Forest Gregg

for teaching `register_output_renderer` to stream it seems like the two options are to 1. a [nested query technique ](https://github.com/simonw/datasette/issues/526#issuecomment-505162238)to paginate through 2. a fetching model that looks like something ```python...

if you went this route: ```python with sqlite_timelimit(conn, time_limit_ms): c.execute(query) for chunk in c.fetchmany(chunk_size): yield from chunk ``` then `time_limit_ms` would probably have to be greatly extended, because the time...

ugh! this is probably related to https://github.com/dedupeio/dedupe/issues/1085

@tigerang22. that’s unfortunate! i used @f-hafner’s example code to debug. can you provide a reproducible example?

@f-hafner looks like you installed it okay. it's a bit simpler to do it like this ```bash pip install https://github.com/dedupeio/dedupe/archive/522e7b2147d61fa36d6dee6288df57aee95c4bcc.zip ``` that's very strange that the performance didn't get better...

playing around with this and the virtual compound blocking map table is often larger then the current style because we usually have compound predicates that have very low cardinality our...

if there is atomic predicate reuse across compound predicates that could make things more attractive. the blocking map would just be the atomic predicate keys and then there would be...

Here's a fast Levenstein search. It uses ternary search trees. https://github.com/mattandahalfew/Levenshtein_search

First need to figure out why tests are not passing. Probably should remove cybergrind's code and isolate what i'm working with. It was trying to subclass mocker.MagicMock to EngineMock.

Could you describe a concrete use case a little bit more. We use lots of web services to get data, but I'm not sure if that's what you mean?