Logs for #trilema-mod6

Go to: #trilema #pizarro #asciilifeform #trilema-mod6 #chainstate #eulora

This is the log of #trilema-mod6 : Contained within are the daily on-goings of The Bitcoin Foundation's (http://thebitcoin.foundation) Reference Implementation (trb) development. Do not, under any circumstances, use any code/vpatch/logic/configurations/ideas or otherwise found in this forum. In fact, you should NOT read this at all. Failure to heed these warnings may result in death or dismemberment! ALL ANNOUNCEMENTS, BUSINESS, AND/OR QUESTIONS INVOLVING THE BITCOIN FOUNDATION AND/OR THE BITCOIN REFERENCE IMPLEMENTATION (TRB) WILL TAKE PLACE IN #TRILEMA: YOU HAVE BEEN WARNED.

2016-12-31 | 2017-2-1

mod6: wb
shinohai: thanx! :D
mod6: i'll post info about testing the latest version of V in here, when it's ready.
mod6: i'm working on automated tests at the moment. so far going good.
mod6: (needed to update the old tests, and to create about 10 new scenarios to cover the wot-variant stuff.)
shinohai: Ah cool, I've got mostly free schedule for anything that needs testing.
mod6: cool, thanks shinohai
shinohai: o7
ben_vulpes: mod6: once https://blockchain.info/tx/f2f63fece2cec444ff2da3494729e7b5e51dc2533031c7bdce30097e765c0deb confirms, i'll have successfully tested privkey tools with begin height flag
asciilifeform: neato!
ben_vulpes: aye, tis indeed
shinohai: O.o
ben_vulpes: shinohai: what dat
shinohai: re: privkey begin height
ben_vulpes: well yes, but what do you mean to say with the emoticon?
shinohai: so begin height does work?
ben_vulpes: https://blockchain.info/tx/f2f63fece2cec444ff2da3494729e7b5e51dc2533031c7bdce30097e765c0deb << yup, 2 confirmations
trinque: very nice
mod6: cool, thanks for checking that out ben_vulpes
ben_vulpes: anytime!
shinohai: -5°C here and sleet/snow
ben_vulpes: shinohai: you still in south america?
shinohai: no i never went
shinohai: situation changed rapidly there yup
ben_vulpes: lolk
mod6: progress is being made on the automated tests.
mod6: got about 5 old ones left to clean up, and 5 new ones left to implement.
mod6: This week has been an otherwise busy week, but I'm still manging to get forward progress on the automated tests.
mod6: Got three more done lastnight.
mod6: And now, all the rest I need to do just need test signatures to get 'em completed.
mod6: Once I've got all that working, I'll introduce you all to the new scenerios and what they do, etc.
ben_vulpes: neato mod6, looking forward to it
ben_vulpes: http://logs.bvulpes.com/trilema-mod6 << sorry for the delay, header warning now in red
mod6: ah, cool. thx
mod6: So in congruance with updates in #trilema, i've made some progress on V.
mod6: I'll get into the details in here tomorrow most likely.
mod6: well, some things came up today... didn't get much of a chance to discuss the stuff I wanted to.
mod6: But I want to do a bit more analysis anyway, probably be a few days yet.
mod6: well, I don't want to speak too soon here... but I think i finally got this optimized.
mod6: I had the solution since mid-last week, but the performance wasn't on par.
mod6: But I've got it working on par with where it was before. Which is nice. Making progress here, Gentlemen.
ben_vulpes: neat. what was the bottleneck?
mod6: so really, all in all, wasn't a performance related problem when it came down to it.
mod6: initially I had my recursive call in the wrong place but was getting ok results, functionally. but bad perf.
mod6: obviously.
mod6: so i refactored my code to, first, put the recursive call in the correct spot, and then clean up my code so it was doing all the right things functionally.
mod6: it basically came back to, keep it simple, keep it simple, keep it simple.
mod6: as it usually does. :]
mod6: i had this one function that kinda did two different things. after breaking it into the two things it should be, i refactored a bit, making it more exact. which got rid of some problems.
mod6: so, ok, trying out some of the same things with v99 just as a point of reference
mod6: (for those who don't follow, there was a #trilema thread about roots today http://btcbase.org/log/2017-01-26#1606798)
