08:11:50 <dan-shearer[m]> #startmeeting 08:11:50 <lumosql-meetbot`> dan-shearer[m]: Meeting started at 2022-10-28T08:11+0000 08:11:51 <lumosql-meetbot`> dan-shearer[m]: Current chairs: dan-shearer[m] 08:11:52 <lumosql-meetbot`> dan-shearer[m]: Useful commands: #action #info #idea #link #topic #motion #vote #close #endmeeting 08:11:53 <lumosql-meetbot`> dan-shearer[m]: See also: https://hcoop-meetbot.readthedocs.io/en/stable/ 08:11:54 <lumosql-meetbot`> dan-shearer[m]: Participants should now identify themselves with '#here' or with an alias like '#here FirstLast' 08:11:57 <rubdos[m]> #here Ruben 08:12:08 <bjk621[m]> #here Björn 08:12:28 <moonshine[m]> #here Gabby 08:13:13 <valldrac[m]> #here Oscar 08:13:31 <dan-shearer[m]> #meetingname Weekly catchup 08:13:31 <lumosql-meetbot`> dan-shearer[m]: Meeting name set to: Weekly catchup 08:13:44 <dan-shearer[m]> #here Dan 08:13:58 <dan-shearer[m]> #undo 08:13:58 <lumosql-meetbot`> dan-shearer[m]: Removed event: 85a1c7b4949840a6997af92bf21ed9be@2022-10-28T08:13+0000 08:14:10 <dan-shearer[m]> #meetingname Bi-weekly catchup 08:14:10 <lumosql-meetbot`> dan-shearer[m]: Meeting name set to: Bi-weekly catchup 08:14:27 <dan-shearer[m]> hmm I don't know what I applied #undo to 08:14:28 <dan-shearer[m]> #here Dan 08:14:53 <dan-shearer[m]> #topic Admin items 08:15:42 <dan-shearer[m]> #info Dan and Björn met in Göteborg to try to advance LumoSQL organisation and funding and tech resources 08:16:05 <dan-shearer[m]> #info Dan is in Amsterdam Tues/Wed next week trying to do the same thing 08:16:19 <MartinaPalmucci[> #here Martina 08:16:28 <dan-shearer[m]> #info There have been significant publications from Martina, Oscar and Ruben 08:16:56 <dan-shearer[m]> and that is all I can think of as general stuff. Any more? 08:20:03 <bjk621[m]> Should we try to conclude the NLnet stuff? Maybe this better discussed outside dan-shearer? 08:20:21 <bjk621[m]> Ie this phase of ... 08:20:52 <dan-shearer[m]> Um I think all reminbursements have been completed. There are some new milestones to claim I know that. 08:21:04 <dan-shearer[m]> What do you mean? 08:22:11 <bjk621[m]> I know we have some milestones left. I was thinking about creating a plan (or so) how to do them. 08:22:20 <dan-shearer[m]> Ok that is a good #action 08:22:43 <dan-shearer[m]> Are you ok to coordinate that? I have input. 08:22:47 <bjk621[m]> I am ok 08:22:58 <dan-shearer[m]> Others have input. But maybe the LumoSQL Labs Director could coordinate that. Thankyou Björn. 08:23:49 <dan-shearer[m]> #action Björn to propose plan for completing and claiming all remaining milestones for NLnet LumoSQL Phase II 08:23:56 <dan-shearer[m]> We also can thank Björn for contributing a powerboard to LumoSQL. Because I forgot my adapter so we created one. Little powerboard, you died for a good reason. You will be remembered. 08:24:01 <dan-shearer[m]> uploaded an image: (205KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/aUJhUCLQigtuTbqxJsamNOUB/IMG_20221025_110949.jpg > 08:24:27 <dan-shearer[m]> Ok that's good. 08:25:49 <dan-shearer[m]> Hmm ok 08:26:46 <dan-shearer[m]> welcome Martina. Try typing #here Martina" 08:27:02 <MartinaPalmucci[> <MartinaPalmucci[> "#here Martina" <- :) 08:27:27 <dan-shearer[m]> oh oops 08:27:32 <dan-shearer[m]> Martina Palmucci: are you ok to do a quick section on your work? 08:28:38 <dan-shearer[m]> #topic LumoSQL Cryptography Documentation and Design 08:30:05 <MartinaPalmucci[> Ok, now we have a prototype of a fine-grained encryption algorithm. The algorithm can encrypt and decrypt any resource based on roles, or attributes 08:30:37 <dan-shearer[m]> #link https://lumosql.org/src/lumosql/file?name=references/LumoSQL-Thesis-Martina-Palmucci-2022.pdf is Martina's thesis 08:31:10 <MartinaPalmucci[> The collusion-resistance of ABE should be implemented to get a “beta-version”, so to be ready to integrate the solution with LumoSQL project 08:32:04 <dan-shearer[m]> #info Part of the thesis is a step-by-step explanation of how we get to ABE from more basic mathematical/crypto building blocks. 08:32:20 <rubdos[m]> ABE minus collusion resistance, according to most popular definitions of ABE. 08:32:35 <dan-shearer[m]> I found it very helpful, although I now have various questions and clarifications but I'm sure that can be fixed over time. 08:34:25 <dan-shearer[m]> So if we can imagine the latex source of Martina's thesis, do we expect that it can continue to be developed into not only background but also accurate description of the LumoSQL ABE+anticollusion underneath PE-SQL, and then to continue to include PE-SQL in a similar level of detail? 