#trilema | 2019-11-13

http://logs.nosuchlabs.com/log/trilema/2019-11-12#1950579 << what exactly do you think that word means?

00:59

snsabot

Logged on 2019-11-12 10:51:30 diana_coman: fwiw my take so far is that trinque wants to provide consultancy at most.

00:59

and why are you soviets so damned allergic to structuring things *like a business*

00:59

if you're asking me, i think it means willing to do work at arm's length.

01:00

and i dunno what the gap is! make it plain, what is it, what's the difference between whatever this is and whatever a business'd be, an' pinpoint the allergy neh ?

01:02

I came in offering to mentor both lobbes and spyked in starting a business, and this is "omg, faggot, do actual work" ?

01:03

and yes, I'm not doing the monastic thing. I went on about this ad nauseum with much rustling of jimmies.

01:04

i don't think we're talking on the same basis here. i dunno how your "omg, faggot, do actual work" maps on the actual occurence.

01:04

http://logs.nosuchlabs.com/log/trilema/2019-11-12#1950644 << the THAT is ill defined.

01:04

snsabot

Logged on 2019-11-12 13:51:42 mircea_popescu: and as to the "bois etc" -- is very well fucking earned. the logs since http://logs.ossasepia.com/log/trilema/2019-11-06#1950066 ie a week ago are sorely missing a "trinque: yeah, I WILL DO THAT". or an equally clear, "no, i won't do that". this is what separates the bois from the men : FIRM committments.

01:04

Logged on 2019-11-06 23:32:29 mircea_popescu: trinque, suppose spyked's game, and what you have to do is run the two fellows + whatever we can meaningfully accrete through the tmsr-ization of cuntoo.

01:04

dunno how "much rustling of jimmie" maps on historical account.

01:04

the last thing I said on teh subj is for those two gents to interview their potential customer and find out what they want, and report it to me

01:04

trinque, so where did you say "i will do that" ?

01:04

link ?

01:06

http://logs.ossasepia.com/log/trilema/2019-11-11#1950544 << here and down, and they agreed, and I have not heard from them in #trinque , is where we stand.

01:06

Logged on 2019-11-11 20:22:50 trinque: http://logs.ossasepia.com/log/trilema/2019-11-08#1950231 << since I've got the botwork already in flight on my end, what do you say you and lobbes tackle making cuntoo into a usable item for mp_en_viaje and diana_coman ?

01:07

anyways. all jobs are perpetually ill defined. that's why they need doing. if they were well defined they also would have been done, past tense.

01:07

yes but they don't remain ill defined, lest years be spent toiling in the wrong direction, or circles.

01:07

I am not a lazy man.

01:07

trinque, there's large gaps between what's actually said and what you construe what's said to mean on the uptake ; and there's similarily large gaps between what you actually say and what you expect it to be taken as, on the rebound.

01:08

i dun think you're a lazy man. currently i think you grew up in a house with a lot of door slamming as a rhetorical device, but that's really neither here nor there.

01:10

now, as to the putative customer approach : i dun specifically want to make a new os for every single app we come up with. i'd like to make one and be done with it, such that minigame can use it, trb can use it, everyone can fucking use it.

01:12

I didn't say sell the OS.

01:12

well so take a minute and say what you meant to say.

01:12

http://logs.ossasepia.com/log/trilema/2019-11-12#1950667 << this dude is literally taking the approach I'm prescribing.

01:12

Logged on 2019-11-12 16:59:38 jfw: dorion and I have a few consulting clients that are learning linux/unix for the first time. We use the distribution ourselves for niches it's able to fill, and we considered it a good learning environment due to avoiding various complexities

01:12

explicity an' at its own length

01:14

heh, where's the quote where you say if you're worth a shit, being poor just kinda brushes off.

01:15

i dunno, like half of trilema

01:18

it is hard from where I sit to reconcile this "the work must be done whatever the cost" with any sense of survivability. so one says either the latter doesn't matter, or one's to fill that gap himself.

01:19

if it doesn't matter, one needs less self-involved shits than myself.

01:19

or one could just you know, chip at it and signal with others to help.

01:19

if it does, I'm gonna keep going "what the hell's the plan for money coming *in*"

01:21

*

trinque 's last signal reads " diana_coman, pls to tell me what the hell you intend to do with this thing"

01:21

oh it did ?!

01:22

there's two portions current minigame strat needs from os. there's a server os, which has to be static-linkable. there's a client os, which has to be portable and support gfx.

01:22

ideally these share as much of a trunk as possible.

01:24

obviously the latter is the ratfuck, multiplied if ya don't specify hardware.

