1 2015-09-03 00:58:26 <esso> What are some tests that need to be run for the Bitcoin core?
  2 2015-09-03 05:47:14 <job_> hi all. quick question -- is there any notion of "starter tickets" if I want to get involved with bitcoindev?
  3 2015-09-03 05:55:50 <btcdrak> job_ browse around the issue tracker for something that interests you, or test and review some existing pull requests.
  4 2015-09-03 05:58:26 <job_> roger btcdrak
  5 2015-09-03 06:00:53 <job_> btcdrak would you say there's demand for contributors? my C++ is pretty rusty, but I could spend some time on it if there's need
  6 2015-09-03 06:01:45 <btcdrak> job_ big demand for contributors, especially testing and reviewing pull requests.
  7 2015-09-03 06:06:45 <Luke-Jr> job_: reviewing is a major bottleneck right now; if you're new, it will be very helpful *and* give you an opportunity to learn the codebase at the same time
  8 2015-09-03 06:11:13 <job_> right on... any gh labels that indicate a ticket's in need of review?
  9 2015-09-03 06:11:58 <Luke-Jr> job_: if it's open, it needs review.
 10 2015-09-03 06:12:14 <Luke-Jr> (although perhaps read the comments first, just in case there's some other reason it's open)
 11 2015-09-03 06:12:49 <job_> hah, that makes it easy :)
 12 2015-09-03 07:05:54 <Luke-Jr> gmaxwell: I'm not Jeff, but I suspect the "idle hashpower" is because otherwise they would have to find blocks slower, which results in even more transactions "needing" to be mined. (I don't agree with the problem existing, because miners can always add *new* hashpower, and if all miners agree on the size increases, presumably the difficulty will adjust with it in mind)
 13 2015-09-03 07:14:15 <gmaxwell> Luke-Jr: it's not true in any case, so long as the formula obeys the invariant that adding hashpower increases the block size faster than it decreases the block rate.
 14 2015-09-03 07:20:37 <CodeShark> is it just me or does anyone else feel like we still lack sufficient datapoints and experience to be able to come up with a solid dynamic max block size readjustment mechanism?
 15 2015-09-03 07:21:08 <CodeShark> seems like we should focus on solving incentives first
 16 2015-09-03 07:23:15 <CodeShark> IMHO, as long as validation cost is externalized we're missing a crucial piece of the economic puzzle
 17 2015-09-03 07:23:18 <wumpus> no, you're not exactly alone in this, I think it has been a recurring theme that we just can't be sure enough yet and shouldn't make rash decisions.
 18 2015-09-03 07:25:37 <CodeShark> in principle, if the incentives are correct and there's transparency, free markets should take care of parameter readjustment
 19 2015-09-03 07:26:43 <CodeShark> until the incentives are correct it seems far simpler to just keep a constant max block size :)
 20 2015-09-03 07:27:08 <CodeShark> or at least one that grows in a very predictable manner
 21 2015-09-03 07:27:50 <CodeShark> throwing in dynamic readjustment means significant complication - with questionable benefits, IMHO
 22 2015-09-03 07:28:41 <CodeShark> and potentially carries risks that have not been fully explored
 23 2015-09-03 07:29:38 <midnightmagic> Plus, bonus, you get to deal with all the altcoin people shoutingt hat you stole their black-hole/warp/adjustment ideas! :-D
 24 2015-09-03 07:30:05 <Luke-Jr> I mentioned the IBD issue on reddit and I was told "Link, or it didn't happen" ._.
 25 2015-09-03 07:30:09 <wumpus> yes, it's the eternal "set simple rules, let the market figure out the rest" versus "set complex rules, hope they're not gamed"
 26 2015-09-03 07:31:40 <CodeShark> I'm definitely more in favor of the former...simple rules are usually better. the thing is that discovering simple rules that solve complex problems is actually hard :)
 27 2015-09-03 07:31:55 <CodeShark> but once it's done it's beautiful
 28 2015-09-03 07:31:56 <wumpus> in any case if your proposal has any predicted sustained exponential growth in it I'm very sceptical
 29 2015-09-03 07:31:59 <Luke-Jr> CodeShark: will you be in Montreal btw?
