Monero Research Lab Meeting - Wed 26 October 2022

[10-26-2022 16:59:51] <UkoeHB> meeting time https://github.com/monero-project/meta/issues/745 [10-26-2022 16:59:51] <UkoeHB> 1. greetings [10-26-2022 16:59:51] <UkoeHB> hello [10-26-2022 17:00:09] <ArticMine[m]> Hi [10-26-2022 17:00:15] <rbrunner> hello [10-26-2022 17:01:51] <one-horse-wagon[> Hello. [10-26-2022 17:02:25] <Rucknium[m]> Hi [10-26-2022 17:02:36] <jberman[m]> hello [10-26-2022 17:02:43] <plowsof> hi [10-26-2022 17:02:59] <xmrack[m]> Hi [10-26-2022 17:03:49] <UkoeHB> 2. updates, what's everyone working on? [10-26-2022 17:04:58] <UkoeHB> me: did a seraphis walkthrough with some monero devs, working on updating the library to support multisig txs with legacy inputs, after that is the very last library component: coinbase tx type [10-26-2022 17:08:01] <Rucknium[m]> Bringing some R code to production level that measures the out-spends for bitcoin CoinJoins. Not strictly Monero-related, but should help CryptoMorpheus with the moneroj.net charts that compare XMR txs with other "private" cryptocurrency options. [10-26-2022 17:08:17] <jberman[m]> setting up a stress test framework on the pool (with an aim toward getting PR 8076 past the finish line), working on fixing a bug in my re-scanning via scan_tx PR 8566 (good spot by plowsof), submitted a PR 8619 for background sync without the spend key [10-26-2022 17:12:19] <UkoeHB> 3. discussion, any topics on anyone's mind? [10-26-2022 17:14:02] <rbrunner> I have collected a number of issues around the coming Seraphis wallet: https://github.com/seraphis-migration/wallet3/issues [10-26-2022 17:14:17] <rbrunner> Some are closely related to that project, but some are also more general decisions [10-26-2022 17:14:38] <rbrunner> So every Monero dev should find something there to leave their opinion [10-26-2022 17:15:44] <rbrunner> It's a bit lonely there so far, with me writing, and then getting some comments mostly from UkoeHB and dangerousfreedom :) [10-26-2022 17:15:55] <vtnerd> hi [10-26-2022 17:16:43] <vtnerd> fixed d++ bug, and focused on serialization branch, will soon replace zmq serialization stuff, so the patch is pretty big now [10-26-2022 17:17:40] <UkoeHB> yeah migrating to seraphis would/will be a very big project so a bit more attention from monero devs in general would be encouraging [10-26-2022 17:18:05] <Rucknium[m]> vtnerd: May be worth a look: BCH devs are discussing using WebRTC as an alternative node communication port to reduce node communication censorability: https://bitcoincashresearch.org/t/webrtc-websocket-based-p2p-networking-for-bch/928 [10-26-2022 17:18:18] <rbrunner> Your walkthrough showed me the size of the task ... [10-26-2022 17:18:32] <UkoeHB> especially considering I don't plan to do a lot of dev work once the seraphis lib is done [10-26-2022 17:19:33] <vtnerd> I have no idea (at a quick glance) why anyone would want to do that for p2p comms [10-26-2022 17:19:46] <jberman[m]> I'm almost there, ready to dive in soon as I finish this slate of tasks [10-26-2022 17:20:04] <vtnerd> I mean, I guess monero nodes running in js-browser thing is interesting, but my goodness [10-26-2022 17:21:36] <UkoeHB> even setting aside seraphis dev, there is an enormous amount of code review necessary [10-26-2022 17:22:19] <rbrunner> That will get interesting, yes. With many thousands of lines of code. Brand new. Dense at times. Promises a lot of fun [10-26-2022 17:23:05] <one-horse-wagon[> rbrunner: Are we going to have the manpower to do it? [10-26-2022 17:23:19] <Rucknium[m]> What skillset is necessary to help push Seraphis forward after UkoeHB is done with his part? [10-26-2022 17:23:23] <UkoeHB> even for me it will probably take 2-3 weeks to do a final cleanup review [10-26-2022 17:23:34] <rbrunner> "it" meaning implementing