#241 - Chuck Peddle

[economist]

Can't wait to find the time to listen to this one. The 6502 probably changed my life. As a young teen, and after getting bored with BASIC on my Apple II, I decided to learn 6502 assembly language. It was a fascinating experience that ended up getting me interested in math and science like nothing else. No assembler. We were lucky to afford just the computer. So it was paper, pencil, and writing out the instructions and then typing them in. I might not have gone to MIT were it not for that programming experience. MIT was the next big change in my life, but I really think a lot of it began with that 6502.

[gildasd]

Really good show.
I've had to listen to it 3 times to get my head around some of it.
This kind of information tends to "reshuffle" a lot of things in one's head.

Brilliant stuff.

[EPLan]

I'm definitely going to listen to this episode, the Commodore 64 was my first computer and I too began learning assembly language as a young teenager. It has always frustrated me that people like Wozniak get all the publicity for the microcomputer revolution but he was standing on the shoulders of Peddle - who most people haven't heard of. 

[BradC]

I'm definitely going to listen to this episode, the Commodore 64 was my first computer and I too began learning assembly language as a young teenager. It has always frustrated me that people like Wozniak get all the publicity for the microcomputer revolution but he was standing on the shoulders of Peddle - who most people haven't heard of. 

Disagree strongly. That's a bit like saying Ford was wrongly credited with starting the mass produced car revolution when he was standing on the shoulders of Otto.

Woz took what was a neat processor and created a finished product around it, with all that requires to facilitate actual day to day use. Peddle created a nice microprocessor, but to the layman it's still a plastic blob with 40 legs that does nothing without all the supporting hardware. Like the 4 stroke internal combustion engine is a neat idea, but it takes a lot more than the engine to get people from A to B.

[EPLan]

Disagree strongly. That's a bit like saying Ford was wrongly credited with starting the mass produced car revolution when he was standing on the shoulders of Otto.

Woz took what was a neat processor and created a finished product around it, with all that requires to facilitate actual day to day use. Peddle created a nice microprocessor, but to the layman it's still a plastic blob with 40 legs that does nothing without all the supporting hardware. Like the 4 stroke internal combustion engine is a neat idea, but it takes a lot more than the engine to get people from A to B.

Well Peddle also created the Commodore PET ... and Apple weren't particularly relevant in the UK microcomputer market as far as I can remember. There is an awful tendency of Apple fans to rewrite microcomputer history in their image.   

[mtdoc]

It was a great show and Peddle had lots of great stories to tell. He obviously deserves credit for many things across a wide spectrum of the early microprocessor industry.  But I also got the sense that when it came to the success of Apple he had a bit of an anti-Wozniak bias.  There must be some untold stories there - perhaps some competitiveness. Of course he's only human so I don't think it detracts at all from the kudos he deserves.

[EEVblog]

Woz took what was a neat processor and created a finished product around it, with all that requires to facilitate actual day to day use.

That's not the way Chuck tells it.
He said Woz didn't even want BASIC on the Apple II, and it was Chuck's team who actually added it (or some such, don't recall precisely right now).
Chuck also says he was the one who was in Job's garage and helped Woz get his Apple 1 design working, being the field app guy for the 6502.

[photon]

A great podcast. I liked the way Chuck started it off. You knew he was opinionated and smart, not afraid to say what he thinks. He is still in the game after all these war stories.

[george graves]

Awesome show. You guys should just send him a case of wine, and a portable audio recorder, and let him go to town.

[nixfu]

Wikipedia has a nice picture of that 6502 demo board that had the PIA on it which he kept mentioning that he took with him everywhere he was demoing the 6502.




I learned assembly on the 6500's.   I believe the 6500's are THE CHIP that made the computer revolution, NOT the Intel 808x.  Without the 6500's we would not have the Commodore PET/Vic20/C64, the BBC/Acorn Micro, Apple Computer(I/II/II-gs), Nintendo(NES/SNES), or Atari(2600/400/600/800XL). 

It is these devices that really paved the way for what we have today as home computing, not the Intel chips with their IBM computers that no one used at home until many years later when the clones took off.  Although I would also give credit to the Z80 of course for the TRS80's and  Sinclair.

[LabSpokane]

What struck me about the show is that Chuck's new invention is .... the RAM disk.  I really miss RAM disks.  There was nothing faster than shoving all the executables and support files that you knew you needed into the RAM and running directly off that.  None of the cacheing / paging schemes seem to really alleviate hammering the hard drive for every trivial task. 

I also was surprised to hear about how RAM can be "repaired".  Are they really fixing broken gates, or are they simply marking known bad ones to be excluded?  Can anyone point me to some references on that?  Googling revealed little, so I'm obviously missing a key word.

[gildasd]

Used a RAM disk about 6 years ago.

