ben_vulpes: not only for mission-critical shit like therealbitcoin, but even my own work.
ben_vulpes: no, that's the point i wish to make. you're conflating github with git.
mircea_popescu: <trinque> the idea needs further development before anyone executes << he has a point, and while attempted executions would prlly be useful iterations, it IS a timesink.
williamdunne: ben_vulpes: I don't know if this exists outside of GitHub (I rather assumed it does) but when you fork a repo so its your own version of it, you make changes to your own copy and then request that your changes are merged into the core repo
williamdunne: One of those is a social issue, the other is solved by not accepting pulls from dickheads, no?
jurov: anyway, primary is there's verifiable autorship and readable content of patches, plus clear repeatable way to combine the patches
trinque: from there, perhaps no existing tool satisfies both
trinque: having the canonical representation plaintext and readable I think is a fine thing
jurov: and i think other was that git stores them in opaque database. you cannot go to .git and start reading
jurov: i don't get it. "by content addressable" i understand somthing like putting sha1 sum in the URL
mircea_popescu: asciilifeform so it has no python ios the idea ?
ascii_field: because this borders on the utterly ridiculous.
ascii_field: mircea_popescu: the apxs ... thing fails with wsgi_python.h:24:20: error: Python.h: No such file or directory and half a meg of other crapolade
Adlai: or did the install not involve running anything?
mircea_popescu: FROM A RANDOM SOURCE ON THE WEB
mircea_popescu: ascii_field obviously you have to first download the mod.
ascii_field: and cannot be retrofitted without being clobbered by cpanel the next time it is used to make any alteration
ascii_field: or for that matter anything other than pgp
ascii_field: mircea_popescu: how -the fuck- does one run a python www proggy on cpanel box
trinque: ah the urbit thing
jurov: and when she was sober, there was no dog concern?
Adlai: the typos resurface thus zzz
Adlai actually spend the alcohols chatting with a dark-skinned lady who was 100% coherent despite also partaking of alcohols
mircea_popescu: jurov i don't see it either, but! gotta ask.
Adlai: ok, you sign some other dude's patch. same diff
mircea_popescu: some other dude wrote.
assbot: Logged on 28-04-2015 20:28:40; williamdunne: Maybe there is a good reason why this would not work, but could you not just generate an XML file that contains the names of all the files, a hash of each file, and then sign the XML? Could be automated fairly easily and would work on top of existing solutions
trinque: jurov: calm down buddy; I don't have the one true way
trinque: relational algebra carries the stink of SQL, and this is unfortunate
trinque: the mistake was probably "non-relational
jurov: go learn from these mistakes
jurov: also, KDE folks tried it in non-relational way, something kinda works but they managed to annoy everyone in the process
trinque: but if the AST of the queries is represented as data you can find that out by joining it against the schema
trinque: obviously if information was lost in the change, you cannot
trinque: lets say you alter a table with some operation; were that operation data, you could programmatically alter views against the old version to be compatible with the new, to some extent
trinque: mircea_popescu: what I was referring to above is like the system catalog in SQL
trinque: the point here is missed; I'm probably not articulating it well
trinque: I can also easily produce a new table which materializes the output of some view
ascii_field: there's perhaps half a meg of generated 'do not edit!!!111!!!' crapolade in every config
ascii_field is blithering at the sheer amount of litter that cpanel leaves in every conceivable aspect of system
BingoBoingo: <mircea_popescu> it requires a hard guarantee you will never want more than 10 categories at any level. << No, only on the top level... After that you wade into the horror of Cutter numbers
mircea_popescu: of the leaves.
mircea_popescu: including the leaves.
mircea_popescu: he makes them all a tree.
trinque: so operations against the system must also be data
trinque: and you can easily with something that isn't braindead SQL imagine a model where the canonical representation can be changed, and new views derived from the operation taken to produce the new representation
trinque: relational algebra is the best model I'm aware of for deriving all needed representations of data
mircea_popescu: williamdunne it's exactly what it iwas, and exactly why i asked him the question.