08:34:50 <dan-shearer[m]> Or is it more of a reference resource for LumoSQL documentationj. 08:35:28 <rubdos[m]> I would say it's a reference 08:35:41 <dan-shearer[m]> For example, when reading it I did some diagrams for my own education about how the different crypto schemes relate to one another. 08:37:34 <dan-shearer[m]> Ok. So that is an action item at some point. LumoSQL needs some kind of documentation that is a design spec for the crypto. Because too much security software exists where there is a large gap between the papers describing eg how an algorithm should be implemented and the actual software implementation. 08:38:08 <rubdos[m]> Sounds like that would go in the RFC, and additional to that some more detailing documentation. 08:38:50 <dan-shearer[m]> ack 08:39:06 <dan-shearer[m]> Ok, so Martina has done a lot of work to give us a resource for LumoSQL project documentation, and her thesis is in the references/ directory. 08:39:55 <dan-shearer[m]> And well done Martina, because you graduated with a fancy wreath like a Roman Emperor and everything! 08:40:23 <rubdos[m]> Yes, congratulations are in place. And parties :-) 08:40:26 <rubdos[m]> 🎉 08:41:10 <bjk621[m]> +1 08:41:32 <dan-shearer[m]> I'll come to the party in Brussels. Italy's great but LumoSQL runs on Brussels time. 08:41:35 <dan-shearer[m]> Good. 08:41:48 <dan-shearer[m]> Ok Ruben how about a topic called "PE-SQL" update? 08:42:11 <MartinaPalmucci[> Thank you so much 😊 08:42:23 <dan-shearer[m]> #topic PE-SQL Update 08:43:29 <rubdos[m]> right 08:43:37 <rubdos[m]> I've been busy traveling to Lausanne, London, and Firenze this month, but last two weeks I've found some time slots for SQL-PE 08:44:04 <rubdos[m]> #info SQL-PE now has some very rudimentary password-based authentication that works with Argon2 and Chacha20. 08:44:29 <rubdos[m]> There's still some ugliness to work out of the API (wrote some tests that fail) when it's really in context of a database, but the base idea is theer 08:44:49 <rubdos[m]> I wanted to finish that up today, but I'll probably have to switch that to Monday because of teaching activities 08:45:35 <rubdos[m]> #info Currently, there's no record encryption yet in place, it's all about role management. The idea is to start incorporating Martina's work soonish. 08:46:05 <rubdos[m]> I think that's about it, modulo questions 08:46:57 <dan-shearer[m]> I have not tried this code. But in principle does it mean that you have a password mechanism ready for you to add additional attribute critera to, attributes that are associated with keys. 08:47:37 <rubdos[m]> I don't comprehend that sentence. The password is for password-based authentication 08:47:40 <rubdos[m]> lumosql_authenticate_user("ruben", "hunter2"); 08:47:53 <rubdos[m]> just like what you do in postgres 08:48:21 <rubdos[m]> #info Additionally, the authentication code is now generic enough to support arbitrary other auth methods 08:48:29 <rubdos[m]> smartcards etc. are all possible :) 08:48:38 <dan-shearer[m]> Quite. But is there any structural connection in your code to the bits we expect to come next, or, is this the necessary step but orthogonal to the more advanced things. 08:48:51 <dan-shearer[m]> ah ha . 08:48:52 <rubdos[m]> Orthogonal 08:48:54 <dan-shearer[m]> Thankyou. 08:50:30 <dan-shearer[m]> We have a slow trickle of interest from people who want "SQLite with better encryption but similar API" who are discovering LumoSQL btw. 08:51:01 <rubdos[m]> that's great :) 08:51:08 <rubdos[m]> I think that's the end of topic SQL-PE ? 