 Welcome to October 19th, Microsoft DevSync. So we're at the start of a new sprint. And we're going to go through and do a quick wrap-up on the previous sprint. And then we'll see what's carried over and what new things we need to add. So, Chris, do you want to share your screen? And we'll take a look at the status of the current sprint. Sure. All right, these are just mine. All right, so it looks like those have been done for a while. So yeah, it looks like we got the database schema for UI elements done. Wow, we got a lot of stuff done. These look like there's just a lot to do with some rollover stuff. I guess I don't see a lot of Mark II prototype stuff. I don't know if that was expected or not. Maybe it's not aligned. OK, so let's go through the stuff that isn't done. I think we can skip the stuff that's in review unless there's anything in there that people are having trouble with. I guess, yeah, let's go through the review stuff real quick and make sure that none of that is actually being held up for any reason. OK, the change wake word file naming scheme, the only thing that's holding that up right now is that there's a configuration, a pilot configuration I need to make to get the CI to work. And actually, the pilot stuff is something I do want to maybe discuss during our summit as well. But anyway, once I get that configuration done, then the test score or the pilot's awful pass and it can be approved. That's the only one that up. OK, do you need somebody else's eyes on the pilot issue? No, I know exactly what it is. What I'm doing is I'm saving a kind of a global pilot file, pilot config file in our DevOps directory, our DevOps repository. And that file can be pulled in using the Jenkins files, can be pulled into each of our Python Jenkins files and then used to do linting. Can we put that in a public review? Good. But I don't know right now if there's no other repo that all of these, the nice thing about the DevOps repo is that all of our Jenkins files reach into there anyway for other things. Yeah, yeah, yeah. We could probably publish the contents of the file somewhere, but the reason it's going in there is just from a processing standpoint I needed there. I might. Anyway, I don't want to get into another thing. I may have started a new repo yesterday. And I might throw it in there. But I'll leave it at that and let your anxiety rise. But no, it's a legitimate good purpose repo. Don't worry. That sounds like something you should be talking about. Michael, while I'm looking at this, can you cut a new ticket for the SJ201 Rev-B enclosure-based scope, which could be the code that's where the board that Kevin is sending me tomorrow? Sure. Yeah. So the next one here is actually, I think the review is done, but did Ken, did you say we're just going to kind of scrap that anyway? Yeah. So what I'm going to do is I'm going to go through and clean it up according to the feedback I got and leave it as Rev-A, even though you don't like using part numbers as a naming convention. I will create the abstract case class though. And then Michael's going to create a new ticket for the Rev-B version of the board. And we can have both of them in our repository, even if we're just using Rev-B version. The difference is we're getting two boards. One is basically jumper configured to work the way the old one did, and the new one is jumper configured to do everything over USB. I have an update on that, but we can talk about it later. OK. Yeah. If I keep the Rev-A code, if we're not going to use it. What's that? What is it? Do I keep the Rev-A code? Yeah. I don't see the benefit of throwing stuff away. In fact, even the re-speaker stuff could be put in there under re-speaker. But yeah, I'm flexible. And all I'm getting at is that this is done only if you consider the fact that it's for the original Rev-A board, and potentially the one that emulates that tomorrow. So I'm just asking Michael to create a new ticket for the new board. Whether we care to keep them around or not, that's fine. We can throw away a code. I don't care. Well, we should talk about that in terms of, you know, maybe it would be better off in a branch. Do we want to handle these things through branches in the code or, you know, separate reposts? I'm not a discussant for right now. The way it's built, especially when it has an abstract days class, is that you just import a ledge, you import switches, you import the class for the volume. And then when you do that, the underlying code, based upon a configuration value, will bring in the board model version code. And it could have all of them there. And so Chris is just saying, I don't want all three versions, ReSpeaker, Rev-A, Rev-B, I just want the latest. And there's no need to fork it or branch it or anything. It doesn't matter whether you have three different sub-directories or just the latest sub-directory. It's fine. I don't care. Yeah. Okay. What's next on our review process here? Roadmap for initial release to community, guys? Yeah. I keep wanting to, like, flesh that out further, but I think we should just publish it as it is. And, you know, it'll get better over time. Has it been reviewed? Do we need to move it to somebody else who needs to review it? It's available there. Like, there's an overarching doc, that public document, draft public document there. So if anyone doesn't want that published, then let me know. Otherwise, it's going to get sent out. Yeah. I've reviewed it. You know, I think, I agree with Gez. We can continue to tweak it forever, but I don't see any harm in getting some feedback now. Okay. Can we move this to Don or are we going to