williamdunne: Guessing some people wanted to kick it old school and not move to the re-write, and there was enough of them that apache updated got called apache2, but thats just a guess
trinque: views and actions against the reflection layer
mircea_popescu: take the dewey decimal system.
ascii_field: because there was a mega-rewrite once
mircea_popescu: ascii_field but while we're doing the "what dope" thing, why the fuck is apache "apache2"
trinque: the answer is probably not for example "build all into postgres" as we were doing
trinque: we were trying to build this at my last job, and fucked it up entirely with braindead concepts from the web space
trinque: in the same way I declare data
trinque: it needs a reflection system; I want to interrogate the system about its structure
trinque: "hm I didn't like that change I made to the system" << rollback
trinque will yet have to be persuaded that the relational model does not solve all this
trinque: why is the standard location not enforced?
ascii_field: *why* is it not in the standard location ?
trinque: the lack of a turnkey solution here is absurd, just a shitty industry
trinque: every bakery should have their own software team because no real scotsman would pay someone else to do that
jurov: i have everything built as nginx OR apache(never both of them) in extra process, PHP/Perl/Python in extra process(es)
ascii_field: i was even about to use the apache (all i need is to proxy to my proggy which has own http stack) but then i find that there is no /etc/apache2
mircea_popescu: the fact that i own a gas station does in no way alleviate the problem that laurel&hardy cars.
mircea_popescu: dude. for the love of budha
trinque: this is why I'd like a small board for the embedded space that does very little, and runs a lisp
mircea_popescu: there ISNT a bmw here. it's varnish or nginx. that's it. not like i'm running unheard-of-software
mircea_popescu: see jurov, the thing is... i have the OPTION to not own obscure exotic cars whose maker went out of business shortly after starting because he sucked.
jurov: i have 14y old daewoo, we can split "never ever" hairs but the end result and frustration is the same
mircea_popescu: "sorry honey, can't come to dinner, i'm apparently unable to find the entry to the brooklyn bridge today"
jurov: oh, they did, until very recently
jurov: there's this saying about inspiration and perspiration, you know
williamdunne: trinque: reasons they gave
mircea_popescu: and well... spend another 4 hours debugging that.
mircea_popescu: ok, so the sad story of my wasted day : first, instlaled varnish in 10 minutes, but then spent 4 hours trying to get ip forwarding.
williamdunne: jurov: "Because potentially it could make money, its an excuse to learn more tech, and there is little risk in terms of what I could lose by doing it"
trinque: on the other side sure, if you've got goons (read: massive govt) to beat payment out of people, loan sharking is great
trinque: aside from the foolishness of this from an investment perspective, it's not helping the guy any
trinque: the recourse for that person is to go find someone for whom to tape boxes shut
williamdunne: jurov: Did quite well on the other platforms until the SEC stepped in, not particularly large time investment needs to be made.
mircea_popescu: mike_c yeah, im a day wasted in here. i could waste another one. sure.
jurov: like, parents invest into you nd then expect to get a "reasonable" part of your income
williamdunne: The entire point is its meant to be risky as shit
williamdunne: Indeed, I'm not talking bout some sort of service where 100 people deposit money and then a "money manager" decides who to invest in
trinque: no that's it; the more you're willing to sell your future the higher risk "investment" you are
williamdunne: And thats part of the risk
mircea_popescu: im not sure you understand the difference between natural and constructive persons
williamdunne: trinque: Because people live on a different timeperiod to the company in many cases
williamdunne: The operator, its not like they're selling 40% of £40,000/year, they're selling 40% of whatever they make, with no guarantees of said figure
trinque: who doesn't let them
williamdunne: That exact problem could exist with regular stocks. Whats the incentive for Exxon to keep generating revenue? Start up, sell out, cash out, bro down
mircea_popescu: this was kinda popular with "artists" thinking themsleves above prostitution on myfreecams and etsy.
