#ossasepia | 2020-03-09

jfw: lolz, illustrating how homework is to be done? :D

03:27

BingoBoingo: framing and links are a good addition indeed.

~ 19 minutes ~

03:46

in other boring statements of the obvious, I shall inform log readers that the whole point of a review is to find out what worked (so you can keep doing it), what didn't (so you stop doing it) and what is somewhere in between (so you improve on it).

03:48

feedbot

http://younghands.club/2020/03/09/jfw-plan-week-of-9-mar-2020/ << Young Hands Club -- JFW plan, week of 9 Mar. 2020

03:53

I've no retort diana_coman, laugh away. Adding that boring statement of the obvious to my (somewhat neglected) quotes file.

~ 19 minutes ~

04:12

as far as I can see I answered everyone that actually used younghands.club today and I even have time to spare, huh; (if I missed anyone/anything, please ping me!)

04:15

I guess I can use the unexpected extra free time for practicing chopping laurels and meditating on their fragrant remains.

~ 3 hours 5 minutes ~

07:20

http://logs.ossasepia.com/log/ossasepia/2020-03-03#1019868 <-- ftr, I've been using ledger 2.6.2 for a while and it's still a monster (cca 18k lines of C++), but probably smaller than 3.x and iirc it doesn't depend on boost

07:20

Logged on 2020-03-03 15:20:56 diana_coman: jfw dorion re clock & ledger - did you genesis them both properly then? because esp the ledger thing, when I just had a look at it earlier today, it's a monster; perhaps you still have at least some less-of-a-monstrosity version?

07:25

anyway, the clock program looks really neat on a first glance, I'm tempted to give it a try

07:25

spyked: well, no boost dep is certainly a gain; and yeah, I'm sure that the earlier the version the less of a monster it is; that being said, I'm either really getting old or something but I can hardly see the point of >10k LOC for what can be done precisely as wanted through <1k lines of cmd line tools, huh.

07:26

nevertheless, if it's indeed in use and useful, a genesis should be made; and possibly from there it might get trimmed in time.

07:27

ftr re clock, my scripts don't bother with "end of task" - there's no end to anything, only start of next thing whether that's break or whatever; since time is spent anyway, the start of a new thing is by definition the end of the previous.

07:28

re. size, that's a very good point. I haven't looked deeper tbh, I switched to it from gnucash, which was even bigger, it came with a whole array of gui 'features' that I never used

07:30

spyked: ah, certainly; and yeah, years ago I used gnucash so I know what you're saying there; and again, the *idea* of ledger is great as far as I can see, right on target; if only there was anyone in there actually stubbornly keeping to precisely that, rather than blindly going with the flow of "adding contributions" and whatnots; such are the sad stories of software over the last 2 decades.

07:35

hm. I think that if anything, the data format (plain-text list of transactions) is worth preserving. the pile of code... not so sure, but either way, I agree it's worth examining if it's useful spending >10k LoC on the processing.

07:37

spyked: sure re plain-text, no argument at all there.

~ 2 hours 51 minutes ~

10:29

diana_coman: Thank you, the context for the plan missing for folks who hadn't been following the conversation had been gnawing at me a bit so it seemed like a necessary addition despite everything else stabilizing in the plan..

10:29

*

BingoBoingo will get to the YH comments, but first today there's a lot of fiatist blood on the market floors.

~ 15 minutes ~

10:44

no rush re comments from my point of view, for sure.

~ 3 hours 30 minutes ~

14:15

diana_coman here's the email thread with the refind guy. he went silent after my reply there. I've followed up with him once since.

~ 45 minutes ~

15:00

dorion: so what's your take on it?

15:02

diana_coman well, I tried to balance giving him context without inundating him with info.

15:03

dorion: sure, the exchange looks ok to me; at most clearly super-friendly on your side.

15:04

I could've cut out the V stuff and kept it to joining #trilema.

15:04

dorion: the q "what's your take on it" is re conclusion to this since you can't wait for him indefinitely or something.

15:04

dorion: neah, you did nothing bad/wrong there.

