

Sorry I meant android/iphone apps and browser extensions. It would seem to me that’s the easier part with all the work done on making the client.


Sorry I meant android/iphone apps and browser extensions. It would seem to me that’s the easier part with all the work done on making the client.


would it be also plausible for say vaultwarden to make it’s own client and just completely fork over if bitwarden becomes less open?


I can give you they had different ways to get to that position. As much as I hate IE, I do have to admit it was ahead of netscape for quite a significant time. But yes google used less monopolistic practices to get in there, beyond like spamming you whenever you went to google. I will admit even now edge does worse in the monopolistic practices "I see you went all out of your way to download another browser, are you sure you really want to switch to it, have you at least given edge a fair shot? Please try it out for a bit longer. (and of cousre it’s worth noting now edge is basically a skin of chrome),
But how they got there wasn’t what I was talking about anyway, The point is web pages now cater to chrome, as that’s what makes up over 60% of the total usage, with about 20% being safari (of which you can pretty much assume almost all of that is mobile), and almost everything at the top is running chromes engine.
So in short, if you are designing a page.
Does it work on blink engine, that covers 76% of users, then does the mobile site work on safari, that covers another 20%,
Point is a monopoly is a monopoly, even IF the reason they are there is purely good. The point of the article is just noting that to not use chrome’s engine, browsers have to take the time to make things work, because the websites themselves have little incentive to do so. for such a small percent of their userbase.


So TL:DR, chrome is like internet explorer was before firefox. It does some things outside the standard, and because it’s the modern day “default”. sites sloppily code to work with it, and other browsers are left carrying the bag because if tiktok doesn’t work on firefox, people will view that as a firefox problem. Even if firefox is the one actually following the standards when tiktok and chrome aren’t.


It’s been a long time, but… wordpress is all sql isn’t it? Shouldn’t it just be, make a backup of the mysql database on the old server, install mysql and restore the database to the new server, and then point the word press servers configuration to the new database?


I do aknowledge that’s always going to be the problem when we have the human + AI driver combinations.
Safest hypothetical is 100% AIs that always follow the same rules… next safest is humans that break the rules, but in a context aware situation (IE everyone going 70 in a 55, is safer than 1 car going 55 and all other cars going 70).
Real danger though is if the AI doesn’t make good judgement calls when doing so. IE rather than deciding based on how fast other cars are going, it’s primary determination is whether the user says they are in a hurry, leading it to sometimes be the one car going 55, but if the person is in a hurry it may be the only car going 70 on a road everyone else is going 55.


Can we like, maybe normalize adding descriptions of the product to these announcements? On the blogs of these apps pages that’s one thing, but on selfhosted forums etc… where we are drawing attention to the product of which many aren’t familiar with… a 2 sentence summary of what the product is before going into the long changelog, would make everyone so happy.


half sarcastic but the overall premise of rigging something in to a local voice assistant, when an arguement starts “Ok nabu record this conversation”. then 2 weeks later on another arguement… “OK nabu search our last arguement for the cabinet”. Would be like having a court transcriber on call.
Honestly my thoughts on how I’d actually use AI… versus OpenAI’s are exactly the opposite. Honestly I could see a world where a lot of people get an on site LLM server. Plug it in etc… I could actually see that as useful, IE keep it air gapped, and then you can train it off of your e-mails, your house etc…, not trust everything to some outside company. From what I’m gathering from the chinese open source models that sounds very viable.
Of course the ultimate annoyance is, the datacenter surge is actually taking compute out of regular price ranges.
Actually annoyingly makes me think of the point in time that electricity itself was at this crossroads, where eddison wanted every house to have a DC generator, while Tesla was pushing for AC transmission to send current from centralized locations