Latest / Oxide and Friends / NeXT, Objective-C, and contrasting histories

Transcript

NeXT, Objective-C, and contrasting histories · Oxide and Friends

• 0:00
  Speaker 1: In terms of the origin here, it actually goes back to our first space that I could that I could not remember if we recorded or not. But we did not record the first one we did of these. I don't like.
• 0:12
  Speaker 2: Yeah. I don't think so.
• 0:13
  Speaker 1: But then I was trying to find in our own, like, in our personal chat history the awesome idea that you had to record it, and I couldn't find it. Was that over?
• 0:24
  Speaker 2: Well, I'm glad you attribute that to me.
• 0:27
  Speaker 1: Well, certainly, I mean, you did have, like, the tweet storm of, like, the 27 step process that you have to take to record this. Like, how could I how could I not attribute to you? Oh, you
• 0:36
  Speaker 2: know, what what precipitated it was I think we were so fired up for whichever, you know, whatever topic we had for the first one, we decided we had to record it. And being the dope that I am, I thought it was gonna be no problem because all I do is like plug one old phone into
• 0:54
  Speaker 1: a computer. I mean, how hard can it be?
• 0:56
  Speaker 2: Right. How right. I mean, I just I just hold one tape, you know, recorder up to the tape player and, you know, copy the tape that way. But yeah.
• 1:07
  Speaker 1: It's hard. Well, I'm really glad you did. So and then, of course, I these recordings are so valuable that I was really trying to find the recording from that first Twitter space because so if you recall, and this just has to be in your in your memory because we have no recording of it. We we talked with the oxide compensation model in that we ended up talking about that. And someone mentioned whose name I don't think we have Oh.
• 1:35
  That Next had done the same thing. That the Oh. That's right. That's right. That's right.
• 1:39
  I do remember that. That the so the Next computer company had paid everyone a flat 75 k, all engineers. Which of course I immediately had lots of follow-up questions. In part because, at Oxide, we pay and we will link to the blog entry, but we pay everybody a $180,000, 180,250, which is basically what 75 k was in 1985. I'm not sure if you've done that math, Adam.
• 2:12
  Speaker 2: I I had not done that math. I was just I was just thinking I get 18250. Oh, cool.
• 2:17
  Speaker 1: I just knew that. Yeah. We got a raise. I should
• 2:19
  Speaker 2: read the block.
• 2:20
  Speaker 1: You sure read the blood about the locust. We all got a raise. Okay. Yeah. It was nice.
• 2:24
  That was the end of last year. But I found that to be almost, like, chilling that they basically Nex had come to the same conclusion, but then it had all gone horribly wrong. And I really wanted to understand why. So I got this book, Steve Jobs and the Next Big Thing, from Randall Strauss, copyright 1993. So just to give a quick next primer, and I'm hoping that there are are folks that if there are folks here, by the way, who work with Next and have exposure to Next, definitely wanna hear from lots of other folks on this.
• 3:00
  But the quick next primer, Steve Jobs leaves apple Apple in 1985. Next is bought back by Apple in 1996. So it's 13 years. It's pretty long. And, the what I did not I just didn't really think about Next, I don't think.
• 3:24
  Because by the time I was coming up, Next was kind of was very clearly struggling and then pivoted away from hardware entirely. But here's the thing I was really not braced for. So I recommended this book. Basically, we had this weird space. The book looks good.
• 3:39
  I buy the book. I also recommend it to CJ who joined us at Oxide. Adam. And because CJ was the the CJ Mendez is asking for, like, I I wanna read some interesting books. And so he I did this thing that I try not to do too frequently where I was recommending a book that I myself had not yet read, because you never know how that's gonna go.
• 3:57
  Totally. And CJ was like, man, I really I loved Steve Jobs and the Next Big Thing. And you're like, you're right. Like, I learned a lot about Sun from that book. And I'm like, what do you mean you learned about what?
• 4:10
  And this book is, I swear, the best early history of sun microsystems. Microsystems. And in fact, basically the entire book contrasts next to sun. So the and with a couple of just amazingly big reveals that leads us to me personally. So and, Tom, you're here.
• 4:33
  I'm hoping you I'm gonna, get you to speak because I I've got a burning I I am really wondering about this because, the spark station 1 was Sun 4 c. Right, Adam?
• 4:48
  Speaker 3: Right. Correct.
• 4:49
  Speaker 1: And do you know what the c stands for? No. No idea. Okay. So I did know this, that the c stands for campus, and it was called campus.
• 4:56
  And so the architecture was called campus. I did not know why it was called campus. I've just never known. And, Tom, do you know why it's called campus? Tom is going to be on mute.
• 5:12
  I'm sure Tom does know why it's called campus, but I did not.
• 5:16
  Speaker 4: Right. Yeah. Here I am. Yeah. And Andy actually started a separate company to go after the education market And that's what that's what Steve Jobs at NEXX were doing.
• 5:27
  And I think that's partly why Andy wanted to go there.
• 5:30
  Speaker 1: That's exactly it. And I And Yeah.
• 5:34
  Speaker 4: And the rest of the sun was like, who needs education until, you know, until Andy came out with the specs for the campus and I was, like, oh, this is a killer machine. So they rolled it back into Tom.
• 5:47
  Speaker 1: So they rolled it back in the sun. Yeah. So I guess and, Tom, obviously, I should have asked you that earlier because I it was always a little bit of a strange code name, campus. And it's because Next was targeting higher ed. Like, the the market that they were gonna go after was higher education, which at a point that this book makes repeatedly makes no sense.
• 6:10
  Like, higher education, not really a I mean, we all know for our own university departments. Like, it's not like there's a lot of extra money sloshing around higher education, although it feels like there should be with what tuition costs. But it's a very strange market to go after.
• 6:27
  Speaker 4: Funny because because a lot of Sun's earliest sales were also to that market. It was to the scientists and engineers and university types who were all the the forefront of the UNIX stuff.
• 6:41
  Speaker 1: That's right. And Sun was and, Tom, I'd love to get your take on this. But definitely, Strauss' take is that Sun was winning that market without having to try very hard. That it was it was just kind of a consequence of winning the workstation market. Sun was winning also the academic workstation market.
• 7:01
  And Next was focused really only on that market with a machine that was pretty underpowered. Is it so, Tom, what was your impression of Next at the time?
• 7:14
  Speaker 4: As soon as we found out what it really was, we were all like, oh, that's not competitive. I mean, the hardware was really lackluster, but damn it. It sure looked good.
• 7:25
  Speaker 1: It did look yeah.
• 7:27
  Speaker 4: So it was the Steve Jobs reality distortion feels like, oh, this is so cool, but just don't try to use it for anything.
• 7:35
  Speaker 1: That's very concise. And and when you say look good, just to be clear, you're talking about, like, the the physicality of the machine.
• 7:43
  Speaker 4: Right. Right. The the design industrial design.
• 7:46
  Speaker 1: The industrial design, which in this thing And
• 7:49
  Speaker 4: and the UI. The UI was great.
• 7:52
  Speaker 1: Right.
• 7:53
  Speaker 4: But, you know, but it wasn't Unix underneath. Not quite.
• 7:57
  Speaker 1: It wasn't. And the so the the and the book is is just loaded with with Steve Jobs' about the so he they wanted the the black paint on this this this 1 foot by 1 foot cube, which they felt had to be exactly 1 foot by 1 foot. For reasons that aren't really clear, I'm not sure. Tom, was that did that have more of a meaning in I I mean, I remember, like, the cube. I just don't know why they thought that was the right form factor.