 30 2015-09-03 07:32:11 <CodeShark> just booked my flight :)
 31 2015-09-03 07:32:19 <wumpus> Luke-Jr: just copy/paste your reddit post into a blog, give a link, voila :P
 32 2015-09-03 07:32:34 <wumpus> Luke-Jr: but yes, people take blogs much more serious than social media posts, even in this day and age
 33 2015-09-03 07:33:57 <gmaxwell> http://www.smbc-comics.com/index.php?id=3846
 34 2015-09-03 07:34:28 <Luke-Jr> CodeShark: cool; I hope I can make it (will probably find out in the morning when I call passport people..)
 35 2015-09-03 07:35:04 <CodeShark> you don't have a passport?
 36 2015-09-03 07:35:14 <Luke-Jr> not yet :x
 37 2015-09-03 07:35:31 <wumpus> gmaxwell: heh
 38 2015-09-03 07:39:36 <jonasschnelli> Luke-Jr: you don't get a passport in the U.S. by default?
 39 2015-09-03 07:40:08 <CodeShark> I've visited at least six different countries in the last year, three times to Europe, this will be my second time in Canada in the last couple months :p
 40 2015-09-03 07:40:45 <CodeShark> many americans never get passports, jonasschnelli :p
 41 2015-09-03 07:41:41 <phantomcircuit> <Luke-Jr> I mentioned the IBD issue on reddit and I was told "Link, or it didn't happen" ._.
 42 2015-09-03 07:41:42 <phantomcircuit> Luke-Jr, ?
 43 2015-09-03 07:41:54 <wumpus> jonasschnelli: same in the Netherlands really, most people get an European ID card
 44 2015-09-03 07:42:07 <jonasschnelli> Huh... I just encountered that I don't need a visa for Canada. Awesome! Less paperwork.
 45 2015-09-03 07:42:27 <Luke-Jr> jonasschnelli: nope, gotta jump through hoops, pay $$$ and wait time
 46 2015-09-03 07:42:32 <wumpus> you likely do need an "Electronic Travel Authorization"
 47 2015-09-03 07:42:43 <Luke-Jr> phantomcircuit: https://www.reddit.com/r/Bitcoin/comments/3jeetv/jgarzik_bip_100_adds_widelyagreedsuggested_1mb/cup0mcg?context=3
 48 2015-09-03 07:43:07 <gmaxwell> About one third of the US has passports, they're much more common than they used to be presumably because of needing them for canada and mexico.
 49 2015-09-03 07:43:25 <jonasschnelli> wumpus: I have checked the ETA. Seems to be mandatory from June 2016.
 50 2015-09-03 07:43:27 <phantomcircuit> Luke-Jr, ah yes
 51 2015-09-03 07:43:29 <gmaxwell> (I think the number was about 5% in the early 90s)
 52 2015-09-03 07:43:49 <wumpus> jonasschnelli: are you sure? I was told to get one. But maybe it's different for NL
 53 2015-09-03 07:44:19 <wumpus> jonasschnelli: but yes I'm happy to need no visa either, always scared of getting the wrong one
 54 2015-09-03 07:44:40 <jonasschnelli> I better do a doublecheck. Don't want to hand around in the airport.
 55 2015-09-03 07:45:50 <jonasschnelli> wumpus: looks like it's not mandatory until March 15 2016: http://www.cic.gc.ca/english/visit/eta.asp
 56 2015-09-03 07:46:06 <wumpus> ok
 57 2015-09-03 07:46:09 <CodeShark> Luke-Jr: no
 58 2015-09-03 07:46:18 <jonasschnelli> Luke-Jr: you laptop
 59 2015-09-03 07:46:26 <jonasschnelli> *Your
 60 2015-09-03 07:46:47 <Luke-Jr> actually, it just occurred to me today that I may have no communications once I leave the US :/
 61 2015-09-03 07:47:02 <CodeShark> yeah, don't forget to get international data roaming :)
 62 2015-09-03 07:47:10 <Luke-Jr> I don't think I can get it.
 63 2015-09-03 07:47:16 <wumpus> Luke-Jr: same here.
