Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
akkartik
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
53 ms
·
811.
▲
by
akkartik
5y ago
I love chatting about Kernel :D Here's my most recent post: https://lobste.rs/s/d0hogq/problem_with_macros#c_nozcrm Thanks for showing me Oh! It really has f-exprs?! I didn't immediately see it in https
812.
▲
by
akkartik
5y ago
Ok that makes sense. Thanks!
813.
▲
by
akkartik
5y ago
I see. Kinda. The 'lambda in the first line isn't quoted. And the eval itself seems identical to JMC. So this is just the substrate level and only when evaluating args? Does that seem right?
814.
▲
by
akkartik
5y ago
By Emacs do you mean Elisp? Your session there is not surprising to me. I don't have emacs handy but what does this do? ((lambda (f) (f)) (lambda() 42)) You might need to throw a 'funcall in there or something,
815.
▲
by
akkartik
5y ago
Why is this quoting the top-level lambda expression arguments? Unless I'm mistaken that's not how JMC did it. For example, to define a binding the cited resource http://www.paulgraham.com/rootsoflisp.html uses
816.
▲
by
akkartik
5y ago
No upvote needed :) but sure I'll post it here. (I haven't done this so far because I find metacircularity to not be very interesting. It was interesting when JMC proved it could be done. Mu's whole reason for existence is li
817.
▲
by
akkartik
5y ago
Sorry to tramp over your thread! Mu isn't too related to SectorLISP, I was just responding to the side-discussion about what we can build in under 1MB. To answer your question, evaluate.mu does support lambda (I call it `fn`). My estim
818.
▲
by
akkartik
5y ago
Assuming you mean static code size, my Mu project currently supports an interpreted Lisp (macros, fewer parens, infix) implemented in a compiled memory-safe low-level language without _any_ C in 450KB. The "runtime library" (mostl
819.
▲
The slab and the permacomputer
(society.robinsloan.com)
80 points
by
akkartik
5y ago
|
24 comments
820.
▲
by
akkartik
5y ago
I used to wonder about questions like this, but somewhere along the line they've become unimportant. I'm not sure why that is, or whether that's a good thing, but my best guess is that I've grown less delusional about ho
821.
▲
by
akkartik
5y ago
In no small part due to Shelley.
822.
▲
by
akkartik
5y ago
I wonder if future generations will totally misunderstand "left to own devices." No, it doesn't mean with your phone, tablet and watch.
823.
▲
by
akkartik
5y ago
I read this, switched to another tab -- and found this sentence staring at me: "Inspired by a script written by Paul Eggert." https://github.com/lcurses/lcurses/blob/master/bootstrap
824.
▲
by
akkartik
5y ago
This article reminds me in particular of my long-open question about how to prepare for solar storms. But now I have a new question. If the way to prepare my electronics for a solar storm is to put them in a cardboard- and aluminium-foil-li
825.
▲
by
akkartik
5y ago
I hope someone who understands Firefox internals and release process is ready to step in with uBlock Origin if/when they finally jump the shark.
826.
▲
by
akkartik
5y ago
You're making a great point that I didn't find in OP. It muddies the waters to conflate holiday shopping with paper towel shortages.
827.
▲
Life After Default
(whitehouse.gov)
1 points
by
akkartik
5y ago
|
0 comments
828.
▲
by
akkartik
5y ago
How soon would we need a fourth?!
829.
▲
by
akkartik
5y ago
Does this support pipes? It's not clear to me what runs in the same process and what forks a new one. Is there a way to run some shell code in a sub-process (parentheses in conventional shell)? Asynchronously (& in conventional she
830.
▲
by
akkartik
5y ago
A Deepness in the Sky is my sentimental favorite scifi novel of all time. I rarely reread scifi.
831.
▲
by
akkartik
5y ago
I remember the good old days when we had offline first by default. We called them just "computer programs". Offline first as a principle is more important than web apps. If today's browsers have trouble with offline first, co
832.
▲
by
akkartik
5y ago
I couldn't stop thinking about this link. Here's a stronger statement of the problems I see with the GNU guidelines: https://merveilles.town/@akkartik/107022123346940723
833.
▲
by
akkartik
5y ago
Regarding "genuinely curious": I've seen some real passive-aggressive comments that say that :) Regarding "asking devs directly": I often find it challenging even to figure out where to ask. Going from a program to
834.
▲
by
akkartik
5y ago
Thanks! Glad you didn't think I was in bad faith after the semi-rhetorical question and disengage. That whole thread makes me sad. Another, more subtle way IP is ruining the world. But that's neither here nor there on this debate.
835.
▲
by
akkartik
5y ago
Thank you! And yes, I admit you were partly right. Questions often have some component of rhetorical quality. It's not always yes/no on whether it's a rhetorical question or not. But either way, I find (based on experience te
836.
▲
by
akkartik
5y ago
I don't understand what you mean. Why can't we just debate the question without speculating about motivations of the people raising/answering the question?
837.
▲
by
akkartik
5y ago
I don't find it so questionable to ask questions that the right google search (out of an infinite number; we're not all equally deeply immmersed in this stuff) can answer. But if we disagree about that I guess I'll stop talki
838.
▲
by
akkartik
5y ago
Does GNU yes need such great performance, though? I'm curious to hear where it matters. Why would we need 10 billion yeses a second, and why would 10 million yeses a second not be good enough?
839.
▲
by
akkartik
5y ago
Their search results used to be great before they started doing this. In fact, they've started to suck more in the past 10 years, though probably for unrelated reasons. As the Conchords say, "what are your overheads?"
840.
▲
by
akkartik
5y ago
I find link mangling to be a great test for when a service becomes too powerful relative to its users, able to add friction to its product that serves no purpose for its users[1]. Google started doing this around 2004[2]. Facebook didn'
More ›