• 8:29
  It's a very weird decision. And Yeah. The It's
• 8:33
  Speaker 4: just a I
• 8:37
  Speaker 1: Well, and they contrast it to the pizza box, to the Sun Pizza Box being like a in some ways, as rigid a form factor, but at a much more, it seems to make a lot more sense. Alright. So so the it it on the, like, the black paint on this thing, I mean, this is, like, feels like such classic Steve Jobs. Like, they are having they wanna have no seams on the outside of of of this cube. And so they end up, casting it, with I wanna say, like, magnesium.
• 9:12
  Although, that doesn't make sense something I'm saying. But they they basically is it the the cube is casted as a single unit without any seams, and then they had to find the right paint that wouldn't it was just like, it's nuts how much time and energy they spent on the look of it.
• 9:29
  Speaker 4: Well, the other thing the other thing that's nuts is they spent a huge amount of money out of the factory to produce these things.
• 9:35
  Speaker 1: Right. So what do you remember when the factory was
• 9:37
  Speaker 4: gonna be? Pump pumping out 100 of 1000. Really just that they're building a huge factory, and I was like, oh, really? It was about the time that people were starting to outsource everything. Yeah.
• 9:49
  Speaker 1: And this is a point that that Strauss makes as well that Sun was not was very deliberately outsourcing manufacturing, and, jobs to decide to manufacture absolutely everything, that Next would manufacture absolutely everything, which made really did that make sense at the time, Tom? That makes no sense now. Sounds like
• 10:11
  Speaker 4: No. You know, app Apple kinda, was doing some of that too. You know? So so for super high volume and and quality control, it kinda makes sense. But there's there's no way Nextiva was gonna get to that kind of volume.
• 10:26
  And, Sun was doing incremental things and Steve was just trying to, you know, do the 100 yard
• 10:35
  Speaker 1: in the
• 10:36
  Speaker 4: past or
• 10:36
  Speaker 1: Yeah. The kind of the Hail Mary. And it there was it it is amazing just the total lack of market focus, like, total misunderstanding of what the market is. Next does not know who they're selling to. Very, very confused by who they're selling to.
• 10:52
  And they think that, like, they decide and, Tom, tell me if this term rings a bell. They decide that, like, Sun is dominating the worst workstation market, and they clearly can't compete with the Mac in the personal computer market because the thing is too expensive. So Atomade targeted a $3,000 price point for this thing, and they're coming in at, like, 10. Yeah. Which is a miss.
• 11:17
  Speaker 2: I mean, it sounds sounds like the the Lisa all over again, not the Apple 2 or the Mac.
• 11:20
  Speaker 1: It is Lisa all over again. I mean, no. I think you're right. I think it is Lisa all over again except without any but but, like, worse in many ways. And Right.
• 11:29
  And it so the thing is, I mean, they don't know so they decide, like, it's too expensive for that kind of that personal computing demographic. We're not competitive with Sun, and Tom and crew are, like, rolling their eyes at the at at how, underpowered this thing is. So we are gonna go into the the personal workstation market, a market that they invent. The problem is that there is no personal workstation market at that time. So it's, it it does not go well.
• 12:00
  Speaker 4: Well, you well, you can tell, you know, Steve Jobs never never understood the power of networking either.
• 12:08
  Speaker 1: Interesting.
• 12:09
  Speaker 4: And, he has this quote. I saw saw somewhere about how the next computer was so that some Stanford student can cure cancer in his dorm room. Amazon. What? Well, okay.
• 12:21
  What about the other 10,000 computers on campus? Who the hell
• 12:24
  Speaker 1: Oh, okay. So this is actually very good. You're making I mean, yeah, this is just dead on because they talk about this over and over again where next refuses to interoperate with anything. So they have this idea that the a next customer is gonna buy all next machines. And they do convince, I think, poor is it Drexel University?
• 12:43
  They do convince one university to go all in next, before but the I with with, like, with Berkeley, for example, they really wanted to get Berkeley to run next. And Berkeley's like, we got we got 2,000 Sun Machines here. Like, we're not going to and then we've got a bunch of other workstations too. It's not just the Sun Machines. Like, you're gonna have to interoperate.
• 13:03
  So, Tom, it sounds like the the at Sun, that was accepted as a constraint. I mean, obviously, the network is the computer. You understood that, like, you had to interoperate with other things. Right.
• 13:14
  Speaker 4: And we made we made good money doing that, you know, with all these interoperability from others.
• 13:19
  Speaker 1: Well, that's it. You know, another point that they they make a lot is that Next was a really proprietary company. This is, like, not a deep this is not a deep thought at all. But where it's really contrasted to Sun being very much an open company and, working within so they in particular, Tom, they call out the creation of the spark clone market by Sun deliberately, like licensing spark designs to create spark clones, which I always use kind of a nonevent. But, I guess it it it gave customers assurance that they were not getting a proprietary solution.
• 13:55
  Speaker 4: Yeah. And it was kind of a a hope that that there would be some market, Some didn't discover that would help the volume and all the spark stuff.
• 14:05
  Speaker 2: And so, Brian, Brian,
• 14:06
  Speaker 1: after I
• 14:07
  Speaker 4: Nothing about that.
• 14:07
  Speaker 2: Brian, after I saw your tweet today, I read the introduction to the book. And one of the things that I I found really interesting was the the Jobs, Gates, Microsoft, Next dynamic. And so to what degree did the did the book go into depth on that?
• 14:23
  Speaker 1: They do go into depth. I mean, where I mean, Bill Gates, volunteers that he would gladly urinate on a next machine. So Apparently verbatim. They are because, I mean, they are attempting to reinvent absolutely everything. So they need all software to be written from scratch effectively.
• 14:42
  And they are really struggling to convince Microsoft to port anything to it. So a big part of Next is going around to these, like, software vendors and basically lying to them about the side about Next's performance, where they're basically I mean, they're engaged in all sorts of things where their their channel stuffing and all kind of all sorts of, like, similar similar misdeeds to convince software vendors that there was gonna be a big market there. And so that plays a big role. Microsoft itself does not play a big role because I don't think it was, like, close for Microsoft. I think Microsoft is just like, no.
• 15:22
  Speaker 2: Well, Brian, did you did you read the the introduction? Because there there was the great anecdote where Bill Gates shows up for a meeting with Jobs and Jobs makes him wait and, like, walks around the office and kind of parades around making a show of talking to everyone except for him, like, thumbing his nose as hard as he could at the the partner who it turns out he needed most.
• 15:42
  Speaker 1: He he did. And so that is, I would say, part of a larger theme of and Jobs says this over and over and over again at Next. He does things to make Next look bigger than it is. And so my inference from that, which the the Gates were countered as well, is that that was not although maybe those all maybe it was a 2 for it. It was also an act of personal domination, but it's as much to be like, I actually don't need Microsoft that much.
• 16:08
  And in a way to look like, more important than Microsoft. And he does this over and over and over and over again. I mean, it's over and over. You're just like, good god. Can you just, like, stop being a dick for, like, 30 seconds?
• 16:23
  And he he has, like, a a meeting with, with IBM, and this is in later next when, he had licensed next had licensed Next Step to IBM. They desperately needed IBM's support. And they needed IBM basically to re up their license because he he had managed to get this IBM to license Next Step software, but no future versions of Next Step software. So it's like, oh, would that oh, yeah. No.
• 16:49
  I'm sorry. If you want, like, the next version, like, that's a different that's a whole different thing. That's Next Step 2 dot o. So he is gonna go out to present to 900 people at IBM. And just like Adam, the slip the IBM analog of the Bill Gates story, There are 900 people who are gonna he's gonna present to you in Dallas.
• 17:07
  He has demanded there are 2 slide projectors for this presentation, and there is only one slide projector available. The next exec who's basically gonna go out there with him knows this and is trying to figure out when do I explain to Steve that there's only one slide projector, not 2? Because I know he's gonna flip his shit when he finds this out. So so he has the idea of, like, I'm gonna wait until we're I I it's like, I actually, do I wait till we're on the plane, or do I tell him before we get on the plane? He's like, you know what?
