log☇︎
20400+ entries in 0.039s
mod6: oh
mod6: your node that you were speaking about ^
mod6: ascii_field: what's your best block height?
mod6: When I do get around to drafting up all of the scenarios, I'll put that in here as well, maybe you guys can think of other tests that I hvaen't thought of yet. Or edge cases. etc.
mod6: Why perl? Because I'm not good with python, and ruby well, lol. ☟︎☟︎
mod6: Of course, there will actually be a separate, more formal testing guide that I'll create also.
mod6: But as far as cucumber, even if I have issues with the steps (so far it's been very easy), if I write out the scenarios in the feature, it should be a decent testing guide in itself.
mod6: We'll see where it goes and if I run into too many problems.
mod6: With some effort up front I might be able to automate the ~20 or so scenarios that i've sketched out for the release. Could save all of us a lot of time by installing/emerging a few things on gentoo and then just running these tests instead of having to do all the testing manually.
mod6: so last 24 hours or so I've been working on getting an automated testing framework setup for this forthcoming release. I've got one scenario working (simple one). Here's what it kinda looks like - although this isn't the entire code, there is a module I left out for now. http://dpaste.com/32GZH0A.txt
mod6: !up ascii_field
mod6: <+asciilifeform> ben_vulpes, mod6, mircea_popescu, jurov, et al: http://dpaste.com/0YSCME3.txt << nice!
mod6: Can only check teh weather if in assbots L2. Which is very temperate.
mod6: Wonderful :]
mod6: http://twitter.com/modsix/status/623222065982607360
mod6: I linked him
mod6: strangely worded
mod6: oh
mod6: wat
mod6: openssl-1.0.1g # grep "my \$no_shared" Configure \ my $no_shared=0; # but "no-shared" is default
mod6: the openssl configure script (perl lel) seems to indicate that no-shared is the default -- but i wanna be sure.
mod6: np. it can wait for a bit.
mod6: asciilifeform: ever get a chance to look at those bins?
mod6: start simple, then build in more complexity as only necessary.
mod6: looks like you can use random-id for some normalizatoin purposes... main thing is here, read the pf.conf man page 2x and then build up your firewall and use the heck outta tcpdump to ensure what your rule set is as you want it.
mod6: im not positive (i don't have access to my pf.conf atm) that i've ever used random-id. i think that's for a very specific problem. but yeah you probably /do/ want scrub all reassemble tcp
mod6: trinque: what are you trying to achieve?
mod6: haha
mod6: haha, no worries. that's fine too.
mod6: the first doc is just the commands you need, so it's kinda cut/pasteable. the second is the commands with their output so you can get a sense of what you should expect to see.
mod6: shinohai: awesome! no rush either. when you get time is cool. let me know how it goes, i'll be around to answer q's.
mod6: here's the second one with full examples for nomultilib: http://thebitcoin.foundation/gentoo-stage3-amd64-nomultilib-guide-wFullExamples.txt
mod6: bah. ignore that second one. that's for uclibc.
mod6: http://thebitcoin.foundation/gentoo-stage3-amd64-nomultilib-guide.txt && http://thebitcoin.foundation/gentoo-stage3-amd64-uclibc-hardened-guide-wFullExamples.txt ☟︎
mod6: ah, yeah, so ... this guide is for a physical box (takes you through using livecd) -- if you wanna set it up in a VM im sure you can do something similar with a livecd, just hvae to attach the iso file to the virtual dvd/cdrom drive.. and probably follow the same steps.
mod6: we need a 3rd person to verify this thing so i can finally send it to the lst.
mod6: :]
mod6: if you're looking for something to keep you busy for an evening, you can test out the gentoo x86-64 nomultilib guide
mod6: actually...
mod6: haha
mod6: ;/
mod6: I'm gonna put up a few too, but gotta wait about 6 wks.
mod6: heheh. well, just make sure if you /do/ put up a node, it is somewhat phasar resistant
mod6: same ip
mod6: is this the one where blkcut error made it fall over?
mod6: 69! yeah, we need more nodes.
mod6: i guess more importantly than that, im gonna do a for(0->366XXX) dump of each block and hash it, compare it to the list that you ended up with ascii.
mod6: <+asciilifeform> i thought everybody knew, at this point, that one cannot expect to end up with same blkxxxx at all times, on account of forklets. << yeah i recall.
mod6: for sure. thanks, we need all the help we can get.
mod6: effort & parts!
mod6: well, thanks for putting in the effort to test this stuff, we appreciate it.
