Software with ZERO bugs: Dave Cutler Interview

Dave Cutler discusses shipping VMS with zero known bugs, versus the practice of closing thousands of known bugs.

[@TurntableTV]

Dave, we want more of these interviews. As an aspiring software engineer, hearing some of the great minds that made the software I use on a daily basis is pretty inspiring. I hope I get to hear Miguel de Icaza or Scott Hanselman as guests.

[@MikeHarris1984]

Yes! These interviews are awesome!!!

[@wandelust]

We're listening to a legend!

[@JohnVella1968]

As a retired IT guy, (I'm now a teacher) I agree entirely.

[@larrycleeton]

Give us Mark Russinovich!

[@goqsane]

As an aspiring software engineer you should consider stop using Microsoft Operating Systems..., and not because they are inherently bad (and that's just my personal opinion), but because using other Operating Systems challenges you and makes you a better developer. Trust me. I'd used Linux solely since I was 12 years old and now at 3X years old I am so far above anyone else in the same boots as I am, only because I have so much more of a breadth of understanding of how systems work, how CLIs work, how pipes work, how the networking stack works. Everything that you won't ever see in your whole life because you're on Windows. Trust me. It'll pay off.

[@JonathanMcCormack]

I was a VMS System Administrator from 1991 to around 2005. It was a great OS, one VaxCluster I looked after had 4 years of uptime and we only shut it down to replace the UPS.

[@EinChris75]

And now we have scheduled reboots (for hours) each month in order to get the windows cleaned. What a shame.

[@kegginstructure]

I ran a personnel system for the U.S. Navy Reserve based on Itanium/OpenVMS, and we tracked up-time. The Navy has strict rules about patches and reboots (couldn't avoid either). But based on the rules they had, I was able to keep our OpenVMS beasts running with an average of 99.93% uptime (as compared to the amount of time potentially available as uptime.) If it hadn't been for a couple of network device and disk device issues, we might have stayed up even more. But they allowed us to not count the down-time for mandatory patching. That machine ran forever and it ran rings around the reliability of the Windows Server boxes AND the UNIX boxes.

[@AlistairMaxwell77]

@@EinChris75 I got asked to look at P2V'ing a W2k server at work a few years ago , the old girl had been up for about 5000 days. You would be surprised how robust windows was if you left it alone, didn't load it up with dodgy software and drivers and didn't do stupid things to it.

[@kittel-dev]

From the bottom of my heart, I would like to thank the previous generation in software development for their passionate, meticulous, hard and smart work. Your work has enriched my life immeasurably. Your work gave me the opportunity to find a deep passion for modern technology, computers and high-level languages. Much of what you were able to do I never learned to understand the way you did. I feel shame and gratitude. Thank you for everything.

[@JeffRyman69]

When I worked for a group at Oak Ridge National Laboratory, that group bought a MicroVAX. It came without VMS installed. The OS was on cartridge tapes and it had installation instructions. I successfully installed VMS and the compilers (FORTRAN and C) following the instructions, got the accounts set up, and was editing and compiling a test program the next day. I was a research engineer who had lots of experience using and programming on IBM mainframes, a DECSystem 10, and MS-DOS but who had 1) never used VMS before and 2) never been a system administrator before! In those days, anything made by DEC rocked!

[@nakfan]

Impressive… Thanks for sharing. BR, Per (Denmark)

[@kittel-dev]

From the bottom of my heart, I would like to thank the previous generation in software development for their passionate, meticulous, hard and smart work. Your work has enriched my life immeasurably. Your work gave me the opportunity to find a deep passion for modern technology, computers and high-level languages. Much of what you were able to do I never learned to understand the way you did. I feel shame and gratitude. Thank you for everything.

[@JeffRyman69]

Two or three years later we replaced the MicroVAX II with an HP 9000 Model 730 running HP-UX. Much faster but had an irritating X-Windows and other OS bugs from day 1. I spent 4 hours on the first day trying to edit a system file because the X-Windows bug made the cursor appear to be on a line other than where the display said it was.

