tmsr - Search: from:asciilifeform

194600+ entries in 0.047s

asciilifeform: but, there is the page table

asciilifeform: yes, you probably won't find anything past 40 or so bits soldered on your mainboard

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181488 << this is doubly lulzy considering that i just pulled my dead tree 'amd x86-64 programmer's manual vol. 2' off the shelf, sept. '02 edition, and it defines 52 bits of valid pointer in 64-bit word... ☝︎

asciilifeform: until not.

asciilifeform: and anyone who is found to be deploying such code, for whatever purpose, ought to answer for it

asciilifeform: you can't even win 'underhanded c contest' with them. they're ~screamingly~ bad ideas. ☟︎

asciilifeform: there is no conceivable reason to ever be doing any of the things described therein

asciilifeform: that article ought to be required reading for shitgnomologists

asciilifeform: gotta love those prison kettles

asciilifeform: 'harvard' is better in the same sense that it is better to be shot with a .22 than with a cannon

asciilifeform: where the latter is able to operate on a tiny portion of the former in any given operation

asciilifeform: 'von neumann arch' traditionally means simply the very idea of having separate devices called memory and cpu

asciilifeform: (or, for that matter, refer to bits in code word unless such was designated as a 'return 8-bit constant' instruction)

asciilifeform: physically had no way to 'jump into' data

asciilifeform: pic16f84, my favourite when i was a schoolboy, has 8-bit data word but 14-bit code word, in entirely separate address spaces

asciilifeform: (the starvation-cheap kind)

asciilifeform: used to be fairly common on micros

asciilifeform: 'On PICs and MCS51s, the two objects could actually be in different data spaces (e.g. RAM vs flash memory).' << correct. at least for the pic16xxx series, which i've used for many years. it is called 'harvard architecture'

asciilifeform: and l0l, it doesn't work terribly well.

asciilifeform: they are hiding from... me & colleagues

asciilifeform: 100% ?

asciilifeform: but hey, 'it works!!111!!1111'

asciilifeform: this has all the appeal of the russian prison teapot (consisting of a mug, two nails, and a mains socket)

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181475 << this is not even the worst known abuse. one book (forget whose) actually suggested 'just in time compilation' on arbitrary (!) machines by snarfing up binary routines from function pointers and memcpying them into a buffer, then (*foo)()-ing in... ☝︎

asciilifeform: (who wants, can write this. i've no strong desire to)

asciilifeform: but why.

asciilifeform: decimation: by emulating a node, etc

asciilifeform: decimation: sure

asciilifeform: when would i have tried

asciilifeform: decimation: i just now wrote the patch, l0l

asciilifeform: i will leave the possible variations on this theme to the alert reader.

asciilifeform: and whatever test (igprof, etc) can now be carried out stepwise (that is, after each added block)

asciilifeform: it is now also possible to elementarily ferret out certain kinds of leaks - e.g., does importing the same bastard block ten thousand times result in an increase in ram footprint ? if so, we have leak ☟︎

asciilifeform: blockchain telegraphy

asciilifeform: for an entirely unrelated and - today - uninteresting - purpose

asciilifeform actually wrote something not unlike ^ a few yrs ago

asciilifeform: (on a netless node)

asciilifeform: mod6: i, for one, am rather curious whether we still get 200M+ footprint when there is ~no~ mempool

asciilifeform: i'll laugh to death if it's from 'blockchain.info' or the like

asciilifeform: or even the commodore-64 rainbow.

asciilifeform: or the rain rainbow

asciilifeform: could be the apple-II rainbow

asciilifeform: will be interesting to compare their blocks with ours.

asciilifeform: http://kryptoradio.koodilehto.fi << via Vexual

asciilifeform: pretty basic

asciilifeform: l0l

asciilifeform: this will also shed some light on the effect of the mempool on resource consumption...

asciilifeform: ben_vulpes, mod6, mircea_popescu, et al: in case this was not clear, it is now possible for the first time to carry out FULLY DETERMINISTIC tests of therealbitcoin (memory consumption, etc) by denying a node a net connection and syncing from disk using 'eatblock'. ☟︎

asciilifeform: now whoever dares, can ask a node to crap out blocks n ... n+k, pgp-sign, bring them to node having blocks j ... n-1, and 'feed' them to it.

