123: Imperfect Signaling Mechanism

Accidental Tech Podcast

Marco Arment

Technology

4.3 • 2.5K Ratings

🗓️ 24 June 2015

⏱️ 87 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Follow-up:
- Bitcode is specific to its originally targeted CPU architecture
  - Bruce Hoult tweet 1, tweet 2
  - C data types
  - Endianness
  - Data structure alignment
  - Steve Troughton-Smith
  - S1 analysis by Chipworks
- The (un)likelihood of ARM Macs
  - Apple's transition from PowerPC to Intel CPUs
  - Intel Skylake
- Swift 2 and compiler self-hosting
  - Mono's compiler history
- Oculus as "the Palm Pilot of VR"
John, his watch, and his nose
- Myke was right again
- Myke was originally right about the iPhone 6 Plus
Safari Content Blocking API in iOS 9 and El Capitan
- WebKit Content Blockers
  - Session video with John and Marco
- Safari View Controller to kill minibrowsers
- GlimmerBlocker
- YouTube ad settings
- Ad-banner blindness
- The ethics of running ad-blockers
- AdBlock Plus accepts money from some ad companies whitelisted as Acceptable Ads
- John's Reload Button extension for Safari
TRIM) on non-Apple SSDs
- New trimforce command in El Capitan
After-show:
- Casey, John, and Marco attempt to modify their houses.

Transcript

Click on a timestamp to play from that location

0:00.0	So this is the show, right? I guess our first sponsor.
0:02.4	If you want to feel free.
0:05.2	It's been 18 minutes.
0:06.8	I guess we might as well.
0:07.7	That's why you're going to cut all this out.
0:11.0	Mirror two or three days ago, we were talking about Bitcode.
0:17.6	How it may or may not allow Apple to change CPU architectures.
0:21.0	I was coming down on the side of mostly not because Bitcode is an LVM thing and LVM IR.
0:26.8	Actually is architecture specific in certain ways.
0:30.1	But I didn't have details on what those ways were.
0:32.8	Bruce Holt on Twitter sent some information about that.
0:37.1	He said it's architecture specific in the same way that
0:40.8	C code might be.
0:41.6	So when you specify int or flow or pointers or whatever and see the C standard
0:47.2	doesn't dictate what size those are, the size is dependent on the target architecture
0:52.0	that you're compiling for.
0:53.1	I mean, these days they're all similar sizes for 64-bit.
0:56.8	Architectures and stuff.
0:57.7	But the C standard doesn't dictate what size those should be, which is why you should
1:02.6	do size of int and all that.
1:03.5	Good stuff and not just assume that at 16 or 32 or 64 are same thing with float and
1:07.6	double or whatever.
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from Marco Arment, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of Marco Arment and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.