mircea_popescu: http://log.bitcoin-assets.com/?date=28-04-2015#1113919 << and what is then your remaining incentive to realise any future income ?
jurov: the greengrocer prolly caused mass development of explosive gases
cazalla: http://log.bitcoin-assets.com/?date=28-04-2015#1113823 <<< but the grocer isn't engaged in things similar to the mafioso such as robbing other businesses, so i don't see how the analogy fits.. if ya gonna throw rocks, steal beauty products and impede mr fireman's ability to put out fires by cutting hoses, why not open up on them too?
cazalla: so Warrick County got back to me regarding the ransomware (in comic sans no less)
ascii_field: i'd be open to a purely gnumake-based thing instead of the sh
trinque: just thinking through it; I have no need to change nor authority to change a process that works for them
jurov: no, btc-dev mailinglist it even renames the recognized signed attachemnts so
mod6: the mailing list
trinque: this is a hypothetical
trinque: to synthesize one point from that pile, I want the computer to strongly enforce relationships for me, not to mindlessly build because I'm too lazy
trinque: so then just a tree of IDs
trinque: the mailing list is a shitty version of this
trinque: ok, this is how I get to the point where I can work on feature X
BingoBoingo: <ascii_field> and the only solution, at this point, is neutron bomb << Better make sure your z-weighted sheilding is completely cobalt free...
trinque: ascii_field: if there were even something which made the relationships between patches data, I would be satisfied.
mircea_popescu: "expose a box to the net just to fetch changes" << what, you want it to divine the changes ?
ascii_field: when i suddenly am expected to care what's in some xml turd, or expose a box to the net just to fetch changes, or similar atrocities - i reach for barf bag
ascii_field: but would never expect other folks to take that kind of radiation damage
ascii_field: for the record, i use all kinds of odd tools internally
davout: trinque: "to the extent that we do it right ascii_field should not have to care" <<< that should be the criteria
trinque: to the extent that we do it right ascii_field should not have to care
trinque: I don't agree with "hidden" re: the state of git
ascii_field: trinque: no, it does not. because there is no hidden state.
trinque: I think hackers here would just like to collaborate via something other than email
trinque: I agree 100% that "canonical" as the outside world is made to see it should not be a git repo
ascii_field: trinque: my opposition to git et al as canonical representations is because they have -any- 'behind the scenes' components.
trinque: pile of hashes and blobs they refer to
trinque: williamdunne: this is what the tree in git already is
ascii_field: the ultimate product is to be a series of -human-auditable- diffs, each small and extremely narrowly focused, starting from classical bitcoind 0.5.3.
williamdunne: Maybe there is a good reason why this would not work, but could you not just generate an XML file that contains the names of all the files, a hash of each file, and then sign the XML? Could be automated fairly easily and would work on top of existing solutions
ascii_field: it is worth repeating precisely -why- i specified 'unix patches' to be the canonical representations
trinque: or if not that some other dvcs
ascii_field: incidentally, for readers unfamiliar with the overall thread, the project under discussion - therealbitcoin - does not belong to me. it is run by ben_vulpes and mod6. who appear to agree with my position here.
ascii_field: so long as the output is a unix patch
ascii_field: i mean, when you write own contributions, you can generate them however you like
trinque: and think you're on the money re: how that should operate
trinque: ascii_field: I would be pleased with a process that allowed for that, but that there is only one route to "released"
ascii_field: i don't have the neutron bomb yet. so we're stuck with flamethrowers
ascii_field: and the only solution, at this point, is neutron bomb
ascii_field: wtf is so hard to understand re: 'the canonical representation of the project must be human-readable' ?
ascii_field: Adlai: go ahead and use it then
Adlai: last thought: a signed-patch-dev mailing list, and some trivial darcs wrapper that consumes attachments from that mailing list to frob any random repository, whether git, darks, or - for mats's's benefit - cvs. left as an exercise to the reader.
