500+ entries in 0.005s
mod6: mp_en_viaje: Ahhh, alright, thanks for the update.
mod6: mp_en_viaje: Hey there, Sir. Now that you're back to CR (glad you had a good trip), plz to process mpex withdrawls when you have an extra few minutes. Thanks in advance.
mod6: Anyway, I haven't even eaten today. So I'm gonna head out.
mod6: I'll just, withold the patch. Make a blog post. If people want it, I guess they can say.
mod6: I'll make a proper post on my blog, indeed. Just want to get this out for folks; now that there is a wedger tool out there, and it's been a number of days of reorccuring wedges.
mod6: Going to push up my logs & notes, and then make a vpatch.
mod6: Took some work and chasing some red-herrings in net.h. I drove through the testing and changes while asciilifeform helped think me through the problem.
mod6: Alright, got a fix for this. *whew*
mod6: If others have this same type of behavior, please write in and say, show us the debug.log if you can. Thanks.
mod6: Apparently main.cpp:ProcessMessage() allows for quite a number of these getdata commands before flagged as 'Misbehaving'. But I've not seen waves like this before now. Quite possible I've just not seen them either tho.
mod6: Happy to provide the 12G log if you want that, or a shorter one that shows on 22nd or 23rd my node starting, then getting wedged after wave.
mod6: When alf alerted about this wedge on the 22nd, I was quick to point out this wave of stuff that I don't typicaly see.
mod6: I have a debug.log that is ~12G that goes back ~3 mos to block #605198, i do see those requests, but in the dozens at a time at most. but not a tidal wave of tens of thousands.
mod6: bvt: Yes, I experience exactly like you describe; right before 'wedge', incoming 30k or so of 'received getdata for: block' in debug.log.
mod6: Yeah, I would, that'd be great.
mod6: I'll throw something in here if it's relevant.
mod6: Thanks for your work on this, I'm following along in #asciilifeform log.
mod6: Thanks for the article & trace bvt.
mod6: jfw: I don't have any information regarding what miners of any size are currently running.
mod6: Your notion is correct on the former, 'more exposed to potentially misbehaving or sybil peers'. As to the latter, I suspect you will be connected simply more prb nodes. It is possible that you'll also find, in a 'subgraph', nodes that are playing nicely, but seems less likely.
mod6: Thanks jfw. You've removed 'malleus_mikehearnificarum.vpatch', noted.
mod6: Heads up to TRB users, seems that nodes have wedged on block 618406. A simple restart of TRB seemed to resolve it. Not sure on the cause yet. Will update with more information as I have it.
mod6: mircea_popescu: re: rss feeds & feedbot; will check it out. sounds like a good way to stay abreast.
mod6: Alright gents, I'm gonna go make some food here. bbl.
mod6: I appreciate the poking re things TRB.
mod6: I like this better than mirrors too, as it seems that those always tend to disappear or end up way out of date without explanation.
mod6: Yeah, not that we would want stuff from nonpersons, just in the case where, as we've seen recently (even with my own blog) where it was down for a few weeks, or months as people are between hosting or such.
mod6: I suppose that's fair.
mod6: (I appreciate this discussion, fwiw.)
mod6: I suppose the foundation should maybe have its own blog then.
mod6: jfw: This might be fine, but what if I'm not the guy? Meaning, let's say that there are threads on my blog, people publish stuff in the comments, or wherever on there, i dunno, pingbacks somehow. If someone else takes over, I get hit by a bus, blog goes bye bye from non-payment, what happens to thread on new guys blog?
mod6: But again, if there is going to be a future solution for this in the form of some kind of existing bot extention, then I'm all for that. At least it would, presumably, notify the gardener (somehow) that there is something to inspect.
mod6: All I'm saying, is that without said possible future solution, it may be somewhat difficult to track through 39 different blogs watching for patch submissions all the time.
mod6: Anyway, sounds like there might be a solution to this between what billymg and lobbes are working on + maybe another added step?
mod6: jfw: aside from a possible mp-wp (future) solution to the issue; several clicks notwithstanding, I mearly worry that somehow, the gardener may simply overlook or miss one.
mod6: Ah, alright. Maybe that's it, just a few moving parts here.
mod6: Ok, I saw this in January, I may need to re-read it, but on first pass I didn't gather that this would do what I needed it to do.
mod6: Now the idea to make a bot that would somehow interact with blogs, is an idea I hadn't thought of before.
mod6: I fully agree, mailman is a ancient artefact back from, probably literally, the pre-september internet days. Yes, it's quite, well, frustrating to say the least.
mod6: yes. a bot that would do this over irc, is basically what I've been thinking about.
mod6: I dunno, anyway, I thought it could be good for people to go ahead and submit a vpatch, and seal to the bot, where it can be checked for L1/L2, and stored for the future, all in one place, making less work for the gardener.
mod6: Not that I love having a centralized bot, either. It's just, I'm not 100% sure what the best way to go is on this. The mailing list was centralized, it worked fine.
