News for hackers. Stuff that matters.

blagund

What does the EU lack? Is it the big corp infra? Or something more fundamental?
IANAL but at least in some EU countries you can't give away all rights preemptively for usecases not yet known at the time. So a blanket giveaway doesn't necessarily include AI training (it is a different question if the people performing that act actually care).
Watch the movie Caveman to find out! Ool anyone?
Re:Learn Datalog Today on Monday January 22, 2024 @01:53AM (#39087035)
Souffle deserves a high ranking. It really is a nice datalog. If only field reference was possible, instead of excessive tuple pattern matching. (Though, one could write a preprocessor for it, it is just syntax).

Differential datalog is also nice, if you have the usecase.

Re:Learn Datalog Today on Monday January 22, 2024 @01:51AM (#39087023)
Note: Rego targets near linear execution speed, so it can't do recursion generally, except limited tree-like cases for sat y acl checks..

That's as intended, just saying don't expect it has full datalog power.

It deals with Json data nicely though (which my favorite souffle does not).

Re:Five Design Principles for Building a Data Pipeline Engine on Tuesday October 17, 2023 @06:32AM (#37912928)
Not op, but maybe the perception is based on relatively few aha moments or new insights in the article. The one thing I would come back to it for is the listing of pipeline languages, which sounds interesting. YMMV.
I, for one, find myself from time to time wondering, what are these subpar search results I'm getting? And then realize I haven't yet set Kagi as default search engine on this device yet.
My hunch is, in any sufficiently large rule set, there will be inconsistencies. Handwaily think Gödel, or just the need for bounded domains in DDD.

Humans (or, well, AI) is needed to cope with inconsistencies.

That said, pointing out the fact of existence of inconsistencies could be very valuable. But a system needs to embrace them, not fight them.

Re:Leaving Haskell Behind on Thursday August 24, 2023 @03:25PM (#37253443)
My notes on using Haskell: mostly agree with the article, but I came to embrace rolling my own tooling, some abstraction on top of cabal. I use my abstractions, not what whoever would force on me. Use Haskell as your typed lisp.. with all the pros and cons.

On production use... Don't even get me started. Will work, but need shedding blood.

I recall someone pointing out, AGPL only affects offering the same core functionality over network. So you cant monetize exposing as is. But you can use as a backend component. Lot of people miss that. I did too. IANAL.