Re: simplerdf & Towards the future RDF library

Hi Nicola,

> I am not sure what you mean by performance gain

For instance, one of the reasons why N3.js is fast is because I use very simple data structures.
I'm pretty sure (= experimentally verified in the past) that using heavier interfaces will slow things down.
Now of course, I understand the interoperability benefits of having a shared interface;
but we should surely think about what this interface should look like.
And I think that library authors, like you and myself, should have a say on that
(regardless of whether RDF-Ext or a new library X becomes the thing we program for).

> In fact, to serve my web app in javascript I can just browserify what I need, instead of browserifying the whole (previous) library.

I can imagine that; but there is a difference between modularization on a technical perspective
or modularization on organizational perspective.

For instance, N3.js is modular from a technical perspective:
you can browserify any of its parts and they will work without the others.
It is not modular from an organizational perspective: it's just one lib/repo.

> However, I think that we - as the rdfjs group - can also bring this and the RDF-Ext spec conversation in an appropriate place as suggested by others

I'd really love that. We need to stick together.
I much rather develop something together with you guys
than continue my own separate projects—which would never have
all the features we could build together.

Summarizing: I'm all for technical modularization,
but we should organize ourselves organizationally much better.
One core lib for basic things like parsing/storing/writing
(think of Jena and RDF.rb) might actually be good;
and people could still modularize that if needed for browsers.

Best,

Ruben

Received on Saturday, 12 September 2015 15:10:29 UTC