• 17:32
  I'm a tell him before we get on the plane. That way, he can really think about how he's gonna change his presentation to accommodate the fact that it's only 1 slide projector. So they're at the airport. He tells him this. Job says, I need 2 slide projectors.
• 17:43
  And if I don't have 2 slide projectors, it's not worth my time, and he goes home. And the and there are, like, 900 IBM execs waiting to hear. And so this guy then flies out by himself to present to to these 900 execs who keep waiting for, like, when does Steve Jobs come out? You're like, no. Steve Jobs is not coming out.
• 18:02
  Steve's Steve Jobs is at home because you don't have a second slide projector. And which is just, like, insane. I think to anybody that that has ever I mean, like, how do you anyway. It it so he Adam, to to answer your question, like, he does that a lot, and it's not it's not clear what he could possibly be thinking, but he does try to make the company look really, really big and important. And one of the way he the ways he does it is by treating people really badly, which is Sounds like Gary killed all over again.
• 18:37
  No. Right. Exactly. Right? I mean, I feel I feel like that that early computing is littered with a lot these kind of personalities that are that try to, like, abuse people into, into cooperating with them, which I don't think that way works.
• 18:52
  I mean
• 18:55
  Speaker 2: So so, Brent, I don't Well,
• 18:56
  Speaker 4: there there is a lot of Sunpro companies that bought bought the vision from Next. I remember, being very jealous about I think, Mathematica went whole hog.
• 19:07
  Speaker 1: Mathematica did go a whole hog on Max. And
• 19:10
  Speaker 4: Instead of fun. And and we had worked with them early on, and then they went in between.
• 19:15
  Speaker 1: That's interesting. So I think, actually, Tom, Mathematica stands as really an exception here in the at least in the book. Did you talk about Mathematica? Mathematica was definitely one of the so I actually did have that question from the book because they talk about Mathematica on next. So Mathematica was not then on other workstations.
• 19:33
  It was effectively only on Next.
• 19:37
  Speaker 4: I believe so for the first year or 2.
• 19:39
  Speaker 1: Yeah. Wow. Well, that would be consistent with the book. So that's that is a a surprising decision. Well, clearly, they came to their own realization that, like, that market was not gonna be as big as they had been led to believe.
• 19:53
  Speaker 4: Yeah.
• 19:55
  Speaker 1: Adam, have you ever seen a NeXTcube?
• 19:57
  Speaker 2: We actually had one in my high school. What? Yeah. So we had a bunch of we had a we had a Vax 3 m s. We had and we had a Next Cube, at my high school.
• 20:08
  Speaker 1: So did you program the Next Cube?
• 20:11
  Speaker 2: No. I I I mean, so this was, no. So I didn't program myself. I like I it was in some backroom. I sat down in it once.
• 20:23
  Definitely did not, like, do anything more than, like, click around at a few things. And it's sort of I remember everyone sort of regarding it as a beautiful useless piece of machinery.
• 20:32
  Speaker 1: That is a very accurate, I think, assessment. Especially, this is now when you're there, this is circa like mid nineties or 95. This is like after they stop making hardware.
• 20:42
  Speaker 2: Yeah. So I graduate from so it was probably, yeah, like 94, 95.
• 20:48
  Speaker 4: I I think Steve Jobs would count that as a win. You know, beautiful.
• 20:51
  Speaker 1: That's that's right.
• 20:53
  Speaker 2: Right. That's what mattered. Right?
• 20:55
  Speaker 1: Well, the it actually oh, in one episode that's relayed in the book helped to explain so my own introduction to Steve Jobs, I never met him, but my own, like, glancing blow with him was I was over at a friend's house. This is growing up. This is in the very early nineties. And, heard, dad was being screamed at in the next room over the speaker phone. And I kinda shot her look like that.
• 21:21
  God, that's like, what's going on there? She said, no. It's fine. It's Steve Jobs. This This happens a lot.
• 21:26
  I'm like, Steve Jobs yells at your dad a lot? And as it turns out, and this is and this is actually related to the book. Her dad is Fred Ebrahimi, who was at the time the CEO of Quark, which made QuarkXPress. And Tom Mathematica was on the net on Next, but they they he could not convince QuarkX to port QuarkXPress to to Next. And, and so he the book talks about, like, him berating Ebrahim.
• 21:54
  Like, hey. I think I'm on the other I think I think there's a there's a young confused Brian Gantra at the other end of that call. The the and, you know, he is, like, you know, he is, you know, using basically slurs of of of Ebrahim's national origin. It's just like it's really but that that tactic did not work for Quark as it turns out. It turns out that that worked very poorly.
• 22:21
  Speaker 2: So I'm I'm sorry. You you you but you know this the story of Steve Jobs trying to sell, NeXT machines to the Brown Computer Science department.
• 22:29
  Speaker 1: I do. I know the story.
• 22:31
  Speaker 2: Oh, this is great. So there was there was this Bake Off and it was like Next, Sun, and Deck. And I think it was like, Tom Depner, the systems professor, advocating for Sun. Steve Rice, the, I guess, programming languages, programming environments professor, advocating for DEC. And then Andy Van Dam, legend of computer graphics, advocating for for NEXT.
• 22:55
  And then John Savage, the head of the department. Advocating for peace. Well, that's right. But then, apparently, having a confrontation with Steve Jobs, where he where he basically said, you know, your your product looks great. I'm just not so sure your company is gonna be around, for as long as we need it to be.
• 23:14
  And then Steve Jobs calling him an asshole and storming out.
• 23:17
  Speaker 1: So that's very on that's very consistent, I have to say. With the that that anecdote would would slot right in with many just like it in this book because a lot of people do question next's viability, and Steve Jobs does not deal with that very well. That question is not the other thing is in order to kinda convince people that it was viable, he next spent very freely. So they were very lavish in their offices. They were very lavish in, like, having things catered.
• 23:50
  And which is needless to say, when when money is short, this might not be the best way to I I'm not sure if it ever would convince anyone of their viability, but definitely shorten their runway.
• 24:03
  Speaker 2: Well, that was that was very I was reading the Wikipedia page, which by the way, one of the pieces they they said, $75,000 a year for all the employees up to some point, and then it dropped down to 50,000. So you so you had like managers making, less than the people reporting to them and and and weird things like that. But the, not that is weird in itself, but but, you know, weird inversions like that. But the other thing they said was, like, Ansel Adams prints everywhere and I'm Pay Architecture, and just just crazy. But but the but money, I mean, money was pretty easy to come by weirdly enough, I think.
• 24:40
  At least venture money, not customer money. Money.
• 24:43
  Speaker 1: So no. That he did this is what the book goes into. In terms of, like, the number like, they went from crisis to crisis to crisis. They got Ross Perot. He did not take VC money.
• 24:52
  He had weird money from beginning to end. So it's Ross Perot who and they think, like, who got basically an unlimited check from Ross Perot, which was true until they spent all of it. And then Ross Perot is very, he thought Steve Jobs is a total genius and then realized that he was whether he was a total genius or not, he wasn't selling any computers. The and then they get Canon to to pony up a huge check. So I think Canon, I wanna say, puts in, like, a $100,000,000.
• 25:21
  And then Canon ends up being really pot committed. And ironically, Ross Perot got into next in part because of, like, some anti Japanese American nationalism coming out of jobs. And, Tom, I'd be interested to know if I mean, certainly, I remember this this kind of zeitgeist as a kid when, you know, Rockefeller Center had been bought by a Japanese holding company and people were very there's a
• 25:54
  Speaker 4: Yeah. The eighties were all about fear of Japan. Totally.