or reviewing or both? [10-26-2022 17:23:49] <one-horse-wagon[> rbrunner: Both. [10-26-2022 17:24:14] <rbrunner> Frankly, I don't know. I guess nobody knows. But probably not a reason to not try it. [10-26-2022 17:25:21] <rbrunner> and maybe we will be able to invoke "special circumstances" if need be. I don't think somebody said back in 2014, after forking CryptoNote, "let's review everything first", right? [10-26-2022 17:26:35] <Rucknium[m]> If there was a clear job description, then MAGIC could put up a job ad. Just an idea. As it stands, we will have to reach inward, but that's limited. I'm thinking for implementation, not review. [10-26-2022 17:26:43] <UkoeHB> not everything is critical, for example it's probably ok if balance recovery has some minor bugs [10-26-2022 17:27:58] <ArticMine[m]> Defining the Job description then would be helpful. [10-26-2022 17:28:04] <hyc> rbrunner: back in 2014, core team commissioned a full code review and attempt to document APIs [10-26-2022 17:28:07] <rbrunner> I think for the wallet proper we do have the necessary manpower secured. But to get Seraphis running there will be a lot of follow-up jobs. [10-26-2022 17:28:26] <rbrunner> hyc: Color me surprised. Didn't know. [10-26-2022 17:29:07] <vtnerd> a job ad for ... code review? [10-26-2022 17:29:11] <rbrunner> There will be changes and adjustment all over the place. In RPC. In the CLI wallet, in the GUI wallet. [10-26-2022 17:29:29] <vtnerd> I mean its probably best if the existing contributors do it, even if its painful and long process [10-26-2022 17:29:56] <vtnerd> er you said for implementation, so IM not sure what [10-26-2022 17:29:59] <hyc> agreed. people should just sign up for chunks at a time [10-26-2022 17:30:30] <garth> I think it's possible that more Monero devs will give this attention as the switch to Seraphis comes closer. As of now it's still sort of peripheral. [10-26-2022 17:30:55] <one-horse-wagon[> The money is certainly there for however many CCS proposals are needed. [10-26-2022 17:31:03] <rbrunner> That's also my hope. So far it's not much more than a murmur in the background. [10-26-2022 17:31:21] <rbrunner> But once things really are in motion ... [10-26-2022 17:32:35] <rbrunner> Every wallet app dev who wants to use the Seraphis core wallet has to do at least 100 of review. How about this? :) [10-26-2022 17:32:44] <rbrunner> *100 hours [10-26-2022 17:34:16] <Rucknium[m]> AFAIK, our options are: 1) Acquire more resources; 2) Re-allocate existing resources (but we have no bosses); 3) Draw out Seraphis dev/implementation for an extended period of time; 4) Seraphis never reaches mainnet [10-26-2022 17:35:18] <hyc> so the question is whether there's enough devs to complete the implementation? [10-26-2022 17:35:20] <ArticMine[m]> I would suggest a combination of 1 and 2 [10-26-2022 17:35:26] <rbrunner> You can sort of mix 1 to 3. [10-26-2022 17:35:52] <rbrunner> Of course not too much of 3 :) [10-26-2022 17:36:24] <hyc> a job ad isn't a bad idea. just like Wolf0 was hired to do an open source miner way back when. [10-26-2022 17:37:03] <UkoeHB> the crypto world probably has more wallet devs than anything else [10-26-2022 17:37:14] <UkoeHB> might be viable [10-26-2022 17:37:54] <rbrunner> Hmm, why not, the Seraphis library is such a nice, mostly self-contained amount of code. Could be reviewed as that. Only problem: It will probably change quite a bit over time. [10-26-2022 17:38:25] <one-horse-wagon[> Wouldn't it be wise to lay out a roadmap of simple steps to the ultimate goal of total implementation. Then ask for CCS proposals? [10-26-2022 17:39:14] <rbrunner> Er, good idea, but believe me, we are still quite far away for being able to come up with a reasonable roadmap. [10-26-2022 17:39:20] <one-horse-wagon[> s/?