mod6: Anyway, so I did a few things. Namely:
mod6: 1. setup just mod6 in the wot
mod6: 2. added all the current trb patches & seals
mod6: (seals just for mod6)
mod6: 3. test a press, works fine
mod6: 4. remove bitcoin-asciilifeform.1.vpatch.mod6.sig from .seals
mod6: 5. check the flow; which as is known, doesn't remove dependants, just the now-unsigned vpatch from the flow.
mod6: 6. test a press with the missing
mod6: which again, as expected fails to press.
mod6: 7. put bitcoin-asciilifeform.1.vpatch.mod6.sig back into .seals
mod6: 8. place foobar.vpatch into patches
mod6: 9. place foobar.vpatch.mod6.sig into .seals
mod6: 10. check the flow; does what my V99995 does; which is, place the "root" at the end of the flow.
mod6: 11. Test a press all the way up through foobar.vpatch.mod6.sig
mod6: All outcomes were expected
mod6: http://dpaste.com/04D9MPA.txt
mod6: When I have implemented the axiom of "all antecedents" must be present into V99994, that one gets dropped out.
mod6: I can try to work around this with a special case for roots perhaps, or maybe another way.
mod6: anyway, just doing some research/digging. will check back later.
mod6: hmm
mod6: so, anyway, aside from the multi-root considerations, this is basically how my V99994 is working now:
mod6: http://dpaste.com/1KKRVD4.txt
mod6: ^ Which shows that given a normal flow with a single signatory, it is correct. Then dropping out a seal to a vpatch in the flow, then removes all descendant vpatches from the one removed (it presses correctly/cleanly as this exact flow too). And lastly, if a second wot entity is added into the wot...
mod6: then we see the mixed flow where 'asciilifeform' has signed 'bitcoin-asciilifeform.4-goodbye-win32.vpatch' but 'mod6' has not.
mod6: this patch is then re-added back into the flow, and the the rest of its decendants re-appear, as long as they are signed by a valid wot entity and all of their antecedents are present in the flow.
mod6: Ok, so lastnight spent some time working towards getting mulitple-roots working in the new version.
mod6: I'm not /quite/ there yet, but on the right track I think.
mod6: Will update more as I have them.
mod6: wb
ben_vulpes: cool mod6
ben_vulpes: keep crankin
mod6: will do, Sir. o7
mod6: asciilifeform ben_vulpes shinohai trinque : did anyone look at this and determine its correctness? http://dpaste.com/1KKRVD4.txt
mod6: I know it's a bit to look at, but if you could take a minute and check it over, that'd be awesome.
mod6: Also, if you have questions, let me know, would be happy to explain anything going on there.
ben_vulpes: idem
mod6: no problem. any guidence with that by mid-next week would be awesome.
mod6: I'm gonna keep plugging away at the multi-roots thing anyway, and I'm /pretty/ sure the wot variant stuff is correct, as well as the implementation of the all-antecedents-must-be-present axiom. Just want to get extra eyes on it before I get too far down the line.
shinohai: ty mod6
mod6: np, how goes?
mod6: good progress today on the multi-root front.
mod6: more later.
mod6: so lastnight i think i've got something working for multiple roots.
mod6: im gonna doing some testing today to ensure everything is as it should be.
mod6: we'll see how it goes.
mod6: alright.
mod6: I have created a second test patchset that contains multiple roots. Three roots actually.
mod6: I have done some intital testing to ensure that this is working as it should. Also tested that if a vpatch is plucked out of the flow, the flow is reorganized as such and that the new flow is also pressable as it should be.
mod6: I'm going to show this off a bit, let me know what you think.
mod6: Everything I'm going to show is produced by the latest source changes to V99994 (forthcoming).
mod6: http://www.mod6.net/sps2_dag.png
mod6: http://www.mod6.net/v-99994-sps2-trace1.txt
mod6: http://www.mod6.net/sps2/
mod6: the last link above contains all of the vpatches in this special set with multiple roots for you to examine, if you wish.
mod6: Salud!
mod6: oh, and I'm putting in a check for these multiple roots too: http://btcbase.org/log/2017-01-30#1609666

2016-12-31 | 2017-2-1