Kasey
Kasey
> I think it makes sense to separate the commit where the transform is updated from the commit where the data is updated by the new transform Okay thank you...
ahhhh. Okay, will need to think about it more. Feels like that needs its own deep dive of discussion.
Glad we've come to consensus around `qri save`: remove the `--apply` flag from CLI & make the default behavior `no-apply` when adding a transform. The main questions surrounding `apply` vs...
has this been closed by: https://github.com/qri-io/qri/pull/1870 ? The PR adds `profileid := profile.IDFromCtx(ctx)` and passes that profile id to the `event in` `bus.publish`: https://github.com/qri-io/qri/blob/d451f6f8a2b247958e80fdd14c79cfc0d4e6b1f2/event/event.go#L153-L170
posting my comment from chat: > Need to look further into this, but this is most likely caused by the way we define and generate previews. We only return a...
Getting a race condition with two different patterns for tests relying on the save path, both touching `dsfs` 1: ``` github.com/qri-io/dataset.(*Dataset).DropTransientValues() /Users/ramfox/go/pkg/mod/github.com/qri-io/[email protected]/dataset.go:244 +0x9c4 github.com/qri-io/qri/base/dsfs.commitFileAddFunc.func1() /Users/ramfox/go/src/github.com/qri-io/qri/base/dsfs/commit.go:69 +0x905 github.com/qri-io/qri/base/dsfs.WriteDataset() /Users/ramfox/go/src/github.com/qri-io/qri/base/dsfs/write.go:221 +0x524 github.com/qri-io/qri/base/dsfs.CreateDataset()...
We will be adding a `/collection/get` endpoint, that takes a `Ref` or `InitID`, and returns a `VersionInfo` with the additional fields needed to populate the `DatasetHeader` Check out this comment...
actually, this has implications for other parts of the API. If we have no way of knowing any authors of a doc (besides ourselves) at the CLI / console level,...
Haven't done a read-through yet, but just looking at the protocol changes: we haven't been using rate limiters for the client yet, but we should talk about what we might...
good point. The export modal should describe this.