Rod Vagg
Rod Vagg
Not quite sure what to do with this yet. This probably isn't the "fix", but I need to get this up before I forget the details of how I landed...
Proposal: IntactPack - storing UnixFS data in contiguous blocks for multi-mode Filecoin retrieval
_This started as a [Notion document](https://pl-strflt.notion.site/Why-Not-Both-Packing-Content-for-IPLD-vs-Piece-IntactPack-e0a8cb7e101f4a08b59b3b025bfc67a6?pvs=4) but it's kind of hidden there, although I've now published it for public viewing and commenting. I don't want it to get lost as...
* Intended to replace github.com/ipfs/go-ipld-format#ErrNotFound * A new IsNotFound() that uses feature detection rather than type checking so it's compatible with old and new forms. Ref: https://github.com/ipld/go-car/pull/363 Ref: https://github.com/ipld/go-ipld-prime/pull/493 ---...
A long-standing discussion is regarding the over-complexity of the traversal and selector paradigm for common cases. There are also serious efficiency concerns with complex or very deep DAGs. Properties of...
Ref: https://github.com/ipld/js-dag-json/issues/91 - same applies here for the dag-json codec, it should return an error if it finds such a case.
(Just a PoC & proposal at this point, for discussion) EmbedMessage sets the second-to-leftmost bit to indicate that there is a length-prefixed dag-cbor message object directly after the header. If...
I _think_ if we allowed for a stored/shared `TraversalResumerPathState` (interface for `pathState`) then we should be able to resume across any selective CAR that has the same selector & DAG....
Currently we run `verifyCompatibility()` on every call to `Prototype()` and `Wrap()` (unless it's inferring a schema for you), and it's fairly expensive. Unfortunately we currently can't cache this because we...
We have this in the `LinkSystem` config for the selective CAR creation utilities in v2: ``` LinkVisitOnlyOnce: !opts.BlockstoreAllowDuplicatePuts ``` These two things should probably not be directly related. Duplicate puts...
Introduced in #232, aside from providing `[]byte` forms (nice), they also handle the `TypedNode` `Representation()` conversion before encoding. But we don't really use them or recommend them anywhere. They really...