Bad idea of the day: an online game where, for one player, it's a tetris-style falling-block game, but for the other player, those blocks form the map for a side-scrolling platformer.

Hot take: gopher is not a failed attempt to invent the WWW, but its own crystallization of a coherent philosophy -- to present jump links between text files, without all the other bollocks. That makes it as valuable today as it was 25 years ago.

@endomain @ekaitz_zarraga @CharredStencil @enkiv2
Transmitting the same bytes over gopher might be slower, but I don't think that's really what we're talking about. (If it is, we shouldn't be.) Webtech as a whole presents a lot of opportunities to transmit more bytes, and a lot of incentives to do so, & web devs usually take every such opportunity. Optimizations like QUIC make this marginally sustainable for top 10% users. Gopher remains as usable as it was in 1992.

@CharredStencil @ekaitz_zarraga @enkiv2
Joel is *very* wrong about rewrites. His logic only applies if you're locked into a design so bad that corner cases are unavoidable (as you are with web-tech). Gopher clients & gopher servers are easy to write because gopher has a simple & consistent design.

Pondering what's inbetween Gopher and the web gopher://zaibatsu.circumlunar.space:70/0/~solderpunk/phlog/pondering-whats-inbetween-gopher-and-the-web.txt

At the end of the day, think outside the box about clichés | Aeon Ideas https://aeon.co/ideas/at-the-end-of-the-day-think-outside-the-box-about-cliches

Wow, fancy that. Web ad giant Google to block ad-blockers in Chrome. For safety, apparently • The Register https://www.theregister.co.uk/2019/01/22/google_chrome_browser_ad_content_block_change/

@ekaitz_zarraga @grainloom @enkiv2 @jamey
Right. If I can't get away with eliminating a dependency & therefore breaking changes are unavoidable, I will usually write a polyfill to make both versions operate the same way (or close enough that my code doesn't care). Most devs don't do that, but I also don't run software written by most devs so it doesn't matter.

@grainloom @jamey
Simultaneously the best and worst thing about Lua is that it doesn't have a rich *anything* library in the standard. (So it fits on a floppy & will run on MS-DOS or RiscOS unmodified.) Python is exactly the opposite: ships with no fewer than 3 distinct irc libraries.

Procrastination is so weird. Now that I promised myself I'd finish the game this month, I've spent all my free time on a mastodon client & on porting my in-progress book's typesetting toolchain from html to latex.

Band name of the day: a politics for alienation

The great Lisp and Smalltalk systems are designed from the ground up with the as... | Hacker News https://news.ycombinator.com/item?id=16586347

Algorithms by Jeff Erickson http://jeffe.cs.illinois.edu/teaching/algorithms/

With the success of Toblerone, Toblertwo is inevitable -- which should keep the Tobler elves extremely busy

TempleOS | Down the Rabbit Hole - YouTube https://www.youtube.com/watch?v=UCgoxQCf5Jg

Band name of the day: rat nausea

@ekaitz_zarraga @Shamar @Wolf480pl @jcbrand

The point of 'big and small computing' is that too often programmers make the computing equivalent of assuming that spaghetti carbonara must taste awful because it can't be canned.

Being aware of the difference in needs between industrial-scale and human-scale processes is a prerequisite for avoiding this kind of confusion.

@ekaitz_zarraga @Shamar @Wolf480pl @jcbrand

To bring back the 'big vs small cooking' metaphor:

We will never see canned spaghetti carbonara that tastes like home-made spaghetti carbonara. The sauce is literally soft-boiled egg.

This makes spaghetti carbonara absolutely useless to the forces of Big Cooking. Spaghetti carbonara cannot be google-scaled.

However, as a home cook, this makes it even more valuable to know how to make it: it's something you can't get cheaply from another source.

@technomancy
Here's a mirror: http://www.lord-enki.net/medium-backup/2018-12-21_Free-software-and-the-revolt-against-transactionality-3a44a1b7f96d.html