#asciilifeform | 2020-02-18

BingoBoingo: http://paste.deedbot.org/?id=H_7q

00:18

BingoBoingo: plz note, there's at least 1 formatting typo

00:19

tyvm

00:19

sorry about the hasty editing, but item is quite fresh and afaik not 'scooped' yet anywhere.

~ 51 minutes ~

01:11

asciilifeform: There might be the potential for something lulzy there, but this is looking fairly mundane to me.

01:13

Yes it's an insane situation, but can you rewrite the copy to hit particular points of wrongdoing? Don't worry about urls, I've got text extracted from the pdfs.

01:14

My eyes just aren't seeing the hook yet.

~ 26 minutes ~

01:40

BingoBoingo: feel free to skip

01:41

BingoBoingo: i'm at the dc again, fighting an apparently cursed raid card ( my box, not customer's ftr )

01:41

asciilifeform: Aite, awk is starting to make some sense. I'm going to report where I've made it on the automation in #trilema-hanbot and get outside to stretch the legs

~ 45 minutes ~

02:27

*

asciilifeform back

~ 41 minutes ~

03:09

*

asciilifeform recently installed, among other things, a thermometer. thought 'let's put it on www' then found, most boring possible graph : it's 21c +/1 0.1 , 24/7 . 4 redundant chiller plants will do that...

03:10

* +/-

03:11

and this ftr was false alarm. ( ftr for dulap-gentoo users : you gotta run 'lilo' after kernel rebuild erry time. )

03:11

Logged on 2020-02-18 01:41:26 asciilifeform: BingoBoingo: i'm at the dc again, fighting an apparently cursed raid card ( my box, not customer's ftr )

~ 13 hours 28 minutes ~

16:39

feedbot

http://qntra.net/2020/02/google-carrying-out-take-down-requests-over-word-did-on-behalf-of-dance-india-dance/ << Qntra -- Google Carrying Out Take Down Requests Over Word "Did" On Behalf Of "Dance India Dance"

16:41

curiousd0g

wtf

16:54

*

asciilifeform in entirely unrelated noose, thought 'why not set up usenet node! wanted for 25yrs nao'. but after 5min of napkin arithm realized that it can't be done with a 100M/s or even 1G/s pipe -- according to various sources (somewhat stale #s, so prolly optimistic) it nao weighs >100TB/day... re 'what if exclude alt.binaries.*' found 0 figs from this decade even. so open q.

16:57

for something that's 'officially dead' -- usenet is surprisingly lively. tho does seem to consist 99.99..+% of warez nao.

16:58

http://logs.nosuchlabs.com/log/asciilifeform/2020-02-18#1006880 <<< "Superpower by 2020"

16:58

Logged on 2020-02-18 03:11:34 snsabot: Logged on 2020-02-18 01:41:26 asciilifeform: BingoBoingo: i'm at the dc again, fighting an apparently cursed raid card ( my box, not customer's ftr )

16:58

gah wring link

16:58

lolwat

16:58

a

16:59

I meant to comment on BingoBoingo 's India takedown Qntra piece.

17:02

for thread-completeness : typical usenet pipe vendor, '6 $ / mo'. and that's ~with alt.binaries.*~. the price point suggests that these are ~all~ subsidized by usg , to work as honeypots...

17:02

Logged on 2020-02-12 23:32:26 asciilifeform: curiousd0g: you know that proxy co.'s are honeypots , yes ? (is this term familiar?)

17:02

( linked example even ~includes~ vpnism. )

17:07

" unless required by law" .... so subjective.

17:08

shinohai: these are in no sense honest commercial shops. they exist to fill the usg 'terrorist pornographer trials' pipeline.

17:08

what verbiage even printed there, is ~entirely uninteresting.

17:09

otoh, a usenet-like animal built w/ support for wot, would be imho interesting.

17:11

but -- just as with the case of btc noadez -- would have to pay somehow for the bandwidth. which is an unsolved problem in the general case.

17:11

(trilema) 2017-03-01 asciilifeform: what are ~all~ of the places where A has the ecstasy, but B does the laundry, where A!=B

17:13

btc node for that matter only requires, in avg., to store <2kB/sec (1e6byte block per 10m, that is). far below even an ultra low-traffic news feed.

17:14

and that's only counting basic intake, rather than relay.

17:15

( e.g. a btc node relaying blocks to 16 peers, still only on avg. needs ~34kB/s. (will leave the arithm as exercise for reader. )

17:16

tx relay (empirically) similar figure.

17:17

so node seems almost 'phree'. but even smallest p2p net w/ human-generated traffic -- orders of magnitude hungrier.

17:29

meanwhile, asciilifeform's cage , as of yest.

17:41

proof-of-rack

17:43

lol

17:44

could still be fake!111 and asciilifeform shits out return packets w/ a telegraph key!11

17:52

meanwhile elsewhere: apparently shtrafbat for BingoBoingo.

17:52

(ossasepia) 2020-02-18 BingoBoingo: diana_coman: If it's not too late, I would like to pledge as a page. This weekend as I turned 34, I was struck by terror as I pondered what exactly is inside my veins. Dread left me paralyzed. For years I've been trying to conquer my shortcomings as an independent, and I have failed.

17:53

*

shinohai hearls the distant sound of safety bell tinkling over grave from previous thread.

