Submind YouTube summaries
Thumbnail for Paladin Community Call - 2026/09/01

Paladin Community Call - 2026/09/01

Watch on YouTube

Video summary

The Paladin Community Call for September 2026 primarily focused on recapping recent progress and outlining the roadmap for future development, with a significant emphasis on performance optimization and versioning strategies. The team highlighted that version 1.1 will largely consist of non-functional enhancements aimed at improving system efficiency rather than introducing major new features. Key achievements include reaching triple-digit transactions per second (TPS) through optimizations such as reducing memory allocations to minimize garbage collection overhead, optimizing lock holding times, and improving data transmission between nodes for co-located queries. These efforts established a crucial performance baseline to ensure that future releases do not regress these gains while preparing the codebase for more architecturally significant improvements in subsequent versions. Looking ahead to version 1.2 and beyond, the discussion centered on two major functional areas: multi-tenancy support and enhancements for private smart contracts within the Pentec layer. Multi-tenancy aims to allow a single Paladin instance to serve multiple users with distinct identities and access controls, addressing organizational challenges where running separate instances for every user is impractical. Additionally, the team plans to address scalability issues in Pentec where single-account state changes involving large maps currently result in inefficient state diffs; future updates will refine this mechanism to make private smart contracts easier to use and more scalable. The community also discussed the importance of clear versioning across different components like domains and domain factories to ensure backward compatibility and transparent sequencing rules for transactions. A notable topic addressed was the integration path for the Java SDK, which has been developed independently but is intended to be merged into the main Paladin release once it reaches maturity. The speakers debated whether the SDK should be released separately or bundled directly with a specific Paladin version, such as 1.2, depending on when the mentorship program concludes and the SDK is ready for production. To accommodate the project lifecycle of the Java SDK without constraining its development timeline, the team proposed aligning the Paladin release schedule with the readiness of the SDK, ensuring that the Maven artifacts are versioned alongside the main software. This flexible approach allows the community to support independent contributions while maintaining a cohesive ecosystem, with further details on timing and integration plans to be discussed offline via Discord. The meeting concluded with an agreement to continue using GitHub issues under the governance repository for community call agendas, replacing the less structured wiki format to better organize minutes, action items, and policy reminders. The hosts emphasized that while this month's agenda was light due to the focus on internal optimizations and planning, the next call is expected to feature a detailed roadmap visualization and further updates on the Java SDK. By maintaining open communication channels and documenting versioning rules clearly, the Paladin team aims to foster a robust ecosystem where various components can evolve independently yet remain compatible, ensuring long-term stability and growth for the decentralized trust community.
Read the full video transcript
Hi, Divian. Great to see you. >> Hi. How is it going? >> Very good, thanks. Yeah, very good. Um, how are you? >> Yeah, I'm good, thank you. We're um uh probably just going to wait another minute or two. We may or may not get other people joining. I didn't put anything on the agenda today uh around the SDK. Um >> yeah. >> Um so if there's anything you do want to mention, then feel free to let me know and I can put it on the agenda, but um otherwise Anna and I were just going to recap a couple of um couple of things. >> Yeah. Okay. as if like nothing from my side also but yeah let's see >> sure sounds good I'll go back on mute and we'll we'll start in a couple of minutes Okay. Okay. So, um it's nearly 5. So, um I think we'll start now. We'll see if anyone else joins. Um I guess mainly for the I guess for the recording. I think everyone in here is aware of the antitrust policy um which all of the Linux Foundation decentralized trust uh community calls are run under. Um uh so just bear that in mind please and also that um this meeting is recorded. Um so uh please stay on mute when you're not talking all those kind of things. Um so there's not not a huge amount this month to cover. We had a really good update from DVM last last month on the SDK progress and I know we got some more PRs um to review from you DVM on on the the next round of improvements. Um uh yeah, like I said, we didn't put anything on here to talk more about the SDK this month. Um maybe next month would be a good time to to do another update on some of the latest stuff and maybe demo some of the newer things if that's okay. Yeah, we'll see. I think next month will be interesting stuff coming. >> Great. That that would be really good. >> Um so the other thing to note is that we've moved to the community call agendas being um under git issues under the governance