Mark S. Miller

The weakmap in question is small and short-lived, right? If so, it should actually be a map. Is this a performance critical piece of code? If the code is not...

We should work through some more examples. But my immediate reaction is that I like the templateGroups approach much better.

Reopening as requested. RegExp.make does not return a subclass. AFAIK the possibility has not been considered. Why might it help? I don't see it.

Yes. The thing about shimming is that the resulting environment ideally does not differ in any observable way from the environment being emulated.

As stated above, the accessors installed by the SES-shim https://github.com/endojs/endo/blob/8e077abc65a644225b70b1d328cde11a44364b0e/packages/ses/src/enable-property-overrides.js#L91 emulate how a frozen data property would work if tc39 had not made the override mistake. It is an imperfect...

See also https://github.com/Agoric/agoric-sdk/issues/9265

@kriskowal , @frazarshad , I co-assigned this to the three of us. @kriskowal I see you merged https://github.com/endojs/endo/pull/2231 despite the two CI failures https://github.com/endojs/endo/actions/runs/8760680740/job/24046036596 . What’s failing there is exactly...

See also https://github.com/Agoric/agoric-sdk/pull/9267 https://github.com/Agoric/agoric-sdk/pull/9273 https://github.com/Agoric/agoric-sdk/pull/9274 and https://github.com/Agoric/agoric-sdk/pull/9267#issuecomment-2067874367 https://github.com/Agoric/agoric-sdk/pull/9267#issuecomment-2067877145

At first I did ask that deep-path notation work for objects with no extra notation. @littledan explained why it was too surprising. Once I got it, I agreed (and agree)...

Since I'm suggesting brainstorming on notation, I'll make a suggest I do not like: ```js const state2 = #{ .....state1, counters[0].value: 2, counters[1].value: 1, metadata.lastUpdate: 1584383011300, }; ``` where each...