Taras Mankovski
Taras Mankovski
Awesome work 👏
A combination of these things, if you're open to a PR, I can get one of our devs to create it.
@benjdlambert, @dagda1 submitted a PR with our first pass at this functionality. Please, let us know what we missed. Once the implementation is mergable, we'll update the docs to explain...
@vinzscam that will load all of the records into memory which is undesirable.
Yeah, I don't really need what #12246 brings. It'd be much simpler to do what @freben suggested, just have an endpoint that takes a list of refs and returns those...
Using entities endpoint for this use case is undesirable for several reasons, 1. dataloader requires passing many refs. The number of refs many exceed URL length limit, for this reason...
> if we have the query string url length problem in the new hypothetical endpoint, it means we should revisit this solution also for the other endpoints The existing endpoints...
@freben that makes sense. We're already using `getEntityByName`, but we're making a call for each ref. We can switch to using `entities` for now. In the mean time, we'll add...
@freben @vinzscam I was reading that code example on my phone. I misread it. I thought it was querying all of the records and then filtering them in memory. I've...
@freben @vinzscam, @cowboyd reminded me that we did try using `catalogApi.getEntities` for DataLoader. It was slower than making an HTTP request for each entity using `getEntityByName`. This is the reason...