15:06

tbh to me it looks pretty much langyel-style anyway ie he wants to "answer your questions" and nothing more really, he's happy where he is and there's nothing else of any interest.

15:07

diana_coman right. http://logs.ossasepia.com/log/ossasepia/2020-01-14#1015230

15:07

Logged on 2020-01-14 21:57:38 trinque: if he works at canonical, I'm somewhat skeptical he'll align himself with burning down the world, but who knows.

15:08

the reason why I asked you though is that you should decide one way or another on it and set it explicitly in that report - ie tried & even followed up once more, nothing useful coming out of it though, end of it.

15:09

diana_coman in my mind there's space to strike the guy out and not his code. i.e. someone could take ownership of it, like musl.

15:09

dorion: heh, right; still, the attempt in itself is useful because now you have a concrete result to point to rather than "what I think it might be"; this is after all the crucial difference, between "did and here's the result" vs "thought it will likely go so and ugh, maybe..."

15:10

right, I see.

15:11

dorion: sure, so state that conclusion there at your tmsr os statement, that's all.

15:11

you're the manager there so that's precisely the sort of decision you need to make anyway.

15:12

diana_coman ok.

15:12

dorion: re that follow up stuff on musl list - do you mean follow up with them? what are you aiming for - ie something you want stated on their board/list/whatever or what?

15:15

diana_coman reinforce what jfw said, pour some sal in the wounds he left unsalted and make it clear that we're forking it.

15:15

salt*

15:16

diana_coman I just gotta do it.

15:16

dorion: nothing wrong with that if you want it done, sure.

15:18

as long as you don't entertain any misplaced expectations re their response, the rest is entirely fine.

15:20

right, it's more on the pick fights for exposure front as a means to grow contributors.

15:20

dorion: not a bad idea at that

15:20

those who have ears to hear will ear

15:20

BingoBoingo: do you have any qntra link for dorion to link there?

15:21

dorion: maybe take care with the grammar & spelling with that, otherwise who comes off as the pot calling the kettle illiterate in the minds of who might be on the fence?

15:21

dorion: apparently hearing ears are in very short supply but again, if you don't try, you can't know if there are any in the audience anyway.

15:22

*

diana_coman notices that jfw offered to spellcheck dorion's message before posting :D

15:22

sure, better than doing it after!

15:24

jfw will do.

15:24

jfw: the minds that discard the message because of less than perfect spelling though might be better off enjoying their fence.

15:25

jfw you subscribe to it first so you can then reply ? is that how it works ?

15:26

dorion, I don't recall if you have to be subscribed, though I am

15:26

diana_coman: projecting my own annoyances with modern sloppiness perhaps.

15:28

dorion: "You can also mail the list without subscribing by sending email to musl #T lists.openwall.com, but messages from non-members may be held for moderator approval."

15:30

( http://musl.libc.org/support.html )

15:30

jfw thanks, was looking just on the openwall site, thanks and subscribed.

15:32

jfw: well, it certainly can't hurt to make it error-free and moreover, certainly better to not leave them that easy option to "disregard", sure; just not much to do with any on the fence, that's all (and I do get your annoyance with sloppiness, oh my, how I get it).

15:32

*

BingoBoingo catching up with thread

15:36

diana_coman: gotcha. And yeah, for me those details make for either extra icing on the cake or extra dirt on the grave, not "wow that's an interesting point but he didn't cross a 't'"

15:37

aha, quite.

15:37

dorion: Aside from musl and refind, what else does TMSR have on the forking plate at the moment?

15:38

~everything?

15:39

jfw: eh, don't be *that* greedy! what, do you want to fork Microsoft's Office now?

15:39

windows!

15:39

yeah, pretty much everything that's needed to support the implicit clients.

15:39

jfw: Well, if I'm recording the news, being able to put the confirmed bozos on one side leaves everyone on the otherside twitching nervously... or not.

15:40

BingoBoingo needs concrete stuff, can't go "everything"

15:40

yeah. well an older linux and gcc come to mind as the biggest

15:40

Right. Linux kernel's bozo'ing is old, but...