/ along the way./ [10-26-2022 17:40:02] <UkoeHB> one-horse-wagon[: if you have some vision for a roadmap, that would be a good contribution [10-26-2022 17:40:05] <jberman[m]> Was planning on doing something like this for my next CCS. Past the hours on my current one already, just want to wrap up the tasks that were in there [10-26-2022 17:41:14] <rbrunner> Of course more people thinking and brainstorming about that are a big win. But it will be difficult to look ahead. [10-26-2022 17:41:39] <plowsof> will bulletproofs++ eventually be used in Seraphis? [10-26-2022 17:42:16] <UkoeHB> plowsof: unknown, we still need code and reviews of the paper (which afaik has not been published in a journal yet) [10-26-2022 17:42:37] <one-horse-wagon[> UkoeHB: I'm going to look into that. [10-26-2022 17:43:43] <UkoeHB> one-horse-wagon[: great, glad to hear it :) [10-26-2022 17:44:42] <rbrunner> Do we already know the approximate sizes of Seraphis transactions, compared with current ones? [10-26-2022 17:44:59] <rbrunner> Maybe BP++ will be very welcomed, to counter the size increase [10-26-2022 17:45:21] <UkoeHB> rbrunner: https://github.com/monero-project/research-lab/issues/91#issuecomment-1047191259 [10-26-2022 17:45:44] <UkoeHB> there haven't been too many changes since those tests [10-26-2022 17:46:01] <plowsof> asking about bp++ because i want to offer myself for doing any legwork where needed for organising it etc - knowing that it will benefit both the existing monero code base and Seraphis is great [10-26-2022 17:46:42] <UkoeHB> jberman[m]: you want to hand off that task to plowsof ? [10-26-2022 17:47:20] <rbrunner> UkoeHB: Thanks, interesting. [10-26-2022 17:49:19] <jberman[m]> yes I do [10-26-2022 17:49:50] <rbrunner> In those graphs, are we now in the "concise" case? [10-26-2022 17:49:57] <UkoeHB> rbrunner: squashed [10-26-2022 17:50:10] <rbrunner> Ah, ok. [10-26-2022 17:50:34] <UkoeHB> hence, txtype squashed v1 :) [10-26-2022 17:50:57] <rbrunner> Of course. How could I forget. Problably information overload. [10-26-2022 17:51:22] <Rucknium[m]> Looks like if our goal is to "offset" anything by using BP++, it would be in verification time rather than size. [10-26-2022 17:52:14] <UkoeHB> plowsof: ok if you want to organize getting a code proof of concept + paper reviews for bp++ that would be great; jberman[m] and I were thinking to approach Cypher Stack for a quote [10-26-2022 17:53:49] <jberman[m]> +1 [10-26-2022 17:54:52] <plowsof> sounds good to me [10-26-2022 17:55:44] <UkoeHB> as a reminder, here is what we know about bp++ https://github.com/monero-project/research-lab/issues/101 [10-26-2022 17:56:29] <rbrunner> Hey, that repository link there is new, right? [10-26-2022 17:57:00] <UkoeHB> not that new I think [10-26-2022 17:57:55] <plowsof> there may be some 'community politics' involved with that company which will require some navigation [10-26-2022 17:58:23] <rbrunner> So those are not those mysterious results from ooo's purported implementation? [10-26-2022 17:59:11] <rbrunner> "navigation". May be, yes. [10-26-2022 17:59:34] <UkoeHB> those are the mysterious results yes, the link is older [10-26-2022 17:59:57] <rbrunner> Although, interesting that Stack Wallet from them did not draw any critics, as far as I could see. [10-26-2022 18:00:10] <UkoeHB> plowsof: it's up to you how to go about it, or to look for someone else to work on it [10-26-2022 18:00:12] <plowsof> received very well indeed [10-26-2022 18:00:33] <rbrunner> Maybe the concerned people did not connect the dots ... [10-26-2022 18:01:06] <UkoeHB> trolls lost their momentum or changed direction perhaps [10-26-2022 18:01:17] <UkoeHB> anyway we are at the end of the meeting, thanks for attending everyone

monero-project / meta

Monero Research Lab Meeting - Wed 26 October 2022 #745

743