mjw1007

67 ms·

721.
by mjw1007 8y ago
Forth was for 16-bit computers, and liked to do everything in 16-bit cells. (That's 16-bit in the sense of "PDP-11" or "IBM 1130", not in the sense of "Amiga or Atari ST", which very much didn't have
722.
by mjw1007 8y ago
I think email has come round to the point where you can, if you really want, run your own SMTP server and send out thousands of emails without having to deal with a mountain of headaches again, yes. Tens or hundreds of thousands, maybe not.
723.
by mjw1007 8y ago
Hm, that relationship is also what you'd get if people were perfect at estimating their own skill while the proxy being used by the researchers was imperfect, isn't it?
724.
by mjw1007 8y ago
I don't think there's a good way to measure Debian's contributions outside strictly packaging its distribution. If a Debian developer, or a group of Debian developers, sets out to solve some problem that isn't Debian-spe
725.
by mjw1007 8y ago
I expect in most cases the same people will stand and be re-elected. One advantage of doing things this way is that it makes it easy for people to step down when they've had enough.
726.
by mjw1007 8y ago
I suppose letting the compiler-writers use new language features gives the usual advantages of "dogfooding" your own product.
727.
by mjw1007 8y ago
That's great. And all of "The Gostak". I think I like these so much because they make me feel like I'm being clever.
728.
by mjw1007 8y ago
Associated blog post here: https://blog.rust-lang.org/2018/11/29/a-new-look-for-rust-la...
729.
by mjw1007 8y ago
A personal opinion: The old site said to me "Here's a programming language made by programmers for programmers.". The new site says to me "Here's a programming language which has a marketing team.". The "F
730.
by mjw1007 8y ago
Principally: start punishing the corrupt vendors; don't assume that reputation mechanisms will ensure that non-corrupt ones will eventually outcompete the corrupt ones. Just as you have to build software for the users you have not the
731.
by mjw1007 8y ago
I don't think it's good enough to say « it's still on the shoulders of the city for choosing the vendors ». If I write a piece of software which is technically capable of meeting its requirements if you read the manual carefu
732.
by mjw1007 8y ago
They said 0.01% of sessions . Feed preview isn't something you expect to use every session.
733.
by mjw1007 8y ago
It's a silly justification anyway, as far as "feed previews" go: you preview a feed when you're thinking of subscribing to it, which isn't expected to be an everyday sort of action. I hope they wouldn't conside
734.
by mjw1007 8y ago
There's a scale from "there's no evidence of abuse and if there had been abuse we're certain we would have seen evidence" to "there's no evidence of abuse but there's no reason to suppose if there had
735.
by mjw1007 8y ago
That internal doc is rather close to the article linked here, though I notice this piece of the rationale didn't make it into the public version: « Additionally, we are working on various initiatives that operate in the same area of fo
736.
by mjw1007 8y ago
I think the feed reader had to register itself, using registerHandler with mimetype application/vnd.mozilla.maybe.feed . See https://developer.mozilla.org/en-US/docs/Mozilla/Firefox/Rel...
737.
by mjw1007 8y ago
They say they're removing three things: - the built-in feed preview feature - the "live bookmarks" support - the subscription UI They give justifications for removing the first two, but not the third. Assuming by "subscr
738.
by mjw1007 8y ago
Maybe there's a problem analogous to "clustering" in opinion polls. I presume every now and again econometrics people have to tweak the way they measure GDP. Maybe if someone proposes a plausible tweak and it results in a sur
739.
by mjw1007 8y ago
I think it depends on whether "took months to years" means there was a team of several programmers working on it full-time for that period, or whether it just takes a couple of people working on and off and most of the time is wai
740.
by mjw1007 8y ago
If you're implementing some kind of style sheet system, you often want to support a tower of overrides (say, theme → document → chapter → explicitly-marked-section).
741.
by mjw1007 8y ago
I mean a bit more concise than that. To catch and rethrow an exception you still need a try and a block, and you probably add a level of indent. Compare Rust, where you can generally spell "I'm sufficiently confident that this won
742.
by mjw1007 8y ago
In general it's hard to distinguish between a good idea executed badly and a bad idea executed badly. I think Java executed badly here. There doesn't seem to have been a clear policy in the standard library for what kind of except
743.
by mjw1007 8y ago
Does anyone know what happened with the lawsuit where Yahoo/Oath were trying to get out of paying? Is it still going? https://wiki.mozilla.org/Mozilla_and_Yahoo_Holdings_Oath_Cou... ends at 2018-01-05. [edit: seems to
744.
by mjw1007 8y ago
I don't think it would have been so bad for them. If I liked Firefox better than Chrome but Firefox didn't let me watch films in my web browser, I'd run Firefox day to day and open Chrome when I wanted to watch a film. In my
745.
by mjw1007 8y ago
The official docs aren't bad. https://docs.python.org/3/extending/index.html
746.
by mjw1007 8y ago
It's a Ten-minute rule bill. Ten-minute rule bills have no hope of passing. You're welcome.
747.
by mjw1007 8y ago
Summarising: they have no guarantee against closure; Google promises to give one year's notice if/when they decide to close them.
748.
by mjw1007 8y ago
I make it 100mm, not 1mm.
749.
by mjw1007 8y ago
Good to see the new trustworthy Microsoft in action. Coming soon: github.com "best viewed in Edge".
750.
by mjw1007 8y ago
Nitpick: the map he shows was drawn by Christopher Tolkien, not Pauline Baynes.

More ›