08:52:17 <dan-shearer[m]> Yes I guess so. It's great to see. 08:53:11 <rubdos[m]> FYI, I'm gone at 1140 because lunch and teaching 08:53:25 <dan-shearer[m]> I think the next topic is a bit of a summary wrapup. moonshine I guess you noticed what Björn pointed out about LumoSQL milestones and I think that can involve you if you want. 08:53:58 <bjk621[m]> I will leave on 1128 because of other meeting 08:54:41 <dan-shearer[m]> You will all have time for multiple coffees because this meeting needs to document some action points and be open for contributions, and then we are done. 08:55:05 <dan-shearer[m]> #topic Other items and next steps 08:55:55 <dan-shearer[m]> #info valldrac has pushed a first prototype of a Signal Server standalone system. This is relevant to LumoSQL in multiple ways. 08:57:35 <dan-shearer[m]> #info The action item Björn took for coordinating finishing the current NLnet grant will mean individual discussions with @moonshine, Ruben De Smet , Labhraich and me for specific next steps resulting in code into repos 08:58:19 <moonshine[m]> definitely, I have looked at Martina's thesis and I think I can put the relevant parts into the documentation, not sure what level of detail it should be, probably the easier to read the better 08:59:14 <dan-shearer[m]> So that is the cool thing, if non-mathemeticians can correctly translate that into general documentation, then that is great. 09:00:23 <dan-shearer[m]> The connection between valldrac's work and LumoSQL might not seem so obvious at first, but here are just some of the overlaps: 09:01:01 <dan-shearer[m]> * Signal clients such as https://molly.im and Whisperfish (hello Ruben) depend on SQLite-with-crypto, but, they also depend on an independent server otherwise the testing and benchmarking will never be good enough 09:01:58 <dan-shearer[m]> * valldrac has been thinking a lot about authentication mechanisms and implementing some code for it, and that involves both client and server. And LumoSQL is thinking quite hard about the theoretical basis for authentication too, obviously 09:02:46 <dan-shearer[m]> * Signal Server needs some storage systems and they are quite a mish-mash mess at the moment. Whether LumoSQL or Lumions or something else is an interesting design discussion 09:02:52 <dan-shearer[m]> Those are just some connections. 09:03:11 <dan-shearer[m]> valldrac does that seem approximately correct to you? 09:06:03 <dan-shearer[m]> ok 09:06:06 <dan-shearer[m]> well let's say that's agreement 09:06:36 <dan-shearer[m]> bjk621 - broadcast email noted 09:06:46 <dan-shearer[m]> Oh yes 09:07:19 <dan-shearer[m]> #info This week Labhraich added RISC-V to the list of architectures that LumoSQL supports. 09:07:58 <valldrac[m]> dan-shearer[m]: Yes, in a very short way, the idea is to make better Signal server a clients integrating LumoSQL. We think it's a good choice to be an early adopter of LumoSQL 09:07:59 <dan-shearer[m]> That is a big deal. RISC-V is something that represents hope for all of us against hardware that has spooky unknown scary stuff deep inside it. 09:08:37 <valldrac[m]> s/a/&/ 09:09:24 <dan-shearer[m]> #info The Molly, Sweet Lies and Whisperfish projects all want to make better Signal servers and clients, including by integrating LumoSQL. We think it's a good choice to be an early adopter of LumoSQL 09:10:56 <dan-shearer[m]> #info Three people in this chat have asked me why we do not have a parallel audio or video channel to this text meeting. 09:12:03 <dan-shearer[m]> The answer is because I have not really thought about it. I suppose it would be an improvement if we can do it inclusively, and I think that is possible. We could include a URL in the heading of the #startmeeting banner. 09:12:03 <dan-shearer[m]> ok I think that is everything? 09:12:13 <dan-shearer[m]> open now for all contributions and things I have forgotten :-) 09:13:19 <rubdos[m]> I assume jitsi 09:13:27 <rubdos[m]> But if we're doing video, let's not do chat 09:13:55 <dan-shearer[m]> is jitsi inclusive accessibe 09:13:55 <rubdos[m]> I would think so, and otherwise patchable 09:14:22 <rubdos[m]> Meetings are long enough as they are, and if we're mirroring the Jitsi to IRC/Matrix, it'll take even longer. 09:14:22 <dan-shearer[m]> Ok. I know Jitsi works, I have used it many times and run a server 09:14:28 <dan-shearer[m]> Ok 09:14:38 <dan-shearer[m]> All right, so I will leave that as an open question 09:14:54 <rubdos[m]> rubdos[m]: This is an assumption, by the way 09:15:26 <dan-shearer[m]> I can definitely see a mumble audio channel being simple and probably an improvement. 09:15:28 <dan-shearer[m]> ok 09:15:50 <dan-shearer[m]> I think that is everything 09:15:51 <dan-shearer[m]> 60 seconds and I will close the meeting! 09:16:10 <dan-shearer[m]> The last two weeks was really good and we should all feel pleased that there has been progress. 09:16:34 <dan-shearer[m]> Including in things committed to repos. And I think there will be even more of that in the next two weeks. 09:17:06 <dan-shearer[m]> #endmeeting