 64 2015-09-03 07:47:25 <jonasschnelli> Isn't wifi now everywhere?
 65 2015-09-03 07:47:26 <wumpus> well we can always bum for free wifi :p
 66 2015-09-03 07:47:36 <Luke-Jr> hopefully
 67 2015-09-03 07:47:57 <Luke-Jr> just need to make sure i have a way to make contact with someone who knows me when I get there I guess XD
 68 2015-09-03 07:48:06 <Luke-Jr> and avoid getting stranded
 69 2015-09-03 07:48:26 <wumpus> so if I disappear after arriving in Canada I've either been black-bagged or I just can't find wifi :-)
 70 2015-09-03 07:48:42 <jonasschnelli> Hah
 71 2015-09-03 07:49:10 <wumpus> at my hotel it's $15 or so per day I found out :(
 72 2015-09-03 07:49:21 <Luke-Jr> wat? hotels charge for wifi?
 73 2015-09-03 07:49:47 <jouke> wumpus: we have a transit flight at washington and we did need an esta for the US.
 74 2015-09-03 07:50:25 <midnightmagic> -- including the hotel I was in where I had to lug 20 lbs of ethernet-over-power gear up some stairs lol
 75 2015-09-03 07:50:43 <wumpus> jouke: oh good point about transit flights. but I have a direct flight.
 76 2015-09-03 07:50:51 <wumpus> midnightmagic: yes, I should have checked, I'm stupid
 77 2015-09-03 07:51:03 <jouke> And we booked a nice appartment with a big living room with airbnb and free wifi ;)
 78 2015-09-03 07:51:39 <wumpus> midnightmagic: I'm not an experienced world traveler like some of you :p
 79 2015-09-03 07:51:41 <midnightmagic> wumpus: No, I mean to imply I didn't know they existed..!
 80 2015-09-03 07:52:05 <CodeShark> I've had some amazing experiences with airbnb
 81 2015-09-03 07:52:21 <midnightmagic> not that you failed in magically knowing to check in advance for every possible convenience that you'd be using that the rest of us just take for granted
 82 2015-09-03 07:53:01 <Luke-Jr> I wonder if there's a company that provides temporary SIMs for events?
 83 2015-09-03 07:53:29 <wumpus> midnightmagic: and indeed, I didn't know either. I've been to some crappy hotels but they always had free wifi of some kind, even ages ago when I was still at uni
 84 2015-09-03 07:54:22 <wumpus> Luke-Jr: well last time I was in the US I got a prepaid sim, but it wasn't usable for data
 85 2015-09-03 07:54:31 <jonasschnelli> Luke-Jr: pretty sure you can get a prepay sim with ~1GB for less than 50$
 86 2015-09-03 07:55:06 <wumpus> (but that may be more normal now)
 87 2015-09-03 07:55:24 <jonasschnelli> http://prepaid-data-sim-card.wikia.com/wiki/Canada
 88 2015-09-03 07:55:54 <jonasschnelli> 30USD/1GB
 89 2015-09-03 07:56:24 <wumpus> wow, sounds useful, thanks :)
 90 2015-09-03 07:57:04 <wumpus> even the frequency bands
 91 2015-09-03 07:57:14 <Luke-Jr> bleh, looks like I have to choose between Rogers 2G and Wind 3G
 92 2015-09-03 07:58:22 <jonasschnelli> But my last trips showed me that most cafes/hotels/event-rooms provide wifi.
