RDFLib 8 Scope

[nicholascar]

This Discussion is about the high-level, major, items of functionality in RDFLib that are new, or to be altered, in RDFLib 8. Non-breaking changes could be folded into a 7.x release.

Here are a list of the potential items, as discussed on the RDFLib mailing list in Oct/Nov 2024, with expected work times and people against them:

Item	Breaking	Dec	Jan	Feb	Mar	Who
1 Python 3.11+ only	yes	x				David
2 Build system improvements	no		x			Ashley, Edmond
3 Updates I/O functions	yes	x	x			Nick, Ashley, Edmond
4 Updates Store interface	yes	x	x			Ashley, David
5 Improved listing of known stores	no	x				Nick, Ashley, Alex
6 SQLite (SQLAlchemy) Store	no			x		Alex
7 Modern K/V Store	no			x		Nick
8 Updated Dataset & Graph API	yes		x			Ashley
9 Updated Turtle Parser/Serliazer	yes/no		x			Edmond
10 Updates to Namespace Manager	yes			x		Edmond
11 Updated DefinedNamespace instnaces	no	x				Nick
11 Improved documentation	no			x		Nick

Other proposed changes which are not critical for an initial 8 release are:

Item	Breaking
x RDF-star handling	no
y improved SPARQL handing	no

[jnicholls82]

Hello,

RDF-star/SPARQL-star support would be really, really appreciated!

Thanks,
James

[namedgraph]

I think that's a stretch. Even the current SPARQL execution is super slow.

[jnicholls82]

I think that's a stretch. Even the current SPARQL execution is super slow.

Is there any realistic timeframe for RDF*?

[nicholascar]

I think we have to prioritise an 8 release in September without RDF 1.2, because we have a long list of small but getting critical fixes. We'll then have to tackle 1.2 as a whole major task. SPARQL performance might also be another major thing, but clearly 1.2 handing should come first!