Algorithms + Data Structures = Programs - Episode 163: Sean Parent on Tilt Five, Metrowerks & Be Inc.

Episode Date: January 5, 2024

In this episode, Conor and Bryce chat with Sean Parent about Jeri Ellsworth of Tilt Five, Greg Galanos of Metrowerks, Jean-Louis Gasse of Be Inc. and more!Link to Episode 163 on WebsiteDiscuss this episode, leave a comment, or ask a question (on GitHub)TwitterADSP: The PodcastConor HoekstraBryce Adelstein LelbachAbout the Guest:Sean Parent is a senior principal scientist and software architect managing Adobe’s Software Technology Lab. Sean first joined Adobe in 1993 working on Photoshop and is one of the creators of Photoshop Mobile, Lightroom Mobile, and Lightroom Web. In 2009 Sean spent a year at Google working on Chrome OS before returning to Adobe. From 1988 through 1993 Sean worked at Apple, where he was part of the system software team that developed the technologies allowing Apple’s successful transition to PowerPC.Show NotesDate Recorded: 2023-12-12Date Released: 2024-01-05Tilt FiveJeri EllsworthHackers ConferenceCommodore 64castAR (formerly Technical Illusions)Compiler Construction by Niklaus WirthHookStar Scrabble TrainerMetrowerks CodeWarrior IDEGreg Galanos (Founder of Metrowerks)Jean-Louis Gassée (CEO of Be Inc.)Be Inc.Pactifc Northwest C++ Users’ GroupNYC++ MeetupC++ ContractsC++ On Sea ConferenceIntro Song InfoMiss You by Sarah Jansen https://soundcloud.com/sarahjansenmusicCreative Commons — Attribution 3.0 Unported — CC BY 3.0Free Download / Stream: http://bit.ly/l-miss-youMusic promoted by Audio Library https://youtu.be/iYYxnasvfx8

And it's amazingly good. You know, I've spent hours playing with this stuff. And where VR, typically I find very isolating and very disorienting, and I get sick really quick with VR. Lots of people do. I could spend all day with Jerry's stuff with Tilt 5, and, you know, it's beautiful. It's not isolating you have glasses on but your your displays are effectively clear and your pico projectors are pointing out welcome to adsp the podcast episode 163, recorded on December 12, 2023.

My name is Connor, and today with my co-host Bryce, we chat with Sean Parent about the stories behind Jerry Ellsworth of Tilt 5, Greg Gallinos of MetroWorks, Jean-Louis Gasset of Bee Incorporated, and more. Oh, yeah. So the Tilt 5 stuff is actually Jerry's second generation. So I've known Jerry Ellsworth a long time.. She created a Commodore 64, like in a little box that you could connect to your TV. So like, I want to say emulator, but, you know, she was basically re-implemented all the hardware and stuff in this tiny little box and put all these escape hatches. And so this whole kind of hacker community evolved around this little device that she she developed and she had been contracted to do it by by by some company. So she made a name for herself on that. And then she got a job at Steam and she was doing research into VR stuff. And her story is she had on her workbench some Pico projectors for doing VR stuff, which would normally be projecting into your eye. And she had them sitting on her desk, facing away from her while she was trying to fix something on them. And while she was doing that,

she happened to have a piece of retroreflective material draped over a chair that she'd just been playing with for a site experiment. And what she got out of the Pico projector was a hologram projected against this piece of retroreflective cloth and and tried to sell this idea to steam or try to pitch it to steam. And she worked at steam and and they didn't want to go for it. And so she ended up leaving the company, but she was able to keep the IP. She believed in this so strongly she was able to talk the CEO of Steam into letting her keep the IP. And so she left with the IP and she did a startup. Boy, now I'm drawing a blank. We can look it up.

But she did a startup between Steam and doing Tilt 5 where she was using this and she was pitching it for kind of all kinds of purposes. Like, you know, if you were an architect, you could, you know, instead of building physical models, you could bring in your clients and you could do this virtual model and your client could walk around and point to things. And so she tried to make a run at it from this startup and ran out of funding and somehow managed to walk away with the IP again and then started Tilt 5. And Tilt 5 now has gotten to the point where they're a shipping product. And it's amazingly good. You know, I've spent hours playing with this stuff and where VR typically I find very isolating and very disorienting. And I get sick really quick with VR. Lots of people do. I could spend all day with Jerry's stuff with Tilt 5. And, you know, it's not isolating you have glasses on but your

your displays are effectively clear and your Pico projectors are pointing out and so your your view of the rest of the world is unobstructed and you got this very crisp display the display that comes back is actually a light field display, which means that the display has natural depth. And so by just changing your eye focus, you can focus on what's close to you or what's far away. Oh, wow. And so it's incredibly impressive. And then she's done a lot of stuff to get visual persistence and high frame rates.