leave it here, Gez, until you actually publish it? Yeah. Just leave it there. It'll remind me to paste it. Okay. Failing user deletes all alarms. Ken, who's that? That was a payout from me. And it might be merged already, I think. Yeah. Yeah. Okay. So this can be done. What's the next one? I might have, I might have already. Okay. There's a couple of times. Yeah. And the process status class. I have not looked at it. It's on my, it's on my list to review, but I have not, I've given some feedback to okay. And I think he's giving me some feedback back. And I haven't, just haven't circled all the way back around. So, um, and this other one here is real, is reliant on the process status stuff to be merged. So it's, we're document backup system for precise data. Joshua is, has that been reviewed by anybody? Well, I read it. Did anyone else read it? Yeah. I put it in the con form. It's a matter of whether you read it. It's a matter of whether you were able to restore from it. It's not, it's not designed to restore. It was just a documentation of the process by which we're backing up along with five new tickets for the things that we should improve in the future. So who should review this? Me since I'm doing the work behind it. It's a one page document, but yeah, go ahead and give it a shot. And you should be able to log into this backup server here and see everything coming through. It's by no means a comprehensive backup system, but I am, I do have a clone of your data. So here, and it is updating itself nightly. Did you see my email about what the different way cord directories before Joshua? I did not. My in, I did not. My inbox has 550 new emails in it today. Okay. I always find it to your initial message. So if you want to find it directly, there's some explanations in there. My, my solution to the problem was just to be a data whore and take everything. That's what that's fine. I think that's good. And I think it's good that we have a ticket to review the documentation. I think it's bad. We don't have a ticket to validate that the data has actually been backed up successfully. All right. That sounds like something that you could do. I would strongly recommend giving it to Chris since he's changed the way everything's structured. And I have to spend some time coming up the speed on it. And I'm a little bit overloaded right now, but if you want, you can give it to me and just set it to a lower priority. That's fine. Is this related to the process status stuff too? No, that one's done. Oh, it's done. Yeah. Yeah. This is another pull request that was merged. Yeah. Okay. So we're down to a smaller, what about this, this patent thing? Is that still being worked on? That was just kicking it to someone for review. I don't know if Ken, you've had a chance to look at it, but or if anyone else wants to have a look at it. I didn't. I commented. Oh, I did not say the comments. Damn, you give the box. All right. Cool. I'll go have a look at that. All right. I'll sign this back to you then, guys. Okay. In progress. Derek, you have a couple of things here for Mark two documentation. Yeah. Well, so I'm still kind of going. Yeah. We kind of roughly identified it. But I want to shape it up. So I've got that in there. And then the longer stating. I mentioned this to Michael. So I just, I kind of want to get it to that. So. Derek, what I didn't see in there is whether you want to, how do you want to identify. That the mute Mike. Is active versus inactive. And are you intending upon. Doing any kind of visual validation. At the device and then a listening mode. Right. Yes. And we do like, so we have some of that already in the mark one. Listening. Because of animation. So. So for the mark two, the intention of being to do something to LEDs. The idea is that when Microsoft's not doing anything. But yeah, the doctor would cover all that. So, you know, say, hey, my problem. L. These light up. And the kids that it's listening. Et cetera. And then after, you know, you say your, you know, maybe if it has to, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, you know, maybe if it has to, you've got a thinking state in the mark one. So that's, you know, if it takes a while to respond to you. And that, that is something that I want us to do with the mark two as well. So after it's done listening, I transition to that. Et cetera. Okay. So we need to get this done quickly because the microcode that's going to be in the firmware. Be in the firmware for the LED controller. We don't currently have a way of reflashing that once it's, once that controller's, you know, been programmed. It can be reprogrammed. It's just that it can't be reprogrammed over the I2C bus. So, so every one of these devices right now, you know, is going to be using that firmware. The more devices we have out there that need to get that firmware reflashed, the more annoying it's going to be. So we need to have a good spec on how to use those LEDs, at least the first pass. That, so, you know, we're not setting up a bunch of devices that aren't going to be able to support this. Nice to see that Kevin's, when Kevin's demoing it, he's running something. Like, you know, he built something. He can do it, but like, you know, if you want to have lights that, you know, pulse slowly or have them all turned on at the same time or whatever, you know, you know, I imagine it's a fairly high bandwidth system that he's using right now. I want to keep it simple, you know, we don't, I would, I'd advocate for, we just do the brightness and the color of the LEDs on the bus and dump a, and just dump a packet, or I don't know how it processes information, but