This article is written around one month after the events that led to break between Mircea Popescu and Stanislav Datskovskiy, so it focuses less on the conflict itself and more on the sadness situation. It is much shorter then its draft was even two weeks ago - others have already said much of the things I wanted to, much better than I ever could. As far I can see, both Mircea and Stanislav are happy about the outcome, so the conflict, after a month of observing, comes off as an improvement. It is easy to see what drives Stanislav in his work, just as how Mircea's approach to work would be almost incompatible with Stanislav's. Thus, I will focus only on a few things that I feel were not well addressed.
The point about constant retreat to downstreams is very real1, though I am not sure whether asciilifeform has anything to do with it - when reading discussions involving him, these things always came up when talking about distant future (when we finally have gossipd, etc.). However it is obvious how it is a problem for TBF2. IMO one of the lessons is that even if the underlying software is rotten, the actually interesting thing must still get it's share of attention and not stop the operation.
I do not think that the solutions by Stanislav were a problem at any point. The problem, however, in the context is: what was not built because of waiting for X? What opportunities were delayed and then lost in the long shelf? Seems to be a purely retrospective analysis sort of activity, I don't expect that anything would have to be torn out as a result of this analysis, because IMO what MP calls the poisonous offerings manifests itself as dead companies, not as shitty tools.
In this regard, I would say that software is useful as a leverage to promote useful changes in the environment, not as an inert isolated object of beauty3; so for example TMSR OS project makes sense to me as a product developed for S.MG, or for potential trinque-managed consultancy, or Trilema servers, or could have been developed for TFB, but not as a thing in itself. To what real-world problem will it get applied, and what are requirements for that use-case? Just building the new distribution "because we can" is not a good idea, and as far as a training is concerned - Gentoo is already producing potential distro-builders at the rates ancient Athens were producing boat rowers, so I don't think anyone here needs more training. What work plan trinque expects to make for it, I don't know, but I half-expect it will be something more just porting ebuilds to Cuntoo and adding static linking4. Anyway, the bigger part of work for distribution might be backporting5.
Finally, as far as 'would you even talk with an unperson?' questions asked in #asciilifeform are concerned, no, I do not mind talking to/buying stuff from asciilifeform, nor do I think that MP would drum anyone out of #trilema for doing so6. I did find the early discussion there to be too heated to participate, but now the atmosphere has calmed down, and the future will show how it will work out.
- "We can't go to war with the enemy, AK-47 hasn't been invented yet!!!" [↩]
- Would it not have been better to do a Keccak regrind of TRB ASAP (for
- But thoughts are given for action’s government; Where action ceases, thought’s impertinent. [↩]
- Some more notes: I haven't seen any statically linking distributive being alive beyond the first year or two for which its authors had energy (or the fly that was biting them had energy) to do the work, I expect in big part because this projects were driven by ideological purity arguments, and not by desire to apply the results somewhere. As far as a NixOS-like structure is concerned, for all its benefits it looks like an attempt to pull an owl of ...whatever it is, over the globe of Unix, so to speak - the human-incomprehensible setup of *[_]PATH is just bleh. [↩]
- Have a much older Firefox work with a bit newer ffmpeg (if X11 is ever added into the mix), fixes to GCC4, etc. [↩]
- Just as I don't mind getting drummed out if the bar gets raised above me or ratchet leaves me far behind - in any status - otherwise the time to leave for me would be now. [↩]
> Would it not have been better to do a Keccak regrind of TRB ASAP (for last year) to restart the flow of patches,
AFAIK Mod6 already had a Keccak regrind posted, but his blog seems off the grid at the moment.
> I haven't seen any statically linking distributive being alive beyond the first year or two
Didn't Alpine Linux had a working statically-linked distro? Anyways, I'm guessing the problem, or one of the problems there is that distro maintainers actually aim to keep up with upstream, which is just nuts. Let's say moving to latest Linux might make sense, if only to get things moving on newer ARM boards; but latest Firefox, GCC (though the same argument could be made there for architectural support) etc. are just as good as garbage.
Other than that I agree, supporting static linking adds to the risk and the time for spending many months on wild goose chases is gone now. I'm very curious to hear Trinque's take on where this whole TMSR OS thing can go.
@Spyked - Alpine is a musl based distribution as of their 3.0 release. I don't recall if they meaningfully did the static link everything deal.
The thing about Linux distros is that all of the ones with longevity have some sorta organizational apparatus behind them. It's kinda why TMSR needs an OS. The closest any OS distro came to sitting at the table was OpenBSD taking Mircea's charity and they went retarded for Microsoft days later.