mod6: TMSR Lords and others seem to publish all their code on their blogs, which, I think is fine. But my hang-up with allowing people to post TRB patches/seals on their blogs instead of sending them in is two-fold: 1) It puts it on me to chase these down. 2) Then I have to place them somewhere for long-term keeping anyway. As we've seen, people's blogs get rather large, hard to find things, or disappear complet
mod6: Now, I've thought that over a bit; the first thing that came to mind was deedbot. However, I think that deedbot probably isn't the right place for developer doodles. Plus, each one costs actual money.
mod6: I find it fairly easy to go and dig stuff up in there if I need to do so. I tried to stand it back up, several times, in fact, since it's home on Pizarro went down. But I haven't been successful there. Upon speaking to jurov about it, he suggested instead that we create a bot for this purpose instead.
mod6: For me, I've always liked the mailing list. It's cumbersome, and it's had it's pain points in the past. Everyone, at one time or another, has had a problem getting things stuck in its queue, etc. However, when it works, it does work pretty well. I like how it checks the WoT on submission, has an archive, and we can all go back and look at it years and years later.
mod6: Regarding your question re IRC bot for vpatch submission: I've thought about this quite a bit. And I struggle with this.
mod6: mod6_phexdigit_fix << This one I'd like to put in for sure, as it seems, at least according to my own analysis that this is a legit bug (does anyone disagree with this analysis?). Just would need a simple regrind at this point. Can do this month as well.
mod6: asciilifeform_whogaveblox << I'm more open to comments on this one too, I don't see a big issue with it being in the tree. I'm also using this one in a test environment (for quite some time now). Haven't seen any issues with it at this time. It would need a simple regrind. Can do it this month as well.
mod6: mod6_excise_hash_truncation << This one in particular is nice, and I've been using it in my test environments for a while. I'm fine at this point with adding it in, it just needs a regrind (again). Can work on that this month.
mod6: (23:28) <+jfw> oh sorry, asciilifeform_aggressive_pushgetblocks is in there, just the other three then. << correct.
mod6: I've updated the How-To document to reflect that, but otherwise, have had good luck on my own with building. Am interested to hear how it goes for others, though.
mod6: Let me work through these here. Glad you're going to try out a build. Please do let me know how it goes. I did just find three 'gotchas' on CentOS (6.10); whereby 'bison', 'flex' and 'patch' all need to be installed manually. Apparently these are not a part of the default system.
mod6: Hi there jfw, thanks for the questions.
mod6: !!v 17FAC8B5F9FCD4B2777C04E1768304D6F8219E2E996264A7D432E874D8622754
mod6: !!v 69D33AF2BEE2D7948DCF9AEF91599BBA2050793B2288E0335E2DBF9173AE6A9A
mod6: !!v E87C94F1499C7735DB00C3A3973170B1C0BD469DE5BBCD07DF2A3780739EFA0D
mod6: !!invoice BingoBoingo 0.018 Winner of Auction #1078 (Foundation's Dell R610 Server in Uruguay)
mod6: Ah, is that how it usually goes? Auction creator invoices the winner of the auction?
mod6: (I then, in turn, will send the sum over to jurov.)
mod6: BingoBoingo: I'm not sure what the exact protocol here is, but I'm guessing you can pay me if you'd rather, since I ran the auction for the Foundation.
mod6: Alright, well looks I am buying 'Server-A' since there were no bids. Will pay 0.01700000 BTC to jurov (Foundation Treasurer).
mod6: That's fine too. I do actually want the server, so if bidding on it works, I'm fine with that as well.
mod6: Any thoughts on the better course of action?
mod6: I guess on the other hand, if the auction goes bid-less, then I'll just buy the machine for the reserve price from the foundation.
mod6: I'm hoping that if I throw in a bid with my mod6_auction IRC registration for auction #1077, this won't be an issue.
mod6: 'mod6_auction' is my new irc account for bidding on the foundation's auction for the dell servers.
mod6: I won't be back around until later today, but i'll check back in later.
mod6: The howto document now requires that you have Ada on your environment before you build, and the builder will also need to build starter_v_2 first, before any other steps are to be completed.
mod6: mircea_popescu: Ok, I have published my trb keccak regrind on the bitcoin.foundation site. It comes with the following: 1. Update to original genesis.vpatch - removes the UTF charater. 2. Added mod6_privkey_tools.vpatch (unchanged fro the original ML posting by myself.) 3. A manifest file. 4. I've also updated the howto document on thebitcoin.foundation.
mod6: thx, for denoting the offline component in scheme.
mod6: dorion: ah, hrm. it's had its ups and downs.
mod6: Upon resoltion of the privkey_tools question, will happily add it to the current archive of signed Keccak TRB Vptaches.
mod6: Also, mircea_popescu, I do have the the entire trb tree (with exception of privkey_tools) signed and ready to go in Keccak. Has been since last January. I havne't been able to find a Lord who will double check my work though.
mod6: I am also aware of the wallet work that has been on-going in python by gbw. Maybe this gets folded into that?
mod6: Anyway, I'm open to another discussion around the privkey tools vpatch. For what it's worth, I think TRB sorely needs it. But again, the whole discussion about the wallet.