Ecosyste.ms: Timeline
Browse the timeline of events for every public repo on GitHub. Data updated hourly from GH Archive.
mhofman created a comment on a pull request on Agoric/agoric-sdk
I believe an alternative we recently discussed is to do the serialization "locally" in a board library, and just send `slots` to the board vat.
mhofman created a comment on an issue on Agoric/agoric-sdk
I'm tempted to say we should not bother fixing this until we work on https://github.com/Agoric/agoric-sdk/issues/10799, especially if we disabled otel-trace in integration tests (https://github.com...
mhofman opened an issue on Agoric/agoric-sdk
Rework tracing slog sender
## What is the Problem Being Solved? We have an `otel-trace` slog sender which creates open telemetry traces by inferring relations between slog events. However it is fragile (https://github.com...
mhofman opened an issue on Agoric/agoric-sdk
Replace otel-trace with otel-context-aware-slog in integration test
## What is the Problem Being Solved? We don't currently use any of the otel tracing reporting (both because it's missing capabilities and because we no longer have a good frontend for it), and I...
mhofman opened an issue on Agoric/agoric-sdk
Classify trigger in contextualized slog sender
## What is the Problem Being Solved? The contextualized slog sender generates a high level trigger type based on the source of the inbound trigger (timer poll, bridge inbound, kernel update even...
mhofman created a comment on an issue on Agoric/agoric-sdk
I think something like this is needed for us to be able to run e2e integration tests on top of the simulated upgrade provided by agoric-sdk. My understanding right now is that nightly e2e tests onl...
mhofman created a comment on an issue on Agoric/agoric-sdk
We still have a remaining leak after the workaround for watchPromise: https://github.com/Agoric/agoric-sdk/issues/10794
mhofman opened an issue on Agoric/agoric-sdk
walletFactory heap is still growing after u18
## Describe the bug Investigating the wallet factory leak in u17 and before (#10706), we identified an issue (#10757) and implemented a workaround for it (#10756), however from preliminary obser...
mhofman created a comment on an issue on Agoric/causeway
In https://github.com/Agoric/agoric-private/issues/235#issuecomment-2569807154 I described some of the breadcrumbs I followed. I think the main aspect is being able to go from syscall send to send ...
mhofman created a review comment on a pull request on endojs/endo
I wouldn't expect anyone to know this, it's a very obscure quirk of `.bind()`
mhofman created a review comment on a pull request on endojs/endo
Maybe I missed something, but I still see this `shouldThrow` which I don't understand the purpose of. Why are we not throwing unconditionally here?
mhofman created a review comment on a pull request on endojs/endo
You do need to redefine `.name` if you want to keep the original `"Proxy"` name instead `"bound Proxy"`
mhofman created a review on a pull request on endojs/endo
Reopened the `shouldThrow` conversation as I don't see anything changed
mhofman created a review on a pull request on endojs/endo
Reopened the `shouldThrow` conversation as I don't see anything changed
mhofman pushed 1 commit to mhofman/ergonomic-exo-classes endojs/endo
- WIP: remove unneeded target prototype change 800d20d
mhofman created a review on a pull request on Agoric/agoric-3-proposals
I have [confirmed](https://github.com/Agoric/agoric-sdk/actions/runs/12571310964/job/35042085407#step:10:5884) the above. This hook should be sufficient to let us do anything needed.
mhofman created a review comment on a pull request on Agoric/agoric-sdk
This works but FYI the preferred flow for manually testing integration between multiple repos for NPM packages is to generate a `yarn patch '@agoric/synthetic-chain'` and check the patch file in. ...
mhofman created a review on a pull request on Agoric/agoric-sdk
Satisfied that this show the hook works
mhofman created a review on a pull request on Agoric/agoric-sdk
Satisfied that this show the hook works
mhofman created a review comment on a pull request on endojs/endo
Not entirely true. Since we never care about the receiver in the case of a plain call, you could use `.bind(null)` which creates a function that lacks a `.prototype` but is constructible if the ori...
mhofman created a review comment on a pull request on endojs/endo
To avoid any potential collisions with trap methods, let's store the original handler as a unique symbol property on `handlerPlus`. After all that's what symbols were meant for :) ```js const O...
mhofman created a review on a pull request on endojs/endo
I've only reviewed the implementation so far. Looks good overall, except I don't understand this `shouldThrow` condition. Also a potential optimization: when adding a proxy to the `noTrappingSet...
mhofman created a review on a pull request on endojs/endo
I've only reviewed the implementation so far. Looks good overall, except I don't understand this `shouldThrow` condition. Also a potential optimization: when adding a proxy to the `noTrappingSet...
mhofman created a comment on a pull request on endojs/endo
> I have a feeling that we can prove an equivalent approach would be to store state on proxy objects only, since by definition ordinary objects can't be trapping. It was obviously too late for m...
mhofman created a review comment on a pull request on endojs/endo
Yeah I definitely considered suggesting either setting the object's trap as `undefined` or as the original Reflect function.
mhofman created a review comment on a pull request on endojs/endo
I thought about that but the main reason I suggested preinstalling traps is for performance reasons for engines that use a hidden class. I'm not sure how a proxy proto interacts with that, but in g...