0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
/
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
/
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
ANNO·​TRICESIMO·​DIE·​DVCENTESIMO·​QVADRAGESIMO·​VITÆ·​POVYA
FirstIsland:TimelineofEvents
First Island : Timeline of Events
Quotes & Excerpts

Well a few days [after getting taking over the role of Avie], I got the first phone call in my new position. That was, I forgot what, some great idea of Steve or that and I realized that there was nothing I could do to prepare for those phone calls. I just had to be me, and so that’s what I decided very quickly. There was no one else [who] I could say, “Avie, please help me.” So it just was me, and that worked out.

BERTRAND SERLET

Yeah, I thought that Objective-C was a handicap and Objective-C was–had good sides, technically. But nobody knew Objective-C outside of the NeXT community. So in '97, the only folks knowing Objective-C were the NeXT community folks. NeXT and then the developers of NeXT, right? A very small group and this was somewhat confirmed when Adobe said in '97, we don’t want to be part of your Rhapsody effort. So I wanted to escape Objective-C, so I pushed towards Java. Java was growing very, very, very fast in those days. I even met James Gosling to try to see if Java could do a few things that would make the Mac better. James Gosling was absolutely not interested, and so we tried Java, but the problem was that Java was huge. It had a huge footprint. We had the same calculators, you know the little calculators that you find on a Mac done in Objective-C and in Java and Java was 10 times larger in resources and so ultimately, Java was not viable. So then moving forward to the iPhone, the iPhone 1.0 was implemented in Objective-C, but was not open to developers and the iPhone 2.0 was open to developers and there was a bit of a debate with Steve Jobs and Scott about, should we open it at the level of web apps, or should we open it at the level of Objective-C, Cocoa? And I was arguing that Objective-C would be a handicap, that very few developers will start writing iPhone applications in Objective-C. I was totally wrong, of course. Steve overrode my objection and said, no, no, no, no, it will work, and he was right and so we went out with Objective-C and at that point, Objective-C was in wide adoption by all Mac OS [X], but more importantly, all iPhone developers. So by the late 2010. But by that time, Objective-C was kind of old, had not changed much. The only thing of significance was the thing that Blaine Garst had done, which is the autorelease, automatic autorelease, so that you can get rid of allocation and free minutia and so we had Chris Lattner, who had done LLVM that we had hired, and he was toying with a new language and I strongly encouraged him, because I thought Objective-C was too old, and we can do much better with modern technologies and then I left. But Chris kept working at it, and that became Swift.

BERTRAND SERLET

Yeah, garbage collection was a pet thing of mine, because I had been raised with a garbage collector. Even when I was programming in the late '70s, I had a garbage collector. So I thought cleaning up by hand when you have a computer is kind of a waste. But the problem with garbage collection is the performance of it and it has a huge performance cost and so the breakthrough was actually this automatic auto-release that brings most of the benefit of garbage collection, but without the cost, and in essence, that’s smart pointers in C++, and that’s how Swift all works.

BERTRAND SERLET
Day's Context
Open Books