for lack of a better word, a packet and have it unpack the, however many lights there are out of the packet. And that way, you know, we can refresh it at 30 hertz or whatever speed we need, and we don't have to worry about the animations being, being built into the firmware. Yeah, we can do that. Cause then we, then we don't, we don't have this piece of firmware sitting out there that, like, we could have this capability if we had just, you know, thought of it now, you know, instead we just write the entire, the entire state of the LED array as one data packet, which doesn't really have to be that big. I mean, you're talking, however many, it's not a lot of big. And then we just, we just, you know, update it periodically. Yeah. That'll allow us to go. Sure. Yeah. Guys, close the door and get out right now. Okay. Cool. Well, so close the door. It's on side. That's simple. It might blink a little bit, might use, you know, this kind of spinning thing. I think for thinking or for loading, that makes sense. So, yeah, but like Michael says, I need to get it. You know, I need to get it down on paper. So. All right. Michael, the Leonard first run of dev kits, does that plan completed? It is not. I need to have a, another discussion with, with Kevin about that. Basically, we need to finalize what our. PCBA vendors going to be, and we had some problems with the, the last one we chose. So. So I need to revise that. All right. Yes, how long before I get three of them? Yeah, good. I put some PowerPoint in there about what's remaining. And so we've, we've got a, chances got a draft. Like an example integration with Minecraft core. So that we can make sure that. There's no breaking changes. Before we actually cut the release. And so. And there's a little update to them, where the skill has to happen. Because of some of that weirdness that. Has previously been done. Like passing. Like equals none kind of thing. So. Yeah. Just got to, I think there's four unit tests failing and. In core at the moment. Which I haven't looked into yet, but I don't anticipate them being a big issue. And. And yeah, we'll cut a final point to release. And then the point three release. And that's in. All right. What about the benchmarking? Well, I've got you. Yeah, I've got a, I've got a few things. Particularly from a DJ. Is. Yeah. Yeah. I've got a, I've got a few things, particularly from a DJ is done. Much stuff. I want to do some extra. Things so that we can actually. Do it as much as possible. Side by side. Kind of comparison. And. But yeah, so it's still in progress. I'd say. Okay. But we should probably also think about, yeah, what's the actual. Ticket number. Precise 18. Yeah. But. Last, the note in the ticket says you're waiting for Chris there to get back with you on. File. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. I'd say it's hard to get back with you on. File. Naming or. Directors stuff. Let me just see what I might have said that. But okay. Anyway, it seems like something that's not on the top of your mind. So let's just carry it forward and. Maybe. I'm sorry. Just JIRA. It just takes forever. No, I mean, I'm probably has to do with waiting for the tag samples. That's. Horrible. All right. Guest news and singing skills air Yeah, these these were Waiting on that process status stuff But we've also we've also done some fixes With some other errors that were happening around tests leading over into Future tests and so I'm gonna take another look in the context of that because Potentially there's something there and now that we've got that the micro plugs and stuff So I need to get back to this it just hasn't happened since I've done that fix so Okay, skill day time randomly failing That's in the same class like it's yeah, it was only failing on the on the CI process and stuff So the PCBA for our state to want it sounds like we already talked about that Michael So you're gonna talk to Kevin about that. Yeah, which ticket are you looking at? Oh? Yeah, there it is. Yeah Yes, exactly. That's still in process Okay camera mode evaluation Derek I haven't finished that yet still in progress Implementing tagger GUI a I was able to implement the changes that we talked about on Friday in our meeting with the exception of How to get from one question to the next where it's obvious that we've switched questions So I'm looking at an angular as an animations Library I'm looking at doing something like sliding one in sliding another one out or something like that So that is the only thing left to do for this GUI that implementation that I can think of Okay, great So what's our process for getting that out in front of the rest of us to test and then getting it out to The community to test and then rolling it out. Yes, I get the animation coded up Then it's gonna go to PR and then once I get to the PR You know I just need to get it into a dev branch and then I need to do a Push and get everything all the database tables all the back end code all the front end code Get it all in the test and then there's there's a I can give you guys URLs to I'll probably need some data too. I mean, I think I know Some sort of test folder in the in the On the NAS that can I can get to test files with So it's a little work to do to get all that set up so that we can test with it, but that's I'm planning on doing this week That so we don't accidentally pollute The data source if something's not working correctly. Is that the thing? Yeah, yeah, well, yeah, it's good to have this was especially we're gonna be doing tag and tag You know, if you're just testing something you could incorrectly tag it, you know, just to