• 25:58
  Speaker 2: How do you thread that needle though between canon and anti Japanese sentiment? Like
• 26:04
  Speaker 1: Well, glad you asked. That's exact and that's exactly the needle needs a thread. And then Canon especially when, Canon wants to start pushing for manufacturing to be done offshore, basically. Yeah. No.
• 26:17
  It's a total needle look. Right? And this is what he he he does thread these needles though over and over and over again and uses this kind of star power to manage to get that kind of that that next round of funding. But, ultimately, they did. Like, ultimately, they had to pivot away from hardware, and they had to get get rid of the manufacturing line that that Tom scrapped.
• 26:38
  And so, Tom, but this the thing that's interesting about this is all of this stuff is used in the book in contrast to what Sun was doing at the same time. So Uh-huh. And, certainly, like, I viewed Sun as, like, a I mean, I thought it was, like, per I I think Sun did not feel lavish from a spending on employees. I don't think it felt, like, cheap, but I don't know, Adam. What do you what do you think?
• 26:58
  I felt felt like I
• 26:59
  Speaker 2: mean, they they they took away our doughnuts on Wednesday mornings.
• 27:02
  Speaker 4: So That's right.
• 27:03
  Speaker 2: So I would not say it's lavish. And
• 27:06
  Speaker 4: Yeah. Like like McNealy doesn't do lavish.
• 27:09
  Speaker 1: I also I mean, I didn't feel like the I don't feel like it was, like, an Amazon like point of making, you know, of, like, making people work on, like, doors on saw sawhorses to, like, they mean you because Bezos had this idea of, like, really telling people and doing lots of things that were really, like, taking frugal to an extreme. I don't think some took frugal to an extreme, but it was not lavish.
• 27:33
  Speaker 2: No. Totally. It wasn't it didn't make you work in weird conditions. But but also yeah. I mean, it it looked
• 27:41
  Speaker 4: Actually, the the other the other fun bit of history is when when Sun was occupying building 1 on the Mountain View campus across the pond and in what became building 5 was Metaphor Computer Systems, which is a whole company you should look at the history of, but they were lavish. So we'd look over there. I was, like, wow, nice furniture. Oh. What's that?
• 28:03
  Another party? And yeah. Very different. Interesting.
• 28:07
  Speaker 1: Well, so so here's something that that it, Tom, that I really want your take on. And I I I tweeted this, day before yesterday because I thought it it was so interesting. But because this again, the the book and, Tom, you you have not read this book. Right? I like yeah.
• 28:22
  I I think you I would love to get your take on it. Although it's telling you a bunch of the stuff that you already know, but it's or you're confirming a bunch of things that it's saying. But the because it does talk about Sun so much, they actually the, Strauss actually calls out why he feels that Sun does not actually he says the the the Sun variant deserves more attention than it is received. Measured by most any yardstick that one could choose. Sun was most one of the most successful companies, stories of the 19 eighties and for all industrial America.
• 28:54
  And he says that, in Sun, we have the the the makings of a terrific tale, which by now should be part of American folklore by now being 1983. It's actually pre Internet, which is kinda funny or pre eternal September anyway. Yet its story is relatively unknown because its founders are not obsessively self aggrandizing like Steve Jobs or Lee Iacocca or Donald Trump, because they freely share credit among themselves, which means there's no single herculean hero, which I I mean, I don't know if I wanna accuse Vinod Khosla of being, like, of of low ego, but maybe it is a fair characterization. Tom, what what do you make of that?
• 29:31
  Speaker 4: Well, I think they all had high ego except except maybe Scott, but just it wasn't turned turned that direction in terms of self aggrandiz self aggrandizement. But, you know, Andy and Bill and Fanon and all, like, are all forces.
• 29:55
  Speaker 1: They are all forces. But you also have some kind of reflection on them. I'm like, and I totally agree with you. These are not, like, low ego individuals, although with maybe with the exception of Scott and I I totally agree with that. You know, it some culturally, the with the with some exception of Build Joy and Java,
• 30:18
  Speaker 2: reality. Right? I mean I think there was there was typically a realization of of actual events going on around us which which seems you know, absent the next
• 30:29
  Speaker 1: certainly. Yeah. Actually, Adam, that's a pretty good point in terms of, like, there is no there's no McNeely distortion field.
• 30:34
  Speaker 2: That's right.
• 30:35
  Speaker 1: Yeah. I definitely certainly. I Tom, I don't know what your take is. I I certainly agree with that. Yeah.
• 30:42
  People are pretty grounded. And certainly, the, it's anyway, I thought that was interesting. I thought it was an interesting take about, like, maybe, you know, maybe we shouldn't be waiting for a book on Sunday because it's not gonna happen because there's there there was no mythos that was created as part of it. You know? Maybe that's the not not that the company should have had its own Steve Jobs, but I think it's kind of an it's it's an interesting take on why maybe it's underreported.
• 31:10
  Speaker 4: Yeah. There I mean, Sun had amazing characters. So it it could be a big movie or something. But, like, John Gage. Oh.
• 31:19
  Oh my god. Do do you know John Gage?
• 31:22
  Speaker 1: I do. I do. And, I don't know. I've not seen him in in in years, but John and have you did you meet John Page at him?
• 31:34
  Speaker 2: He's Yeah. I did around the around the Oracle acquisition, actually. Because didn't he go to Oracle?
• 31:39
  Speaker 1: God, he did he? He is he is spellbinding. He's someone who I felt was very, I don't know, Todd, what your take is on, John, but I always felt him to be a great storyteller for sure.
• 31:52
  Speaker 4: Well, he he knew every scientist, rock star, politician, and news person on the planet. And he he was just one of these connecting nodes of everybody and always had amazing stories. And, you know, just because he was at Sun, Sun Sun became interesting to huge numbers of people.
• 32:13
  Speaker 1: That's interesting. Yeah. So he left with in 2008 according to to Wikipedia anyway. And, then went to to VC, of course, like any like any self respecting former Sun exec, I guess. Tom Lyon accepted.
• 32:32
  So the the only thing I that I that was interesting in terms of, like, the the technical contributions of Next. Because that I've I mean, I don't I don't know. I've Adam, like, did you spend any time with with Next step? I tried to install it once, and it was very slow. But I didn't
• 32:49
  Speaker 2: Not with Next step. Like, only with the sort of, early versions of Mac OS 10, like the pre release versions that that had not yet kind of shed the vestigial next step components, and kind of take it on its its full fully realized macOS form. But it was it was only kind of briefly in that period.
• 33:12
  Speaker 1: Okay. So did you read this this objective c hopple paper?
• 33:17
  Speaker 2: I did. I did read that. I did read that.
• 33:20
  Speaker 1: Okay. Have you written any objective c?
• 33:23
  Speaker 2: So I'm gonna I'm gonna say basically no. But I've written like a teeny bit just because I needed to interact with some components of Mac OS. But like, I'm I'm not fluent at all. I don't claim to understand any of the idiosyncrasies of it.
• 33:38
  Speaker 1: Yeah. I mean, I haven't written no objective c. And other and I've always been kind of like, I don't know, curious, but I've always also viewed it as like having I mean, it's basically a and small talk together, which doesn't seem like it can end well. But, so this oh, Adam, what'd you make of the HOPWA
• 34:04
  Speaker 2: Well, so I I it's hard to to not read that hopple paper kind of playing back an alternate history. And I was wondering about about the, the book as well. Like, if Apple does not acquire Next. If Apple, you know, cheaps out and acquires B instead. And I know there's lots of what ifs, but but, like, does does next matter?
