I started out with film and took my camera all over everywhere with me for years. I switched to digital as soon as I could, before it was really even practical, fully embraced it, and ran with it for a long time. But I eventually got bored, and stopped carrying a camera at all for while.

Last year, though, I got back into film, and I'm having a ball! The point of the retro process is not that it's better, it's that I'm enjoying the time I spend with it. The constraints are interesting. The technique is challenging. It's not so much about the photos as it is the photography: I enjoy the practice of making images, and dealing with the challenges of vintage equipment is part of the skill I'm practicing.

It doesn't actually matter whether I take any of these photos or not, you know? I'm not a professional; I'm not making unique art, or documenting historic events. I'm doing this because I enjoy watching the light, looking for interesting frames, and trying to capture them. Right now, the most enjoyable equipment for that purpose happens to be an all-mechanical, medium-format film camera.

It seems more likely to me that the Windows API will become the de-facto Linux gaming SDK, and the idea of porting a game to Linux will become meaningless.

A startup I worked for twenty years ago used that approach: we shipped an update once a quarter, every quarter, no matter what. We'd begin with a week of planning, build as much of the plan as we could, then cut anything we hadn't finished and release whatever was left. Of course the trick was to build the high-priority items first.

I loved it and I think it was one of the more productive development methodologies I've ever seen. It made sense, it was honest, it required no heroics, and it improved our long-term design work by forcing us to break every grand plan down into a series of incremental deliverables.

There is so little context here that I cannot tell what this author was trying to measure, what their categories mean, or what conclusions they might draw from this experiment.

Generically I like the idea that you could identify attributes a text has and have a classifier attribute those attributes to documents and make some kind of report on it -- but this is so vague I just don't get it.

It'd be nice if we could take a few years off from having a president at all - just rest, recover, and start to clean up the mess.

America needs a Snow Leopard. No new features, just bug fixes.

That would probably be the best. But I can't see how one would even begin to introduce that in the current top-heavy presidency-driven system.

I used `killall vi` from another terminal for years before I finally memorized the proper way to do it.

How did you save the edits you made in vi?

I didn't try to make any edits in vi; I just ended up there, on occasion, when something would invoke $EDITOR.

That is a remarkably pedantic nitpick. If you're going that far, you should really be complaining that it's not "KiB".

I don't think it's pedantry, it's looking for clues as to the age of this.

nah, it's pedantic

Well, this sucks.

The fact that I can sideload whatever I need and stay out of Google's ecosystem is the whole reason I use Android. Given the miserable choice between two fully locked-down platforms, why would I pick theirs?

That's how the internet worked back when we were all excited about it. Giving things away for free is easy on the web; irritating people badly enough that you can squeeze money out of them is what takes effort.

> Then the older buildings say, "Well, we can raise our rents 20% and still be the best deal in town,"

They can only get away with that when there is a housing shortage.