And so in the stuff that I did for Lightroom Mobile and Lightroom Web and now Photoshop Mobile and Photoshop Web, I frankly stole a bunch of ideas from her to get to get smooth, smooth pan and zoom working in those those products and very high frame rates. She has some really nice software techniques to keep everything locked. So it's interesting, even if your computer freezes, there's enough intelligence in the glasses that you still can move around and see different angles of the scene that's frozen. Huh.

Huh? So it's, yeah. So for your listener, I strongly recommend they go out and buy a Tilt 5 system. Yeah. I mean, link will be in the description. Let's check how much it costs. It's much cheaper than a, actually, I don't know how much the, what is it? What do they call it?

Oculus. Do they still call it the same thing? But I know the Apple one is like thousands. Yeah, that's, Tilt 5 did a great ad where they're like, you could buy the Apple system or you could buy a Tilt 5 system for you and all your friends. Yeah. It says here that there's a bundle for like $6.59. I'm guessing this isn't Canadian because my browser should recognize I'm in Canada.

But yeah, I will definitely be looking into this. And two things we couldn't remember the name of. So Technical Illusions, I think, was the first startup, according to her Wikipedia page. And then also Compiler Construction, I believe, was the name of the Nicholas Wirth book. When I Googled Nicholas Wirth Compiler Book, Compiler Construction came up, which says it was initially published in 1996. But yeah, links to all of this stuff will be in the description.

First published in 1996. Okay. I found a 96, a 2005, and a 2017, and definitely it wasn't the latter ones. But okay, we'll find the initial publishing name. But this is very cool. And I feel like this guy, you know, Bryce isn't interested, but I feel like this kind of, you know, this combined with like a D&D type of thing, you can still use your imagination. But still, it would be very cool to, and it kind of like removes like a ton of limitations on, you know, or maybe, I don't know, maybe in Bryce's opinion,

it constrains you because now you have to design everything. I would try it. I would try it. Bryce is, Bryce is, I was saying I would try it. I would try it. It's, it's, it's very cool in, you know, kind of a group setting. You can get, you know, three or four people with the glasses looking at the same table, and you all see the table from your own perspective. Right. And then you've got the virtualness so that you can have the dungeon master who can see additional information on the table.

Yeah, yeah yeah yeah can't see um but it's one of those things where where you kind of have to see it before you appreciate it because like you can stand up and walk around the entire table and see all the sides yeah so it's it's the closest thing to um uh you know the star wars holograms that yeah another thing that I I mean this was a like a pipe dream but like I've I've written a Scrabble training program called Hookstar which basically is a combination between Family Feud and Scrabble that every time let's talk about the name let's talk about the name though uh it's It's just a misspelling of my last name. But hooks are a very common term in Scrabble where you build one word by spelling that word, but then hooking off of another word, which like the most common hook, obviously, is an S. You can hook an S to make something plural or to change the tense and then spell another word. But there's a bunch of other way, way less common hooks,

like if you are changing the word hat by adding a prefix C to make chat. And then there's a bunch of crazy hooks, like a very common one that all Scrabble pros know about is EAU is a valid Scrabble word that is just the French word for water, but it shows up in English dictionaries, so it's playable. But the plural is also a very common Scrabble word, E-A-U-X. And a lot of people, a lot of Scrabble pros know that, like, beginner Scrabble players don't know that word, so they will specifically put the E-A-U with a triple word or letter score after the E-A-U so that they can get, like, a triple X in two different directions

and turn that eight points into, what is eight times six? 48 points for a single letter. Anyways, going on a tangent here. I've always thought it would be cool to play a like holographic Scrabble that shows you these kinds of like word hints and stuff. And then also too, one of the things about Scrabble is you have to spin the board because it's easier to find words if it's facing you. But with a holographic one, just as you mentioned, you can change the perspective.

And technically, I think you could have just like each person looking at the board, you know, the correct way. Like it sounds like it would be capable to basically just have a first, you know, player point of view for each person. And then when you're placing, it'll just show like the reflected version

to the other person. And I always thought about building like a hardware version of this, but I was like, that's way too involved. But now it sounds like this wouldn't be too hard to take like a Python 2d application with some knowledge and like translate it into sort of this 3d augmented version,

which would be super cool. They have a really nice SDK too, that, you know, helps with the, you know, so you don't have to become the 3D expert. Right. So, yeah, so definitely check it out.