01:24

I'm not saying you should, just pointing at it

01:25

well, "portable", aka everything

01:25

but, yes.

01:26

considered statically compiling the whole game, and letting the kids run wherever, or not desirable?

01:26

statically compiling graphic stack is currently impossible.

01:27

it's actually the broken glass we're trying to extract through all this process.

01:28

yep, just enumerating. what about bundling the necessary libs alongside. afaik this is what steam does, static linking for idiots

01:29

steam does not work.

01:29

i don't even mean this in any kind of theoretical sense, i've been testing it.

01:29

it's a crapshoot.

01:29

it will half the time "upgrade" itself into unrecoverable broken state.

01:30

ie, at any given moment some significant percentage of hw configs are unsupportable, and there's ~no way to predict which.

01:30

I wasn't proposing using actual steam, but the approach, which probably would still suffer from what you say.

01:31

if i knew of something giving a leg to stand on i'd push it, but atm got nothing there, so yeah

01:33

who is the user?

01:33

this is their first exposure to an adult operating system?

01:33

well, vidya right

01:33

so then oldschool boot cd, and you can play from it, and you can install if you want?

01:34

the idea is to construct funel towards republic, so he needn't be dumb. but he might be new

01:34

trinque, install ~on what~

01:35

i have no problem with "Boot from it". but it hasta boot. which... is not trivial, given one wants to support 3d gfx.

01:36

mhm, wads of firmware and drivers needed depending on gpu

01:36

efi complicating the shit out of things these days

01:36

and no static linkage in any case.

01:36

taken to the sad extreme, one could argue to target the browser

01:37

stupid idea. you just get more of the same crap.

01:37

the notion browsers are portable is a lot like the notion women are whores.

01:37

I'm just marking off the limits here

01:37

~not if they can help it~

01:40

so the target is a tower of some kind a sharp kid might have, or be able to build

01:40

or maybe a laptop he has

01:40

so this is nvidia, radeon, and intel drivers

01:40

in fairness, game as it stands works fine w/o acceleration anyways. so it's a lot more portable in practice than the theoretical extension.

01:40

walking through the list of firmware one needs for these isn't insurmountable. I've got a radeon box sitting here that runs cuntoo

01:41

trinque, i agree ; it just needs the doing.

01:41

I think it's beneficial to have the thing able to run as a livecd, with option to install.

01:42

the cuntoo bootstrapper I produced was misery in that it either produced a booting drive, or fuck you peasant, learn moar.

01:42

what is it that would even be persisted to the disk?

01:43

gpg keys? assuming the kid gpgs on same machine

01:44

lots would persist on the disk. keys, configs, and the data cache.

01:44

potentially gbs

01:44

k, so not much fun to be had in the livecd.

01:44

meaning that you can't play in there, just install

01:44

the idea is that the client runs on its own responsibility, asks server for all items it dun have.

01:45

so it's a sort of perma-bootstrapper of the game

01:51

k, so this is little else than creating a signed tarball of the cuntoo bootstrapper's output, and building a livecd/usb that can unpack that item onto a fresh system, bundled with some driver firmware and a try-all-modules kernel.

01:51

this I am willing to actively manage spyked and lobbes to produce, if they will. others also may apply.

01:51

not necessarily. a dvd is what, coupla gb. a recent client could just pack its data disk there.

01:51

if you want newer gfx, install and remake dvd

01:51

if you don't care, play from dvd

01:52

if that's possible, it opens up kids throwing the thing into w/e school computer, like we used to do with quake back in the day.

01:53

sure.

01:53

same image in both cases, just one's readonly root

01:53

i don't see why it wouldn't be possible. doesn't work atm, but yes.

01:53

http://logs.ossasepia.com/log/trilema/2019-11-11#1950552 << this ftr was not some kind of alfian goalpost-moving.

01:53

Logged on 2019-11-11 20:26:06 trinque: http://logs.ossasepia.com/log/trilema/2019-11-11#1950494 << this one's an experiment, not ready for public use, but feel free to discuss with me in #trinque

01:54

oh ? cuz it sounded just like it

01:54

I am working on that other item, and I was telling the guys *not* to worry about it for now.

01:54

it doesn't supplant using gentoo any time soon.

01:54

it's a derivative of something I've used in embedded machines for a long time, and I'm seeing if it can be made into an adult OS

01:55

aite

01:55

if so, it'll have human-sized readable patches for every item running.

01:55

but it's just not further than speculative, and I don't want people chasing after it

01:55

also fuck not acting because tomorrow there'll be even better icecream

01:55

yes, well, with all that context a whole different thing!

