If you switch IPs you can usually get a message or two through during these outages. This has been a consistent issue with Gemini for the past few months, but has increased in frequency and duration of outages, especially since 3.5.
Early on this outage it was just the 3.5 pro models affected, then flash started having issues, and now flash-lite is not being responsive.
Slightly related, but has anyone else using the gemini cli noticed that over the past few months, usage of the Gemini 3 pro models frequently times out or hangs? Antigravity cli has been better about this but the quotas are just so low with the 5-hour window bs.
On the other hand, seperating concerns by process boundaries leads to more secure, composable and stable code. By not reinventing the wheel, you avoid a whole class of problems. Of course a stable API or library might be better, but convenience always wins out.
The difference is apple doesn’t let you modify your device to use other services. Their contractual obligation goes beyond the service itself. That’s why EPIC won this case.
The reason I brought up roomba wasn't to talk about Lidar or vision necessarily. It's more a story about how the first-mover in a technological space became entrenched in what works and became resistant to investing in newer technologies. The result was rival companies taking away marketshare from a market roomba once defined. Roomba has since incorporated lidar and other innovations after being stagnant for a decade, but its too late - their competitors now dominate the market.
To complete the analogy, Tesla is invested in vision-only technologies, while its competitors are making gains with Lidar and other tech that Tesla refuses to acknowledge. It's very reminscent of Roomba in the mid 2010s.
A fun anecdote - a lot of people may remember Roomba from forever back with their automated little vacuums. Roomba's market share declined significantly because they failed to adopt Lidar technology as quickly as their competitors, instead they depended on the bumper for as long as possible. This put them at a disadvantage in navigation and efficiency as their competitors started using Lidar. Combined with aggressive pricing from rivals, expiry of its patented roller in 2022, a weird insistence to not combine vacuum and mopping into one device, Roomba (or iRobot now) is a just little fish in the sea it made.
Hey, just wanted to say that this looks like a really cool and useful project. I work with a few medical databases and sometimes I just need a very quick breakdown of specific data and while I usually just write a short script, the utility and portability of this code looks fantastic to me. Which brings me to a question, how well does this program handle moderately large databases (~100GB-1TB) in your (or anyone else's) experience? In other words does it try to load everything into memory, or does it query as needed when given a database?