--- Log opened Thu Apr 02 00:00:41 2020
06:55 < adiabat> hey did you find anything with the rev files?
06:56 < adiabat> I'll stat looking again in an hour or so
06:56 < adiabat> I thought of something that might be related
06:57 < adiabat> I have 1 blk file and 2 rev files.  So it reads blk0.dat and rev0.dat, rev1.dat
06:57 < adiabat> so maybe it's looking at the beginning of the first blk file and the second rev file
09:10 < kcalvinalvin> adiabat oh the lastIndexOffsetHeight should also be -1 of the actual block height
09:10 < kcalvinalvin> because that's the offset height
09:11 < kcalvinalvin> yeah... we should fix this...
09:11 < kcalvinalvin> So... I think it's right
09:16 < adiabat> hm ok if that's it then cool...
09:16 < adiabat> but.. I don't think that's it
09:16 < adiabat> because it sometimes gets the right amount, but an empty pubkey
09:17 < adiabat> this is just if I have mainnet blk0 and rev0
09:17 < adiabat> I added rev1 because I don't know if there's a 1:1 mapping between blk files and rev files
09:18 < adiabat> probably, in that they both go up to 2021
09:30 < kcalvinalvin> hmm I haven't come across any nil pubkeys
09:31 < kcalvinalvin> In line #147 in rev.go, there's sanity checking for nil pubkeys
09:31 < kcalvinalvin> I'm doing some printf in rev_test and I don't see any empty pubkeys up to block 700,000 on testnet3
09:34 < kcalvinalvin> oh wait hold up
09:42 < adiabat> maybe try mainnet?  What I did was just the first blkand rev file from mainnet
09:43 < adiabat> that might have messed it up somehow. but really it shouldn't
09:44 < adiabat> or maybe something else I'm doing in my PR is messing it up
09:44 < adiabat> try running that and see if you get the same ... if it works OK outside of the PR it could be somehting else
09:50 < kcalvinalvin> I found the bug
09:51 < kcalvinalvin> RevOffsetfile is returning all 0s for revblocks that aren't from rev00000.dat
09:51 < kcalvinalvin> I'll push a fix
10:05 < adiabat> huh ok that could do it
10:14 < kcalvinalvin> ugh remember when I said the revoffset was building so fast
10:14 < kcalvinalvin> It wasn't building anything
10:35 < adiabat> aha :)
10:35 < adiabat> that speeds things up 
11:52 < kcalvinalvin> Ok I fixed the revoffset problem but found a script decompress problem
11:53 < kcalvinalvin> It'll take me a good day.. so I won't be able to push today
12:02 < adiabat> ok
12:02 < adiabat> should I look at it? or work on other parts for now?
13:00 < kcalvinalvin> whoop I think I might have got it
13:00 < kcalvinalvin> Better to look at other parts for now...
13:00 < kcalvinalvin> But I think it's working now
13:30 < kcalvinalvin> adibat
13:30 < kcalvinalvin> lol Github seems to be down
13:30 < kcalvinalvin> Can't push :(
13:40 < adiabat> github's down..? yikes
13:40 < adiabat> hmm I can see it
13:41 < adiabat> if you can post the diff to pastebin or something... or put it anywhere else on the internet I can try it out
13:42 < adiabat> hm yeah the *git* part of github isn't working, but all the web stuff seems to
13:42 < adiabat> well if you can post it anywhere I can try it, if not I'll make a hack levelDB utxodb for now and continue that way, then replace it with rev files after
13:43 < adiabat> so no big deal either way
20:36 < kcalvinalvin> adiabat I made a PR with the fix
21:15 < adiabat> kcalvinalvin: cool, looking at it
21:15 < adiabat> seems like it works
21:16 < adiabat> both the rev blocks *and* the regular blocks are off by 1 so it matches up :)
21:16 < adiabat> also rev data exists for utxos that are created and destroyed in the same block
21:17 < adiabat> which you don't need, but I guess it's easier; don't have to have any special code for that
21:20 < adiabat> nice, yeah this looks like it works, so everything for the add / del / sig check is there
21:20 < adiabat> should I merge?  ysangkok pointed out some comment inconsitencies
21:20 < ysangkok> i think they are fixed already
21:22 < adiabat> oh, yup they are
21:22 < adiabat> ok cool merging
--- Log closed Fri Apr 03 00:00:41 2020