~ 18 minutes ~

02:14

lobbes and/or spyked: you around?

02:16

I'd like to approach this in phases, documentation preceding every phase's work product.

02:17

I'd like to chat before any of this, and come to an agreement on how we work together

02:19

http://logs.ossasepia.com/log/trilema/2019-11-11#1950546 << this stands, and what I meant wasn't the bastardized american version, where you feed the client shit til he notices.

02:19

Logged on 2019-11-11 20:24:11 trinque: I'm thinking you and he could turn this into a proper consulting gig if they like what you produce, and I'm happy to manage / advise such an endeavor.

02:19

meanwhile the new kids got sucked into this also.

02:20

http://logs.ossasepia.com/log/trilema/2019-11-12#1950667 << seems like there's a lot of overlap, yes

02:20

Logged on 2019-11-12 16:59:38 jfw: dorion and I have a few consulting clients that are learning linux/unix for the first time. We use the distribution ourselves for niches it's able to fill, and we considered it a good learning environment due to avoiding various complexities

02:20

sounds like not very sophisticated clients, not that I'm throwing shade at your hustle jfw

02:20

potentially there's a good half dozen hands here, if ave1 & bvt can be seduced into it.

02:21

*

trinque strains, yet tits do not erupt

02:21

trinque: I am around (and have been following the conversation)

02:21

more the merrier, to a point. so long as it's *directed* effort

02:21

right. directed an' structured an' documented and os on

02:26

lobbes: do you have a machine with a decent GPU?

02:26

just to be clear, my I believe role in this thing (at least to start) should be at the bottom of the totem pole. I have much to learn still but am willing to work. Though I will need management on what exactly to work on. I don't think I even could have conducted the client interviews on the level trinque (or jfw, for that matter) did. I simply do not have enough experience yet

02:26

trinque: I might have one. How do I determine decency of gpu in this context?

02:27

plus i don't remember you throwing a hissy fit for no apparent reason

02:28

lobbes: if you have that GPU someone else will also, eh?

02:31

trinque: I'm not following you

02:31

w/e GPU you have is fine. if it ends up not being fine, I'll just send you one.

02:31

kk, works well enough for me

02:32

and ty btw, for this opportunity

02:36

trinque: I also want to make sure you know that I still have some items to knock out this month before I can devote full focus, but starting in December I should be ready to roll

02:37

Not that I can't start chewing on things early, of course

02:37

I don't want to burn anything out. I think it's upon me to specify this thing, receive feedback, and then delegate tasks.

02:38

*burn anyone

02:38

sounds reasonable

02:39

but I want to know who wants to eat the thing with me, and start talking to them on a regular basis.

02:42

spyked: what about you, have a usable GPU?

02:44

!!v E08EF01B1D517D3DE9D9A8C0688CC0EBBAE8B67C16F81B8ACE5941B4EEABC504

02:44

deedbot

trinque rated lobbes 1 << cuntooist

02:44

^ not being frivolous here.

02:47

lobbes: starting when in december?

02:52

the first lol, i'm not dragging the whole mp-wp logger on forever.

02:55

^ what he said

02:58

I'm aiming to get remainder of mp-wp logger complete in under two weeks. Then rest of Nov I want to tidy up auctionbot. So Dec 1st should be a good starting point for me

03:05

lobbes, http://trilema.com/2019/proper-html-linking-the-crisis-the-solution-the-resolution-conclusion/#comment-132249 << in other lulz.

~ 44 minutes ~

03:49

mircea_popescu: nice bug findin'. I'll make sure it is rolled into the patch

~ 3 hours 51 minutes ~

07:40

trinque: ask me if there's any info you need from me; and are you going to answer this century re http://logs.ossasepia.com/log/trilema/2019-11-12#1950568 ?

07:40

Logged on 2019-11-12 03:53:35 diana_coman: http://logs.ossasepia.com/log/trinque/2019-11-03#1000013 - any eta on this?

07:40

even better, just publish the damned thing already, will you?

07:41

other than that, great to read a trinque-heavy log for once, ha.

~ 1 hours 7 minutes ~

08:49

http://logs.ossasepia.com/log/trilema/2019-11-12#1950779 <-- around 10 hrs, plus-minus 2. it's not much time, but I aim to, as much as possible, stick to that 10hrs/week figure and properly break down/manage the tasks I can fit into this time.

08:49

Logged on 2019-11-12 17:55:51 mircea_popescu: spyked, what sorta time could you commit on a weekly basis, if this tmsr os megaproject actually gets going ?

