@damian, I think you doing it right, but you might replace

dao.dependsOnTable.select {...}

with

dao.searchQuery{...}

. And ofc making query with order by and limit is much faster and effective than loading everything into memory, but don’t forget to add index to updated_at if you have large table.

thanks. i've added the incides. i've never used

searchQuery

before, could you give me a pointer on how i could solve this with

searchQuery

?

searchQuery

do the same: returns Query with slice on dependsOnTable. This function used when you use find/getById functions. in special cases someone could override that function to add additional conditions to every select, so using searchQuery directly instead of using depensOnTable is prefered.

so it'd be

dao.wrapRows(dao.searchQuery(op.build { row = myCondition }).orderBy()...

? i don't really know what the

wrapRows

does, i think i haven't found a good documentation of this yet. i saw it in a project on github and its working, but no idea why.

wrapRows create lazy iterable of entities (dao in your case) over resultset from database.

do you think it should be on docs?

I think information about overriding dependsOnTable/column and searchQuery should be present in docs, but it's "advanced technics", which must be described with good real-world examples. But i think, that having in docs the way how you can create entities from Query is very plesant