And also any of the systems you purchase are the developer system, so you don't need a separate system to do development on it. And you can just buy the system and download the SDK and reach and this is gonna this is the this guy's unlimited now i'm thinking if i can just carry like it's even better than carrying a scrabble board around because those things are pretty hefty all you need is these two glasses and some power source or i'm sure there's batteries on it you could take this on a plane and then you know sitting next to your friend just both be sitting forward and then play a board game virtually yeah all right that would be pretty cool that'd be pretty cool you do need the little you do need

the retroreflective mat so you're going to need like uh you know talk to jerry about making like airplane tray sized man yeah yeah so so you can do that that's i have no idea how we went on this hard hard left or right tangent we were talking about hardware. You know, hardware and how is it easier or harder for somebody to get started in hardware these days than it was when Nicholas Virich was building a Lilith machine. And I think it's a little bit of both, right? There are some things that are easier and some things that are harder. And Jerry's my example of uh you know the amazing things a small company can do wow do you know if is i guess this uh ip is patented

i'm i guess by tilt 5 so there's no chance of uh like microsoft hololens uh like stealing this and uh you know yeah not till the patents expire so whenever that lasts interesting i had no idea about this definitely check it out it's uh it's pretty cool yeah where do we i mean where so if i if we skip all the way back to uh i'm trying to find where's the tab i've got like 40 tabs open now with the uh the you know wikipedia profile of jerry ellsworth and all the the different things the links will all be in the show notes um i'm trying to think if you know you mentioned metroWorks and that you had written some projects in Modula 2. And then I think, yeah,

you posted something earlier about, I guess we sort of covered that, yeah. C++ was used to replace Object Pascal. So I guess, yeah, like to wrap things up, because I know this is kind of like the time we were aiming to end at, you know, is there anything more to say about the, because I have absolutely, you know,

no idea what the MetroWorks CodeWarrior IDE was, or, you know, you mentioned the Modulate 2 projects, or any of those things worth mentioning? Like you linked to the Lilith computer, so we'll definitely throw that in the show notes if folks want to read more about it. Was there anything special about those IDEs? So CodeWarrior, I kind of view as the

precursor to every modern IDE. I mean, it was really, even though it... When did it come out? I don't know offhand. So it would have been maybe 93 or 94, right in there. Could be 92. Wikipedia says December 23rd, 1993. So you're exactly correct. Yeah. So Greg Gallinose just on his own wrote a Modula 2 compiler and he was selling it to a handful of Mac people, myself included. And so, you know, when I got stuck in Modula 2, I would pick up the phone and call their tech support and it would be Greg Gallinose. Rest in peace, Greg. Greg passed away here recently. And then when Apple switched to PowerPC, that left this compiler gap in the space. And Apple's compiler at the time was pretty horrible. Most people were using, I think it was called Lightspeed C at the time for actually doing Mac development, but they had a 68K backend and no

PowerPC backend. IBM had a good compiler, but you needed a power workstation to actually run it. So that wasn't very approachable. And so Greg Gallinos decided, hey, I know all the compiler people in the industry. I could go build a virtual company if I don't ask these people to all move to one location. If I just call them up and say, hey, let's build an IDE and build a compiler, we can do it. So they created the, you know, MentorWorks Code Warrior was their IDE, and they had a C++ compiler. And almost all of the PowerPC projects on the Mac at the time were built on top of it. And it was, you know, an incredibly successful project. And, you know,

at one point, they purchased some technology, which was called Latitude, which was because they were trying to branch out onto Windows. And so they shipped their compiler for a while on Windows. They purchased some technology called Latitude, which lets you run Mac software to a limited degree on Windows. if I mentioned in one of our previous talks, but at one point after Steve Jobs came over, retook over Apple, and there was this gap where Apple was pushing Rhapsody and no company was looking to support Rhapsody. I cooked up this idea with Eric Trout, who Eric Trout had done the dynamic recompiling emulator for 68K code to run on power pc at apple and then he did a virtual pc at connectix and some other products at connectix a virtual playstation at connectix so you could you could emulate a playstation 1 on your mac he and i had this idea of if Apple isn't going to support traditional Mac anymore,

why don't we talk to the guys at Bee? And we think we know enough people in the industry, and we could get together a plan to build a Mac competitor. And so we talked to our engineer friends and invited them to a meeting. And somehow we ended up in a meeting with Jean-Louis Gosset, who was the CEO of Bee. For people who don't know, Bee made their own computers based off a 603 PowerPC chip running their own OS called BOS at the time. Jean-Louis Gosset had come out of Apple. And so they were building this machine. And Apple had started licensing macOS. And so there were a lot of people

who were going to build Mac clones, including IBM, who at the time IBM built ThinkPads and IBM was going to build a PowerPC ThinkPad running macOS. And then Steve Jobs came back and squashed all the clones of those ideas. Smart. And so IBM was very far along that pipe. So we end up in this meeting with a bunch of CEOs and executives at various companies to pitch this idea.

