james anderson

don't try. just code.
...
technician @datagraph; lomoramist; inveterate new yorker in absentia;

[Datagraph GmbH / Dydra.com @dydradata]

To conclude the series on using #NDB interpreted code (IC) programs with #RonDB I decided not to write more blog posts but rather two formal research papers, which I presented at The 19th European #Lisp Symposium (#ELS'26), Krakow, 11-12 May 2026. #Dydra

https://t.co/fYJ0bCfkHj

[Datagraph GmbH / Dydra.com @dydradata]

Read our new article:
Scaling Dydra with RonDB: Toward a Trillion-Triple Store

September 22, 2025, Datagraph GmbH, Berlin, and Hopsworks AB, Stockholm.

https://t.co/j62zeL3X79

[Datagraph GmbH / Dydra.com @dydradata]

That’s exactly the kind of challenge we’ve been tackling with #Dydra, a revisioned #RDF graph store, and #RonDB, a high-performance, clustered database built for scale.

https://t.co/j62zeL3X79

[Datagraph GmbH / Dydra.com @dydradata]

What if your #database could remember everything—every edit, every state, every version—across time, while scaling effortlessly beyond the limits of a single machine?

https://t.co/j62zeL3phB

[Datagraph GmbH / Dydra.com @dydradata]

After more than a year of work, we just released a much enhanced version of our #CommonLisp bindings to the C++ #NDB API of #RonDB.

So, welcome cl-ndbapi for RonDB 24.10!

https://t.co/uw3K2mXV5G

We are currently working on a series of articles, how we use it to scale #Dydra.

[james anderson @lomoramic]

... that's nice. even nicer is that they are so polite, as to "apologize for any inconvenience, and ... appreciate [our] understanding and patience".
compensation for our engineering hours in the middle of the night, on the other hand, is unheard of.

[james anderson @lomoramic]

i just read the https://t.co/WIClc7EubV terms and conditions. section 14 implies that the can suspend services whenever it suits them. in fact, on 20.2 we can expect them to interrupt network access to all of our services at random intervals over four hours. that's nice...

[james anderson @lomoramic]

@namedgraph ... if you do not have a revisioned store available, put the two states in distinct datasets and federate to the from sparlq to compute the diff.

[james anderson @lomoramic]

@namedgraph depending on the nature of the changes, a https://t.co/yvW176FVCR repository can offer that. if the changes are just additions, the repository history has links to diffs. if there are also deletions, sparql can be used to diff two revisions.

[james anderson @lomoramic]

hanging around on the acropolis

[james anderson @lomoramic]

@wohnjalker watch for their copyright 😛

[james anderson @lomoramic]

Barcelonas Wahrzeichen: Zwei neue Türme der Sagrada Família eingeweiht https://t.co/TyKJlO0Apn via @derspiegel

[james anderson @lomoramic]

@linkedktk @LearningSPARQL does xsd:decimal casting not yield the intended result?

[Adrian Gschwend @linkedktk]

So much yes

[james anderson @lomoramic]

@rubensworks @LearningSPARQL at less than 20€ a day? how long does a test protocol take?

[james anderson @lomoramic]

@rubensworks @LearningSPARQL if your "own hardware" include aws infrastructure, it's just a matter of whose clock is ticking.

[james anderson @lomoramic]

@rubensworks @LearningSPARQL no, the only thing it makes difficult is to account for the differences in terms of implementation contingencies. aside from which, the existing publications would even be sufficient to carry through abstract algorithmic comparisons.

[james anderson @lomoramic]

@rubensworks @LearningSPARQL ... with respect to availability, https://t.co/6IlaCFVPJU had our offer to install a service and similar offers were made elsewhere, but somehow not having the source has been an insurmountable obstacle.
if you want to carry the infrastructure costs for amazon, it can happen.

[james anderson @lomoramic]

@rubensworks @LearningSPARQL on the contrary, while it may make it hard to review the source code, in order to benchmark, you need just create and account, load your data and run your queries.

[james anderson @lomoramic]

@rubensworks @LearningSPARQL ... you mention the 2021 paper. yes, i am aware of its content and attempted to improve its details when it was published. curious is, does its account - even at the abstract level, not contradict the claim in this latest demo document?