repo. So um at least one other LFDT project started using this mechanism kind of scales better than the wiki which doesn't have a lot of structure at least the default wiki doesn't have a lot of structure in in GitHub. Um so it's just be becoming an endless list and this way doing it means we've we've got a template. So when we create new community calls in the future, um you can choose the community call agenda template which adds in the antitrust policy links uh the reminders and so on. Um and this is a good place for people to drop comments. We can drop any actions minutes into here and then close the issues. So um I've put a note in the wiki pointing at this first one. Um so um uh yeah, we'll use this going forward and see how we get on. Um, so we just put a few small updates on here for this uh month. Um, a quick update on performance. So Anna mentioned this briefly last month, but um, we thought it' be good just to mention a little more, just a smidge more detail, just a couple of minutes um, on that. A bit of thoughts around what we might want to get into 1.1. So what we're kind of waiting for. Um, this this title here I think is kind of a bit too simplistic. Let me just edit. I think um I think we've got I'm just going to tweak this. [snorts] Uh I think that is a little better. Um uh so yeah bit around 1.1 when might we get it and what are we waiting for and then what what might be the features coming next. We haven't got um like a road map chart for for these two yet, but maybe next month we'll have a road map roadmap picture with a bit more detail on um and then I just wanted to mention some discussions Anna and I have been having it having in person because I think it's good for the community to hear some of that where we don't have them in the open because we happen to be kind of working colloccated. Um so Anna, do you want to just give us a smidge of an update on performance in a bit more detail than we had last month? So I think I mentioned last month I've been looking at some noto scenarios. So specifically the performance of no mints and no transfers. This has all been stuff we're looking at within a single domain instance. Um we've largely wrapped up that work now. Um there's still some code working through the pipeline. Um but at a rough target for what we've been aiming for here, we've hit triple figures in TPS. Um which is a big step forward from where we were. It's probably worth a lot of the Paladin code. This is the first time that we've actually performance tested Paladin code. Um so this round of testing and optimization has really just been about um establishing a good baseline. I'm sure in the future we will have more work to do where we do more architecturally significant work to raise performance. Um yeah, so a lot of the changes this time around have been areas like really reducing allocations so that we don't have the garbage collector hampering our performance. I would work around optimizing when we're holding locks. >> Um um work to reduce how much information sending between nodes to co um allow querying to happen colllocated where the information is that's assembly time um and yeah that's y some really good results. So now moving on to trying to achieve the same sort of effects for pente um and at the end of this we should have some baselines established which mean that as we go with future releases, we're making sure that we are not regressing these levels. >> Great. Thank you. Um, and that kind of leads into to a little bit because I think we certainly wanted version 1.1 when it comes out to benefit from a bunch of this performance work, which I think it will do. Um, and that's why I think functionally if we move on to item two a little bit, functionally u 1.1 is probably going to be kind of it's going to probably seem a bit light in in the amount of time between v v1 and now or when we put it out. Um, but I think that's because a lot of it is kind of non-functional performance enhancements and um, uh, refactoring existing logic so it performs and and is just structured better I think in a lot of the code. Um, so the main thing I think that we will probably be getting in for version 1.1 is ZTO's multi- version support and backwards compatibility with existing ZTO tokens um to get it in line with noto um there are some fixes around not coming in as well but but they're really kind of bug fixes business as usual to an extent um and then so so stuff that we'll talk about in more detail as we have time to kind of design and plan around it a bit more for like maybe a 1.2. So this is kind of item three here. So the I think two of the main items we've been talking about are multi-tenencies. So if I have a an instance of Paladin and I have several users who um are going to use different identities or they um have no need to have visibility of each other's uh transactions but there is a kind of a a tenency benefit to just running one Paladin instance but having different uh users uh with different um access control, excuse [clears throat] me. um that's kind of all hung under this banner of multi-tenency. At the moment, we've had some users kind of express an interest in that um where they have a large number of discrete users who um running an an instance for every one of them would be kind of organizationally challenging. Um so that we are going to be giving more more thought to it. Um and we'll have maybe some some thoughts around