test something so I don't Point it's a production data test the tagger Well, when we deploy it for internal testing we should be using it Well, we may want to test some malicious use as well, right? So we'll have to Yeah, and if anybody has the has the bandwidth to do it point your Point your device at the test API as well So you can you can submit your own samples as part of part of this change is the you know the collection of submitted samples as well so That would be useful to test I have been pointing mine at a completely non-existent URL to test out the you know Hey, my car that works and what is your line? Just before I address The personal pronoun use my personal pronoun both singular and plural I should finish this today. Hopefully and move on to getting everybody Set up so they can test it themselves Hey, my craft What is your writing update service from work to you Joshua? My next work IP address is still working on that hand to core has built our update service for us and I've been from pretty significant I'm working side of Belina in the last week and still haven't heard back from from a boon to about some of the questions that I asked So it's still open. All prototypes should have an enclosure and core version in the chemical place settings Let me talk about this. Let's Just add us on this Ken so this is the Ticket where we're going to try to capture More about the surrounding environment than we do right now when we pair a device. Is that correct? Yeah, we pair device. What are the enclosure and core version? Yeah, and there's nowhere to report that information to right now So until such time as we have an API that will accept that additional input I haven't moved forward on it That probably was in progress I would put it out of in progress because I don't see that API being available to sprint Well, what the API that would take that information is the for example the wake word submission API, right? No, no, no, no This is the enclosure. This is the reporting of what environment we are running in What version of core what's the operating environment is the mark one and mark two is at your laptop? Whatever problem is we don't have an API to handle that extended enhanced information right now That would have to be done during pairing I assume since it doesn't change but until that API exists There's no need to do the actual work to post to a non-existent API So my recommendation is to take it from in progress and put it to do Okay, let's take let's put it into do and set up a short meeting to discuss that because I think that it's something that we can implement But we we need to design it first Yeah, we just need a back-end API that will take that information If we if we're only talking about the platform like a platform designated like mark two and then a version number for the platform And then core version number all of those things already exist in Selene correct Okay, but if you look at the ticket, there's more information than that if I'm not mistaken Right, right, right. Okay. I'll set up a meeting to talk about that this week when it's where that skill is a bit of a nightmare It's like it's a terrible thing. It's just lots of people working on it over the years So I Have started refactoring it a bit. I don't want to I know we don't want to like do a huge fix of skills but Anyway, the problem comes down to like the way that we're we're pulling in data and Trying to get a forecast for a previous day Anyway, I'm working on it still The assumption is forecast previous days are always 100% accurate Yeah, it's just that the API is like we don't have a forecast for yesterday because that's not a forecast, you know Okay Can the next two are yours The view architecture of core with respect to boot up and drivers Can you open the ticket? I'm sorry. I just don't have my large screen. I was supposed to be back at the house and I can't read it Yeah, this is still in progress, this is this is a piece VK test for wakeward tagging so Yeah, so I've been vk testing but Not for tagging because to the best of my knowledge Tagging is not ready to be tested yet Not nor is upload did I hear earlier that the upload wake word is API is working in test or no? It's the code is merged into dev. I don't know if it's in test or not off the check So Yeah, what I've been doing is and I'll just give you my status. I guess while I'm here This is what I've been working on but I refactored the stuff we had for precise I broke it out in the three things precise performance precise operation precise upload Everything's working except for precise upload which once I turn the URLs on they should work So that's what I had done but leave this ticket open because I'm also going to be adding the wakeward tagging stuff into those tests as well I'm assuming and I don't know that this is a valid assumption that the tagging tests Will run in a different location then our current VK integration tests which in core is that correct? Yep And is there a reason for that? Because the codes on Selene Okay, but Selene is a back-end service that can be tested from core, right? Not that tagging stuff is nothing to do with court That's true. Okay. So yeah, then that'll be located in different place once it's ready to go But I haven't started those yet. So leave this open. Okay, and Josh, you're still working on the third-party Management, so what's the difference between the canonical Mark 2 171 and mark in my C4 155 One should be a subset of the others Okay, the canonical is a subset of the brother ticket. It was created before it though Yeah, what are the you know the one you were just asking about