08:49

http://logs.ossasepia.com/log/trilema/2019-11-12#1951050 <-- this sounds more than reasonable to me

08:49

Logged on 2019-11-12 22:22:17 trinque: I'd like to approach this in phases, documentation preceding every phase's work product.

08:57

http://logs.ossasepia.com/log/trilema/2019-11-12#1951078 <-- I have an old thinkpad with an intel 3000 (works decently for games that don't ask for the latest shaders & crap) and perhaps I could get my hands on other GPUs. the problematic point is that I'm spending a considerable portion of that allocated 10hrs in saeculum, without physical access to the hardware. so would have to work out some routine for

08:57

Logged on 2019-11-12 22:47:54 trinque: spyked: what about you, have a usable GPU?

08:57

testing, or maybe get some sort of remote kvm access installed.

09:00

ftr, I was considering doing my initial cuntoo boostrapper install on a qemu. I've gotten disk images and kernels prepped for that environment in the past, it's pretty easy due to the absence of odd vendor-specific drivers. this is probably not appropriate for gfx work though.

09:03

^ imho would be worth doing this either way, if only because my familiarity with cuntoo so far is limited to reading the bootstrapper scripts.

09:17

http://logs.ossasepia.com/log/trilema/2019-11-12#1951029 <-- just to be as pedantic as possible: plus all the deps needed to run the eulora client. from what I understand, would have to produce static builds of these, which touches new ground. not sure if e.g. http://www.eulorum.org/Gentoo is up to date, but we'd be the first to attempt doing a static build of nvidia-toolkit.

09:17

Logged on 2019-11-12 21:56:47 trinque: k, so this is little else than creating a signed tarball of the cuntoo bootstrapper's output, and building a livecd/usb that can unpack that item onto a fresh system, bundled with some driver firmware and a try-all-modules kernel.

~ 20 minutes ~

09:38

spyked: the eulora client build guides are still up to date, yes.

09:47

http://logs.ossasepia.com/log/trilema/2019-11-13#1951092 << i think it's plenty.

09:47

Logged on 2019-11-13 04:54:57 spyked: http://logs.ossasepia.com/log/trilema/2019-11-12#1950779 <-- around 10 hrs, plus-minus 2. it's not much time, but I aim to, as much as possible, stick to that 10hrs/week figure and properly break down/manage the tasks I can fit into this time.

09:47

a LOT can be accomplished in ten hours, provided one doesn't pay any being an engineer rents with them.

09:47

most high level tasks are "5 minutes' worth of not being fucking stupid"

09:58

http://logs.ossasepia.com/log/trilema/2019-11-13#1951098 << that's silly. can't jus ttake the laptop over with you ?

09:58

Logged on 2019-11-13 05:03:13 spyked: testing, or maybe get some sort of remote kvm access installed.

09:59

http://logs.ossasepia.com/log/trilema/2019-11-13#1951099 << possibly worth doing at least once anyways

09:59

Logged on 2019-11-13 05:05:55 spyked: ftr, I was considering doing my initial cuntoo boostrapper install on a qemu. I've gotten disk images and kernels prepped for that environment in the past, it's pretty easy due to the absence of odd vendor-specific drivers. this is probably not appropriate for gfx work though.

10:00

also, an attempt at a static nvidia-toolkit would be interesting regardless of results. afaik nobody even tried to any significant degree

10:02

*

mircea_popescu has 0 problem saying "oh, eulora only works on nvidia" or w/e limiting in this vein.

10:03

i'd rather it not be "only works on intel" if at all possible, but otherwise...

10:03

fwiw I actually ran it without that toolkit but the resulting quality of graphics is visibly low.

10:03

aha

10:03

and we intend to muchly broaden the extant gfx basis anyway, hence all these self-curing client genuflexions

10:04

a lot of the deps there are not fully mandatory in that if one wants, one can run the client anyway but at the extreme this goes to "well, if one WANTS TO, one can make their own bloody text client already and run that"

10:04

right

~ 15 minutes ~

10:20

diana_coman, so listen, e3-1241 or e5-2620 x2 ?

10:21

the latter has significant larger ram aperture (3x the speed, 16x the address space) and significantly more threads possibly (24 vs 8)

10:22

myeah, I was just having a look at the specs

10:22

they're intels tho.

10:22

that's the thing, that I'm not sure that it makes all that big difference really esp atm.

10:22

you gotta tell me.

10:23

it's pretty much all e3/e5. i could perhaps squeeze a celeron if pressed, but there's no amds.