15:41

BingoBoingo: effectively crystalspace and cal3d can be added to the list since it's not like anyone other than S.MG working with those versions /doing anything with them anyway.

15:41

cal3d theoretically is "active"; in practice..

15:41

diana_coman: tyvm

15:42

I invited the guy from cal3d to #eulora and there was a talk after which he faded silently away and that was that;

15:42

jfw dorion: Most useful are the door's you've already knocked on and you can swear Zher Modesty Bozo the clown answered.

15:43

BingoBoingo: here's the discussion with the cal3d guy in #eulora

15:43

(eulora) 2019-06-13 diana_coman: hi mp3butcher

15:43

BingoBoingo, as of now, musl and refind are the only doors that've been knocked on.

15:43

diana_coman: Thank you dorion: Aite.

15:45

if spyked or jfw genesis that ledger, I guess that would be therefore added to the list too.

15:45

BingoBoingo: for Gales I forked openbsd's ksh

15:46

jfw: Ty, because with the Bob Beck thing, the bozo'ing there's been done and can be linked

15:47

and since iirc djb was already declared dead we can add his works, of which I'm using daemontools, qmail, ucspi-tcp and djbdns

15:48

heh, how the list grows once one actually looks.

15:49

BingoBoingo: mpwp for that matter...

15:49

well, "wordpress"

15:49

python 2, officialy EOL now I believe

15:50

diana_coman, BingoBoingo the way I'm seeing it now is since everything in TMSR OS is under V, even if the refind guy -or any other project maintainer- expatriates, he'll be forking his current git/cvs/svn process to V.

15:50

*and* the king stepped down so who would we even approach on it

15:51

dorion: certainly.

15:52

jfw: python 2 indeed; the thing with this growing list is that all those should be genesised really, huh.

15:53

I did say so!

15:53

Logged on 2020-03-03 15:30:09 jfw: There's a long list of things I use but don't have genesis, but might as well chip away.

15:54

jfw: I know you know what you should do!

15:54

My plan is to start next week with creating a 'gypsy code signing' key then I suppose get the list together and into some kind of order.

15:55

This is also where the list of dependencies comes in.

15:56

jfw: how/why the gypsy code signing key? ie you are using them, aren't you?

15:57

diana_coman: because I have not read, much less understood them, which to me is what a V signature means

15:57

that "some kind of order" suddenly illuminates dorion's earlier vision of jfw the order restorer

15:57

Logged on 2020-03-03 16:43:09 dorion: pictures jfw breaking out ruler to strike ball throwers and restore order.

15:57

heh, I just meant ordering by priority or category or something.

15:58

jfw: I can see the point; it's still a V signature though, if with another key, heh; and still using them.

15:58

anyways, not to say to not do it, it's your choice; just to consider if and whether it's worth the bother of yet another key.

16:00

I think it's the right sort of bother: given all the code I ~have~ read or written it seems an important distinction as far as which category someone chooses to trust from me.

16:01

ok.

16:10

http://logs.ossasepia.com/log/ossasepia/2020-03-09#1020381 - I have no idea why it needs that much code either, but what appealed to me was the documentation (though iirc they tried to rewrite the manual for v3 and lost a lot) and flexible multi-commodity system. Not sure whether they found the best design for that, but it was more than I could have cooked up with a beginner's grasp of accounting.

16:10

Logged on 2020-03-09 08:43:19 diana_coman: spyked: well, no boost dep is certainly a gain; and yeah, I'm sure that the earlier the version the less of a monster it is; that being said, I'm either really getting old or something but I can hardly see the point of >10k LOC for what can be done precisely as wanted through <1k lines of cmd line tools, huh.

16:11

hey lobbes, re the reclaiming time from the saltmines, it occurs to me the coronavirus scare could be good opportunity to push for more work from home. 'never let a good crisis go to waste'.

16:11

*

BingoBoingo will put a draft here before publication for jfw and dorion to review. Will try to formulate some questions for further clarification as well.