Every distro interesting for the cleanliness they propose int heir marketing is defuct for a lack of organizational sanity.
And Gentoo is not a distribution. It's a build tool chain.
IIRC Alpine is musl-based, but not statically linked by default for everything. Buildroot seems to have an option to statically link everything, but 1. it also has a smaller scope (embedded systems), 2. this option is conflicting with others (C++ support IIRC).
That said, with a well defined scope, static linking should fly, perhaps even with only minor changes to Gentoo eclasses/GCC specfiles, but the cost of moving the remaining 10-30% of software to static linking may be not worth pursuing.
spyked is correct. The TRB Keccak regrind was published in January of this year. My wp-mp blog has been down since Pizarro, but am currently working on a replacement. Meanwhile, I have all of the blog articles posted to a temporary place here: http://www.mod6.net/blog.html
The TRB Keccak regrind article can be found here specifically: http://www.mod6.net/2019/January/13/keccak_regrind_noUTF8.txt
@mod6: I apologize, I missed your post. Though this is something that would make sense to have on thebitcoin.foundation as well - all the instructions are still using old SHA512 vpatches.
@BingoBoingo: Yes, organizational apparatus is important, true. Definitely puts TMSR OS above other attempts. Re Gentoo - IMO it is somewhere in the middle between more classical distros and toolchains like buildroot.
And yet, for all the self-evident ridiculousness of that statement once written out (and therefore read, as a foreign object), it is the very substance of every serious Civ player's gameplay. Isn't it ? Isn't this what the whole game of Civ is, "wait on war for research x" ? Sid Meyer's Parasitosis Of The Delaying Gratification Mechanism, aka "we can't have kids until we build the nursery-palace-planet-galaxy".
If it gets to live in the warm, embracing, moist recesses of the brain where the brain's cunt is, if it survives attempts at examination, if it manages to avoid ever being written out and there laughed at on its desperate, simply desperate attempt at finding its way back in...
These days I am rather persuaded this and nothing else is what's at the root of most contemporaneous English speakers' notdoings.
His approach to life is substantially what turned the 16th century Spanish empire (period title-deed holder of all land west of the Azores, with proper paperwork sent mediately from Christ-the-king himself) into the sad remnant that couldn't defend Florida from Jackson's merry bands of drunks and loudmouths : fierce defensive during the good times followed by "inxeplicable" failure in crisis. He will sit on any opportunity, indefinitely, and think of ways to defend it from threats he imagines, rather than fruit it ; if he were alone on a deserted island with a pubescent girl who was, by virtue of how nature works, very much in love with him, he'd lecture her on hygiene (lest improper palm leaf wiping ruin his future girlfriend / wife's twat), he'd build watchtowers to keep an eye on any possible rapists/pirates (and dedicate significant portion of his waking hours to guarding a deserted island) and he'd sort out the professional life of her eventual fifth born, preemptively. Then one day some half-drunk bucaneer will stumble onto the beach, and upon despatching the idiot the very next minute discover a twenty-something virgin, half-crazy with... what ? What's she half crazy with ? Ten years' worth of daily compulatory impregnation with the seeds of insanity, something like that ?
Sure, he sticks with like-minded people, so it's not clear if maybe it was not-him. It wasn't not-him.
I think this is both true and eminently well put. Let me add that trading shitty tools for dead companies is no kind of improvement! If there was a hospital that had a perfect process but which sadly also came with the guaranteed death of the patient, it'd be called a morgue, not a hospital -- notwithstanding how little claim any particular hospital found in the world might have on calling itself a hospital in the first place.
This begs the question, yes obviously things are useful in the way they are useful, not in the way they aren't useful. I would agree that software is chiefly interesting as a tool to change the world rather than as an embodiment of aesthetic perfection ; but I'm not quite prepared to throw out the coda altogether.
It seems to me software could be an object of beauty ; it is self-obvious that the makers of such objects of beauty may, by the very defintions of the terms, expect no kind of payment whatsoever to be earned through their performance -- if anyone pays anything for a piece of art, it's because of reasons at best indirectly related to anything like the author's "work". Not because making art's not laborious, but specifically because it is not work ; it further seems to me that in the case of software any personal claim to such beauty is thoroughly spurious and eminently ludicrous, much like would be the case for any proverbial Joe (for good historical reasons we're well persuaded any such moron'd have to be "American") deeming the beauty of the dodecahedron somehow his own. It's a purely Platonic sort of beauty, so to speak.
This is actually a strong point ; I shall have to think about it.