Adlai is making too many typos to continue participating in the discourse
trinque: well, the social experiment would be instructive, would it not?
ascii_field: but this is not what git aficionados want. they expect -me- to use it if i want to participate
davout: but by the wot
Adlai: trinque: this applies to the use of silicon that you didn't forge yourself
trinque: shouldn't use them because an idiot will just end up without an arm?
trinque: ascii_field: doesn't this apply to the use of power tools?
ascii_field: and creates the expectation of using git
ascii_field: davout: problem is not just the build
davout: according to Adlai it isn't really practical to do with git, however, if darcs can output actual clearsigned diffs that would allow one to re-construct the same source tree by hand and check the sigs manually
davout: ascii_field: would you have no problem with a tool that you can bypass, and build the source with the actual signed commits ?
Adlai dogwalks and drinks away the shittiness of modern computing
trinque: the process of managing a bunch of feature branches and shit, I want a tool
trinque: so the process of establishing trust should be human and involve eyes
Adlai: williamdunne: the problem here is more one of convincing ascii_field that there exists trustworthyness outside of his own skull
Adlai: trinque: darcs is even simpler, but has the disadvantage (which was previously mentioned as an advantage!) of allowing people to submit changes which cause malicious behavior when combined with previously-signed changes
trinque: but the data model itself is easily understood
trinque: all things are blobs, even the aforementioned
trinque: commit points to other commits and a tree, tree points to tree entries, which point to blobs
ascii_field: git is pernicious for safety-critical code because it - however slightly - reduces the expectation that every line of diff is attentively read
Adlai: either approach lets you determine which fingers inserted which code
davout: Adlai: they're functionnaly equivalent, not identical, the functionality in this case being the identification of the head to chop off
Adlai: they're not! the darcs model would've let mircea_popescu submit a single fix to apache, whereas the git model requires him to sign the entire apache source tree
davout: they're functionally equivalent, but i guess that if darcs is a better fit, why not
ascii_field: not the git thing again
Adlai: ... but whence the sed binary?
Adlai: git has the advantage of a single signature covering the entire current state; darcs has the advantage of letting a single signature cover changes alone. it's really a question of use case
davout: Adlai: you're probably right, i think the difference isn't that important though, the point of the wot is to make an identity valuable
Adlai: so you either want darcs with every single patch signed, or git
Adlai: davout: an "individual commit" is still just a hash of the commit message + merkle root of the source tree
davout: now, how you verify them, and how you enforce mandatory signature is something else
davout: Adlai: well, from what i understand of the docs, you can sign -individual commits- which is a new feature because originally it could only do what you say
Adlai: there's stuff to be said for either side... tbh, it seems like signing merkle roots is what yall want
Adlai: davout: the "sign commits" feature still essentially consists of signing a merkle root of the source tree; darks lets you sign the patch
Adlai: fwiw, darcs is much more friendly to the "signed patch" model of modification; git is better for 'signed code'; and neither are perfect
trinque: why have factories in the US when China will produce everything for us cheaper?
williamdunne: Not like they can modify the code or take it from you, just delete it from their servers
trinque: if I find the time maybe I will
Adlai: stop waiting for other people to take initiative
mircea_popescu: jurov not sure if worth the hassle forjust this
williamdunne: mircea_popescu: Still getting; "Looks like you tried to comment off a stale page. Reload the article, count to three and try again."
williamdunne: Ahh, I assumed MP meant the varnish code
williamdunne: mircea_popescu: GitHub is the accepted norm
mircea_popescu: jurov how do i contribute signed non-bitcoin code to the great codex ?
williamdunne: lobbes: Potentially so, although of course that assumes that you have the capacity to do so, and he has the capitulation to get to that point
lobbes: <williamdunne> ... If I believe paymium will possibly fail, but that you're going do something great in the future, I would be better off investing in you than paymium << In this case, wouldn't you just wait until he creates the project you think -will- succeed and invest in that?