[@perfectionbox]

In a job interview with MS back around 2000, the interview asked me what do if a program had bugs. After a minute discussing ideas he said "you ship it anyway." Man, my respect for him was like an anvil falling off a tall building and I wondered, how could such a mindset happen?

[@kRySt4LGaMeR]

I love Dave Cutler's opinion of PMs - "doesn't meet the bar", just to be thrown under the bus when customers complain about said bugs.

[@doubleHLabs]

Can't wait for the full interview!

[@zzco]

It's going to be sweet!

[@Syscrush]

I love this grumpy old bastard. No filter, just an intense dedication to doing stuff right.

[@G5STU]

A true great, I feel privileged just watching these clips, keep 'em coming Dave.

[@dilzaaaa]

Thank you so much for these interviews. Just finished 'Showstopper' last week and it's great to see him in the flesh.

[@colinsphoneemail]

I’m really loving these interviews. This is some great content Dave it’s wonderful to see and preserve the history of the Dave’s. Loving it

[@thatjpwing]

I loved working in VAX/VMS when I worked at Digital. I had a Rainbow 100+ hanging off a VAXcluster and it was amazing how efficient the OS was. And version numbers!

[@InconspicuousChap]

Versioning was quite a feature. Too bad it didn't make it to Windows NT. I guess Files-11 was too advanced to be understood by Basic fans (and still is).

[@jej3451]

This guy reminds me so much of the best engineers I've worked with over the years. I think it's a personality type.

[@wtmayhew]

A rite of passage for programmers should be to have to do some significant project on a Teletype. I can still recall the aroma of warm motor windings, ink and oil when I think about it.

[@louishiggins8881]

Remember getting our first VAX in ~1978/79 (in the UK) running VMS 1.6 - it wasn't buggy, but it did lack a lot of features, though 😂. But, we were able to learn the system and start our software port from the IBM mainframe. Really happy times (the whole VAX/VMS ecosystem was sooo elegant, pleasure to work on) 😁.

[@neatodd]

I miss VMS. I wrote a lot of Fortran 77 using it back in the 80's.

[@LTVoyager]

Likewise.

[@jamesbond_007]

These interviews are great -- I have several friends, both ex-DEC and @Microsoft who have worked with him, but to me he's been an unseen mythical figure behind a number of interesting technologies. It's fantastic that you're providing a "voice" for him to provide his oral history of some of the developments he's spearheaded.

[@paultrayers494]

Showstopper! is a fascinating account of the development of WinNT that shows just how driven Dave and his engineers producing the most stable ever version of Windows to that

[@codingbloke]

Nice insights. I used work for Schlumberger many many years ago, huge users of DEC equipment back then, PDP-11, large VAX clusters, and then "microVAX" all over the place. Cut my coding teeth on VMS (Assembler and Fortran) and doing mental stuff with DCL. Unix/DOS always felt a bit alien to me so I was very pleased when I heard the news that Dave Cutler was bringing his expertise to MS. I found WindowsNT internals had a familiar feel.

[@kittel-dev]

From the bottom of my heart, I would like to thank the previous generation in software development for their passionate, meticulous, hard and smart work. Your work has enriched my life immeasurably. Your work gave me the opportunity to find a deep passion for modern technology, computers and high-level languages. Much of what you were able to do I never learned to understand the way you did. I feel shame and gratitude. Thank you for everything.

[@wtmayhew]

I supported VMS and Berkeley UNIX on Vax systems for quite a few years. Off hand, I don’t remember VMS ever crashing due to a software error in the operating system. On the other hand, I do recall kernel panic messages from UNIX and reboots with quite lengthy mandatory disk fsck (we joked that was a misspelling of the actual word intended). That was in the days before journaled file systems.

[@kittel-dev]

From the bottom of my heart, I would like to thank the previous generation in software development for their passionate, meticulous, hard and smart work. Your work has enriched my life immeasurably. Your work gave me the opportunity to find a deep passion for modern technology, computers and high-level languages. Much of what you were able to do I never learned to understand the way you did. I feel shame and gratitude. Thank you for everything.