mod6: yikes, threw a bering 'eh
mod6: it bricked the drive?
mod6: haha, pogo chat server?
mod6: thanks for reporting back on that
mod6: good deal man.
mod6: did it sync the same way from block 0->current in the same manner, i.e. by finding nodes all over irc?
mod6: nice, what block are you on?
mod6: 0.5.3.1 ?
mod6: height=205332
mod6: my openbsd one is crawling along.
mod6: height=351462
mod6: haven't gotten there yet on nsl's node
mod6: i'll be revisiting this not too long from now with some hashes from mp's orig chain
mod6: as a reminder tho, those hashes of my .dat files were from "the wild" sync.
mod6: asciilifeform: nice pics. protocol? maybe i don't get what you're asking exactly, but it says "D8027G1"
mod6: don't challenge me to a spelling contest, i'll lose.
mod6: yah that :]
mod6: *indefiniately
mod6: gonna have to test prune by hand, and then test and ensure that these empty dirs will persist indifiniately.
mod6: yeah. well, i think i'll just prune by hand if I do at all. like i was saying earlier, im a bit paranoid now that some script might later accidentially remove the dirs if they're rendered empty.
mod6: there's a bunch of lines that reference it actually, and no, i was hoping to get rid of these barnacles without having to change other files.
mod6: i've tried this thing 48 ways from sunday. context diff, unified, truncated files, removed files with/without -E smh
mod6: <+asciilifeform> we will need more, more of these. << indeed
mod6: <+phf> ben_vulpes: because makefile.unix expects them, (search for obj/nogui/%.o: %.cpp, etc.) << ben_vulpes yeah this stuff ☟︎
mod6: cool!
mod6: s/simple/clean/
mod6: sure there are other options. nothing quite as simple as "just blow away the .gitignore files"
mod6: <+mod6> so... yah, a bash script or removal by hand. would be fine i'd think. but now i gotta test it a bit harder. if we leave empty directories in there, im worred that patch might come along at a later time and be helpful again, removing those object output diretories. << so if i do remove them by hand, have to ensure that these empty dirs wont get nuked later on accident
mod6: <+punkman> mod6: ben reminded me to add a patch that removes the 5 .gitignore files. << I dunno why patch was needed for this, just remove them manually in next release << yeah, i mentioned that after the fact lastnight. but was discussed here: http://log.bitcoin-assets.com/?date=19-07-2015#1206370 << this makes sense. ☝︎
mod6: that'll leave the dirs in place
mod6: maybe the best option is to add a one line shell script that just does: `find . -name ".gitignore" -exec rm '{}' \;`
mod6: the only thing in the obj-test directory before applying that patch is '.gitignore' so why wouldn't it be removed like obj/test and obj/nogui?
mod6: now im really confused.
mod6: well, thats weird. it seems to remove obj/test and obj/nogui, but NOT obj-test which is left empty?!!?
mod6: i might just be dumb. build a test env with http://dpaste.com/23VKWD8.txt and see if anyone can get the rm_gitignore.patch file to apply without removing the obj/test obj/nogui or obj-test dirrs
mod6: anyway i'll work on it. just don't use the rm_gitignore patch.
mod6: if i want an empty dir, i should be able to have an empty dir
mod6: unless our solution is to add .keep files in there, which i think kinda sucks.
mod6: so... yah, a bash script or removal by hand. would be fine i'd think. but now i gotta test it a bit harder. if we leave empty directories in there, im worred that patch might come along at a later time and be helpful again, removing those object output diretories.
mod6: smh
mod6: "why would you want an empty directory!"
mod6: lol, it's trying to be helpful.
mod6: yah, and when applied `patch` ends up removing the dirs along with the .gitignore files because it renderes those directories empty.
mod6: here's the patch: http://therealbitcoin.org/ml/btc-dev/2015-July/000125.html
mod6: ben reminded me to add a patch that removes the 5 .gitignore files. seems easy enough, so crated a quick patch, attempted to apply, worked fine -- I didn't notice that it blew away the directories as well though. : obj/test obj/nogui obj-test
mod6: but now I'm scared that even if i /do/ remove them by hand, they might get accidentially pruned by a downstream patch (later in time) causing the makefile to puke. ☟︎☟︎
mod6: lol, yeah it's obnoxious. and i think the only way to solve this is to remove them by hand.
mod6: yeah, it basically zero's out the file, and then it seems to remove the file with -N (diff) and if there is nothing left in the directory, it zaps that as well.
mod6: check out `man 1 patch` -E