Haystaq API responses and questions

[jrod01]

Feedback from @hoolamillie:

We are not recording survey results on our end yet (and may need a format change on your end there), so we will have to re-pull everything we have pulled, but what I'm pushing back accurately denotes if we matched it to our DB or not. Over time we might be able to get more matches...

I do have an additional question:

Up until today, every time I would restart the program it would start pulling from the beginning, which is what I expected from the API. Today, it seems to be picking up from where I left off (which makes debugging harder). Did that behavior just change?

I tried using the offset API to force it to let me re pull previous records, but again I'm using the .pdf from a couple of weeks ago and it doesn't seem to work.

[krainboltgreene]

Couple of things:

1. We assumed that you were storing the matches locally and would only hit the index for more matches. In order to make the process easier we only show new matches. This is why the offset doesn't force it to start at whatever the id you use. It will show, starting at id x, the next 10 matches that are new.
2. You should no longer be using the PDF. The api documentation now lives on this repositories wiki: https://github.com/Makindo/api/wiki

Here's what we can do:

1. For now just hit us up and we'll rebuild the matches list, from scratch, so you have a chance to play with the API.
2. We can start listing non-new items in the list, but that requires you paginate through the list every time. From what I understood your client only wants to consume new matches, is that correct?
3. We're going to up the count of matches per response, since we've improved the performance of the API. This will let you avoid too much pagination.
4. What kind of survey format changes do you have in mind?