the? Mark 2 171 ticket is that's the overarching one, right? We need to find a you know find a system that's we're gonna use for updates And then I believe he's got separate tickets for each different system that he's evaluating Okay Okay, so this is super exciting I Can already tell that we have either bitten off more than we can chew or kind of gotten distracted along the way here I Think that we should just roll this sprint into the next one. Let's close this one out and move all these tickets over into number 17 and Rather than trying to add new things And try to fill out the the two weeks. I think let's just take this week and try to close out all of these tickets And anything that's not Closable this week We should You know rethink Maybe it's maybe it needs to be split up into you know separate tickets or maybe it's just Things that need to be kicked back into the to do But you know if we're not making you know enough progress on these things there's got to be a reason it's either you know lack of lack of time or Something blocking them Yeah, I'm wondering if like especially with the the tagger for example, you know That there's hundreds of lines of code I've written for that that probably could be splitting up That's a smaller tickets to really show my progress you know, and I wonder if that's the case for Some of these other things that we were making the tickets big enough that You know They're spanning multiple sprints Just a Observation, yeah, I mean, I guess I'm less worried about Demonstrating progress. I'm assuming everyone's doing good work, but I am more concerned about Things getting lost in the shuffle because there's too many things open and people are working on too many different things at one time so let's just close some things out and include some path ahead of us for getting more stuff done I Had a chance to figure out what we were talking about earlier when I couldn't see the ticket That was the automate thing and that's simply that that's ready to go except I'm waiting for the API to be able to grab the files to Determine whether we want to continue training or not. So that ticket's remaining open until that code is done Yeah, and then I can actually close that out by validating that it made it through a run Okay, that sounds like another thing you should connect with Chris there on Just to go through Check your assumptions regarding that API and see if what he has implemented will satisfy those or you need to be something else Yeah, Chris is that API ready or not yet? I mean, I it'll be have a PR here in a little while So if you want you can bet review the PR first and then And then yeah, like I said this week I plan on getting everything into our test environment soon all right if you can just Matter most me the PR request link I can take a look at it and see what the API is providing and That was Cool So I'm I'll provide just a little update on the mark to Before you go Michael add the ticket for the new mark to enclosure to the sprint. Yeah, I've got a list of tickets I'm gonna add a minute Okay Okay, I completed 616 or now on 17. I'm gonna go ahead and start it so we can start a lot more up to it So Okay, so a quick update on the mark to the sd201s that we sent out for the US based manufacturer and paid a pretty penny to get done in very fast turnaround Have not been delivered yet not even shipped. They're running into some problems with their double-sided assembly and Kevin's currently debugging that fortunately Kevin also very cleverly ordered a backup set of boards from a Chinese manufacturer and they came in first and so over the weekend He was able to get get it up and running so we've got some he sent us some demo videos of that and so the system is fully working and He's sending those to Ken Right, you know today, so it can't you get them, you know within a day or two And we can start getting that enclosure code working so So Ken You've got you got a couple days to clear your schedule for Getting to work on that Before the meeting and they'll be here tomorrow morning and he's sending me one of each Jump one jumper to behave the old way and one jumper to behave the new way So So we're supposed to be able to use the boards we're getting from the more expensive us being your back to just get them later, right? Yes Unfortunately, right? Yeah, yeah, we go and after them for the for the additional rush fees and everything else I get all that stuff back if I mean the value is you can deliver it quickly if you can't deliver it quickly We'd expect an appropriate price I we have to I have to figure out what the exact problems are it may have been related to the Of the double-sided assembly and not just to you know the turnaround Because all the boards that we've gotten working so far. We've only had the PCBA You know factory assemble one side and then Kevin's hand soldered the second side So they may be running into some design problems with our boards You know that they just don't show up when you're only assembling one side on the automated automated line So if we can if it comes down to that then You know, then it's probably not their fault but But yes, I will definitely have a very keen interest in making sure that We're not paying for something we didn't get so that being said we should have ten of those boards Out by the well in the next couple of days Kevin's actually got another project. He's got to work on So by Wednesday, I think he's gonna send out eight or ten more boards to the rest of the team Well, probably to Derek to get turned into things. Although maybe we'll send half to Josh to Depending on what what how your 3d printers