16300+ entries in 0.108s
mod6: But since that was an utter failure,
I'm just trying something different because of /dev/disk/by-uuid/ or whatever it is.
mod6: The days leading up to the blog post,
i used nothing but names: /dev/sda{1,2,3} and eschewed UUIDs totally.
mod6: Oh,
i totally felt the same way.
mircea_popescu: from what
i gather actually setting the correct path (ie, /sda3 or w/e it actually is) should really do it.
i don't expect you can just unilaterally set uuids, gotta make them work from the other end too first.
mod6: So perhaps im kindof on the right track - use UUIDs where
I can, and build an initramfs... however, haven't had a lot of success with building the initramfs yet.
I may have to fight through that. But that's the latest update.
mod6: But
I keep getting failures that say "Failed to compile the "all" target. No matter what
I do it seems.
mod6: Then
I got on the track that perhaps
i do need to have an initramfs.
☟︎ mod6:
I think that
I might be on the right track here though, and
I did try a few other things after reading some documentation. For instance, after the above,
I changed /etc/fstab to use only UUIDs instead of '/dev/sda{1,2,3}', and then tried that. Same problem essentialy.
I also found about 'PARTUUID', which is supposed to help in certain circumstances. Nothing has worked yet...
mod6:
http://p.bvulpes.com/pastes/BoTML/?raw=true << Ok, so after installing Cuntoo,
I did what
I said
I'd do, which was test editing the append section and throwing a UUID in there instead of 'root=/dev/sda3'. It didn't work,
I did get a kern dump.
mircea_popescu:
i'd like a purpose made item, fit in head as such, can keep for later.
mircea_popescu:
i've been thinking about how to correctly construct a calling test for this purpose, but
i confess nothing
i have yet is passing muster. if anyone wants to step in.
mircea_popescu:
i'll still want her to fill in the rest of the table ; but tentatively yes. looks like
i'll be withdrawing my objection to bvt 's original comments (
http://btcbase.org/log/2019-02-12#1895233 ) sometime tomorrow. in which case all the better, we just do that and good riddance.
☝︎ mircea_popescu:
http://btcbase.org/log/2019-02-13#1895836 << so on the basis of this better table neatly presenting data
i'm concluding that a) serpent run indeed takes 2.8 us or so ; b) timing data converges within 1/3 s test runs or so ; c) these statements equal to foregoing earlier items which are thus retrospectively deemed correct and finally, and most importantly d) tentatively it seems sjlj adds no measurable time delay on running co
☝︎ mircea_popescu: asciilifeform
i suppose ; but worst time of all for moments of clarity.
mircea_popescu: elderly father who whole life hammered kid with better notions and interests to "you must be emperor of rome", if eventually presented on death bed with circumstantailly obtained, wholly worthless emperorship much likelier to declare "he was right all along" than to see "holy shit what an idiot
i've been"
mircea_popescu:
i suspect though that it's unlikely to really happen. yamamoto went down indignant like the anti-chucker "never ~really~ went down" etc.
mircea_popescu: no no,
i didn't mean re crypto.
i meant, whether he realised he fucked up.
mircea_popescu: there's also some possible in erly irc finish punklands ; my fin-enabled green brasiers are scant,
i'm not excavating it very quickly.
mircea_popescu: possibly on bbs ; though
i think you beat me to punch, dug out naggum.
mircea_popescu:
i don't really expect it, because you need specific technologies to enable it.
mircea_popescu: item was deeply lulzy from a lot of perspective ; but
i will point out and udnerscore that no one was well equipped to appreciate that lulz prior to this here republic's senate discussions.
mircea_popescu: asciilifeform anyway, the fault for "there's exactly 1" lies not-with-me!!!
i do not as much as keep it secret ; or not make a point of it ; or anything else.
mircea_popescu: so, yes,
i claim first mover, and for good (and well fucking documented) reasons. cuz it's what it is, what.
mircea_popescu: as these things usually go ;
i shall limit self to pointing out that time the ny times copied my paywall model.
mircea_popescu: in any case --
i claim dibs for trilema re leadership in this whole "defection" business. WAY before femstate media reported on such a thing, it was on trilema.
mircea_popescu: why, because then you'll do what you didn't do last
i build you the list ?
a111: Logged on 2018-08-28 22:27 mircea_popescu: and
i do mean ~every single last one~.
i talk to every single chick on fetlife, meaning EVERY SINGLE ONE. that's the job of existence. nothing else passes muster.
