the fun is disappearing from programming

(see http://infowatch.livejournal.com/86171.html (in Russian))
Basically, the article says that the *implementation* of software products is being shaped as the conveyor belt. But the game dev (except for the sequels and sport games), and startups are mostly out of the question.
Or, we may say that there is finally a distinction between the research & production.
Same thing with the artists and designers (make it “feel cool” is more art/research, but “respect the accessibility” and “paint 50 icons for different actions” is clearly a production)

That’s why those who bored should innovate and materialize innovations in the form of startups. At least, my (little) experience shows, that it’s much easier and productive to find answers to unsolved (technically) problems than to follow existing guides and instructions.

UI done right

i’ve always dreamed of bringing Linux-style switchable desktops to Windows, –even Microsoft has one, but they’re all slow and buggy.

and you know what? seems like they fixed this problem in Win7. Really. I use Win7 more than six months already, and the more i get used to the new taskbar, the more i like it. cool.

it’s hard to describe why it’s so convenient. just it’s hard to describe why the Google Desktop is convenient, if you don’t know that you can press Ctrl-Ctrl, type 3-4 letters from the document name and just open it, don’t involving Recent Files, Explorer and other pile of junk (actually, i’ve already ceased using it because i don’t like Google messing with my files, and the Windows Vista and 7 start menu can do the same, but you get the idea – there are things which can’t be explained, they only can be tried)

aesthetic needs

wondering why i’m always so, um, worried in the absense of designers/artists for the thing i’m developing? isn’t that the thing that attracts so many geeks to command line interface and therefore linux/unix?

i.e. to get the feeling of completeness and satisfaction i require from the thing not to *just work*, but to be nice and convenient. that’s why lots of things are never shown to the public, if only as the screenshots. and that’s why i *don’t like* the user interface of most programs – either graphical or command-line.

internet map

I think that different visualizations of the internet aren’t so successful, because internet does not have a notion for ‘location’, and a notion of ‘time’ is somewhat different from the real world one, too.

So, firstly, human instincts about the land mapping are not applicable to cyberspace (it’s difficult even to define the concept of ‘distance’), and automatic visualizations aren’t based on any real (perceptible) facts, so they don’t bring in any useful information and are considered of little use.

If we could do search using successive approximations approach, then an average web user could hold a significant part of the Net in his brain in the form of associations “here i need to turn this way”. But now we’re tied to search engines 😦

In this sense the ancient Gopher and current FTP protocol are better than HTTP, because they provide at least tree-like object arrangement, but unfortunately internet is far from being tree-like…

Hyperlink graphs seem to describe the Net structure pretty precisely, but they aren’t reversible and opaque. Also, they are vulnerable to spammers, outdated and poorly reflect the ‘non-page-like-objects’ (user profiles, emails, instant messages, etc)

But there should be some kind of a solution! 🙂