Meeting minutes
[Aki provides her background, including previous role as co-Chair of TC39]
Aki: I am also working on the Cyber Resilience Act. Happy to chat with people about that.
[About ECMA International]
<mvsamuel> Where should I be scribing to? Writing into a text file for now.
Here.
<mvsamuel> Aki was an engineer, now engineering manager, secretary of TC39.
<mvsamuel> Ecma is a bit different.
<mvsamuel> Ecma standards are freely available not RAND.
<mvsamuel> Starting discussion of History of ECMA
<mvsamuel> (Aki rabbitholed on ECMA history)
The meaning of conformance to standards
<mvsamuel> Early OCR in ECMA 3. Gets into details on how tall characters must be to be recognizable
Focus on outcomes over process.
Technical committee (26?) on noise pollution standards for computers
Standards on how to do business. Trying to nail down what is greenwashing / malicious compliance
Standards on acceptable levels of electromagnetic emissions
The CD tray shape (aka cup holder). That shape was standardized by ECMA. Along with varieties of those disks CD-ROM vs CD-ROM RW
Lots of other information storage device work incl. holographic storage
NFC (Near field communication).
ECMA is not all European. There is a technical committee that is almost entirely based in Japan which does a lot of NFC work
> 100 CST (Computer supported telecom) standards
ECMA has a fast track for standards like those
Up and coming: CLE (Common Language E???) coming in november
History of document formats as archival formats follows
Microsoft was running afoul of antitrust in the US
EC: Microsoft you must publish your document formats
While sun was working on ODF (open document format?)
Ecma did not have a not-for-profit tier, awkward working with universities. Large corps could afford dues, others not so much.
MS: railroaded Office Open XML standard (6000 pages) through standardization. Noone could implement it.
ECMA invented ISO JTC fast track. ISO brings small standards bodies output to international attention.
JTC = Joint Technical Committee
ISO: Ok, ECMA your standards are nice, so you can skip line and bring for votes quickly.
MS's office format goes before ISO ...
(intermission, Aki hunts furiously for missing slide)
Visual: street demonstrations to demand ISO pull OOXML form the ISO fast track
MS delegate: Sun delegate cannot participate. Not enough chairs. Very sorry.
ECMA offices have a room with stuff for posterity. Aki would love to contribute anything about this saga there.
New topic: 2/3
New TC (technical committees)
- Natlang interaction with AI agents protocol
- HLSL high level shader language
- structured data and schema languages (approved, not chartered)
<Ian> Here are some W3C specs around data and schema => https://
Discussion: what are schemas. RDF? Others. Aki hesitant to name them prematurely. The TC will be royalty free which requires finalizing some IP concerns before promoting.
Discussion: Is ECMA choosing winners from among ecosystems in the same space? Sometimes industry coalesces around one because sometimes having an agreed upon standard is better than having none. For example, early ECMA standard: 6b character encoding. HLSL might be a modern example.
For example, two SBOM standards are in competition but also fill different needs. Both may persist long term.
Q: Is there a risk that ECMA biases towards interests of member organizations over technical merit. Yes, that is a risk in any membership driven organization. ECMA cannot prevent members from making terrible decisions, but typically it's in their interest to propose good standards.
Open participation model increases odds that people will produce standards that get used. Ian: would love to hear about models that lead to successful standards with wider net cast. Aki: ECMA does have mechanisms for invited experts, and incorporating smaller organizations. When voting, ECMA orgs get one vote. ETC does not do things that way;
their interest is allowing members to publish.
Ian: W3C has horizontal review as a process choice. These choices and the culture shape decisions.
Q: W3C process focused on getting things suited to browsers with process requirements like dual-implementation gate. On the schema question, W3C's process might have some baggage. Might W3C learn some lessons on ECMA specifically for thngs like schemas?
cwilso: Yes, I said TC39 is its own standards body.
W3C is suited to building a platform. It needs to be a coherent platform; it needs to fill gaps so W3C spins up a gap filling project. ECMA does not think about things that way.
<Ian> (CWilson point on W3C creating a coherent platform [and how that might influence process])
ECMA has constraints that TC39 process needs. ECMA bylaws talk about standards and technical reports (TRs) and how they proceed to general assembly for voting, and has organizational structure for TC/TG (technical committees, groups) and notes and documentation requirements. Most else is left to the particular committee/group.
W3C: process constrains "here's how groups are run"
For ECMA a lot of constraints are about "record things so obviously no antitrust/collusion" and some aimed towards encouraging productive collaboration.
ECMA new TC process: declare what they do and how and how to avoid overlapping other groups work before chartering. Took two GA (general assembly meetings) to iron out scope for (???) technical committee. The secretary helps keep people within process requirements; they work with chair.
aki: in theory my technical knowledge of the TCs I am secretariat for does not need to be deep, but in practice I have a pretty good handle. But e.g. for the schema group I don't know much except that I used to e a programmer. Know enough to be able to point out "this is unclear, or this conflicts with that. are you sure that's what you want to
say?"
New topic: Programming languages
ALGOL, COBOL, FORTRAN, BASIC, C#, ECMASCRIPT (JavaScript), Dart, Eiffel, Common Language Infrastructure (CLI) platform
New topic: how can Ecma support W3C?
ian: cwilso and I have talked about (e.g. webrtc) efforts that cross bodies. Is there any model for standards bodies to harmonize, or sign an MOU, so that a groups with similar missions can agree on things like "people in this particular groups can produce things that are usable by multiple bodies?" Some other bodies skeptical of open ended
committments.
aki: ECMA has a policy on MOU (memorandum of understanding).
cwilso: ECMA patent policy is actually a TC39 specific thing. Could agree on IPR (not speaking for Google here) but that means that agreements on IPR that were scoped to one working group.
IP policy could be Katamari ball of goodness given to world, but that's not what organization members have signed up for.
Transitive nature of MOUs would be a sticking point that lawyers would not agree to.
cwilso: e.g. W3C has an MOU with Whatwg. IPR covered because of what the two were doing together. The context made it workable. Having people in both groups was a solution to challenges in this case but does not generalize.
aki: the way that makes lawyers least antsy is for working groups to have specific goals but that stifles innovation. Creative lawyering could help.
aki: seems that patent behaviour has largely transitioned from offensive to defensive.
ian: in W3C we're focused on the platform which simplifies things, but ECMA as a "we do lots of independent things" might have a hard time building on that shift.
ECMA standard common: minimum common web (???) standard.
The formatting requirements are kind of crazy. Partly driven by desire for brand consistency, but also to meet ISO requirements. For ISO, everything needs to be translated into many human languages.