 93 2015-09-03 07:58:47 <jonasschnelli> Just don't forget to enable openvpn on a trusted server/VPS :-)
 94 2015-09-03 08:00:48 <wumpus> jonasschnelli: of course :-)
 95 2015-09-03 08:01:00 <jonasschnelli> And use port 443 to bypass firewalls
 96 2015-09-03 08:01:30 <Luke-Jr> iodine
 97 2015-09-03 08:03:20 <Luke-Jr> Wind has no coverage in Montreal.. so I guess 2G it is
 98 2015-09-03 08:03:26 <wumpus> jonasschnelli: that's already my ssh port! :p
 99 2015-09-03 08:04:24 <jonasschnelli> wumpus: hah. Use a different ip than. :p
100 2015-09-03 08:05:13 <CodeShark> I rarely have problems at hotels with port 22
101 2015-09-03 08:05:51 <Luke-Jr> port 22 sucks because idiots try to bruteforce it, but screw up the SSH protocol so it logs to my critical syslog and annoys me
102 2015-09-03 08:05:58 <CodeShark> hah
103 2015-09-03 08:06:32 <Luke-Jr> (changing the port is a trivial workaround for this)
104 2015-09-03 08:07:11 <jonasschnelli> Changing ssh port is probably the first thing to do when you set up a root or VPS
105 2015-09-03 08:52:45 <CodeShark> miner voting just seems like such the wrong approach for approving tweaks in economic parameters :p
106 2015-09-03 08:52:59 <eligigi27> true
107 2015-09-03 08:53:15 <eligigi27> but big crash will coming due of it
108 2015-09-03 08:53:29 <CodeShark> unfortunately we lack any other algorithmic consensus that cannot be easily sybilled
109 2015-09-03 08:53:36 <eligigi27> look at signalcoin graph's http://www.signalcoin.com
110 2015-09-03 08:56:04 <CodeShark> I don't even think there's a general agreement amongst core developers as to what the objective in raising the max block size really is
111 2015-09-03 08:56:12 <CodeShark> what is it we're trying to accomplish with it?
112 2015-09-03 08:56:15 <wumpus> talk about crashes belongs in #bitcoin-pricetalk
113 2015-09-03 08:57:05 <CodeShark> what would be the "sweet spot" for the max block size?
114 2015-09-03 08:57:10 <wumpus> please...
115 2015-09-03 08:57:37 <CodeShark> what model should we use to compute the optimal block size (assuming we could make the appropriate measurements)?
116 2015-09-03 08:58:43 <CodeShark> it seems like it ultimately comes down to a tradeoff between fees and security
117 2015-09-03 08:59:10 <CodeShark> unless the incentives can be fixed
118 2015-09-03 09:00:33 <CodeShark> how much validation are we willing to sacrifice to get lower fees?
119 2015-09-03 09:00:40 <CodeShark> it seems like that's what it really comes down to
120 2015-09-03 09:01:00 <wumpus> depends on the 'we'
121 2015-09-03 09:01:51 <CodeShark> can we all at least agree on the notion that block space is inherently scarce? :)
122 2015-09-03 09:02:18 <CodeShark> we can only raise it so much before the network breaks
123 2015-09-03 09:02:27 <wumpus> that's where the shoe wrings, really, it's a tragedy of the commons issue. But again, I don't think we should be spamming bitcoin-dev with this, it's been done, it's been discussed to death, people just won't agree.
124 2015-09-03 09:03:02 <CodeShark> I know it's been beaten to death - I'm just dumbfounded that supposidely technically competent people would find so much controversy in this
125 2015-09-03 09:03:37 <CodeShark> I still can't believe it :p
126 2015-09-03 09:03:46 <wumpus> because it's not - really - a technical issue. Different choices would end up good for different people.
127 2015-09-03 09:04:26 <CodeShark> I'm not even talking about specific choices here - I'm just talking about a basic principle
128 2015-09-03 09:04:36 <CodeShark> bigger blocks -> less validation
129 2015-09-03 09:05:31 <CodeShark> there's an s such that if blocks are larger than s the network breaks
130 2015-09-03 09:06:18 <wumpus> you'd first have to define 'the network breaks'
131 2015-09-03 09:07:28 <wumpus> I agree that bigger blocks would result in less validation, and less *distributed* validation, there is no point at which it completely breaks (unless you push it to an absurd point no hardware, no matter what can handle it anymore), but the properties will change
132 2015-09-03 09:07:29 <CodeShark> yeah, I guess that might be where there's disagreement
133 2015-09-03 09:09:13 <CodeShark> the second issue that also doesn't seem all that controversial is that flood networks have inherent scaling problems
134 2015-09-03 09:09:32 <wumpus> I'm not even sure 1MB is a sweet spot. I am sure the width of the channel needs to be limited, a broadcast-to-everyone channel is inherently a scarce resource.
135 2015-09-03 09:09:39 <wumpus> right.