16:12

BingoBoingo cool, I'll be near.

16:13

*

BingoBoingo not unhappy with the local situation where the dollar's down, but the peso's down even more for the time being

16:13

http://logs.ossasepia.com/log/ossasepia/2020-03-09#1020383 - I quite agree from experience; there's no point in having two record types there. I was building off the time log format from ledger but would do it differently if from scratch.

16:13

Logged on 2020-03-09 08:45:13 diana_coman: ftr re clock, my scripts don't bother with "end of task" - there's no end to anything, only start of next thing whether that's break or whatever; since time is spent anyway, the start of a new thing is by definition the end of the previous.

~ 16 minutes ~

16:30

lobbes: While you are getting pinged, I have been wondering... What do you want the landscaper to do and what is your USDA zone? re: http://younghands.club/2020/03/02/ejb-review-of-feb-24-mar-1-plan-for-mar-2-mar-8/

~ 1 hours 13 minutes ~

17:43

dorion jfw: Here's a first draft http://paste.deedbot.org/?id=UhCO I'm still thinking on some questions for you both so feel free to suggest some

17:57

*

dorion looks

18:03

BingoBoingo trb seems like a big one to include in 2nd paragraph.

18:06

also, even though bvt first genesis'd and patched 4.9 kernel and 2.6.32 isn't done yet, could include that as well.

~ 1 hours 13 minutes ~

19:20

ah there was also Eucrypt, part of which was a liberation of gcrypt from Werner Koch & co. right?

~ 25 minutes ~

19:45

jfw dorion: What does it mean that these things have been liberated? What have they been saved from?

19:52

BingoBoingo: "progress" I suppose

19:54

historical versions may quite literally be saved from destruction if they were in the hands of those who don't care to preserve their history, for the simplest part

19:57

Well, can you elaborate. The consequences of MP-WP's are visible in the field. It does what wordpress says they do, but it doesn't suck while doing it. The hammering of Cal3D and Crystalspace is rather well documented on Ossasepia down to the bones that are points.

19:58

another layer would be that in the world we're breaking away from, old code tends to stop working after a while due to churn in its environment, 'bitrot', which if you don't control the environment, forces you to accept whatever new changes have been piled on in the 'update'

19:59

(dynamic linking is one of the bigger mechanisms through which this works though not the only)

20:01

jfw: Alright, now take http://logs.ossasepia.com/log/ossasepia/2020-03-09#1020496 make it a stand alone statement (you can add in context from Gales) that can be used to bring in the Kernel's forking and the eventual forking off of the rest of the legacy devs that don't want to come along

20:01

Logged on 2020-03-09 21:15:47 jfw: another layer would be that in the world we're breaking away from, old code tends to stop working after a while due to churn in its environment, 'bitrot', which if you don't control the environment, forces you to accept whatever new changes have been piled on in the 'update'

20:02

Then, the focus that V puts on the changes / process and identity of the changer realigns incentives to favor smaller and more important changes

20:04

*

jfw pondering that request

20:05

jfw: How about I restructure it

20:06

Could you repackage the very good point from your last few lines into one quotable paragraph, bonus points if it cuts to the insanity approaching code any other way

20:11

...apparently not readily. What makes something more quotable / what's missing here?

20:11

jfw It isn't that I can't repackage your points. It's that it means something less if I do that, because coming from you it has authority.

20:12

jfw: All that's missing is something that fits nicely between blockquote tags

20:16

But mostly that the points are assembled into a paragraph.

20:23

In addition to building new, simple and comprehensible tools, we are moving to preserve historical codebases, keeping them maintained and operational, adding missing functionality (that which solves actual problems), and permanently revoking the power of previous developers to tax the users by forcing unwanted change.

20:25

BingoBoingo: does that capture it?

20:27

Prologue from http://trilema.com/2016/ideological-history-of-the-republic/ gave some inspiration there.

20:30

jfw: That does, there's one addition that can strengthen it. Append another sentence of the form: "We did it with... , We're doing it to... , and we're prepared to do it to anything else necessary"

20:34

There's probably a better way to do it, and it may not be necessary, but if the listing comes from you it... has greater authority than if I'd typed out the exact same sequence of letters.

