I was the engineering lead on that product, and built a SW platform from scratch for it (Microsoft provided an SDK to Cortana which they developed in parallel.)
The internal build could actually run Cortana, Alexa and Google Assistant simultaneously and you could e.g. set an alarm with one of them and query it with another, and they could interrupt each other based on priority. Obviously nobody wanted that feature, but it was hella cool that it worked. Oh, and you could make Skype calls from across the room, and the microphone array lived up to Skype's tough certification requirements which took weeks of testing in Microsoft's anechoic chamber for the DSP/algorithm team to fine tune.
I tried to push for open-sourcing the platform but it was tricky because 1) the director of engineering in Harman didn't know what open source meant and for a hardware focused business to understand the value was a hard sell, 2) it used a HW module that came with a SW stack I mostly got rid off but a few parts were remaining that would need to be replaced which would require additional resources, 3) I was burned out at that point and had limited energy left to fight the good fight. Really too bad, it could have been a cool voice agent development platform, and I honestly think it would have sold in large volumes as a developer-friendly device.
Glad you like it, sorry about the remaining Bluetooth bugs nobody got around to fix, since it basically flopped instantly.
Disagree. I think adults with ADHD are well aware that they are not "normal" with respect to many of the symptoms, even if they don't know it's due to ADHD. At least that's my personal experience.
Few examples off the top of my head:
- being unable to pay bills/fines on time even with plenty of money in the bank.
- being physically unable to keep sitting in a chair after a while in meetings.
It's easy to see that most people around you don't exhibit such traits.
This is achieved using Acoustic Echo Canellation (AEC). This essentially subtracts the output of the speaker as well as reverberations from the room from the microphone input. Here's a youtube video explaining the basic principle: https://www.youtube.com/watch?v=bJKGrheOoY4
Eww, I wasn't even aware of that. I thought the filename UI issue was actually kind of a subtle fail and I was proud to find it, but that one is terrible.
Good memories from a more innocent time. The ICQ client truncated long filenames in the UI, so you could send "image.jpg (50 spaces) .exe" which would open an embedded picture and install a back-door while just seeming like regular picture.
I do miss casually texting with people on the computer rather than the phone, and I don't think it's only due to nostalgia or having more leisure time back then:
- If someone was online, it would typically be a good time for a casual, interactive chat. Texting someone on the phone is (at least for me) rarely "live", because it usually happens at an inconvenient time for one of the parties.
- Much faster to type, and easier to copy-paste stuff from other places. Can communicate almost as effortlessly as a spoken conversation.
- Easier to multi-task in case of a slow reply.
I don't enjoy texting on the phone. Millennial logging out for the last time. AFK BRB.
I think you are right regarding limiting English teaching to prevent the population from consuming "enemy" media. Definitely also happening in China where they massively cracked down on English language tutoring two years ago. Probably due to the governments ongoing effort to ramp up nationalism and spread increasingly improbable narratives about current events and history.
I was the engineering lead on that product, and built a SW platform from scratch for it (Microsoft provided an SDK to Cortana which they developed in parallel.)
The internal build could actually run Cortana, Alexa and Google Assistant simultaneously and you could e.g. set an alarm with one of them and query it with another, and they could interrupt each other based on priority. Obviously nobody wanted that feature, but it was hella cool that it worked. Oh, and you could make Skype calls from across the room, and the microphone array lived up to Skype's tough certification requirements which took weeks of testing in Microsoft's anechoic chamber for the DSP/algorithm team to fine tune.
I tried to push for open-sourcing the platform but it was tricky because 1) the director of engineering in Harman didn't know what open source meant and for a hardware focused business to understand the value was a hard sell, 2) it used a HW module that came with a SW stack I mostly got rid off but a few parts were remaining that would need to be replaced which would require additional resources, 3) I was burned out at that point and had limited energy left to fight the good fight. Really too bad, it could have been a cool voice agent development platform, and I honestly think it would have sold in large volumes as a developer-friendly device.
Glad you like it, sorry about the remaining Bluetooth bugs nobody got around to fix, since it basically flopped instantly.