• 34:26
  And certainly, does objective c matter? Like, you know, the the Hoppell paper talks about, you know, the customers and the market, but really lands on next as the only folks using this thing in anger. So, if next drops to a relevance or if, you know, the stars don't line up for the iPhone and so forth, like, does, does it, you know, how much do we care about objective C? And I think it was hard for me to read that with read the paper without that thought in the back of my head, which was like, you know, it it it all fell, this one event.
• 35:01
  Speaker 4: I had similar thoughts about the mock micro kernel. I mean, after all these years, has it actually made a difference?
• 35:08
  Speaker 1: Yeah. I think you're both right. I think that that both of those these technologies are in a vessel that has that survives despite the odds. Because the by the time they're acquired in 1996, I personally had forgotten about Next. And I think, like, just forgotten that they were alive at all.
• 35:30
  I mean, I think the company was in very dire straits. And the the the when and actually went back and just reread the chapter on that on the acquisition next in the Isaacson biography. Just gonna wanna remind myself of that. And the it what Isaacson says anyway, the the biography of Steve Jobs, is that Emilio was basically told whichever company you buy, that's gonna be the future CEO of Apple. And Emilio really did not like Jean Louis Gasse.
• 36:03
  Apparently, hated Gasse more than he hated Jobs. And Jobs also hates And Gasset, I think, feels more ambivalent ambivalently about, about Jobs than Jobs feels about Gasset. But but I was amazed that after having read this biography of Next effectively, the Isaacson biograph biography of Jobs damn near doesn't talk about NeXT. He talks about the founding of NeXT, and then he talks about the acquisition of NeXT by Apple and does not talk about the intervening 13 years at all.
• 36:34
  Speaker 2: Well, that's what the introduction does such a great job of that. Somehow, Next was always a company that was 2 years old, even when it was 8 years old. The press would write about it like it was, you know these up and comers give them a break they've only been around a couple of years why would they have any product or have sold any
• 36:49
  Speaker 1: product? That that's right. That's right. It was a and he was just yeah, masterful at that about kind of being perpetually young, which of course although but in so in the end, and the answer to your question is from my perspective is, yeah, I don't know that objective c makes it at all if next perishes in 1996. I'd love to get someone else's take on that, but I I I don't think it's the and the paper itself, did you find it awkward that 2 of the authors of the paper are are
• 37:26
  Speaker 2: Yes. Yeah. Are are are talking about themselves in the 3rd person and their great accomplishments. And and I think that probably the way it played out is that, Sue was interviewing Cox and Naroff. Right?
• 37:39
  So they weren't necessarily writing these these statements about themselves necessarily. But, which I don't know, maybe that's too generous, but I think took some of the edge off of it for me. But, you know, when they you know, if if, if I wrote a biography, if you and I got together and wrote the biography of observability and described ourselves as legends of tracing, You know, I'd I'd feel like that was maybe a little much.
• 38:07
  Speaker 1: Maybe a little much? Well, how about do you like when they have, totally different beliefs about why this company, PPI, which renamed itself to StepStone? The board or orders itself to rename itself because it's what is it? Personal Productivity Incorporated or whatever it is? They're like, no.
• 38:24
  You have to rename yourself. It's like, okay. We'll be Stepstone. Like, I don't know. Maybe we'll go back to PPI.
• 38:28
  Like, how about we you do a little bit better than that?
• 38:31
  Speaker 2: Oh, well and you have Naraf who was the employee, describing, some of the actions of Cox, the executive, and how, like, the they got a bunch of investors and the investors, you know, insisted on getting a new CEO. So that's a story. I'm sorry. You can't just gloss over that. You can't just say, well, that's that's just a normal thing that happens.
• 38:52
  Like, that that's reflective of some really deep misgivings.
• 38:55
  Speaker 1: Oh, they're definitely deep misgivings. And I like the line again, this is like one co author talking about another. Cox concerned himself with doing his own research, but collaborated only minimally with the development staff, treating his job somewhat like an academic appointment. I'm like, isn't he a he's a wait a minute.
• 39:10
  Speaker 2: Like, the the
• 39:12
  Speaker 1: the same Cox that coauthored the paper?
• 39:14
  Speaker 2: I've I have the exact same, like, line underlined with a star next to it, and I just wrote, wow. So yes. Agreed. It was
• 39:24
  Speaker 1: I mean and I'm like, are you know, I've I've not that I've read a ton of Hopple Papers, but I don't think Hopple Papers spill the tea to quite this much. I mean, this is
• 39:34
  Speaker 2: No. And and there there's there's the bit about how, you know, object oriented programming and GUI are are basically interchangeable terms. And it's like, you know, I I I get that, you know, early on that that OOP was a very useful paradigm for some of these object oriented. But just, I don't know, it felt it it felt like a lot. It felt like they were making a lot of very bold claims.
• 39:59
  Demeaning a bunch of other languages like disparaging c plus plus and Java for having, parameter so for lacking the ability to name parameters.
• 40:09
  Speaker 1: They make a big deal about that. And I Oh. Yeah. What do you think about it? I because I kinda I view the name parameters as actually a bit of a a a bit of an anti pattern, actually.
• 40:20
  I'm sure that if folks disagree, like, hop in here if you but I think name parameters can be kind of a gateway drug to functions that are really hard to reason about.
• 40:30
  Speaker 2: Well, I get it. When you think about when you think about small talk as this message passing where the ordering of components of that message might be independent, I think it may make sense in, as you sort of narrow the focus there. But, but I I did think it was interesting and I was wondering how triggering this was for you, Brian. That they described, objective c as a soldering iron, meant to connect these other components rather than, I I guess rather than the language that everything would necessarily be built of built in, you know, in and of itself.
• 41:04
  Speaker 1: I will tell you that was not triggering. Do you think that triggers maybe are you are we are we making fun of the fact of my poor soldering ability or is this with a spec? You could. I I I am I am getting better at soldering slowly. And I'm injuring myself less.
• 41:23
  Speaker 4: My only my only attempt with objective c was looking at doing add, drivers in Mac OS. It was like, I get there. I was like, oh my god. Why the hell did they do this to the kernel?
• 41:35
  Speaker 1: That is a very good question. And so and then this is where I do think that, Adam, the part that was triggering for me was and just tell me your your observation that you're doing this in the kernel. The the part that was triggering to me was all these references to the software crisis.
• 41:50
  Speaker 2: Oh, yeah. That was really interesting. I I I I need to go dive into this. But the software crisis being that software is rising in importance. And that kind of software is unbuildable through enough quality, and being able to manage the process.
• 42:09
  Speaker 1: Also programmers are shitheads. Don't forget that. Like, there's a they they say, like, they like, this often justified such programmers high salaries and contributed to their sense of unmanageability. Like, that sentence was triggering for me because
• 42:23
  Speaker 2: Yeah. The fact that these were correlated too. Right? The more productive a software engineer might be, like, the the, like, more unmanageable they were they also are.
• 42:33
  Speaker 1: Exactly. And and the the the the the these kind of, like, entitled prima donnas and the hardware is the real thing and the the software and it it did remind me that this was a really this was a prevailing zeitgeist in this era where and
• 42:52
  Speaker 4: but, you know, the what the the the way that those two things can be true is if managers are trying to do the wrong thing.
• 42:57
  Speaker 1: Exactly. Yes. Exactly. It's like we would be a lot more manageable if managers were trying to do the right thing a little more frequently, by the way. Yeah.
• 43:06
  I know. That's exactly right. And why also so it's, Tom, I would love your take on this especially because I'll tell you I don't reading that. The I was reminded about that. Oh, you have a software crisis and, like, we are not gonna be able to build softwares, this idea.