have been working But we'll figure out how to send me a bareboard Okay, yeah Yeah, so I my my 3d printers are working my 3d printers are working. Okay, though. I Haven't started. I'm printing enclosures as fast as the little printer can print and I put a new print head on that guy After six years, the other one was Derek. It was this nasty black mess of gross and it's printing. It's printing great Here's a here's a new enclosure from you or for the For the mark to I'm writing a memo right now on Strategy so if we definitely need to talk about that, but I'm super excited to have everybody out and Actually like like walking out of this this get together with a working mark to like hey It works because there's a couple of blockers on the sales side of things that that we need to resolve and in order to Resolve them we need a working mark to just one, but we need one that works okay, so Let's uh, let's just like I said, let's spend the rest of this week working on the tickets that you that we've inherited from last sprint and We will Let's we'll check in on Friday and See see where we are and After this print That'll also give us some time to know where the mark to is as well This meeting wasn't nearly long enough, so let me bring a couple of points up the first is somebody mentioned something about firmware and burning the Current model or the next version of the board we have coming out as a separate Microprocessor on it and the only way to change the code on that after it leaves our factory Is Additional hardware You know it is not programmable in the field And there is no easy way to make it field programmable at the current time without additional hardware and Speaking to your point Josh If you want to leave after their summit with working mark to We should probably figure out if we're going to use the kivi or qt image And get somebody working on a soft keyboard if that's the solution Wi-Fi setup Indicative of the fact that it is not going to be a trivial one-day issue Yeah, so the Wi-Fi setup we do have a solution for that So if we can get into that and the kivi versus qt thing I sent a back channel message in the middle of this meeting to Michael and that is entirely on Michael's death So I will I will defer if you were to have me make a decision today I'd be happy to give it to you, but it's not my call No, the all I'm getting at is that that decision is going to drive How your soft keyboard is going to be implemented and at some point in time somebody should probably evaluate Whether in our environment the soft keyboard will work at all I started down that path I don't really I don't honestly don't care if it's not keyboard works I'm firmly in the Chris Bayer camp of we should not require a keyboard for our thing to work We are a voice company and I do agree with requiring a second device But Molina publishes a Wi-Fi setup that we can use right out of the box You know, we're probably gonna have to spend right around a day Hacking on it to get it to work where we would you know provided our chips efforts, but it's designed for Raspberry Pi 4, so I don't anticipate that being an issue on the kivi qt stuff you already have an answer on that but that's the Getting a working mark to working on the sprint is not a like hey, that would be nice sort of thing It's like a hey if we're gonna have jobs next year. We need to make this thing work like we've been goofing off with with this I Maybe that's probably not the right word But we've been working on getting something out for a really long time and the places where we need to be focusing our efforts According to the financial model that Johnny and I ran this morning is on adding value added services to get people on to our If we cannot get to the point where people where we're adding value and people subscribe then we're done We're done so we need to get the mark to out because that's the blocker for Description that's the blocker for getting out in the retail That's the blocker for starting to focus on mobile devices There's all these other things that are blocked by this 40 really need to get on block and I have every confidence But if we put everybody here in this shop, I'm probably need some more tables But we put everybody here in this shop and just rent like I think we can get amazing things done I have a lot of confidence in everybody. Yeah, I just want to make sure I understood you correctly So you're saying the Wi-Fi set up to trivial issue. Don't worry about it. You're on top I Trivial issue I have confidence that once we we approach the problem as we're implementing an existing library That you can get it done Well, I would feel a lot more comfortable if you forward me some documentation. Well, I could potentially sign up for that Well, so if we get this if we get the audio stuff working on the SJ201 There's a lot of stuff we can look at, you know, we can look at the Qt image that guys has been working with The guys well, you know, I've been kind of in there too, but working with the guys Blue systems to see what can be paired down to make it run faster The soft keyboard works on that we can evaluate all kind of We're putting in a Wi-Fi password somebody has a solution that's just already to go out of the box I just sent it to you and what's that I just sent it to you and what's that have a look Davis Jarvis is often helped to pull the mark one enclosure stuff out too. So I think we should definitely take him up on on that offer Yeah Just means we need to we need to have a good idea ourselves about what that enclosure looks like. So which is already a ticket in there. So What do you mean by pull it out? Well, I