log☇︎
1100+ entries in 0.094s
trinque: wherever you introduce the manifest, if it is not the root, in order to involve it in the flow of the tree, that patch must also edit some other file.
asciilifeform: mod6: i see it as moar of a bugfix ( a la mircea_popescu's fix of the db locks constant ) rather than troo patch, fwiw
mod6: I might not get to it this week, probably sometime in early july, but I'm going to roll in your aggression vpatch to the foundation patch set as well.
ben_vulpes: http://btcbase.org/log/2018-06-23#1829108 << the alternative, which would be a smaller patch, is a "setchangeaddr" RPC function. i'm leery of changing the call signatures of sendfrom and sendmany, but doing so might be The Right Thing nevertheless ☝︎
asciilifeform: Mocky: v rejects the traditional concept of 'merge'. therefore in order for a patch to continue through time, it must be not only correct, and simple not only to understand but to manually reintegrate ( see l0gz re 'regrind' ) .
asciilifeform: fella had enthusiasm, even some talent, but could not grasp the essential idea that a patch gotta be compact and touch a strict and justifiable minimum count of moving parts
Mocky: well not build order but patch ordering
Mocky: asciilifeform, yes. I'm still wrapping my head around v. My understanding is attaching name and trust to every patch with an explicit dependence tree and build order. But I've not grasped the details yet, or used except to build trb, or understood the src
asciilifeform: when ave1 comes back with patch for cross-x64 , it will be time to genesis the thing
asciilifeform: http://btcbase.org/log/2018-06-19#1826979 << phf i for one would not be opposed to 'rewind to 19 and patch as-needed', like we did with trb. ☝︎
a111: Logged on 2018-06-13 01:32 asciilifeform: phf: speaking of which, consider the easiest winner , if the anti-patch condition is absent -- a google shitmonkey who knows the hole already and 'wins' on the monday right prior to patch tuesday.
asciilifeform: phf: speaking of which, consider the easiest winner , if the anti-patch condition is absent -- a google shitmonkey who knows the hole already and 'wins' on the monday right prior to patch tuesday. ☟︎
asciilifeform: formulated simply as 'to get other half of prize, it will be the case that vendor update month from submission day, does not contain patch for hole'
phf: i don't think "can still buy and diddle of amazon in a month" is adequate test for "didn't leak the patch to google". but i don't think there's a procedure to test the goal in general (see absence of evidence above). perhaps you could restate the goal, but then whatever restatement i'm not sure it will be under the control of the participant. in fact as mircea_popescu pointed out, a restatement of this particular goal simply introduces a random element
mircea_popescu: "what is mgm, with a patch on the eye on the corner begging for nickles ? they repackage free shit idiots gave them in the hopes of '''making name for selves''' and pretend like they;re in business"
asciilifeform: (given as enemy will immediately patch when they get hold of the boojum, as crapple does)
phf: mostly a convenience since <patch name>/hunk/<path of hunk> works for all patches now, e.g. http://btcbase.org/patches/genesis/hunk/bitcoin/src/init.cpp
phf: i added a link on patch pages to "split hunks", which allows to render any patch the way mp-wp genesis is being rendered, e.g. http://btcbase.org/patches/genesis?inlinep=false
deedbot: http://thetarpit.org/posts/y04/072-trilemabot-i.html << The Tar Pit - trilemabot [i]: introduction and self-voicing patch proposal
phf: i can obviously fix btcbase to be more useful (i.e. continue to aid the patch exploration) in cases where a patch is big, but in general a 9mb patch seems to go against the whole fits in head
phf: right, that was the original idea, as evangelized by ascii. btcbase patch viewer is designed with that idea in mind
phf: i'm not sure how practical that makes patch page though, it definitely doesn't open on my x60. i could perhaps introduce some split mode, where patch page only lists the hunk filenames, and you need to click on hunk to see the contents
phf: http://btcbase.org/log/2018-06-03#1820367 << there isn't one, there is a peculiar bug that seems to only manifest in my lispworks dev environment. things actually work on production, though the patch page is 20mb ☝︎
asciilifeform: after that : to patch the gcc , to cure http://btcbase.org/log/2018-05-30#1819935 ill. ☝︎
mircea_popescu: another thing that would very well go into the comment section there would be proposals for patch naming conventions.
trinque: actually looking at it, could probably use the patch's name.