• 43:19
  That, like, software is becoming so complicated, and it takes so long, and so many of these things run over budget. And to me, the thing that broke the back of that was actually open source. And the not you so people had this idea that, like, object orientation would break the back of that, and you would have these, you know, the these software ICs as they called it here, software integrated circuits. And, you know, one of the things that one of the authors of the paper wanted to do was charge only $300 for the compiler, but then $30,000 each for the libraries effectively. And that makes software on that would make software on build.
• 43:58
  I mean, if you can imagine it, we if we were deprived of all open source software right now, we would would be kind of back in a very different era.
• 44:08
  Speaker 2: And they they talk in the paper about how that's an unviable model. They talk in the paper about how taking some compo like assembling a bunch of components that you're soldering together is unviable because, like, if there are bugs in those components, they are unfixable. If there are design flaws in those components, it's unfixable.
• 44:27
  Speaker 1: Yes.
• 44:28
  Speaker 2: And, like, so there's this dichotomy of the the the raison d'etre of objective c being also invalidated by the market concerns that they described. And one of the things that was so interesting in here, beginning bringing it back to next, was this description of how how PPI was building these software ICs, these these extensively reusable components across different operating systems and environments. And the they weren't that good. And the ones who were the the folks who were building them, the the good ones, was next because they were also using them. And I think this this harkens back to a theme that we've talked about a lot of times on these spaces of, like, using the things that you're making.
• 45:07
  And, like, building a framework without actually
• 45:17
  Speaker 1: feuding couple element of this paper is Naaroff and Cox, 2 of the authors disagree over that role at PPI. Naaroff has a opinion that that because we we weren't actually building anything real with this, NEX was the one that was actually building real things, and they were the ones that were actually hitting the real issues and PPI wasn't.
• 45:37
  Speaker 2: Well, here's a great line. They say, thus despite significant early problems with objective c, next decision would
• 45:54
  Speaker 1: Alright. So I know that that I I can see that that that Rick is here. I know Rick, you've actually dealt with objective c in a way that certainly Adam and I have not. I would love to know your take on to what degree objective c achieved its kinda aim of composability because that's kinda the big question.
• 46:14
  Speaker 3: Yeah. I mean, I I wrote a lot of Objective c, during my time at time at Apple. And, like, there's there's certain aspects of it that carried through and and were really, really nice. Like, the way it works into a UI framework, definitely, there's there's a lot of nice, fit there for a lot of the dynamism. You you don't know exactly what message you're sending to.
• 46:40
  And as long as you have the correct interface, it's all good. And you you can create a UI infrastructure, which is kinda how Next step made it work, where you can do things like entirely switch out a theming layer or an implement like, a large implementation in a fairly seamless way. That said, there's a lot of not not so nice parts about objective c with, like there's there used to be a a routine exercise of searching SDKs for the longest objective c method names. Oh.
• 47:14
  Speaker 1: And I
• 47:15
  Speaker 3: just pulled up a list right now from 2010 where in 2010, the longest method name found now this is method name. This does not include the class name. The method name was 202 characters long. Right? Because because, ultimately, when you consider what's happening, named parameters actually become part of the method name.
• 47:39
  Speaker 1: Oh. And that string
• 47:41
  Speaker 3: that whole string, so the the base name of the method, which is really the parameter or the the name of the first parameter, is all appended together, and then that is actually what's issued down into objective c message send for the actual function dispatch. So imagine that you're you're dispatching like, your your your equivalent of a c plus plus v table is actually you taking a 202 character string, somehow hashing that, and then, you know, dispatching based on that. That's what's actually happens internally in the language, which means that, you know, when when we're writing performance analysis tools, obviously, message send is the name of the method, would often turn up, which basically just meant you're making a lot of Objective C method calls.
• 48:25
  Speaker 1: Interesting.
• 48:25
  Speaker 3: And that was usually an indication that you had made a mistake because everyone knew that if you got into any sort of performance critical code or where you're doing any sort of tight loop, that the actual message dispatch cost was way too high. This led to a lot of, you know, hybrids where the c part was used for writing the high performance aspects, and then the objective c was to provide a nice, interface into the rest of the ecosystem. We did this in, like, the symbol lookup libraries where you actually have a large data structure. You're mostly doing small operations and and searching through data structures. You can do all that in c very cleanly and then box it up and hand it back in an Objective c wrapper, which is fairly lightweight.
• 49:11
  Speaker 1: Yeah. Interesting.
• 49:13
  Speaker 3: It also led to the even stranger world of Objective C plus plus?
• 49:19
  Speaker 1: They talk about Objective C plus plus in here. I had not heard of Objective C plus plus until this paper. Okay. Yeah. So did you have to use Objective C plus plus at all?
• 49:27
  Speaker 3: We did. So so one of the things that we ended up making was a dictionary that mapped ranges, so, like, address ranges to objects. And and you can imagine why this would be useful in, like, a debugger performance analysis case. Right?
• 49:42
  Speaker 1: Like Absolutely. Yeah.
• 49:43
  Speaker 3: You're looking up you're trying to do queries of what symbol is this address. Right? But you wanted the infras like, the wrapper of objective c for hooking into larger parts of of the infrastructure and, like, the UI, pieces. But the implementation needed to be fast, and so that, ultimately, I had a coworker who wrote a wrapper that then used c plus plus STL. It's like, map class to or the the map template to implement an actual range based dictionary lookup.
• 50:17
  And so that kind of stuff was pretty common where people would bring in, you know, existing c and c plus plus platforms and just wrap it with enough objective c to make the UI interactions work.
• 50:29
  Speaker 1: Interesting. Well, so in a lot of ways, like,
• 50:31
  Speaker 3: while it worked well as a UI framework, it that's kind of where it stayed, and most people just treated it as, like, an unnecessary evil.
• 50:42
  Speaker 1: Well and so they performance is very much a nongole for objective c, which they make clear in the sample paper. They do not they are I mean, it's a it's a they view it as an objective c c hybrid or small talk c hybrid, and they're much more concerned about the composability of software than they are about its performance, which I think Rick gets to exactly that's exactly your experience. It's like composability, pretty good. Performance, pretty bad. So you have to be careful about the way you construct the hybrid.
• 51:13
  And it it I mean, it's it's I don't know. Adam, I'm sure you had the same reaction readiness. It's really impossible not to think about Rust as it kind of delivering it all here and allowing us to get it all.
• 51:26
  Speaker 3: Yes. Swift is Swift is also a really interesting Yeah. Yeah. Right. Because that's that's the logical successor to objective c.
• 51:34
  And in a lot of ways, it it delivers more on the performance side while providing some of the niceties of the Objective c kinda goals. But they had
• 51:43
  Speaker 1: to go about it in
• 51:44
  Speaker 3: a wildly different way. Right? The idea of applying small talk to this just does not work at the performance that they wanted.
• 51:52
  Speaker 1: Right. They actually talk about so they did talk definitely talk about Swift and kinda Swift inheriting kind of the mantle of that. They also Adam, did you find strange the kind of the comment that Naroff hired Lattner Yes. At Apple? But then he goes out of his way to point out that he then reported the Lattner later.
• 52:07
  I was just it's like there's a lot of, like, organizational, like, ephemera in
• 52:10
  Speaker 2: this paper. Absolutely. There's a lot of weird, like, like it's like, mute, like, self reflexive Wikipedia editing, it felt like on some of the stuff. Just like, I don't I don't care what your org chart was. I don't know.
• 52:23
  Like, it it it it did feel like it went out of its way to to name drop some of
• 52:27
  Speaker 1: these things. To name drop, I know.
• 52:29
  Speaker 2: But it just And it also felt like a a sort of a mixed opinion about Swift. Like, that, that Swift sort of, you know, was, was taking the mantle, but unfortunately, you know, really the Swift idioms looks nothing like objective c idioms. So it it seemed a little mirthful, almost. I mean, a little little regretful almost on on some these things.