asciilifeform: 'Using 'dumpblock' and 'eatblock', it is now possible to move blockchains around via various direct channels (e.g., pigeon.)'

asciilifeform: quoting from the ml msg,

asciilifeform: (before anyone asks: doesn't require the weird asciification that using 'getmemorypool' to import blocks would)

asciilifeform: '(EXPERIMENTAL) Raw Block Eater'

asciilifeform: http://therealbitcoin.org/ml/btc-dev/2015-June/000105.html

asciilifeform: ben_vulpes, mod6, mircea_popescu, et al:

asciilifeform: achtung, panzers!

asciilifeform: (if anyone here is a regular reader of these snakepits and knows of evidence to the contrary, it would be interesting to see it)

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181377 << my current understanding is that the blackout in 'reddit' et al. re: therealbitcoin is more or less total. ☝︎☟︎

asciilifeform: the presumption is that a fella who is facing a return to life in hard prison might have an arse grenade, etc., or otherwise disinclined to 'go quietly'

asciilifeform: esp. if no witnesses

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181365 << my understanding is that this is sop in usa when the escapee is considered 'armed and dangerous' (tm) ☝︎

asciilifeform: http://lenta.ru/news/2015/06/29/flibusta << mega-l0l. ru dept. of net censorship blocks ~the~ w4r3z library, but not for just anything, but for hosting... bradbury's '451 fahrenheit' !

asciilifeform: (but i distinctly recall a 'pear' thread ages ago, where somebody pointed out that the two types can be distinguished by the presence of 'cervix pliers' on the business end)

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181360 << l0l, must've missed that thread! ☝︎

asciilifeform: (easiest method is to actually understand the changes, as one ought to in any case, and re-apply them by hand, then produce new patch)

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181387 << my igprof patch for therealbitcoin is almost certainly b0rked by the 'dumpblock' addition. but fixing it ought to be trivial - good project for any of the less-experienced folks here ☝︎

asciilifeform: and thereby reveal their position

asciilifeform: they can't resist

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181363 << even if rms no longer writes (and i'm informed that this may be the case in recent years) he is still tremendously useful as a 'fire magnet' for shitgnomes ☝︎

asciilifeform: afaik gcc does not support multiple ~levels~ of macro expansion (these are braindamaged c macros after all, not a lisp macro system with actual control over evaluation)

asciilifeform: http://log.bitcoin-assets.com/?date=30-06-2015#1181371 << almost inevitably the result of this simple experiment is quite unreadable. anyone who wants to try is welcome to: gcc -E ☝︎

asciilifeform: nah

asciilifeform not only reads, but sometimes starts with footnotes

asciilifeform: http://log.bitcoin-assets.com/?date=29-06-2015#1181273 << this sounds like a terrifyingly bad idea: if i run sync after mircea_popescu plugs his node back in, and end up with conflicting blocks, the very last thing i want is for it to silently 'fix' ☝︎

asciilifeform a bit biased, as he uses vertical displays

asciilifeform: http://log.bitcoin-assets.com/?date=29-06-2015#1181315 << very spiffy ☝︎

asciilifeform: !up ascii_field

asciilifeform: ;;later tell mircea_popescu http://www.loper-os.org/pub/turdsums/readme_s.txt http://www.loper-os.org/pub/turdsums/sums.txt.gz http://www.loper-os.org/pub/turdsums/sums.txt.gz.sig ☟︎☟︎☟︎☟︎

asciilifeform: ;;later tell jurov please nuke the size limit on turdatron

asciilifeform: haste makes waste (tm)

asciilifeform: erratum: where i wrote 'O(N^2)' ought to be O(N). complexity of fetching ~all~ blocks is O(N^2).

asciilifeform: the 'words' thing is asinine

asciilifeform: i still can't see any reason

asciilifeform: wtf does #b-a search not behave exactly like grep ?

asciilifeform: http://log.bitcoin-assets.com/?date=29-06-2015#1180703 ☝︎

asciilifeform: nah, here