Juan Benet

@jbenet said (in the filecoin community slack) --- > For example if you tried to store the filecoin blockchain this way, it would include all the state trees, and even...

cc @warpfork -- relevant to your interests

(first responses, more coming) > the go-ipld package should replace github.com/ipfs/go-ipfs/merkledag package. Is that right ? that's right. > If so, I'd think we should modify both the old merkledag...

@mildred i think your analysis re the Node pointer is correct. and agree with your thoughts. the pointers are used all over importing, i believe, to make mutations of whole...

i do like the idea of keeping the thing serialized and only pulling out what you need, this in general makes programs faster, because sometimes can avoid deserializing, or can...

cc @whyrusleeping to take a look and provide feedback, as he manipulates nodes a lot. @whyrusleeping -- do you prefer the "convert to map" approach, or the "keep it all...

i think having the pointers on the object is very useful to write algorithms that traverse the graphs. it makes thinking about them much easier to be able to do...

> 8:47 jbenet: I suspect that JSON-LD is going to be slower than our current IPFS objects. not by a ton. we could benchmark this. i expect a constant increase,...

### More notes @diasdavid and I resolved that the main feature we want/need from JSON-LD at the moment is just the `@context` aliasing, and that it's a great place to...

thanks @mildred! all lgtm. > How do you know when to stop down the hierarchy ? > Does it stops when it finds a non escaped @context (or other directive)...