hugh_kagi

5 ms·

1.
by hugh_kagi 2y ago
Question is the same as it ever was, can you locate your data and your compute close to your users. Issue with search is there's a lot of data, and a lot of compute to run over it, so you end up replicating PBs just to reduce latency.
2.
by hugh_kagi 2y ago
It took a few weeks for our developers to learn basic sewing techniques, but once we did that we were able to move onto automating the looms fairly easily. Looms were the original computers so it was easier than you'd think. https:&#x
3.
by hugh_kagi 2y ago
We're working on something that should separate the concept of 'have you paid us' from 'who are you', at which point you'd be free to use (the existing) bitcoin payment method and have a pretty private experien
4.
by hugh_kagi 2y ago
We're definitely keeping the extensions going, just need to prioritise a few parts internally. You can just git clone the new chrome extension and 'load unpacked' if that's of any use. It's extremely simple. https:
5.
by hugh_kagi 2y ago
We face a number of challenges simply letting our paying customers change their search engine: 1. On iOS the list of allowed search engines is simply baked into OS, we have a fiddly extension that hooks outbound calls to /search and re
6.
by hugh_kagi 2y ago
Not yet but it's something we want to look into.
7.
by hugh_kagi 2y ago
> Are they finally applying filters and operands to image searches? That was a bug, apologies. It should be fixed now.
8.
by hugh_kagi 2y ago
Hello, there's a Chrome policy around extensions that change the default search engine, which the Kagi extension does. Our additional features, e.g. the Summarizer seem to require their own separate extension for this. https:/&#x