Since we launched our multi-check API yesterday we've been hard at work improving performance and squashing bugs. Today we'd like to share with you some progress.

Firstly there were some bugs with the IPv6 VPN detection with regards to Google address spaces. This has been corrected in both the /v1b/ endpoint and in the back-ported code which is running on our main /v1/ endpoint.

The second IPv6 bug we had was with VPN detection. If you had not set the VPN flag to on but checked an IPv6 address it would be checked against our VPN data and a positive VPN result presented. This has also been corrected today in our /v1b/ and /v1/ endpoints.

The third bug we dealt with today was with dashboard statistics. Under certain circumstances you may have had a discrepancy in your total API queries reported at the top of your dashboard compared with the graphed breakdown of your query statistics. This was caused by underreporting on some negative detection scenarios. This bug only affected our /v1b/ endpoint.

Apart from fixing these bugs we've also corrected some very specific edge-case bugs. For example when performing a single IP check and the address entered into whitelisted IP ranges you may have received a response without the IP Address being repeated back to you in the JSON response. This behaviour has been corrected.

New functionality wise we've improved how we handle invalid IP Addresses. So previously you would simply receive a vague message indicating that one or more addresses were invalid but it didn't list the actual addresses you supplied. That's not an issue when you're checking a single address but with multi-checking you need to know which addresses in the data you've sent to be checked were invalid. To that end we now display that information back.

We're also planning to improve the 100 check per query limiter code. At present it simply stops processing your addresses once it reaches 100 and does not output to you a list of unchecked addresses. We'll be changing this behaviour soon to indicate which addresses were unprocessed due to hitting the limit. We'll also be changing statistic behaviour to account for this, at present if you send us 500 IP's you'll have 500 queries registered to your API Key even though we only processed the first 100.

So that's a quick update of where we're at. The new API is coming along steadily, we've already improved the performance since yesterday and we're squashing all the bugs we find as quickly as possible. We're on track for an early January 2018 rollout to our main API endpoint address.

Thanks for reading and have a great day.