that for future calls. And then one of the other things that we'll probably look at because I think this has come up on Discord and we were kind of aware of it but as as people start to to use Paladin a bit more in earnest it's going to come a bit more um is around excuse me again um the way that uh private smart contracts in Pentec changes um are implemented. Um so a single account um which means a single contract and its storage um is represented by a single state. So changing one entry in a map um where there may be hundreds of other entries in the map uh result in one very large state change diff um as the input and output UTXO behavior. So that doesn't scale very well. Um and I think that's something that we'll be looking to improve. Um, we've had some minor kind of chats about it in person. Um, but I think that's probably one of the functional changes that will make Pente much easier to use in certain for certain types of contract. Um, okay. And then yeah, the only other thing that this item four here was really um again partly just to relay some of the thoughts and discussions we we have in person because you know as an open source community um we benefit from Anna and I benefit from from being loc. So some of the things we've been talking about which may may kind of come up in a more concrete fashion in the future of our PRs and issues are just generally around versioning. One part of that is um forwards and backwards compatibility between in versions of Paladin. So we have version one we have put out a statement around the point at which we will endeavor to have uh new versions of Paladin always compatible with previous versions for sequencing of transactions. Um so that's one area. Um but also as we've been having conversations we realized we're talking about many different categories of version be they versions of domains um versions of domain factories um how that how they relate what kind of what rules exist today for our existing domains in terms of supporting for example noto tokens that were created from an older version of a factory as well as creating new tokens with a newer version of that specification. and so on. So I think we're going to just try and partly maybe some docs, partly some guidance around what the different versions relate to. Um not all domain implementations will follow the rules that the the reference domains have used. Um but it's useful I think to be really clear in the documentation um the kind the different versions that exist and the current behaviors we we endeavor to have for the reference domains at the very least um so that we can have conversations that are nice and clear which ones we're talking about. So just general versioning um discussions because there are lots of not not plugins in the traditional sense yet but there are lots of discrete components that could arguably um be used independently of each other and they all have tended to have versions that sound very similar but actually maybe quite different version one being a good example. So um fairly short but sweet. Um maybe we will aim next month as we said Divvam to have maybe another update on the Java SDK which I know is kind of going going really well. Um uh I don't think I had anything else for today. Um any other business either of you aware of? Okay, great. I think that's uh that's good. >> Hey, I have a like question to ask. >> Yeah. So, uh since like you guys are uh have you know discuss discussion of like versions on this call. So this thing came into my mind uh like the Java SDK that I am working on like we'll be launching it independently uh or like that will be included in the next version of Paladin only. >> Yeah, that's a good question. Um so currently the SDKs are shipped within um um as part of Paladin's build and release process. So my thinking around this have been that when we're happy that the release branch is is pretty close to complete, we would align that with a release of Paladin to merge the release branch into to main. Um >> my my kind of gut feel is that that might be around the time that we would cut 1.2 of Paladin. Um, no, no, we we we could try and align a release just around the timings of the one of the Java SDK being ready because I think it would be kind of a substantial new contribution to Paladin that also because I know your your mentorship program runs for a period of time and and we don't want to necessarily we don't want to pin your kind of your project life cycle to other work we're doing necessary. So it might be that we plan for let's say Paladin 1.2 is released when the Java SDK is ready and then the SDK would be just part of Paladin 1.2. I think that's that's my thinking and and then the the Maven artifacts would be kind of versioned along with the Paladin version. Um >> okay >> does um but I may maybe um offline of the call if if you have any kind of kind of comments or questions or thoughts around that Divian do do give me a shout on Discord if um if you want to discuss that in a bit more detail particularly like the timings of when the mentorship program ends when you think you you're going to be aiming to wrap up and then we can maybe pin pin down some proposed dates a bit more closely to help you out with your planning. Yeah. Yeah, sure. >> Great. Okay, cool. Thanks everyone. I think that's everything for today. Thanks for joining. Um, and we'll catch you on next month's call. >> Yeah. Bye. >> Bye for now.