "

@mcr: python uses a single word to do both include and import:
\nfrom foo import * <- include
\nimport foo as bar <- import

", "time": "2024-03-22T05:43:01Z"}, {"author": "Orie Steele", "text": "

newer languages like Go and Bun, and Deno consider imports by URL

", "time": "2024-03-22T05:43:21Z"}, {"author": "Orie Steele", "text": "

a package manager is in our future.

", "time": "2024-03-22T05:43:33Z"}, {"author": "Michael Richardson", "text": "

Christian Ams\u00fcss said:

\n
\n

the import/include is confusing to me too; main motivation AIU is that it's geared to even work with linters after conversion to CDDL1.

\n
\n

include, is like C, #include. it just pastes stuff into the current file.
\nimport is more like python, gives you namespaces, and recursive importing, but only of stuff that is used.

", "time": "2024-03-22T05:43:44Z"}, {"author": "Michael Richardson", "text": "

I would use include for stuff in my local module, which I want to keep in multiple files for various reasons. Such as because I want to {:include them into different parts of my kramdown.

", "time": "2024-03-22T05:46:25Z"}, {"author": "Orie Steele", "text": "

TypeScript also has import *

", "time": "2024-03-22T05:46:46Z"}, {"author": "Michael Richardson", "text": "

I don't object to the suggested change.

", "time": "2024-03-22T05:47:06Z"}, {"author": "Christian Ams\u00fcss", "text": "

python sounds good for me

", "time": "2024-03-22T05:47:14Z"}, {"author": "Orie Steele", "text": "

import * as COSE from 'RFC9052',

", "time": "2024-03-22T05:47:41Z"}, {"author": "Brendan Moran", "text": "

import * from RFC9052 as COSE

", "time": "2024-03-22T05:48:10Z"}, {"author": "Orie Steele", "text": "

^ python vs TS but yes, same concept

", "time": "2024-03-22T05:48:30Z"}, {"author": "Orie Steele", "text": "

regardless, I prefer to not have distinct \"include: vs \"import\" even if both are conceptually possible.,

", "time": "2024-03-22T05:48:50Z"}, {"author": "Brendan Moran", "text": "

^ I was making a proposal. Python would be:
\nfrom RFC9052 as COSE import *

", "time": "2024-03-22T05:49:21Z"}, {"author": "A.J. Stein", "text": "

Overall, these are interesting cool changes. I am excited about these developments in CBOR, nice work everyone.

", "time": "2024-03-22T05:49:37Z"}, {"author": "A.J. Stein", "text": "

Oh XML Schema time datatypes and timezones, flash backs to work work projects, I feel your pain.

", "time": "2024-03-22T05:53:46Z"}, {"author": "A.J. Stein", "text": "

So mcr is talking about this on the queue?

\n

https://datatracker.ietf.org/doc/draft-ietf-anima-constrained-voucher/

", "time": "2024-03-22T05:58:41Z"}, {"author": "Christian Ams\u00fcss", "text": "

yes

", "time": "2024-03-22T05:58:53Z"}, {"author": "Orie Steele", "text": "

Is application/yang+cbor a thing ?

", "time": "2024-03-22T05:59:10Z"}, {"author": "Chris Lemmons", "text": "

The good news is that if we use packed CBOR, we can say their name 3 times very efficiently. :)

", "time": "2024-03-22T06:00:27Z"}, {"author": "A.J. Stein", "text": "

Orie Steele said:

\n
\n

Is application/yang+cbor a thing ?

\n
\n

Lazy analysis: explicit search on Google found nothing but with prefixes you don't see it IANA no?

", "time": "2024-03-22T06:00:30Z"}, {"author": "Orie Steele", "text": "

its yang-data+cbor apparently

", "time": "2024-03-22T06:01:33Z"}, {"author": "Orie Steele", "text": "

https://www.iana.org/assignments/media-types/application/yang-data+cbor

", "time": "2024-03-22T06:01:47Z"}, {"author": "A.J. Stein", "text": "

I assume there is a reasonable back story, but netmod being last resort for YANG work versus their published charter has me ... confuzzled.

", "time": "2024-03-22T06:04:07Z"}, {"author": "Chris Lemmons", "text": "

If everybody uses the same packed model, we can all share the same libraries.

", "time": "2024-03-22T06:17:26Z"}, {"author": "Christian Ams\u00fcss", "text": "

... and still fill the tables with common population mechanisms if need be

", "time": "2024-03-22T06:18:42Z"}, {"author": "Chris Lemmons", "text": "

Also, it's worth not underestimating brain capacity. People that understand packed cbor understand it everywhere. People that understand bespoke encodings only understand exactly them.

", "time": "2024-03-22T06:19:15Z"}, {"author": "A.J. Stein", "text": "

I think this could be super interesting as a next-gen Parquet and why sometimes the packed model could be preferred: you could remotely stored larged packed objects in a remote blob store and now how to intelligently query and retireve segments. I don't know how Parquet operates like that but primitives like this could be great to surpass this if it is naive.

\n

(So Parquet just does the naive compression so this validates my guess about an interesting future, save the annoyance of coordinating model info. https://github.com/apache/parquet-format/blob/master/Compression.md)

", "time": "2024-03-22T06:19:18Z"}, {"author": "Chris Lemmons", "text": "

I think this slide meant to say \"Except for the title and also this footnote\"

", "time": "2024-03-22T06:30:34Z"}, {"author": "Maria Mat\u011bjka", "text": "

see ya!

", "time": "2024-03-22T06:30:41Z"}, {"author": "Martine Lenders", "text": "

See you all in Vancouver or online!

", "time": "2024-03-22T06:30:53Z"}, {"author": "Chris Lemmons", "text": "

Safe journeys to all of you!

", "time": "2024-03-22T06:30:57Z"}]