-- Leo's gemini proxy
-- Connecting to gemini.tuxmachines.org:1965...
-- Connected
-- Sending request
-- Meta line: 20 text/gemini;lang=en-GB
Tux Machines
Posted by Roy Schestowitz on Aug 26, 2023
A Gemini client* is needed for the following links.
> What if you could relive a favourite childhood memory—but with everything upgraded so it’s as awesome as you remember?
> I’ve just finished playing Black Mesa, and ... well, that’s what it is for me.
> It’s a remake of the original Half Life.
> Dedicated fans—particularly, I think, movie fans—wince at the word “remake”, because remakes are often awful: missing the point of the original, adding nothing.
> Black Mesa on the other hand truly captures the spirit of the original while not being shy about making changes. Weirdly, it ends up closer to how I remember Half Life than the actual game.
> In the summer of 2021 there was a discussion on an email list where someone mentioned the tendency to invent technically advanced solutions to engineering problems instead of simpler, more robust ones. Of course, engineers need to make themselves needed! Someone linked to Cheapskate's guide to computers and the internet, and there I found a mention of this new gemini protocoll. I took a look through a web proxy and became fascinated. There were already many gemini browsers available, but only one that I managed to install. It was amfora, and I've stuck with it since then. Although amfora may lack some features, the experience is so dissimilar to browsing the web that you can't forget for a second that this is not just a stripped down html homepage but something quite different.
> From this point of view, anything that you can't easily program yourself is effectively not a computer. Effectively, it's an appliance (at best).
> But why the pedantry? Why would you program anything for yourself? After all, there's a veritable army of professional programmers out there, clambering over themselves to get you to run software that they've written! And their programs are *polished*; way better than you or I could accomplish on our own.
> Automatic differentiation is a method by Seppo Linnainmaa for quickly computing the partial derivatives of a function defined by a straight line program. This method is very important in machine learning because it makes it easy to implement gradient-based optimization methods (which are, among other things, used to fit neural networks to data; did you know that neural networks are just straight line programs?).
Gemini links can be opened using => https://gemini.circumlunar.space/software/ ↺ Gemini software
. It’s like the World Wide Web but a lot lighter.
-- Response ended
-- Page fetched on Thu Jun 13 11:27:30 2024