• 52:49
  Speaker 1: Yeah. That's interesting. Well Yeah. Sorry. Go ahead.
• 52:51
  Speaker 3: And any any mention of, like, bringing in Latner for Swift is sort of missing a big piece of the the actual history, which was that, Lattner and the LLVM folks were really brought up as interesting for an alternative compiler infrastructure because Apple had gotten in their head that GPL 3 was going to be a terrible thing, and they needed to excise anything that would be potentially GPL 3. And then much later, Swift came about.
• 53:20
  Speaker 1: Oh, interesting. Yeah. They don't talk about the licensing at all. They do they mentioned LVM and Lattner being part of LVM, but not in the licensing context. I think it's really in the objective c 2 dot o context, before moving on to, to Swift.
• 53:38
  It would Swift, presumably, would be the subject of its own, HOPWA paper. So we'll get I I I would imagine Swift seems to to merit it. Of course, there's a there has to be, I believe, a 10 year latency between, the the the things that are being described and the Hopple paper, if memory serves. I have to say, like, having read this paper, I can see why if, like, everyone is gonna, like, just just unload on their former colleagues. I guess you wanna make, you know, the the the longer the latency, the better.
• 54:09
  Speaker 2: We suppose an interesting one just in that like, you know, Apple is able to push, and and the paper makes this, the hop up paper on objective c makes this point that, you know, folks, flocked to objective c because of the introduction of the iPhone and the App Store. And so, you know, Swift will be significant. And and again, when when asked the question, you know, absence this necessity, you know, if it were an open ecosystem, would it would it be as significant?
• 54:38
  Speaker 1: Yeah. And I and it's certainly certainly my takeaway from both the book and from, honestly, the Objective c story is just the, the importance of open source and and open not just open standards, but actual, like I mean, I just think it's it it is, it's a big part of of both Next and objective c was the the degree to which all of this was very proprietary, which made it I it mean the idea because it feels to me that objective c, at least a PPI, is built as much around a proprietary software business model even though they don't appraise it as such. Adam, is that a Yeah.
• 55:16
  Speaker 2: Yeah. Profoundly. And and as I said, you know, it it it feels like this, you know, buried in here, it makes the case for open source and the necessity of open source, in terms of assembling these these components.
• 55:29
  Speaker 1: There was one, line that, Rick, in here that you would love where I mean, love in the, the guffaw at where they want to be to software what Intel is to hardware. It's like, dude.
• 55:46
  Speaker 3: I this doesn't really surprise me. I mean, there there's I was I was on Apple at a very interesting stretch of history. Right? I I was there from, like, 2,000 3, so started the end of PowerPC through 2009, which was just after iPad launched.
• 56:06
  Speaker 1: Yeah. That's a real inflection point.
• 56:09
  Speaker 3: There were there was always this divide between the core OS group, which is kind of led by the former Next Step folks who really had a view that the hardware didn't matter. If you, if you created good abstractions and you, you could basically do everything in software and, and the hardware folks were just there as a, a means to make the hardware or the software something that gets executed, but they shouldn't trouble themselves with any details. And the hardware folks, of course, were all, that's nonsense. Like, we have to devote a huge amount of resources to make your software run well in order to hit performance targets, etcetera. And I mean, to be fair, I was right in the middle of this and that I was on in the Mac hardware group writing performance analysis
• 56:56
  Speaker 1: tools. Right. So you're right on the coal price of it.
• 56:59
  Speaker 3: Well and and that was as a response to the performance tools on the team on the core OS side, which had their own tools that refused to dip into the hardware.
• 57:11
  Speaker 1: That's very interesting. And so and and so you could definitely still see those kind of those fracture lines from the next acquisition. Those folks are I mean, is there a there are a a healthy group of folks coming from next in that or at least culturally?
• 57:26
  Speaker 3: I mean, there were definitely a lot of folks that in high in sort of the leadership positions that had a lineage back through Next and and, all the you know, you have, some various folks from the mock community too that were very influential in how the system should work.
• 57:46
  Speaker 1: So one question I have for you on that on that note, Rick, is because the I I mean, I know Apple had secrecy as a a tradition when from before jobs left. Next definitely quadruples down on secrecy and uses secrecy a lot. And in part because the truth was that Next wasn't doing very well as a company. It wasn't selling very many. So they, like, necessarily, there was a lot of secrecy.
• 58:12
  How much of that was culturally brought back to Apple? Do you think Apple got more secretive as a result of the next acquisition?
• 58:23
  Speaker 3: Apple was always fairly secretive from what I heard. Right? Like, I I I got there well after many of these events. I I can certainly say that my time there, I would have to pass through 1, 2, 3, 4, 5 different badge readers to get to my desk.
• 58:46
  Speaker 1: Oh, Jesus. Wow.
• 58:50
  Speaker 3: And around the time that iPhone started, they started doing NDAs per project. So being on a hardware performance tools group, I had an NDA per project that my software was targeting, which was everything.
• 59:10
  Speaker 1: Is is every engineer assigned their own lawyer to negotiate the NDAs they need with various projects? And how does this world work?
• 59:17
  Speaker 3: Well, it it actually leads to a very different cultural situation of, you don't know what other projects your friends are working on. And so there there was a whole admit during my time there, there was a whole cultural aspect of negotiating of, you know, sort of a negotiation protocol of which projects are we both disclosed on so that we can talk about something?
• 59:46
  Speaker 1: Without disclosing them. You always say you have to, like you you have to describe them abstractly enough that you're not,
• 59:51
  Speaker 3: Right. I have learned that at some point since I left, this is now an internal tool where you just enter the other person's LDAP, and it comes back and tells you what project you have in common. So
• 1:00:05
  Speaker 1: they have it for
• 1:00:06
  Speaker 5: surprised they don't I'm surprised they didn't learn from the, VFX world where you would have a code name for a project. So, like, I worked on Harry Potter and the Deathly Hallows part 1, but internally, there was a code name, which is, like, extra time. So whenever we talked about it in public, we would be talking about extra time so people wouldn't know that we were working on, you know, Harry Potter. But for yours, you would just be like, oh, yeah. Are you working on extra time?
• 1:00:28
  I'm working on extra time. And then you would know. Right?
• 1:00:32
  Speaker 3: Well, that was mostly the thing. There was a long time where the power PowerMax were all, you know, p and then a number. So p51, p67. And and that covered all the laptops and the desktops. But then as they got into the Intel machines, those became the Ms.
• 1:00:54
  And then the iPods became Ns. And then some of the Ns were actually iPhones. And you could guess what products were coming because of the pattern that they were using for assigning code names. And so that's where, you know, they created their own cultural problems by,
• 1:01:13
  Speaker 1: you know it's a big one. Secrecy is so much work.
• 1:01:17
  Speaker 6: I I was actually I I I I couple of years ago, I I read the blog post about how how Apple did all of that. Someone probably blog blogged about it. And my first reaction was, this is so useless. Why would anybody ever need this in a company? And then maybe a couple of years passed by, and then the, the NSA in Armenia asked me to run a project for them.
• 1:01:39
  But the deal was none of the army officers who were coding had to know what the project was. And my first reaction was, oh, okay. So this is where this could be ever handy.
• 1:01:53
  Speaker 3: And and for Apple, it's all about maintaining the secrecy of it so that you own the initial presentation of it. It was all about being able to walk on stage and dramatically drop something that was going to be life changing in some way.