davout: williamdunne: my main point is that this is already what you're doing when buying a BTC-denominated asset that's run by someone and that your idea removes more value than what it adds to the stock exchange value proposition
davout: mircea_popescu: good good, let the broken http requests flow through you
williamdunne: davout: Basically its no longer a bet on whether or not the project is retarded, but whether or not you're retarded
williamdunne: davout: If I believe paymium will possibly fail, but that you're going do something great in the future, I would be better off investing in you than paymium
davout: that's kind of a basic mistake here, assuming you can invest in things, if I buy a bitcoin asset I'm really investing in the people running it, not in some ethereal abstract "thing"
Adlai: ascii_field: on the contrary, the money is there, but it's steered by people who prefer to fund http://itsthisforthat.com/api.php?text
ascii_field: (to a first approximation, afaik, there isn't)
ascii_field: wake me up when there is actually the money floating around to give anything serious at all its minimal take-off velocity.
davout: what other metric than money did you have in mind?
williamdunne: X Y and Z have a number of other metrics you can use to attain success i.e revenue growth
davout: "i didn't even get to try projects X and Y because the market told me it was retarded", on the other hand Z was pretty successful"
williamdunne: I'm hoping it just increases the velocity, so shitty stuff fails sooner and good stuff succeeds sooner
davout: in this case a stock market would have the advantage of making shitty stuff fail earlier
williamdunne: If you believe that someone has the potential to do something great, but do not know how many attempts it may take them to do so, it allows you to bet on their success rather than the success of the current project
williamdunne: Thats the point
davout: williamdunne: because if your work consists of projects X, Y, and Z you can decide to raise capital separately for each of them separately on a stock exchange, not only as "here's me, throw money at me"
Adlai: this is also why the only sane approach to programming depends upon homoiconicity... there's simply too much code for a single human to type by hand
mircea_popescu: now back to the issue : the "backported" thing compiles. apache has no care in the world.
mircea_popescu: the other way to look at this is that everyone involved in computer software should be sent back to third grade
ascii_field: mircea_popescu: one way to look at it is that division of labour never really happened in computing. to the useful point where you can actually safely hand something off to a pro. instead, we're all in a situation not unlike that ru doctor in the polar expedition who had to self-appendectomy
mircea_popescu: "there's free as in beer, free as in freedom, and free as in your time has no value"
davout: yeah well, either you code due diligence or you don't, up to you, never mentioned anything about "reviewing random code for free"
mircea_popescu: moreover, if i do know some guys who trust him, the cost of obtaining a new Takashi Takizawa is not equal to s/Takashi Takizawa/Takashi Takizawa/
williamdunne: What difference would it make if he had signed it? Its not like you know the guy enough to trust him
mircea_popescu: incidentally, this is what the stack of shit is built on. "oh, little step X can't be too hard"
mircea_popescu: so what's the wot vote, do i thank " Takashi Takizawa" for his work and download random code offa da github ?
davout: and wrt to the reference index thing, it's kind of weird that the same stuff would be defined at multiple places anyway
davout: i don't get how the structure can be "dynamic modular" and then "cause problems" when stuff isn't built right into apache
williamdunne: Or whatever the variable is called
mircea_popescu: Apache 2.4 uses a dynamic modular structure by default. This potentially can cause problems if a LoadModule directive calls a module that was not built into the Apache binary.
mircea_popescu: among the changes from 2.2 to 2.4 (current) ? alf will appreciate this
davout: mircea_popescu: i haven't been using apache for a long long while, but maybe mod_remoteip does what you want, override the source IP by some IP it finds in an HTTP header
mircea_popescu: With your preferred file transfer method, upload the custom module's archive file to your server's /var/cpanel/easy/apache/custom_opt_mods/ directory.
mircea_popescu: Download the custom module archive file to your computer.
mircea_popescu: To install a custom module, perform the following steps on the command line as the root user: