 Thanks everybody. Thank you for coming. I'm gonna do something a little different this year in the past I've done a lot of state-of-the-Union with key cad this year I'm just gonna ask you if you want to keep track what's going on with key cad go out look at the version six roadmap Works in progress now. So we're working on that. I want to talk a little bit about I've been kind of musing in my mind about How key cad's role has helped the open hardware movement and where we're going with that in order to do that We've got to go all the way back to the beginning of the open source heart software movement and see how that progressed And I think what you'll notice is for I look around and I'll see a lot of people in this room older than I am So there's probably a lot of you won't remember or some of us You know where we started with the open source software movement You're gonna see a lot of similarities in the open source hardware movement even though it's much much earlier stage of that so In the beginning before there was anything that resembled a personal computer people worked on mainframes and everybody sat in terminals And it was connected the mainframe and the only time you worked on software was when you were connected to that mainframe So software development back then was the only people that worked on software software collaboratively with people who either worked for the government Universities or large large corporations like IBM so But it was pretty obvious even back then in very early days of computing that software copyrights were common So the big companies lobbied the major Governments and in 1964 Software copyrights were recognized in the US and pretty much everybody else fall the rest of the world followed suit pretty quickly Okay So there was kind of a lull there while they were still sorting out the microprocessor world So the small 8-bit Micros that were in design between the early 60s into the early to mid 70s. It took a while So there wasn't really any software development open to the what you would call the general public, right? By the end of the 70s the PC was starting to take root. So these startup companies of Apple and the IBM released their PC in this area, right? So early really early in this movement only only enthusiasts went out and bought computers for personal use, right? Nobody else had them. They weren't on everybody's desktop Everybody's now they're in everybody's pocket So there was a lot of sharing of code early on people would say look at this thing I made and they'd get there would be like these conferences like this Where got you know people who were writing software for computers would meet up and share ideas But it didn't last very long because this grew so fast that people realized that well There's money. There's not just money to be making made selling software for mainframes There's money to be made for selling software on personal computers So the that rustered in the dark ages basically Once the sharing stopped all the licenses went proprietary there was there was no longer any concept of open source software at all They were restrictive proprietary. You've we've all seen those licenses before they've been around for you know decades Early on even commercial software quality was not what it is I mean today if you've never used the if you're never part of this and Having to live with because you lived with this because software distribution was by meet some media hard media Like there was no internet. It was either a floppy disk or some other, you know, that was how Software got distributed So bugs persisted until you upgraded to the next version, right? So you had to buy the next version upgrade to get the bugs fixed But then there was always some new bugs that came with the new features, right? Some of the things that also happened during this time that was not very nice was work working Applications were purposely broken for competitive advantage. I won't call anybody out But if you know the history of that, you know who did do the did those kinds of things Basically, if you were a user or a software user like like I was during this time You kind of felt like you kind of felt like that you were just an afterthought, you know It didn't feel like they had your best interest at heart So what happened is a large group of professional programmers were kind of getting frustrated with this whole idea of proprietary software and so In 1980 The BSD released the first what we now know what we now know is the first open truly open source license, right? And MIT followed not long after that In the mid 80s, so Most of this was done at the university level most of the open source software was done by educators and and Their the staff, right? But there was a couple of big time projects during this period that were Because because of the permission of permissiveness of these licenses They were the product there were two projects that comes to mind BSD sockets and Kerberos and they were hijacked and then you know The point of those two projects was to unify things, right? We were going to bring everybody together, right? Well, they were broken for competitive advantage and so a group group of people realized that these licenses were not necessarily The spirit of these licenses was always to to make your code freely available, but it didn't specify that so Later in the 80s the GNU project was formed and they released the first version of the GPL, right? and so that opened up the Foundation so the GNU project worked on all the core tools compilers librarians They basically had clean room versions of all the Unix tools that people had used in back in the day right and What helped what helped spur that was the internet right the early internet even the good old dial-up days of 19.2k and your your phone line would scream at you when you know you logged on to it It was a way for people to collaborate and things started to grow pretty quickly at this point, okay? So now things were so we still lacked a kernel so in 1991 some guy from Finland created a kernel under the under the GPL to okay, and that that really spur development the things really took off at this point so Within a couple of years by the mid 90s. There were probably a dozen packaged distros of of Linux so you had red hat and Debian and Mandrake and Susie and so there was a whole bunch of them so you didn't even need to do any compiling anymore. You could download install and and run Linux and this is when I got in. Oh, sorry. Just I like to move around as this This is when I started getting involved in key in in open-source development and at the same time right around this same time There's some guy a professor named Jean-Pierre Chirac that worked in a university in France started working on a project called key cat. Okay, so that's how old key cat key cat goes back to the very very You know just when the whole open-source thing was really starting to take off What really changed the tide was high-speed internet access once high-speed internet at net access became readily available The effective cost the distributors software went to zero right so now open-source was really development was kind of in full swing at this point By the late 2000s the small form factor devices and smartphone industry It was everywhere now you could you couldn't hardly use a device without touching open-source software You may not have known that you were using it, but you were using it. So at this point Floss one, you know, we basically it's there's not any Debate here about whether or not Floss was the dominant Development model for software right But also during this time we started to see hardware design tools on key cat project start went Even though it was always public. It was never really publicly developed and we that that was Happened in the in the late 2000s group groups of people started actually working on key cad to collaborate and improve it Okay So also during this time There were some moves. So now we had some tools and they weren't super powerful compared to the commercial hardware development tools But we had them so we were starting to see some growth in the idea of open hardware because people wanted the same arguments for Open-source software you could almost make for open source heart or open hardware, you know, it's freely available You can learn you can extend it. There's lots of good reasons to do that, right? So what happened here was The CERN released their open hardware license in 2011 and they also joined the key cat project because one of their thing One of their mandates is is their intellectual property is put they want to be public and they realized that they were Going to use a proprietary piece of software. Well, how open was it it technically their hardware designs were open practically less so Okay, so that's why they joined and started to help make key cad functional So during this time that it's it's pretty obvious from some of the talks that open source design tools are increasing in power usability In some cases they're Approaching a lot of commercial products So as you see there's an interesting thing here that Recently board vendors have been supporting key cad board files directly You don't have to go through the processing step of generating Gerbers and the way we've done it for you know, 40 years now so Obviously we're having some influence there in in the hardware space because people would otherwise because if there wasn't enough for Reason to do so they wouldn't so that's that's kind of where we are at the moment so Well also happening at the same time which was really interesting and I honestly didn't see that one coming was risk 5 development So we now have an open in the last few years risk 5 has been Implemented by quite a few people. So this is an open Tape out part and a design for people to build upon right That I think that's I don't think you can undercut the significance of how important that is because You know, we're getting an open heart when we're getting open boards and open electronic circuits designs But we really don't there's still things we don't know about of our electronics, right? That's all hidden from and if you like in the past the hardware was always kind of approached as well It's a black box. You just have to trust that it works But given the last few years like with things like Specter and meltdown We now know that our black boxes might have problems and so One of the things I'd like to see happen is die-level development tools start to be opened and moved in and moved into like he had for Development so now I'm not naive enough to believe that the hardware will quite be the same way as a software Because obviously the cost will never be zero for hardware. There's always going to be some cost involved But will one of the questions I have is well Will the will the stack from top to bottom ever be fully open and I I don't see the reason why it can't be I mean I there was going to be resistance just like there was resistance to open source software back in when it was developed so so The only thing it's left is is will the cost of that? Determine how the rate of Open open hardware and you can see like it's we've all so you think about 1980 is the first open source License software license. That was 40 years ago. We're there, right? It's only been nine years since we've had an open hardware license a true open hardware license that where will we be in? 30 more years from now in the same time frame. Well, will the entire stack be open? I hope so because and I think key cad plays a prominent role in that as we'll continue to try to increase the the debt the breadth of things you can do it with open tools over time and That's one of our goals. So Whenever you can use an open design use it It sends a message to the the people who aren't open that You would you prefer open because that's what happened with the open source software Basically the message was if we can use open and it suits our needs But that's where we're gonna go and so people vendors notice that kind of you know when their business starts going away So if you can use open, please do Okay, now a lot of things if for those of you have don't follow the project closely I this is just a quick year and review since last Fossum right after Fossum last year we released 5.1 is the stable release. I got hired to work full-time on key cat. Yeah Well, that is it. That's this is another side to that story We had the first ever key cad conference in Chicago in April last year And if you didn't make that you missed a great con a great conference probably one of the best I've ever been to so if you have a chance to make it this year. I highly recommend you make it And I'll talk a little bit more about that later Key cat services corporation was formed this year one of the the dings against key cad was well I can't get paid support for key cat. Well now you can there's commercial So if you're if you're a corporate user and you just got to pick up a phone and call somebody you can do that Now there's and in the fair and fairness of open Disclosure I work for key cat service it was formed by Seth and I thankfully he invited me to be part of that and in October with the company that Promised me that we were going to make key cat fly Drop me so So so boo So that didn't last long enough to get anything significant done really was a kind of a bummer The other thing that happened in November of last year We joined the Linux Foundation and now what that means is we have a two-pronged Set up for Donations the the CERN donations will continue as they always have but one of the reasons we joined Linux Foundation is CERN has Pretty strict rules about how we can spend that money. So really for software development The Linux Foundation allows us to do things like expense pay our developers to go give talks So that's the Linux Foundation gives us a lot more flexibility how we spend our money. It's not just software development So if you were interested in donating to Linux through the Linux Foundation, you can just find the links there and in the My presentation So I joined key cat services in December We also moved another big thing that happened We moved from Launchpad to get lab at least just the source for now Eventually, we're going to try to migrate the whole project there because there's a bunch of stuff that get hub right now But I'm really happy this moves worked out really well I'm surprised how easy it was and how much we're really I think it's gonna help develop our workflow just a lot better Yesterday we had the first ever new key cad Developers meeting so there was about what 10? 15 people who aren't regular key cad developers and Three of us that three of the lead development team were there and we work we introduced them to the code base and Help them work through some pet bugs that they were interested in fixing So that's significant because it's always nice to be able to you know bring in new talent into the project And you know because the key cad code base is pretty intimidating So it's not it's not something that you'd probably want to take up on your own now a couple announcements key con 2020 If you missed last year's you definitely want to make this year's it's going to be at CERN in Geneva, Switzerland From September 11th through September 13th the tickets haven't gone on sale yet But keep an eye on the either the blog post on the key cad website or you can go I have the link right now It's just it's just a placeholder image file. There's no information But I suspect the tickets will probably go pretty quickly to this one so if you if you're if you're in Europe in in September and Going to go into CERN to do to talk about all things key cad sounds interesting to you Please please come to that conference The everything is happening is Monday and Tuesday and Wednesday of next week The folks at Eisler and I'm gonna butcher this. Ahen. Did I say it right? Ahen, ahen, okay We're going up the the there's quite a few members of the lead development team here We're gonna go up for hackathon for three days and work on key cad So it's we get a chance to all it's rare when we all get a chance to sit down and work together It's usually email and that whole process is less than ideal So and just an announcement about we've already I put in an initial goal I had no idea what to really put down so I put 40k down for one year at Linux Foundation We're already about a third of the way there since November. So the donations are good. I thank you for All the time and so just what I always it always I always say thank you to the development team because I get to stand up here and People know my name by the fact that I'm the project leader, but I'm just one guy of many It's a big team of people that writes write source code develop libraries documentation translations package builders the whole nine yard, I mean there's Probably what 50 80 people combined it takes to pull you know Contributing their own little in their own little special way to make the project work I always thank to say thanks to our donors. It's that keeps key cad Going and it and it helps it tracks interest to Other people who might be interested in developing key cat. It definitely is a plus we I always say thanks for that Thank you for your interest in key cat Um, we're going to continue to do our best to get to provide the best possible electronics design application that we can and I don't see us slowing down anytime soon. It's just going to keep growing and growing and growing It's getting it's getting more feature rich. It's getting more impressive every year the things you can do with key cat these days are your There's very few limits. I mean, it's not a it's not a stripped down four-layer board layout package. It's a 27 layer 800 pin FPGA monster board layout one of one of our users who's like our Performance test guy because he's got some incredibly complex designs. So key cat can handle something like that You're probably not there's I don't know too many people doing board designs that are more complex than that And a special thanks to Seth in hobby ours absence Seth stepped up and did the dev room So thanks to Seth for stepping up and and helping out there. So Thank you. So No, not if you mean the relationship between myself and Jeff. No, it wasn't that it was I mean I'm not it. I'm not at liberty to talk about it because it's not it's I signed an NDA So I can't talk about anything with company wise, but it was not my decision It was it was based on other other factors If you you could probably guess if you were there that night, you can probably guess what happened So I'm just gonna leave it at that Verse I knew somebody was gonna ask them they say something about version six. That's not really a question But I'll say something about version six Things are starting to roll into the master branch now I'm working on the new file formats for the symbol libraries and the schematic and So that should be done here hopefully not too distant future because there's a lot of features new features that kind of sit on top of that but There's a lot of work going on. There's some configuration work that's going to change We're gonna start one of the things I heard somebody earlier complain about that, you know that when you know that we used to save Are we currently save like some of the? The visibility states of the layers and the board file so when you change them Even though you didn't change the board the board file changes. We're ripping all that out. We're taking that out So that that'll be a separate config file. So we got somebody working on that So there's a lot of things happening. I mean where it's key features curve traces Groups and rooms What oh yeah past acts we have a man we have a layer manager now that you that's already done That's been in there for a while that you can you know set layer color you can set the the all the definitions of Between layer so there's a layer stack up at it or now you can you can sign it a color a silk screen color mass color The the thickness between each each copper layer. So yeah, there's a there's that what else are we working on anybody? What am I missing guys? Yep, 2581 export will be in there Somebody's actually working on an Altium importer whether that'll be ready for whether that'll be ready for six or not I can't say but I mean it's Where is here? Is he oh, there he is? Yeah, so who knows you know I'm sorry John's bus work. Oh, yeah, that's right. The there's There's real-time net list Management so the net list will be updated real time as you make changes. So that was done. Yeah, sorry Sorry, John But yeah, it's nice because we have the bus unfolding now that if it's a feature if you you know You create a like a data bus or an address but a complex bus using the you know standard text definition You you can go and break out when you right click you can break out the bus each net and it'll automatically connect It's pretty slick some neat stuff going on. There's some people who want to work on orthogonal dragging With for the the schematic editor so that you know the wire stay nice square, you know, right angle lines instead of when you drag Now you get you got to break them yourself and you make them square again So yeah, there's a lot. I mean there's lots of things going on right now I you know, I know somebody's good the next question is going to be when and I would say I would say Given the current rate I'm hoping we're like in a feature-freeze and beta around Key con We'll have a better idea key cop, but I'm hoping we'll be close to that so that maybe by the end of the year We're close to a 6.0 release Yep Well, I there's actually I know now there's actually people working on trying to get an open an open PDK so That's work in progress. I those are things that I'm not at liberty to talk about But I know people are trying to do that and I think of it like I said, I think eventually It's it's the one it's kind of the one domino theory if you can just get That one even if it's not perfect and even if it's but it's you know It meets the needs of a large enough group of people Every people who want open will start to merge that way and people will see their You know other customers will or other, you know fab guys will see their business going elsewhere And it's kind of how the open source worked, you know It was there was there was a first implementation and people started to go that way and know somebody noticed Hey, we're all these guys going right so I suspect that like I said I don't I don't I'm not naive enough to believe that it's quite the same as software because like I said the internet makes Software zero cost zero. There's never going to be a zero cost hardware solution the question the question the question is is when So for key cat services corporation if somebody pays for a feature or a specific fix Does that affect it doesn't affect the development it only affects the It only affects the order in which things get done Normally, I have like a progression of things. I'm working on so as long as nothing gets in the way of that That's kind of how I go But if somebody comes to me and pay it's going to pay me and say I want this done first Well, I'm not gonna say no, you know, it just I just slot that in and then everything else But but but a lot of times it's a feature that it all get it all gets good It all goes back into key cat proper, right? So it gets merged into the main key cad master branch So I there's I don't see a downside to it. I really I just don't see a downside It's not like we're just making those features for like one person and we're hiding it because you can't do that The key the key cat code basis Excuse me gpl3. So we're kind of tied to that What do I think the biggest thing holding back key cat development? Ah That's a good question a lot of it's just time, you know people were most of us up until recently or Very part-timers, you know when you have a full-time job and family and you know You want to have at least something that resembles a life, you know Which I didn't for a long time My wife was here. She'd tell you that So, I mean, I think that's just you know what it boils down to is is if we can get back into the way Ways some way to monetize it so we can pay people to work spent spend more time just working on key cad instead of working It on it in their spare time. So So actually development pace has real has picked up in the last couple years We have we've at most of the lead development team seems to be able to find enough time to spend To get that get worked on so I mean obviously it's all you'd always be nicer if you could do it faster But it's actually pretty it's moving along pretty good right now Any other questions The question the question is is trade-offs between usability and complexity There's always trade-offs. We try to keep the core features That you would use to do simple board layouts fairly straightforward and key cat A lot of the times if you don't need to use any of the advanced features, you're not necessarily exposed to them So like say you would need to do matched length tracing you tuned to length tracing, right? There's there's some configuration you have to do there. You can't get out You can't get away from that because otherwise You there things there's parametric things you have to set up And there's no way to avoid the complexity of doing that But if you're just doing simple two-layer trace boards and fill in some zones that stuff's all fairly straightforward I can't imagine you know like to just to do simple board design layouts that key cat is like so complex that you can't wrap your head around it It just it's fairly straightforward some well The question was it's the problem with the symbol libraries Which has always been a problem since the original design of key cad and the answer that is yes These were the problem is is to get from where we want to where we were to where we want to go There was a couple of intermediate transitions and they were somewhat painful because one of the things We always had a bug in the old library lookup that was based on the library ordering. So if you had two symbols Or identical symbol names in two different libraries What mattered was the library ordering That was a if you pick the library from one the one you thought you got and you had the library ordering wrong You got the other symbol and that was a problem. So that what yet that the table where you map it to a table now They're all hard to find. They're absolute. There's no relative to The symbols are and so there's a we had to get from there to there The next in version 6 the symbols will actually be embedded in the schematic like the footprints are on the board So you you won't need any cache library anymore You'll just pick this the schematic up and take it with you and Every now you won't you may not be able to if you don't have the symbols on your the other machine You won't be able to update the symbols that are in the board with the latest versions of the symbols in the library But you'll still have those you still be able to work on the schematic open it It'll be completely portable. That's all part of the new file format stuff that I'm working on Gotta go