This is a very strange and unusual September. As you know, most of our team is based in Belarus and since August the law has not been working in our country. It’s now like a dashed line — for some cases it works and for some — it doesn’t. Hundreds of people are arrested on the streets each day, even random ones. And you never know if your day will end up at home or in jail. So it’s a really hard time for creative people to keep generating new valuable content. Now it is more about concentration on existing processes like support, finalizing some long-running tasks, and creating the protection against hundreds of negative events that may occur in the country in the near future.
By the way, so far only one of our developers got to jail for 15 days and is now back to work. So the hit rate is just 10% — not as bad as it could be.
Starting from August we are constantly keeping a backup team outside of the country. Just to eliminate possible risks and continue providing a high level of support in any circumstances.
In September our uptime for the EU region was 99.9999% with a small 7-second downtime due to a network hiccup. And our RU region kept running with 100% uptime. You know — fewer changes — fewer upgrades — fewer possibilities for anything to go wrong.
Still, we need changes to go forward. That’s why we introduced two important API changes in September. One is related to /gw/protocols/ REST API call which will be renamed to /gw/channel-protocols/ API call on November 23. Another change is related to the handling of tokens ACL entries which will occur on October 8. Please double-check both links above to ensure your current operations and tokens will not be affected.
There are a few minor changes to our frontend tools. Toolbox received a map feature to visualize location from coordinates in the message; and in the date/time picker control it is now possible to enter values with timestamps.
We migrated almost all our protocols to the new version of the PVM engine. Only three protocols are left: wiatag with special hacks for outgoing command processing, omnicomm with extensive byte staffing, and the largest and our most popular protocol — teltonika. By the way, we have extremely good coordination with Teltonika engineers — they are using comprehensive flespi debug tools for some internal tasks and even sometimes test new firmware versions with our backend. So it’s not uncommon that the integration level of Teltonika devices into flespi is far ahead of other platforms with new and experimental IO ID and features supported. Once teltonika protocol is upgraded to the new PVM engine we will be able to introduce two important features: SSL encryption for the TCP connection and remote firmware upgrade feature directly from flespi SetBox.
In our closest plans are the introduction of custom quickly engineered stream protocols, so that flespi can be used not only as a universal input engine but also as a universal output engine. And we are very close to completely replacing the GUI for the REST API documentation with our own version. It is a long project that took more than a year and we are almost ready to release it.