just you know, how there's all a mark one code like sprinkled through my crop core Trying to pull that out into an actual enclosure Code, you know, it's only enclosure code Yeah Does Jarvis just does Jarvis want to Come visit Hawaii for a couple weeks Why don't you why don't you put that on the table? We we can't pay We can't pay salaries and stuff But if he wants to come sprint with us for 12 days and agrees to kind of focus on the focus areas where we're focused on He can be our virtual gas interface. So You can be here and and do some work. But yeah, I'd be I Certainly would be open and I think that the The amount of input that we could get from having another developer who's thoroughly Thoroughly familiar with core far outweighs the cost of tickets to Hawaii and lodging for 12 days. So You want to you want to clear that with Michael Michael you not to put you on the spot You know what Michael and I had a little conversation about about oh, he just said yeah, okay So yeah offered up. I don't know where he is. Where is he? Portugal But yeah, I Yeah makes the offer and then you know, I let him know it wasn't good faith that I didn't realize he was in Portugal if he was anywhere Yeah, it might not be doable Yeah, and be sure to just put a put a tweet out there to the real Donald Trump to thank him for the fact that he can't go to Hawaii Yeah We did also talk about implementing doing using the plug-in system for enclosures and whether that could be a useful idea and So just starting a bigger discussion about what enclosures look like Well, there is a an interesting idea about creating a new branch of a new Experimental branch of my cross core that is you know free for our community team to Just merge Whatever, you know features they want into it so that they can like go to town on on playing around with things Yeah, it would be a centralized fork Yeah Hey The whole idea would be it would be like everything in here can break at any moment. It might be reverted at any moment Is it like it would either be it lives in our repo or it lives in like Jinx is jinx or Jarvis's repo or something like that, but I guess the benefit of doing it out Yeah, I mean if it's gonna live in anybody's repo, I'm very open the host to get internally if it's gonna get done anyway I'd rather I'd rather keep it in house my thinking is that it reduces the chance of a a noncom of an actual fork of the project and then you know them trying to Run their own thing and oh, anyway, I Don't want to get into that discussion right now though. I just thought I'd you know sure something to keep the meeting going Jarvis playing a hostile takeover without your telling me No, they intentionally don't want to They like they want the ability to you know add in all these things that they're playing with that They understand that we need to test things properly and that we have limited capacity to test all that sort of stuff and They don't actually want the project to fork and become you know separate and then become the kind of thing where you can't actually Become you know diverges so far that things no longer work from one to the other and they know that that is bad for the project so Yeah, I think it's a you know Because master is fully state is our production branch dev is going to be production We don't want to merge things in there if they're gonna break and so this would be like, you know You could like they could they could merge things in there without without one of us reviewing them and It breaks and everyone that's using that Trace the hands up in the air Yeah anyway This code you this this link you sent me from Belina as soon as we have a working access point Chris Is there a working access point currently in the? Mark to image Keep in keep in mind Ken that the Belina code is written for Raspberry Pi 4 So it's that there's some questions in there of like hey does your platform support like if it's a feature in there It works on pie, so you don't need to ask the question it works on pie But if we were working like some other rando board, yes, you that would be a good question to ask It is what micro OS users which is now filled the open voice operating system Chinks is Yeah, micro based OS. That's what they use so it can work Sounds like everybody's more confident. You want to make sure that's that's our friend who was using the brand with with Where we hadn't published the brand standards yet, and so he he was he was experimenting with our stuff I do want to I do want to make sure and I don't know if we made the offer You know we can officially endorse like help with that side of things. We just can't have them use the brand Yeah, we have our brand our brand use guidelines. Okay, like it. We're not sure. I want to make sure that Very happy with it. He fully understood Okay, and we'll write a blog post about how awesome he is in the whole bit like I'm more than happy to do that Yeah, yeah, I thought we'd add it to our community releases section on the get my craft even essentially Anyway He's doing really cool stuff in that in that project though So it is Qt base, but it doesn't it doesn't have even have a plasma session running and I think he's running directly to the frame buffer somehow, but Anyway, this is unintentional delaying the end of the meeting now. So I Mean I'm concerned more concerned than everybody else I sure would like to have a meeting on how we expect this Wi-Fi stuff to work it doesn't have to be this meeting but I think it would be helpful to have that meeting before we get out to Hawaii I I show you a concern I just didn't didn't think right now is the time Go chiefs Go chiefs