a111: Logged on 2018-12-12 19:26 mircea_popescu:
i expect if attempted it'll immediately run into the same problem pizarro is encountering, whereas bois will do ANYTHING WHATSOEVER, no matter how patently stupid and laughing impending beheading in the face, just as long as it's NOT "talk to a lot of people".
mircea_popescu:
i understand why the "
i'd like to talk to as few people as possible" is appealing to the fucktard^H engineer mindset. nevertheless, the only way to win this is to have the decisionmaking atomized as much as possible.
mircea_popescu: asciilifeform just about. plus a ridoinculous flavour whereby "white sox won cup because
i, joe fatass, was screaming in a bar!!!"
a111: Logged on 2019-02-13 18:41 asciilifeform: historically
i found that o3 is moar or less unusable for benchmarking anyffin, it risks to 'optimize away' unexpected pieces ( for instance, quite possibly your loop-and-mod-2 )
diana_coman: no hurry,
I'll get back to it tomorrow morning anyway
diana_coman: it would have at most the path
I suppose, if it's hardcoded somewhere
diana_coman:
i.e. where/if the trouble is not somehow just on my machine (though
I got it on 2 machines...)
diana_coman: re paths
I suppose a simple grep -r "libmpc" would show perhaps duplicates if nothing else?
diana_coman: at least there are no more surprises of huge differences in timings; but
I'd still test also with some exception handling since that's supposed to slow sjlj down
☟︎ diana_coman: grrr,
I did NOT compare between diff machines
a111: Logged on 2019-02-13 18:41 diana_coman: anyways,
I'll compile the dataset and publish it in a bit
diana_coman:
I guess until now
I always built ave1's gnat on a machine that had adacore's gnat installed with all the paths to libs like that quite standard so possibly that's why it never failed
diana_coman:
I'm trying to rebuild it because current build doesn't have sjlj
diana_coman: t shows that it is looking for them in the wrong place and more specifically in the place where they were on the previous system where the working gnat was built! so hm, where and why does it store that path and how do
I point it correctly?
diana_coman: ave1 or anyone else more experienced in rebuilding ave1's gnat with a previous incarnation of same:
I'm trying to build using the scripts in ada-musl-cross-2018-09-24.tgz on a machine that has as only existing and perfectly working !) gnat a previous ave1 gnat version;
I ran as the readme says simply ./build-ada.sh absolute_path_to_dir but the whole thing fails because it doesn't find some libs such as libgmp.la; a closer look at the outpu
diana_coman: anyways,
I'll compile the dataset and publish it in a bit
☟︎ diana_coman: at that specific shot
I was exploring esp given the previous issue of "too long time"
diana_coman: mircea_popescu,
I found the key misunderstanding there (in addition to comparing different machines and all that): the 1.25value is when serpent is fully optimised for time
i.e. it goes from ~8s to 1.25 for full 22 loops 1..10, mod 4; sorry for the confusion there,
I was still in exploration mode.
diana_coman: anyways, back to it: for one thing you were comparing there numbers from different machines and
I get the impression you got the "4 runs" in the 2nd report to mean different loops while they were not (it was same thing, repeated)
diana_coman:
I got stuck here in traffic earlier too, hence my being late, sorry;
diana_coman: mircea_popescu,
I think it was a mis-communication really
mircea_popescu: well now
i have to step out ; diana_coman do some more experiments, get some more data, we gotta figure out wtf this is.
a111: Logged on 2019-01-03 19:24 stratum: Right now, for the billions,
I think it is probably better than nothing, just like easily popped household locks.
mircea_popescu: asciilifeform incidentally and unrelatedly, since
i'm fixing this other article : do you happen to recall the trilema piece where "middle age crisis" ie, dood getting rid of older wife hooking up with younger girl is explained in terms of "well, she won't be as fucking annoying" ?
mircea_popescu: asciilifeform basically, we found
i can't math ; that aside, we found that in one context serpent takes ~3us, and in another ~0.3us.
diana_coman: because the clock thing is not very precise; on 4 runs of same thing
I got: 1.25, 1.27, 1.29, 1.28
diana_coman: given that it's still not that long,
I'd do at least an avg
mircea_popescu: asciilifeform
i think her work above provides us with some useful data : it turns out it is reasonable to expect timings converge for measured intervals of at least 1/3 of a second, on the basis of the above.
mircea_popescu:
i might be the weakest person in the world for basic arithmetics.
diana_coman: so
I'll post the testing code with full set of loops in a bit and then go and set up another run too
diana_coman:
I think a gradual approach (
i.e. run it & record results gradually increasing the number of loops) might be useful meanwhile
a111: Logged on 2019-02-12 23:41 mircea_popescu:
i'm still waiting for diana_coman to return.