10:23

mircea_popescu: hm, listen, can this wait until I meet those dc guys tomorrow?

10:24

it can wait.

10:24

at the end of the day, if I get a rack or whatevers next to me, I can just as well plonk a server in there

10:24

and I can put whatever I buy.

10:24

sure.

10:24

I'll take my camera tomorrow and hopefully find some time in the evening to blog it up too

10:24

very nice.

10:25

ok so then ima put this on ice for a few.

10:25

thanks.

10:25

*

mircea_popescu still wants eulora prod server sorted this week.

10:25

noted.

~ 3 hours 49 minutes ~

14:15

http://logs.ossasepia.com/log/trilema/2019-11-12#1951034 << the more i think about this, the less a problem i perceive. i mean, nobody even uses cd boot anymore, they do the hybrid thing so you can use the image on sticks too. well... there's 32gb sticks for ~free, you just set it with a data partition and that's that. in the very unlikely case someone actually has enough artwork to fill 32gb, um... you just tell the pl

14:15

ayer to buy a 256gb one or w/e. or alternatively... to only save the MOST USED 32 gb.

14:15

Logged on 2019-11-12 21:58:37 trinque: if that's possible, it opens up kids throwing the thing into w/e school computer, like we used to do with quake back in the day.

14:15

srsly, this can't become a problem.

14:17

Even down here the little shim for converting laptop optical disc bays into generic 2.5" SATA drive bays are common enough finding one wasn't a headache

14:30

BingoBoingo: pizarroisp.net shows a lot of warrnings/errors spew.

14:32

diana_coman: Right, I put in on the 'anyserver' but haven't done any tweaks to it other than making it clear the operation is closed/closing. I'll give it a couple minutes and see if I can silence the spew.

14:32

hardly worth the bother huh

14:33

Well, I didn't think so, but I'll give it some minutes now that I know someone else is trying to read it.

14:35

And the spew's been silenced

14:35

#nomorespew

14:35

BingoBoingo: alternatively if you can point me directly to how you calculated http://logs.ossasepia.com/log/trilema/2019-11-12#1950906, it would even save some time.

14:35

Logged on 2019-11-12 19:35:26 BingoBoingo: !!pay diana_coman 0.10491609

14:36

diana_coman, on the basis of smg experience, he prorated by time

14:37

Rockchip 0.08691609 + Shared 0.018 on http://bingology.net/2019/pizarro-isp-pre-closing-statement/

14:38

I split all of mircea_popescu_popescu's things for their being used for different ventures, wasn't sure if anyone else wanted that in the paying out.

14:39

But yes, everyone got prorated by time to figure out their customer equity.

14:40

it's ok, now I see where I had it wrong, got confused re start/end on the RK; it's fine as it is, thank you BingoBoingo

14:41

diana_coman: You're welcome and thank you for checking.

14:46

*

BingoBoingo priviledged the deedbot returned ledgers and invoice statements over the notes in mathing, though everything was math'd. Repeatedly. Still I admit the room for error to creep in when doing math on calulator and will entertain claims, but preferably sooner instead of later.

~ 27 minutes ~

15:14

meanwhile in commercialization lulz, https://sissybrides.com/

~ 4 hours 31 minutes ~

19:45

In local news, the day of the last Presidential debate... The Argentina's president-elect came for a big photo shoot with Uruguay's sitting president and the Fat Forehead nominee. I am having serious trouble conceiving a situation where the commies aren't trying to throw the contest.

19:53

feedbot

http://qntra.net/2019/11/medium-com-now-fetid-cesspool-of-spam-and-malware/ << Qntra -- Medium.com Now Fetid Cesspool Of Spam And Malware

~ 56 minutes ~

20:49

feedbot

http://trilema.com/2019/jasons-folding/ << Trilema -- Jason's folding.

~ 1 hours 6 minutes ~

21:55

bvt

http://logs.nosuchlabs.com/log/trilema/2019-11-12#1950799 << yes, sure. the article with measurements will definitely come tomorrow -- most of it is written, measurents - finished, bugs - fixed.

21:55

snsabot

Logged on 2019-11-12 22:03:27 mircea_popescu: bvt, once done fixing the kernel rng, you wana join in this fray by the way ?

21:55

neato

21:55

i'm really excited about this!

22:07

bvt

ty, i guess after the article it will become possible to decide what to use for each of the hashes; after that one more patch - user-settable key for hashing, and rng work will be done

22:10

bvt

re tmsr os - i am curious what work plan trinque will come up with, esp wrt static linking.

~ 52 minutes ~

23:02

bvt, me too.

23:05