a111: Logged on 2018-05-01 16:25 phf: ight have time to sit down with v.pl before mid may. i can also just remove the right hand side of vtools for now, since this new complexity is coming from an experimental v graph anyway. i've no idea though if people are using a sha512 vtools in preference of awk vdiff / gnu patch.
a111: Logged on 2018-05-30 15:14 asciilifeform: !Q later tell ave1 your newest tarball barfs : http://p.bvulpes.com/pastes/5FS6l/?raw=true << gmp demands a patch that can't be found, hangs with prompt
lobbesbot: ave1: Sent 3 hours and 12 minutes ago: <asciilifeform> your newest tarball barfs : http://p.bvulpes.com/pastes/5FS6l/?raw=true << gmp demands a patch that can't be found, hangs with prompt
ave1: asciilifeform, I'm sorry for this. I ran the build multiple times so I'm not sure how the patch found it's way into the last release. Fix is on it's way. (fix is to remove the patch, gmp-4.3.2-2.musl.diff from the patches directory)
a111: Logged on 2017-05-19 02:04 asciilifeform: '2017-02-21: The patch is surreptitiously snuck into the Monero codebase in pull request #1744. It is kept secret to prevent it being used to attack other CryptoNote coins.'
asciilifeform: !Q later tell ave1 your newest tarball barfs : http://p.bvulpes.com/pastes/5FS6l/?raw=true << gmp demands a patch that can't be found, hangs with prompt ☟︎
phf: ave1: ^ patch and signature links 404
trinque: get the defpackage in there, test, then when you're satisfied, post a genesis patch on your blog.
a111: Logged on 2018-05-25 15:47 phf: so an intermediate step that someone else could perform is to take your rockchip gentoo, generate new rsa pair, sign the kernel with pub, patch google's uboot with priv and get a clean booting rockchip gentoo setup, without accidentally bricking the device? (while still retaining known amount of google in the system)
trinque: spyked: so sounds like you're going to patch your items (in as many small, easy to read patches as possible) atop ircbot?
a111: Logged on 2018-05-25 13:29 phf: back in the canonical log days, i'd patch the name in the logs, but since we're past that i think it can be left as is, unless there are objections as far as log authenticity.
trinque: sure, but whether the same patch hunk ended up in two places is computable, interesting in a patch-viewer sense, not in a vtron-operation sense
spyked: mircea_popescu, then if they're separate branches and if I want to make purely hypothetical spykedbot that does both? would end up with patch with two antecedents, should v press that? should I regrind?
a111: Logged on 2018-05-02 11:03 spyked: speaking of which; to all ircbot users: I have a patch proposal for ircbot (and possibly logbot). the problem: nickserv authentication makes a distinction between "nickname" and "user". this allows e.g. to group multiple irc bots (with different nicks) under a single username and cloak. so my proposal is to add a new *optional* "user" slot to ircbot and use it for auth instead of "nick" when available
mircea_popescu: it means this : that instead of explicitly communicating state, a. "i r making rss bot" ; b. "hey trinque wtf, ircbot doesn't do voicing ?" ; c. "o well, stopped making rss bot, making voice module for ircbot to be able to make rss bot later" and THEN 1. pushing a voice patch to ircbot BEFORE d. "back to making rss bot"
phf: so an intermediate step that someone else could perform is to take your rockchip gentoo, generate new rsa pair, sign the kernel with pub, patch google's uboot with priv and get a clean booting rockchip gentoo setup, without accidentally bricking the device? (while still retaining known amount of google in the system) ☟︎
phf: back in the canonical log days, i'd patch the name in the logs, but since we're past that i think it can be left as is, unless there are objections as far as log authenticity. ☟︎
asciilifeform: let'em get raped plenty while birthing patch.
a111: Logged on 2018-03-01 13:52 spyked: anyway, comment was that I managed to compile and run vdiff with small mods; error: http://p.bvulpes.com/pastes/BiBTI/?raw=true and fix patch: http://p.bvulpes.com/pastes/9mOiz/?raw=true (tested this with the generated vdiff); I can try to link this reply later in a comment to test.
mod6: in a pinch -- you could always try to just : `patch -F 0 -E -p1 < mp-wp_genesis.vpatch` in some directory and just inflate the thing.
trinque: esthlos: there's also no sense in giving the thing a version number. the "version" is the patch the operator pressed to.
