verisimilitude: The code I was writing was repetitive, even with generics.
asciilifeform: i've yet to witness this marvel with own eyes, but will believe
verisimilitude: What I was doing was trying to take the same approach to a problem I already used with Common Lisp and heavy metaprogramming, but I now feel that was foolhardy.
verisimilitude: I have a quick question concerning your website, asciilifeform: You have a number of different websites in the ``People'' section; what criteria do you use for deciding the contents thereof?
asciilifeform: verisimilitude: 2 types, ones i know in person, and others that write or at one time wrote (updated erry coupla yrs) something interesting.
verisimilitude: Would you be interested in taking a look at my website and seeing if you'd add me to yours and me add yours to mine?
verisimilitude: Anyway, that main project is why I decided to learn Ada; I sought to reimplement a simpler version of it and being able to survive memory exhaustion, use little memory, and other things were a great concern, so Ada seemed a good candidate.
shinohai: Also asciilifeform has honor of being the first rss feed to come across my bot, so now I know it worx lol
shinohai: re: your discussion on portage in #t asciilifeform ... thus far I have had success running own local portage host w/ packages I need, dunno if option for your use case. (as binhost so I dont have to compile same packages 1000 times)
asciilifeform: shinohai: that's half the point of cuntoo -- fixed portage mirror.
asciilifeform: as for packages, the only use case i see for those is bring-up of minimal set needed for building. i do not use canned packages anywhere else and don't intend to start.
shinohai: I started this on own recipe, because like you said emerge sync would break everything with my USE flags if I didn't
asciilifeform: shinohai: the eventual goal is 'cuntoo disk has vtron , compiler, & kernel, errything else comes from mirrors where sourceballs are signed by l1'
verisimilitude: I've started reading your FFA articles. I've yet to do so seriously; what I did was scan through it all to get a feeling for an Ada program organization convention.
verisimilitude: I'm glad your articles are available. I anticipate the next time being a serious read through and complete understanding and then going on up through the following articles in the series.
verisimilitude: I do have a question for you about something. Why do you not use child packages in cases such as Word_Ops, FZ_Type, FZ_Arith and whatnot?
asciilifeform: verisimilitude: what would be gained from lining'em up in child packages ?
verisimilitude: One could argue the hierarchy, where one exists, would be a tad more clear; a child package also doesn't need to with its parent.
verisimilitude: Mostly, I'm just wondering if this was a stylistic choice on your part.
asciilifeform: as i see it, it'd make sense if ffa were sectioned into chunks that are likely to be used selectively. but how often do you write a proggy that, e.g., needs addition but not multiplication ?
asciilifeform: imho ffa is simply too small, conceptually, to be divisible this way. hence flat package.
asciilifeform: and yes deliberate choice, like errythingelse in ffa.
asciilifeform: verisimilitude: feel free to comment on my www also, btw, as you read these.
asciilifeform: ( there's a ~day or so delay in the mod queue, but generally i empty it errynight )
verisimilitude: I was looking at your organization to get a feeling for how I'd organize my own program, but I believe I'll be using child packages for now, given my considerations; I can always change it later, anyway.
asciilifeform: i started out writing a 'conventional' bignumtron, with 'growable' integers. and quickly concluded that the problem is ill-posed, there is no way for an arithmetron with growable (i.e. non-fixed-bitness) ints to run in constant time.
verisimilitude: I've seen an unfinished proposal to add bignums and bounded bignums to Ada 2020.
asciilifeform: verisimilitude: they're actually in there (but not exposed to user)
asciilifeform: take a look at the gnat src some time, it's a horror
asciilifeform: ( contains not only this, but, e.g., an entire SNOBOL , and many other things )
verisimilitude: Do you have any thoughts on that? From what I've heard from others, this organization isn't very privy to outside comments.
verisimilitude: In my earlier years, I was, of course, looking for a language to learn and I'd heard about C and so sought to learn it first.
verisimilitude: However, I learn a language by reading a standard and a relevant book and whatnot.
verisimilitude: So, in contrast to so many C programmers who don't know C, I was learning C to find a hideously complicated language with none of the qualities ascribed to it. Even the syntactical rules were disgusting.
asciilifeform: i know of currently exactly 2 highlevel langs where you can actually write nontrivial proggy using only what's in the paper standard -- cl and ada