136 2015-09-03 09:10:50 <CodeShark> but I think you're right about defining 'the network breaks'
137 2015-09-03 09:11:42 <CodeShark> some people seem to think that full blocks (and therefore unpredictable confirmation times and fee complexities) is 'the network breaks'
138 2015-09-03 09:12:50 <CodeShark> others think if validation isn't sufficiently distributed and accessible to typical users 'the network breaks'
139 2015-09-03 09:14:56 <CodeShark> the main problem I have, I guess, is that while I don't think the former can be avoided I think the latter can
140 2015-09-03 09:15:44 <CodeShark> so trying to avoid the former at the expense of the latter means neither is avoided
141 2015-09-03 09:16:20 <wumpus> the other thing I'm convinced of is that hardcoding exponential growth is a bad idea - if the crash of 2008 has teached us anything it's that nothing grows exponentially forever and relying on it leads to danger. No matter if it's based on Moore's law, general economic optisism, population growth, Kurzweilian tech halleluja, etc
142 2015-09-03 09:17:19 <CodeShark> on the other hand, we can always scale back with a soft fork if we went too far
143 2015-09-03 09:17:25 <wumpus> which 'we'?
144 2015-09-03 09:17:33 <wumpus> there's not one group here
145 2015-09-03 09:17:35 <CodeShark> the ubiquitous 'we' :p
146 2015-09-03 09:17:56 <CodeShark> it's sort of like 'them' except that you and I are a part of it :p
147 2015-09-03 09:19:20 <CodeShark> the 'we' in the case of soft forks really is miners
148 2015-09-03 09:19:27 <wumpus> it's not easy to brake if those in front think everything is okay while the back of the train is falling apart :)
149 2015-09-03 09:19:52 <CodeShark> or mining pools, in any case
150 2015-09-03 09:20:50 <phantomcircuit> wumpus, the more i look at things the more im convinced that the only possible safe continuous growth function is to add a constant amount per re-target to maintain worst cast linear growth
151 2015-09-03 09:20:52 <wumpus> which already presupposes centralization of mining - yes, it's realistic :-)
152 2015-09-03 09:21:17 <phantomcircuit> which will be ok as long as bandwidth/cpu time continue to grow exponentially
153 2015-09-03 09:21:35 <wumpus> 'as long as' is the problem, it's contingent. And for whom?
154 2015-09-03 09:22:16 <phantomcircuit> wumpus, yes if we dont make that assumption then the only safe course is to increase through a number of hard forks based on estimated cost
155 2015-09-03 09:22:20 <phantomcircuit> which is uh
156 2015-09-03 09:22:22 <phantomcircuit> "fun"
157 2015-09-03 09:22:55 <wumpus> maybe the only safe course is to keep this chain as it is, and innovate and try to push limits somewhere else
158 2015-09-03 09:24:41 <CodeShark> that's the approach I would prefer, wumpus...to be quite honest
159 2015-09-03 09:24:59 <CodeShark> but I obviously inject my own personal preferences into that opinion
160 2015-09-03 09:25:36 <CodeShark> to me, bitcoin's greatest strength is as a censorship-resistant network with very strong settlement guarantees
161 2015-09-03 09:25:42 <CodeShark> and I don't mind paying a little more for that
162 2015-09-03 09:26:04 <CodeShark> obviously others disagree
163 2015-09-03 09:26:36 <wumpus> yes, people want to fork. I wonder if that can be done with as little damage as possible.
164 2015-09-03 09:27:35 <wumpus> a "responsible fork" would at least change the network identifiers, maybe change the transaction format incompatibly, to avoid crosstalk
165 2015-09-03 09:27:54 <CodeShark> that's called an alt :p
166 2015-09-03 09:28:13 <wumpus> no - it shares history with bitcoin
167 2015-09-03 09:28:27 <wumpus> no alt does, afaik
168 2015-09-03 09:28:39 <phantomcircuit> wumpus, er CLAMS does
169 2015-09-03 09:28:44 <CodeShark> despite my concerns regarding validation costs, to me it isn't really so much about this particular fork - but setting a precedent
170 2015-09-03 09:28:48 <CodeShark> that's the real danger