And basically our idea was, look, we can take the Connectix has something that they called Speed Doubler, which was a very fast dynamic recompiling emulator for 68K. Because at the time to run Mac, you still needed to emulate some amount of 68K, even on PowerPC. So we've got the emulator. We've got Greg Galanos here, and he owns Compatibility Shim for macOS, which actually just implements a whole bunch of macOS from the ground up. We've got the B operating system, and we've got Jean-Louis Gasset, and the thinking was, and then we have hardware manufacturers who are sitting waiting to ship PowerPC Macs, and we said, well, instead of shipping PowerPC Macs, how about we build Mac clones and sitting on top of the B operating

system and let you run Mac software and ship this new hardware. So I was involved with Greg in that. It ultimately fell apart largely because everybody said, well, we need high profile apps to ship on this thing. And I went to John Warnock and John said no, that he, although he didn't believe in what Steve was doing with Rhapsody, he was not going to, you know, undermineve and take him on head to head with a mac compatible clone and so so without a high profile app that was committed to this uh uh hybrid machine that effort kind of fell apart and and it was very complicated anyways we had you know so many companies and people involved and all with their own take and and, you know, who was going to license what to who and his Windows business to Nokia. And he sold that off.

And then like two weeks later, Apple announced that Apple was moving from PowerPC to Intel. They gave him no heads up. And that basically killed MetroWorks and CodeWarrior. I was at the Apple's WWDC, you knowDC with Greg when that was announced. And we were kind of in a back room. And yeah, he was in tears. It was like the end of his company in one fell swoop. He went on, just not like he was doing poorly.

He went on to be a venture capitalist for SoftBank for the rest of his career. But I guess it's less about whether you're doing well or not. It's, you know, if you built a thing and you want to see it succeed. Exactly. You know, a horrible thing, and it comes about, you know, I always think

Apple is just way too closed and doesn't communicate well with their developers. And that's one historical example of if they had been communicating with their number one tool vendor in advance, then he would have been ready day one with an Intel compiler. But Apple was building their own Xcode compiler and wasn't talking and blindsided him. Just like on Survivor. You've been voted off the island.

Yeah. So that's my MetroWorks story. That's a pretty good one, I've got to say. Greg was a super nice guy. I mean, I still work. Eric Scouten at Adobe was one of the MetroWorks compiler guys, and he's now doing the content authenticity work that I mentioned earlier. So, kind of small world well probably we've kept you for an hour and a half now and uh yeah we gotta let you we usually do yeah this i know i mean i mean i feel like i think this is the second time we i looked i i looked back when we had interviewed we i think we did two or

three times in 2021 and then we did zero times in 2022. You know, we upset our listeners to no end. And you, as we would mention every other episode, oh, we'll have Sean on and we'll reach out to him at some point. And we never did. And then this is the second time in 2023. So we will make sure, we promise the listener, we will not go through 2024 without at least once,

but probably multiple times. I mean, hopefully, I'm not sure what conferences if you've said, when are we going to see you next? What are your? I don't know, right? I have no conference commitments right now. I've got an invite to a user group meeting. What is it Northwest user group meeting up in Well, hey, I hear that there's a bunch of Adobe C++ people in New York now and we have a meetup in New York

you can come here and visit us and kind of visit me Who are all the David didn't move down to, I guess he's got a big family David Sankel's here and then recently Cool, well let me know when you're I guess, yeah, he's got a big family. David Sankel's here. And then recently... Cool.

Well, let me know when you're heading to town. I will take you to get good food. Sometime in the spring, I will be in the Seattle area giving a talk at the Northwest meetup. I've got to figure out dates for that. And Adobe has an office in the Seattle area, so I want to spend some time there. We didn't talk at all. Dave Abrahams and I gave a talk at CppCon on contracts. We can talk about that next time. Yeah. How do you think C++ contracts are going? How do I think C++ contracts are going? I don't know. Me neither. Yeah. I don't know me neither I don't think they're completely off the rails

but they're also not quite what I want yet some of what I want they will support I'm just more opinionated I would enforce as opposed to support and then some of what I want they don't support yet so we can talk about contracts next time. Cue that up. We will definitely cue that up.

