marcusb 18 hours ago

My favorite anecdote from Soul of a New Machine is about Josh Rosen, who designed the MV8000's arithmetic block. According to the story, after troubleshooting various timing and layout issues he left a note on his terminal:

  "I'm going to a commune in Vermont and will deal with no unit of time shorter than a season."
zabzonk a day ago

I love books about writing/creating new software/hardware. One I can recommend, about Windows NT is: https://www.amazon.co.uk/Show-Stopper-Breakneck-Windows-Gene...

Does anyone here know about any book(s) on the VAX, or indeed on any of DEC's machines, or IBM's stuff (which were - probably still are - beautifully engineered) come to that? I'm aware of "Big Blue", but that is mostly about IBM business practices.

  • sillywalk 15 hours ago

    Inside the AS/400 by Frank Soltis, is a deep dive into IBM's AS/400 and its predecessor the System/38. Quite interesting. There is a 2nd or 3rd edition called Fortress Rochester: Inside the iSeries, but it tends to be hard to find and expensive.

    [0] https://archive.org/details/insideas4000000solt/

  • ghaff 14 hours ago

    That's probably the other product development book I'd most recommend after Soul of a New Machine. Also Mythical Man Month but mostly for historical reasons and probably a couple of the essays provide the most important flavoring.

  • pinewurst 21 hours ago

    IBM's Early Computers A Technical History by Charles J. Bashe, Emerson W. Pugh, John H. Palmer and Lyle R. Johnson

    IBM's 360 and Early 370 Systems by Emerson W. Pugh, Lyle R. Johnson and John H. Palmer

    Computer Engineering: A DEC View of Hardware Systems Design by C. Gordon Bell (Author), J. Craig Mudge John E. McNamara

canucker2016 a day ago

One of the main technical nuances about the new Eclipse box mentioned in the book was the management edict for "no mode bit" in the machine architecture/CPU.

The book never explains how the team solved this restriction.

Someone asked that exact question and Carl Alsing himself (head of the Microkids) answers the question, see https://cs.stackexchange.com/questions/44915/data-general-mv...

  • peterfirefly 14 hours ago

    One of the tricks was that exceptions/traps/interrupts were handled in a neat way.

    The microcode looked at the first instruction of the handler. If it was an old instruction, a backwards compatible stack frame was created. If it was a new instruction, a new-style stack frame was created.

    Data General published a couple of papers about their new 32-bit architecture.

spcebar 6 hours ago

One of my favorite books by my favorite author. Tracy Kidder gets to the heart of everyone he profiles like no other. Strongly recommend his book House if you like Soul of a New Machine.

Funny anecdote: I've read Soul of a new machine many times over the years, and love it dearly. I stayed with my aunt and uncle last year and somehow the topic of mini computers came up. I mentioned Soul of a New Machine and my uncle said "oh yeah I worked on that computer." I was blown away. Apparently he had worked at Data General and he was on the software team (which is briefly mentioned towards the end of the book) that wrote the software the Eagle was released with.

forrestpitz a day ago

"Soul" is one of my favorite books. Kidder writing is phenomenal. He explains complex topics with ease and the storytelling drags you in from introduction. I also think it might be the best example of what great management looks like. Tom West isn't the focus of the story but without his hand none of it could have existed. Cannot recommend highly enough.

  • habosa a day ago

    I just finished this book a few weeks ago, it was excellent. I don’t normally read nonfiction, and when I do I avoid reading about business or technology. Despite this book being all 3 of those things I highly recommend it.

    One of the most thorough and well written pieces of journalism (it reads like a book-length magazine feature) that I’ve ever encountered. The characters and company felt so alive.

  • ghaff 14 hours ago

    I also like his House. Unfortunately the topics of his other books haven't grabbed me sufficiently to pick them up.

TomMasz 15 hours ago

I read the book when it came out and it was indeed a great story. Kidder did a remarkable job explaining technical topics that even non-techies could understand. You could just feel the pressure the team was under to deliver.

The Eagle seemed like a neat design but I never encountered a DG machine of any kind in the various companies I worked for. VAXen seemed to be everywhere, Sun and other UNIX workstations followed only to be shoved to the side by PCs in the early 90s (except for some applications like CAD and that was only for a while). Funny how such a vibrant industry segment could die so quickly.

klelatti a day ago

Author here (of the post not the book!)

A kind reader pointed me towards an interview with Dick Sonderegger who worked at Data General at time of the development of Eagle [1].

His view based on discussions with people who appeared in the book is interesting:

'They did not have flattering things to say about Tracy Kidder ... This is a historical novel, this is not what actually happened'.

'To a man [they said] he didn't know anything, we could have told him anything and he could have bought it. And reading the book that's accurate'

DG section is at approx 21 mins

[1] https://operationcode.org/podcast#dick-sonderegger-link

  • ghaff a day ago

    I don't know.

    I worked at DG for 13 years (albeit a few years post-Eagle) and I've never heard anything along those lines. And the company liked the book as far as I can tell--they regularly gave copies away at the executive briefing center.

    I have no doubt there were embellishments for the purposes of narrative. And people always have different takes on all sort of things that they participated in at a company. But I've never heard it was substantially inaccurate and I knew a number of people in "the book" pretty well.

    • klelatti a day ago

      Thanks for sharing this.

      I definitely didn't intend to endorse this view. I think if it was really true that they had purposely misled him and he'd taken it all in then there would be some major technical 'howlers' in the book. As far as I can see there aren't any.

      Maybe they didn't like the way they were portrayed - which I can understand - and there was a collective, organic 'well of course it wasn't like that' soon after the book was published.

      • ghaff a day ago

        Yeah, I was product manager for a number of subsequent 32-bit MV minicomputers and, later, a number of Unix-based systems including the big NUMA servers. It's been a while since I read the book but I certainly don't recall anything that caused me to go "That can't be right."

        No first-hand knowledge of all the internal politics that shook out in the creation of the MV/8000 (Eagle) but, again, never heard anyone say it was BS and I knew many folks both in Westboro and RTP quite well.

deterministic 10 hours ago

One of my favourite books. Read it three times.