It was to help with Photoshop:
The shop were printed from floor to ceiling and had a constant fade out of colour over 2.5m high, times the length of the wall.
Producing the files was really pushing it (no way to compress without getting weird VISIBLE artefacts) but printing was really hard, very much at the limit of the 4 to 8 gigs of ram we could put on a machine back them. Print files would often reach 250 to 500 megabytes, and the actual output (to the printer) be in gigabytes.
And, this particular printer (5m wide), its dedicated computer and the supplier's work flow was not mac compatible.

Anyhow, with a 8Gig RAM disk as a Photoshop primary scratch disk (and nothing else), all our troubles went away.
These things really work, I have no idea why they are not more popular.

[SeanB]

The repair is to use an onboard ( to the drive board, not the chip) microcontroller that on the production line does a memory test of each cell, and then generates a bad block table to enable masking out of dead data areas, then writing the table to an onboard EEProm so the table is available on power on in normal use to transparently ( to the outside) appear as a single contiguous block of memory, even if the inner side is using a whole load of scrap chips with a bit here and there that work, and large chunks of chip missing but still having intact IO capabilities.

This works but needs to have spare capacity on chip to have some basic ECC ability so the micro can see if some block has gone bad during data retention, then reconstruct the data and deliver it ( without showing the outside world it has done that in most cases) then either fix the data or move it to spare space and update the table.

Basically what a modern hard drive does but in RAM.

[Towger]

The repair is to use an onboard ( to the drive board, not the chip) microcontroller that on the production line does a memory test of each cell, and then generates a bad block table to enable masking out of dead data areas,

I thought they all do this anyway. He seemed to be implying he did something 'special'. To me it sounded like he/they were producing dodgy ebay flash memory from memory the manufactures were rejecting as they were marginal/had to many faults..

[SeanB]

He probably was the first to do this, and at the time it would be novel to think of making a dodgy piece of silicon appear to the outside as perfect, hiding all the errors and remapping data on the fly. remember at this time disks had a bad block table you entered either by hand or by using a mapping program, which was visible to the OS.

[robrenz]

Most excellent show   Just one correction, some master machinists are still around

[LabSpokane]

I will have to relisten but I thought Chuck said "fixing" which to me is very different from marking.

[SeanB]

Fix so the outside sees it as perfect silicon. Inside it is held together with bits and blobs, but the appearance of the interface is the only thing you see.

[NF6X]

I loved this interview! It was quite fascinating. My early computer experience was mostly with the 6809 (TRS-80 Color Computer) followed by the 68000 (Dad's Lisa, followed by my Amiga 1000), but now that I'm interested in retrocomputing I'm backfilling all of that missed experience with the 6502, Z80, etc.

I'm intrigued by the ramdisk product that he's now working on, but it's not clear to me how it would perform much differently than a traditional SSD with aggressive caching on the host computer. Maybe I'm missing something? I can see potential benefit from the capacitive power backup to allow it to flush cached data out to flash on power failure, since embedding that function into the drive might be more reliable than performing it the traditional way with a UPS for the full system and proper power failure monitoring... particularly since I suspect that most home users don't bother with a UPS, and getting the power failure monitoring working properly isn't trivial for the non-geeky.

[Rigby]

Most excellent show   Just one correction, some master machinists are still around

I was thinking of your videos when he said that.

Got any apprentices?

[SeanB]

His son.............

[LabSpokane]

I'm intrigued by the ramdisk product that he's now working on, but it's not clear to me how it would perform much differently than a traditional SSD with aggressive caching on the host computer. Maybe I'm missing something?

It would be interesting for someone with experience in caching algorithms to chime in.  Because, technically, we shouldn't need a RAM disk.  We should just be able to stuff the machine full of RAM and it would automatically become faster with a few setting changes, but I think most folks experience is that is *not* the case. 

This could also be a case of misguided nostalgia for the ramdisk of yesteryear.  Maybe my glory days of being able to personally manage everything with a few batch files weren't so glorious after all...   

[Goober_in_CA]

This was my favorite show to date.
Bring Chuck back please.
Daryl

[I wanted a rude username]

Thanks for reposting this informative and inspirational episode! (PubDate: Mon, 30 Dec 2019 03:22:44 +0000)

Libsyn only seems to list the latest 100 episodes, so new podcast listeners ordinarily wouldn't encounter it.

[Ed.Kloonk]

I wanna know what was discussed after the recording stopped.

 

[fastguido]

Thanks for reposting this informative and inspirational episode! (PubDate: Mon, 30 Dec 2019 03:22:44 +0000)

Libsyn only seems to list the latest 100 episodes, so new podcast listeners ordinarily wouldn't encounter it.

Thanks for the repost, iTunes only shows the 2019 podcasts.

A question though, Chuck mentioned that Radio Shack, killed Heathkit. Can someone please explain that statement? I never went to RS to buy a kit, just parts and assembled items.