• 1:02:11
  Speaker 1: So that and, Rick, I'd be really curious for you to read this this history of Next because I I have to say that, like, that is definitely, like, part of it for sure. And maybe even, like, what jobs would claim is, like, it did it for that kind of pop. But it really does feel like there's this darker side of it where, transparency is prevented to to kind of prolong this distortion field and get people to because in particular, they would you know, the way that jobs got people to a lot of hours out of jobs was by kinda creating crises And by also promising that, you know, there are these massive orders or they mean there are a lot of promises that were not actually backed up by facts? And it's hard to not see that the the secrecy was being used to to manipulate people at at some level. Maybe maybe not deliberately.
• 1:03:05
  Although it feels pretty deliberate. I don't know. That that maybe be too dark a read.
• 1:03:11
  Speaker 3: No. I I think that'd be fair. I mean, to to give some examples
• 1:03:17
  Speaker 1: from my time there.
• 1:03:17
  Speaker 3: So, the x86 shift, people had to be read in to that project, so you had to have your NDA. And one aspect of that was that they had set up an entirely separate OS to x86. So, like, Darwin had always been ported to x86, and so that had been maintained for a very long time. But the upper layers of the stack all had to be ported. And instead of just, you know, tasking the teams that owned various pieces of the infrastructure, They literally just set up a entirely parallel organization that was then charged with making a snapshot of the source repositories and working through the issues, and they were not allowed to talk with the original authors or the actual maintainers of those components.
• 1:04:05
  And that got them through up to the point of publicly announcing t one, the transition kits. A similar thing happened for iPhone, where iPhone started as a project. They created an entire little parallel org. They made an entire copy of the build infrastructure and all the source code repositories and just started hacking away at it. It was very interesting for that one specifically because they came calling to my team 3 months after they had started
• 1:04:35
  Speaker 1: Right.
• 1:04:35
  Speaker 3: Because the tools were not working well.
• 1:04:37
  Speaker 4: Right. Right. Exactly.
• 1:04:37
  Speaker 1: It's all broken. It's like, what's all broken? Oh, the shadow thing that I created. Right.
• 1:04:43
  Speaker 3: And it turned out that they had made copies of of our code repositories and not really understanding the hardware aspects of it at all, commented out a bunch of things and basically, you know, disabled most of the functionality. But also, what they didn't realize was that they had copied source code repositories that we had explicitly locked down because they were covered by 3rd party NDAs with processor vendors. And we never really got a resolution for that. But it was just, like, the the need to create these parallel orgs to have the secrecy to drive folks for, you're building the next big thing, and you're doing it in secret. And we're not even gonna read in the people that have the best knowledge of this.
• 1:05:23
  It certainly was around building a culture of, like, a a small team is gonna get this done. And once it's released, then we'll bring everybody else in and you'll get a break, except that never happened.
• 1:05:34
  Speaker 1: Well and so I have seen one another thing that that is again over and over again in the the book on jobs and at next is his oscillation between everything was either great or shit. And there's nothing there's no kind of nuance in between. And I did laugh at the his would be, in the paper, Adam, Jobs talks about c. And it I think he, he's, like, he he tells them to make it great, because it's currently shit, which I felt was very, very on brand. And then I love the fact that they've got this, yeah.
• 1:06:13
  He urged love to folks on making the core language great and to stop, quote, wasting time on the IC package, the libraries. I loved the the Steve Jobs meeting had a profound impact on Nirav. His ability to articulate problems with objective c was impressive. Am I the only one that reads euphemism into that?
• 1:06:31
  Speaker 2: It it was I I it was so weird. It was so weird because, it it was so
• 1:06:35
  Speaker 1: sycophantic too because he referred
• 1:06:36
  Speaker 6: to him as, like, the CEO
• 1:06:36
  Speaker 1: in what
• 1:06:43
  Speaker 2: CEO, the chief engineering advocate. Like that.
• 1:06:45
  Speaker 1: The CEA, the chief engineering advocate.
• 1:06:48
  Speaker 2: Yeah. It's hard work. I mean I mean, this guy clearly got snowed by a person who I mean, like, all respect to Steve Jobs, but I I I doubt that he was getting down into the nitty gritty of of, of, like, whatever Naraf found objectionable or or great about objective c.
• 1:07:05
  Speaker 1: That's a I'd be interested to know. I I get the sense that he is obviously not as technical, but it but was that that didn't stop him from really weighing in on things. Certainly, I get from the at least from again, from the the next book. It does remind me of the the Sun CTO, Greg Papadopoulos, described, press once, that he did I don't know. Describing because he was working with Gates a lot, on the the Java integration with c sharp, I guess.
• 1:07:35
  And every time he met with Bill Gates is when Sun and and Microsoft were definitely enemies. Gates would just berate him for some very small technical detail, like, the first 30 minutes. And he just had, like I just have to kinda, like, let him go, and then you can, like, actually have a meeting. But he has to, like, I don't know. It feels like jobs has the duration with not necessarily the the the same mastery of technical detail.
• 1:07:58
  Yeah. I know we would want sorry. Go ahead.
• 1:08:03
  Speaker 6: Thank you. I did want to ask, since I'm out of topic a bit. Is objective c being used today anywhere outside of the Apple ish ecosystem? And, I assume also, Ganostep.
• 1:08:17
  Speaker 1: I think the answer to that is broadly no. But I would is It it
• 1:08:21
  Speaker 2: would be hard to imagine because this this paper, again, was was so aggrandizing that it, it, that I, I'm not sure why they would have omitted a, a reference, however minor, to someone using it. Is that is that unfair?
• 1:08:36
  Speaker 1: I I it doesn't feel no. It doesn't feel too unfair. Although, I do feel that, like, I mean, Apple is certainly the I I don't think it's broadly it is not used outside of of Apple. I, and, and, you know, step I do not I mean, I remember I I got a new step curious at some point but couldn't get it to work, and that was the end of it. I don't know.
• 1:08:53
  And, Charniak, have you actually used Canute Step?
• 1:08:56
  Speaker 6: I actually do because I do use WindowMaker, the the desktop environment on my free b s d machine. So I I sometimes I even, like, write couple of small objective c codes. You know, those are small things like, what what do they call it? It? Applets?
• 1:09:11
  At at the at the left or the right side of the screen. Like, it could be like a duck that would start boiling when the CPU is too hot.
• 1:09:19
  Speaker 1: Right. Right. Right. Yeah. This is, like, this is, like, why people learn Ticl so they can interact with their EDA tooling.
• 1:09:24
  So, yeah. You you may be it. That that that may be the well, I know we wanna keep this to about an hour. Thank you everyone. Sorry to binge read this over the weekend.
• 1:09:45
  Definitely, it was a, it was a fun read. Theo, I I saw you wanted to get in here. You wanna you wanna give us some parting words? Theo may have actually tried to hang up on us and accidentally hit the request button. I don't know.
• 1:10:02
  But, Tom
• 1:10:03
  Speaker 7: Oh, sorry. I'm here. I'm here. Yeah. I was gonna contribute that there is a significant populations that does agent based modeling using objective c still.
• 1:10:16
  Speaker 1: There we go. Agent based modeling. The that that that community is alive and well with objective c. So
• 1:10:22
  Speaker 7: it's alive. It came
• 1:10:23
  Speaker 1: out of
• 1:10:23
  Speaker 7: the University of New Mexico, but it still exists in some high high brow consulting companies that do, like, genetic model genetic algorithm modeling of business business logistics and things.
• 1:10:37
  Speaker 1: Alright. Well, another place that one can take one's objective c, aptitude, I guess.
• 1:10:47
  Speaker 4: Yeah. Cobalt There
• 1:10:49
  Speaker 1: you go. Exactly. I'm sure it would all survive in fertility. The but, thank you very much, everyone. Tom, thank you, especially for being able to fact check some of the stuff about Sun.
• 1:11:00
  But, this this was a lot of fun, and, we'll look forward to to seeing you all next week. Thanks, everyone.
• 1:11:07
  Speaker 2: Thanks everyone.