trinque: it'd be bad form to try to throw a genesis patch "over the wall" for this without taking the time to have the necessary threads.
hanbot: danielpbarron, hey, why not patch on it?
hanbot: hey danielpbarron (or any other mp-wp users with 'classic' theme goin'), assuming you're using mp's antispam, wouldja mind pasting your themes/classic/comments.php for me so i can see how you're handling the $suffix? differs from the 'default' i'm using, but i'd like to include a fix for it in a comments patch.
hanbot: and that'll be mp-wp's first patch, eta this week.
trinque: if I tell the thing to press to a signed patch, it should press, history file edited or not.
a111: Logged on 2018-01-23 14:28 mircea_popescu: esthlos it is not standard procedure ; the emerging consensus is to have a dedicated philosophy file which a) all patches must touch (by protocol) ; b) contains comments as to the patcher's state of mind and c) contains one line per patch uniquely identifying it, machine generated. the format's not fixed yet, but as phf is working on a new proper vdiff it's probably going to coalesce around a variant of whatever he uses.
trinque: sections of a given patch will yes, be parented on the previous, in a straight line, but this does not at all mean that the patch doesn't *also* have other parents, following the antecedent lines of the other files
trinque: it is a fact about correct operation of a V, and perhaps a V client does something to help the operator along "hey you didn't edit the history file, wild patch!" but does absolutely nothing differently. it's just another file.
trinque: this oughtn't be enforced by the V implementation. operator might fully intend to *not* include a patch in the formal history of the project
esthlos: now let's say that the new philosophy file contains hash of all directory contents in lexographic order, onse per patch. As far as i can see, this forces the tree of vpatches to be strictly linear, since latest patch depends strictly on single previous patch
lobbes: hdd only, but I got the aggressive patch applied, so will see if it can hack it
lobbes: I tried to press on the first pass with asciilifeform_aggressive_pushgetblocks.vpatch, and it -seemed- to work. I essentially followed the trb how-to ONLINE build, but then downloaded the aggressive patch and seal from btcbase/patches
diana_coman: lobbes, indeed; pushgetblocks patch seems to keep hdd node at the top without any problem
lobbes ended up getting a cheap dedi out in kansas or w/e for 25 bezzlebux. Soley for trb-ing. Drawback is it has hdd, but going to try and use the aggressive pushgetblocks patch. Seems like diana_coman had nice results with similar experimentation
a111: Logged on 2018-01-23 14:28 mircea_popescu: esthlos it is not standard procedure ; the emerging consensus is to have a dedicated philosophy file which a) all patches must touch (by protocol) ; b) contains comments as to the patcher's state of mind and c) contains one line per patch uniquely identifying it, machine generated. the format's not fixed yet, but as phf is working on a new proper vdiff it's probably going to coalesce around a variant of whatever he uses.
spyked: current patch draft is at http://lmogo.xyz/v/patches/ircbot-nickserv-auth-add-user-slot.vpatch ; and sig http://lmogo.xyz/v/seals/ircbot-nickserv-auth-add-user-slot.vpatch.spyked.sig . I'm not very sure that my changes to make-ircbot are proper, so I am eager to hear your feedback
spyked: speaking of which; to all ircbot users: I have a patch proposal for ircbot (and possibly logbot). the problem: nickserv authentication makes a distinction between "nickname" and "user". this allows e.g. to group multiple irc bots (with different nicks) under a single username and cloak. so my proposal is to add a new *optional* "user" slot to ircbot and use it for auth instead of "nick" when available ☟︎
a111: Logged on 2018-05-01 16:25 phf: ight have time to sit down with v.pl before mid may. i can also just remove the right hand side of vtools for now, since this new complexity is coming from an experimental v graph anyway. i've no idea though if people are using a sha512 vtools in preference of awk vdiff / gnu patch.
mod6: instead of $P, you use $patch
phf: ight have time to sit down with v.pl before mid may. i can also just remove the right hand side of vtools for now, since this new complexity is coming from an experimental v graph anyway. i've no idea though if people are using a sha512 vtools in preference of awk vdiff / gnu patch. ☟︎☟︎
a111: Logged on 2018-05-01 15:05 trinque: yeah, could remove press head entirely and press all it can, and operator controls via patch, seals, and wot, eh?
