It’s just over 2 months ago that I reported 10.000 active devices with CanZE installed (Android stats only). We went over 11.000 on December 9th . 10% in just under 2 months, it doesn’t seem much, but a compound growth of 77% annually is nothing to sneeze at. I doubt we will reach that BTW; compared to December 2020 the real compound growth was 42%. Still great I would say. And that’s up from 29% the 12 months before that.
A big thank you to all users, new and existing, of CanZE!
Last weekend we handily crossed the 10.000 “active installed devices” benchmark (on Android). It’s just an arbitrary number, but hey ho! Thank you all for using CanZE!
Growth has slowed down a bit but that doesn’t mean we’re not baffled to find our app on so many devices! This week we broke through CanZE being installed on 9000 devices. Google’s definition is: “The number of active devices that the app is installed on. An active device is one that has been turned on at least once in the previous 30 days”.
Quick update. Last week saw a huge 3.4% uptake of users, or more precisely, “devices that have been turned on in the last 30 days and have CanZE installed”. This is twice the usual uptake in a month! As a result, we’ve handily past 8.000 users. I have not taken the effort to investigate if there is a specific reason such as a review in a popular channel, but in any case: Thank you!
Allow me to also do a mild vent. Reviews on the Google Play are very often super positive which we of course appreciate a lot, but now and then somebody slips in a 1 point rating with a “It doesn’t work with my car” review. No details, and never a response to a reply (which we always do). I checked, and Google Play does a notification on the device if a reply by the developer is issued. It’s a very frustrating endeavor. So, if you ever run into an issue, please report it on github: we do take pride in ironing out kinks! Feel free though to direct praise to Google Play 😉
Speaking of kinks: almost a full week with zero crashes now. Fingers crossed.
Thank you “Silent Energy”
While we’re certainly not done untangling the ZOE ZE50 and new ZE40 (together called ZOE Ph2), we’re pretty confident we’ve cracked the basics fairly well.
ZOE Ph2 is a significantly different car than the old models and required additional code to access the car’s computers. We had to make changes in the CanSee dongle and the ELM327 (KONNWEI) dongle driver has much improved, also performance wise. And once again, lots and lots and lots of testing in cold winter nights on driveways, city streets and even road trips. While all this was going we’ve also been doing some internal refactoring to facilitate the different models in a less cluttered way, together with an improved process to do our research, read, finding the fields we need to make CanZE do it’s thing. All this despite the modest outside looking the same, so maybe we’re more impressed than you are!
Having said all that, our short term effort is clawing back on decreased stability that was introduced by all the code changes made in the last few months. Crashlytics which anonymously provides post-mortem information is a huge help and as I write this, all known crashes are tackled either in the current production release or ready for the next release, so please do update if you haven’t already. You, the thousands of end users are the heroes doing the cold harsh field testing. Thank you for that.
So what’s the speculative part? Well, with the Twingo EV now on the market we’ve had a bit of time to take a look and we can possibly make it work. This is not a promise: history taught us we couldn’t make the ZE50 work until enthusiastic and knowledgeable team members actually owning these cars were co-developing. But with above changes and already one enthusiastic tester in the field, we might just crack it. Still, it’s a tough cookie with slower research cycles. But who knows? If it works we could maybe add say a Twizy the same way? Even more speculation!
Anyone who has played with the Research facility, or any log file really, knows the results can be massive and hard to process.
Massimo Ceraolo made a small post-processor that splits up the data per field, timestamped from start-of log. The output files (one per field) are pretty easy to convert into plots with Excel or other programs.
His software can be found here, the PDF has many more details.
Thank you all for installing and running no less than 5000 CanZE instances on Android! A humbling number for a total amateur project.