20:40

We have a version control methodology built around the concept. We did it to key cryptographic tools including Bitcoin and GPG. We did it to the Linux kernel. We've proven the concept with a full statically linked Linux system bootstrapped from source. We're doing it to compilers, standard libraries, even firmware and bootloaders. And we're prepared to do it to anything that stands in our way.

20:45

lobbes

http://logs.ericbenevides.com/log/ossasepia/2020-03-09#1020465 << lol, this is a point. I think they even sent out one of those mass email memos on it

20:45

ericbot

Logged on 2020-03-09 20:18:48 dorion: hey lobbes, re the reclaiming time from the saltmines, it occurs to me the coronavirus scare could be good opportunity to push for more work from home. 'never let a good crisis go to waste'.

20:45

lobbes

http://logs.ericbenevides.com/log/ossasepia/2020-03-09#1020471 << I only want someone to come and cut the grass every two weeks or so. I'm in the Southeastern Zone so the grass comes in quick and thick, and to me the cost is worth not having to do it myself. I tell you, I've come to hate lawns (before I didn't care one way or the other). Next house I will just have, idk, clover or something. Dirt maybe.

20:45

ericbot

Logged on 2020-03-09 20:37:50 BingoBoingo: lobbes: While you are getting pinged, I have been wondering... What do you want the landscaper to do and what is your USDA zone? re: http://younghands.club/2020/03/02/ejb-review-of-feb-24-mar-1-plan-for-mar-2-mar-8/

20:48

jfw: tyvm

20:48

yw BingoBoingo. I only hedge with 'proven concept' on Gales because it's not properly V yet.

20:49

lobbes: Ah, I though maybe you were trying to resurrect the lawn in order to shop the house around

20:53

lobbes

BingoBoingo: yeah I thought about making it a more 'marketable' lawn, but I'm betting on it being a negligible difference between well-landscaped-lawn and simply-alive-grass

20:58

lobbes: Well, a little bit of nitrogen can go a long way, but if you're only cutting every other week... You're playing roulette with the rain.

21:06

http://logs.ossasepia.com/log/ossasepia/2020-03-09#1020514 - there you go. betcha they'll take your word for it and as long as you perform, they'll probably appreciate the proactive self-quarantine.

21:06

Logged on 2020-03-09 22:03:10 lobbes: http://logs.ericbenevides.com/log/ossasepia/2020-03-09#1020465 << lol, this is a point. I think they even sent out one of those mass email memos on it

21:06

ericbot

Logged on 2020-03-09 20:18:48 dorion: hey lobbes, re the reclaiming time from the saltmines, it occurs to me the coronavirus scare could be good opportunity to push for more work from home. 'never let a good crisis go to waste'.

21:14

http://logs.ossasepia.com/log/ossasepia/2020-03-09#1020508 http://logs.ossasepia.com/log/ossasepia/2020-03-09#1020513 - that's money. thanks for helping to draw it out BingoBoingo

21:14

Logged on 2020-03-09 21:40:46 jfw: In addition to building new, simple and comprehensible tools, we are moving to preserve historical codebases, keeping them maintained and operational, adding missing functionality (that which solves actual problems), and permanently revoking the power of previous developers to tax the users by forcing unwanted change.

21:14

Logged on 2020-03-09 21:58:14 jfw: We have a version control methodology built around the concept. We did it to key cryptographic tools including Bitcoin and GPG. We did it to the Linux kernel. We've proven the concept with a full statically linked Linux system bootstrapped from source. We're doing it to compilers, standard libraries, even firmware and bootloaders. And we're prepared to do it to anything that stands in our way.

21:18

dorion jfw diana_coman: I'll leave the latest draft here till the morning http://paste.deedbot.org/?id=BMzH

21:24

BingoBoingo: does Jacob Welsh get a WoT link? ;)

21:26

though you did link my blog, which DOES now have my own contact info, so nbd.

21:27