Fork me on GitHub
#datalevin
<
2022-01-03
>
Huahai02:01:21

There is a LRU cache for any storage access for the datalog engine

Huahai03:01:12

I can add an option to allow user to tune the size of it

dgr16:01:25

What’s the current size? Is it limited?

Huahai01:01:22

it is very limited for now, 1000

Huahai01:01:14

1000 storage calls, but one storage call may return lots of data, or little bit of data

Huahai01:01:08

in the future, we can also specify the size by bytes, but it will be more complicated to implement

zeitstein08:01:38

Will you be incorporating https://github.com/tonsky/datascript/releases/tag/1.3.0 improvements to pull?

Huahai15:01:20

Eventually. We have not even finished feature parity with Datascript 0.8.3 yet. This particular improvement should be easy to incorporate, maybe simply copy the code over? PR welcome.

Huahai15:01:59

Most of these Datascript improvements are easy to incorporate, for they sit on top of IDB abstraction. Most of Datalevin’s work is below that and doesn't change too much things above it. Even after our query engine rewrite, the pull part would remain the same as Datascript

zeitstein22:01:57

Thanks, that is good to know. I'm not familiar with the code base, so I'll have to find some time to take a closer look.

denik19:01:50

these types improvements also make a great case for a clojurists together application https://www.clojuriststogether.org/open-source/