~ 1 hours 32 minutes ~

19:25

On the topic of those who lead and those who follow, do you've any thoughts regarding that Common Lisp Style Document I wrote, asciilifeform?

19:31

verisimilitude: this one ?

19:31

Logged on 2020-02-16 00:34:53 shinohai: Common Lisp Style Document <<< ftfy

19:31

Yes.

19:32

it aint unreasonable, but i always put space after comment chars (in all langs.) also wondered why verisimilitude is fond of 1970s-style all-caps

19:32

I just use that for display.

19:33

It's nice for display in some cases, I think.

19:33

i assumed that since item is proposed as 'style guide' , all aspects then (incl. case) are illustrative of the desired style

19:34

btw verisimilitude , imho if you want folx to consider a cl styleguide, gotta publish a modded commonlispmode/slime to go with it

19:34

given as elementarily most folx who write ~any amt of cl, use the robotic stylization of the above

19:35

I considered pointing that out explicitly; it's still in-progress, so I'll explicitly note the case is for display purposes only.

19:36

Interestingly, I put a space after the comment introduction in APL and Ada, but not Common Lisp. I write more comments in those two than Common Lisp, as well.

19:37

As for mechanical enforcement of this style, I'm too lazy to dig around to accomplish that is one way to phrase it. The mind and the body are both unwilling here.

19:39

Is there anything important you believe I've omitted, asciilifeform?

19:40

aside from what mentioned above, not afaik

19:41

this is where i admit that i'm not terribly preoccupied with cl style. it is imho somewhere on the bottom of list of defensible priorities for a cl aficionado .

19:41

'chair arrangement on titanic'

19:46

Sure, I don't disagree. Someone in my group wanted a good style guide and, lacking writing ideas, I thought I'd write one distilling my thoughts.

19:46

Besides, the guides which do exist truly are poor, in my eyes.

19:47

Well, those which existed before mine, anyway.

19:47

verisimilitude: the issue is that no matter how beautiful 'style guide' you make, it aint gonna bring cl on pc back to life

19:47

this'd require a gigatonne of much more unpleasant and less 'sexy' work

19:48

I yearn to begin work on my own Common Lisp implementation this year, but it will have its own goals reflecting mine.

19:48

verisimilitude: what sorta goals do you have for this ?

19:49

I want it to be memory efficient, rather than time efficient.

19:50

i.e. using classical 'stop the world' gc method ?

19:50

So, making it use as little memory as is reasonable is the main goal, as opposed to making it fast. I'll stress I won't be cheating by implementing DISASSEMBLE and friends as naught, among other things.

19:50

Sure, asciilifeform.

19:51

seems to go w/out saying that you'd leave out compiler , if want strictly compactness

19:51

Regarding what else you wrote, I'm usually rather uninterested in what others are working on or what's popular. I'm currently learning REFAL, because I found it following links on your website lately, but were I only concerned with what's available or popular or whatnot, I'd discard this neat language without a second thought, and be worse for it.

19:51

That's another way I won't be ``cheating''; I want a full machine code compiler, along with an interpreter.

19:52

seems to contradict 'memory- rather than time- efficient' neh

19:52

It would be memory efficiency without eschewing other nice things.

19:53

Anyway, I'll probably only start on the Common Lisp side of things, this year. One aspect of macroexpansions under the popular implementations I vehemently dislike is how they don't expand to pleasant Common Lisp.

19:53

this is indeed true on e.g. sbcl -- they certainly do not

19:54

You want to see what LOOP, say, expands to and you get calls to what may as well be LOOP-PACKAGE::IMPLEMENT-THE-SPECIAL-LOOP-THING-DONT-TOUCH-FUCK-YOU.

19:54

afaik this is on acct of having the compiler tho. otherwise all cltrons would use the loop given by g. steele in cltl2

19:55

Now that I mention this, I should add this macro note to my style document, certainly.

19:55

there's a very real engineering tension b/w 'readable' and 'machine-optimal'

19:55

Yes, but I think it can be walked better than it currently is.

19:55

By the by, can you guess why it's a ``style document'' and not a ``style guide'', asciilifeform?

19:56

nope

19:56

It abbreviates to ``Common LSD'' this way.

19:56

lol

19:59

will add to this, that much of the 'ugly' of e.g. sbcl, comes from the fact that it is very, very difficult to compile anything efficiently on pc arch. (yes, even c.)

20:00

verisimilitude: ever had occasion to write a compiler ?

20:04

I've not written a compiler to machine code yet, no, but I've written other compilers.

20:06

Architectures I don't like or care for is what the interpreter would be for.

20:10

My machine code work is still preoccupied with that development tool I've been lazily working on. I'll probably work on a MIPS targeting, next, once I finish that which I'm currently focusing on having working and well.

20:11

The Intel architecture isn't even suited to assemblers, so it's not a target I intend to aim at.

~ 29 minutes ~

20:40

zeder

there's obviously a lot of backstory here that I've missed, but it's not immediately clear to me what "Intel arch isn't even suited to assemblers"

20:43

zeder: i suspect he meant 'not friendly to hand-written asm' but will let'im clarify

20:57

In considering writing a development tool for Intel architectures, there's too many instructions and variations thereof for my model, but even an assembler is unpleasant to use, due to the sheer number, prefixes, and other such things. It's my understanding Intel has a library specifically to make it easier for others to write an Intel assembler, because it's so complicated.

20:57