mircea_popescu: mod6, they're "unrelated" in the sense that not all files are touched at the same time, by each patch -- principally because we favour small patches.
trinque: yeah, could remove press head entirely and press all it can, and operator controls via patch, seals, and wot, eh? ☟︎
trinque: why did you have to comment in files unrelated to the makefiles patch?
trinque: your solution in the makefiles patch was to comment in unrelated files, which was inelegant.
a111: Logged on 2017-12-27 04:33 mircea_popescu: how about a convention whereby all new genesises must contain a manifest.genesis file, which file will be constantly patched on each patchj, no exceptions, by adding a line which reads : "This is patch #x and the codebase hash is blabla".
asciilifeform also curious re why the port required 300kB of patch, incl to .adb/.ads of gnat, but possibly this will be answered later
mircea_popescu: "here's a black box", howsoever implemented in practice, "really lengthy patch you don't have the time to read" or "my poker bot" or whatever, is not-this.
a111: Logged on 2018-04-29 14:58 mircea_popescu: yes, it was. to summarize, careful what you do with your patches, in the following sense : if diana_coman doesn't CONTINUE your patch, it will have to be reground later. as she's a human being with other problems and interests than folding in your alt universe, it helps IMMENSELY if you make your patches small and readable.
mircea_popescu: yes, it was. to summarize, careful what you do with your patches, in the following sense : if diana_coman doesn't CONTINUE your patch, it will have to be reground later. as she's a human being with other problems and interests than folding in your alt universe, it helps IMMENSELY if you make your patches small and readable. ☟︎
esthlos: aha, so if I want to patch "as close to genesis as possible" the mpi lib, AND I want this patch to be used in diana_coman 's proggy, the ONLY choice is to patch current leaf of diana_coman 's proggy
mircea_popescu: esthlos, it's very centrally how the v mechanism is intended to / works : you get to opt which patch trees to continue, much like in the case of bitcoin.
mircea_popescu: esthlos, you wouldn't make it a patch on genesis unless you were trying to signal your disdain for diana_coman 's work so far.
esthlos: if I make decrudification a patch on the genesis, probably requires no changes to eucrypt vpatch tree. but if it's a new genesis, tree breaks, no?
esthlos: cool, I may make a patch which rips it out.
phf: yeah, it was a combination of factors. his logging facility patch received a lot of criticism, though i now regret voicing mine. it's a bit of a bike shed problem, everyone has their own idea of how to do it "right", but it's not done to this day.
mod6: <+trinque> mod6: I've got a sendrawtxn patch sitting over here << actually, i should mention that i /did/ create a trb sendrawtransaction (just this one rpc call) vpatch late last year, but never sent -- really wanted the rest of the gang 'create', 'sign', etc.
trinque: mod6: "patch exists" is no reason to not go exploring on your own
trinque: mod6: what was wrong with polarbeard's exactly? my patch is more or less his.
trinque: I'm blocked on not having the manifest, so next on my plate is to regrind every patch with a manifest entry.
trinque: (got the walletless patch too; trb node on pizarro is walletless)
trinque: mod6: I've got a sendrawtxn patch sitting over here
deedbot: http://qntra.net/2018/04/microsoft-again-introduces-new-vulnerabilities-via-update-this-time-in-a-meltdown-patch/ << Qntra - Microsoft Again Introduces New Vulnerabilities Via Update: This Time In A "Meltdown" Patch
mircea_popescu: asciilifeform, well, ideally patch random.c into a fg streamer.
mircea_popescu: so you can waltz into dc installed "linux", patch it, and it either has /dev/random perma-blocked or else fg-streaming.
mircea_popescu: well if the kernel can't be patched then a patch won't help.
mircea_popescu: yeah. anyway, i'm thinking the best approach would actually be a kernel patch to destroy the extant random/urandom bs and replace it with fg
diana_coman: is there a tmsr patch/recipe for feeding /dev/random from fgs? my search of logs yielded only http://btcbase.org/log/2018-02-09#1783118 ☝︎
lobbes: I was having issues with the logbot init step (wasn't pulling the patch into patches)
trinque: mircea_popescu: the patch of dirt and wait was exactly my idiot mother's plan
mircea_popescu: so your plan for the ~rest of your life is to sit on a patch of dirt in ok and basically wait ? ☟︎
ckang: hmm so its like a patch ?
trinque shipped a patch that'll retry fetching the key