And I mean, depending on how schedules work out, but I assume, so does the New York C++ meetup, do they have a set location or does it hop around to whoever's hosting? Sometimes it's at Adobe. So that's what I was going to say. I'm going to guess that it's going to be held at Adobe.

So maybe I'm assuming Bryce, pending he's in New York at the time, will be attending. Yeah, which is not at all a given. Yeah. So if Bryce is going to be there, maybe I'll, depending on what my schedule is, I'll hop down because Toronto's like...

Well, even if I'm not going to be there, because you can always just stay at my apartment. Oh, yeah, that's true. But I mean, it would be cool if you were there. Yeah, it would be cool there and then we came down and then we could do like uh you know we didn't we wouldn't have to record for a whole extra hour and a half down there but we could do like a little mini thing if i'm assuming there'd be some kind of social afterwards and uh presumably yeah yeah i mean usually some folks go out for

for um uh for drinks at some bar afterwards um It depends on who's there. And also I suppose the, the weather. There you go. You can, we can do a drunk podcasting again, like at. Yes.

Yes. That was fun. That was fun. CPP on C. Yeah. Yeah. Yes. Yeah.

That was, you know, like I've said it on multiple times. It's gotta be one of the high. Oh yeah. Our get ready. Bryce,, that was, you know, like I've said it on multiple times. It's got to be one of the high. Oh, yeah, our Get Ready, Bryce, our new year, you know, reflecting on 2023, looking ahead forward to 2024, which was one of my favorite episodes is when you came up with the Star Wars analogy. Definitely in my top five will be the C++ on C recording. I guess I got to start thinking about what my top five is.

I mean, obviously, there's the road trip. That was pretty good. Oh, yeah, yeah, yeah. It's going to be hard to match that. It's going to be hard to beat that, yeah. I think Dave Abrahams might be at C++ on C this year. Okay. I don't know if they're open for submissions yet.

I don't think so because I think Phil listens now and again to these and if he is he'll probably be nodding his head or not nodding his head. We're so lazy. We can't even send an email to somebody. It's just like, hey we'll put this message out there on the podcast.

He'll listen and then they'll contact us. I monitor the CF the cfps and uh i know right now that it's not open for c++ on c because there is at least one and i think i saw a headline that said there's two other conferences that phil is running um in in along with c++ on c and the first one is a virtual one that's happening i think in early 2024 like in february so i think most of his attention right now is focused on running that virtual conference.

That's cool. I think the CFP has opened and closed for that one. That's an online C++ conference. I think it's just called C++ Online or something like that. Anyways, notes will be in the, link will be in the show notes. And yeah, hopefully, if not in February or March at some other conference later in 2024, we'll all be able to meet up again.

That would be fun. And yeah, thanks for taking your time to chat with us once again. Obviously, the listeners are thrilled. You're probably half the reason why our podcast has taken off. That's definitely true. Your stories are, you know,

they stand far and tall above all the other ones. You know, at some point I'm going to run out of old stories. I have a hard time believing that because, you know, it's just going to take us migrating to a different topic to jog your memory of to be shouting at the podcast. Why aren't they mentioning this? And then poof, then we'll have a whole new set. And if not, we're just going to have to meet up in person and you know have drinks and just just chat and you know i still think those are wildly entertaining to the listeners so we haven't even gone through all of the stories that i have that i have heard of sean's in my few years of knowing him i i feel

like we're gonna be i feel like we'll have plenty, plenty of content for the anticipated lifespan of this podcast. Well, the lifespan is, you know, I thought you said forever. Okay, fair point. Until one of us murders each other. Maybe I was just making a commentary on how long I think it'll take for us to get there. What's the latest on Better Code, the future book? I've got an entire team making as fast as progress

as I was making on the book before. That's a very interesting statement. You know, that could be interpreted as a good thing, but if you know how the progress was, the speed of the progress before. Yeah. So, yes, progress is not going as rapidly as I would like.

But we are kind of doubling down, at least on our commitments for this next year. Well, that's good. We will see. We have not abandoned the effort. I think increasingly we all believe in the effort. All right. Cool. Well, on that note believe in the effort. So, all right. Cool. Well, on that note, until next time, thanks for coming on.

And we will chat in the future, probably next year. Be sure to check the show notes either in your podcast app or at ADSB the podcast.com for links to anything we mentioned in today's episode, as well as a link to a GitHub discussion where you can leave thoughts, comments, and questions. Thanks for listening. We hope you enjoyed and have a happy new year. Low quality, high quantity. That is the tagline of our podcast.

It's not the tagline. Our tagline is chaos with sprinkles of information.