[@wtmayhew]

@@kittel-dev Thst is very well stated. I had the chance to meet programmers who had worked on vacuum tube machines, IBM and Univac, of the late 1950s and early 1960s. Just dealing with the machinery was a large part of programming back then. One of my friends told me about having to work out the timing of machine instructions so they were interleaved on drum memory so that when one instruction completed that the drum would be in position to read the next instruction. It was a relief when an automatic optimizer, SOAP, was written so that programmers could think more about problem solving than machine quirks. We’ve come so far that we now can tackle problems symbolically at a high level with barely considering the hardware or operating system supporting the work. That is so far from the days when it was necessary to manually code multiple steps just to print a single character on an output device.

[@SwarmerBees]

"ASR 33" For the young folk in the audience, when Cutler mentioned how hard it was to press keys on the "ASR 33" - He was talking about a Teletype machine. These were great for loading programs into early microcomputers that had no storage drives, because you could load programs from paper tape- a kind of really narrow piano roll if you know what a player piano is. Anyway, you could save programs to paper tape then reload later. Some versions had an integrated acoustic coupler attached so you could dial up the central computer at 110 bps. OK, so the post keys that Dave mentioned actually weren't that hard to press at speed if you learned on manual typewriters. Those who learned on electric typewriters tend to exagerate how much force it took to press them. My grandmother could move along at 60wpm with no errors on a manual, and would have found the teletype a little mushy, but requiring less force than a manual typewriter.

[@glasser2819]

thank you uncle Dave for inviting your nemesis Dave C. two titans swaping notes 👏 It's interesting when you two chat about system stress ! This is where live timing bugs come into play - Hardly a way for architects developers to predict these race conditions under load. I've improved NTK performance 15% - Live system is the truth. BRILLIANT WK. I worked on DEC VAX VMS for pharmaceutical Mfg development 👏

[@ashfagg91]

I’m unreasonably excited for the full thing. I read “Showstopper” recently, and have really come to admire Dave. Very interesting guy indeed.

[@PrestoJacobson]

Yeah I'd rather listen to the whole interview than these separated videos.

[@dameanvil]

00:00 🐞 Dave Cutler shipped VMS with zero known bugs, emphasizing the importance of bug-free releases. 00:24 🐛 A contrast is drawn between shipping a system with 20,000 known bugs and striving for a bug-free release. 01:17 🕰️ A last-minute race condition bug in a disc driver was fixed just before shipping, but there were zero known bugs at that time. 01:46 🗓️ Success for the company was tied to selling hardware, necessitating a timely operating system release. 02:12 🏗️ Early Vax Hardware was provided to key partners like Slumber for development and testing. 03:23 📊 Base levels of functionality were planned and achieved before moving on to additional features, ensuring a stable system for users. 04:04 💥 The system was usable and reliable for about a year, enabling development and testing with minimal crashes. 04:30 ⌨️ Improved terminals with CRT displays and responsive keyboards replaced older models, enhancing user experience.

[@LTVoyager]

Who breaks down a 5 minute video? 😂😂

[@wtmayhew]

@@LTVoyager Looking back, that was a lot of ground covered in just five minutes.

[@dannutley931]

In 1988 I was working for Boeing on the Peace Shield program in the Green River Office Park. We had rooms with two or three dozen (!) as-yet-to-be deployed microVAXen. We had three large VAX clusters. Even then you could already do an 8-way symmetric VAX cluster. Ours were 3-way. State transitions, which were very rare, were imperceptible. I worked most often in a server room that had an actual dial lock, like on a safe, on the door. And I had a terrific boss, named Jerry Jennings.

[@xlerb_again_to_music7908]

My doctoral project ended up with 106k lines of python, and no known bugs. How? Every part was prototyped in the small, debugged, integrated, tested etc. building up to a full system. Development was super slow, super tedious and the method only worked as I'm a pedant who does not mind having everything under a microscope (background in military systems where 100% QA is mandatory). If it's wrong, it has to be fixed - or else I would not be able to sleep at night worrying about corner cases and how they could happen for "valid" reasons. Why focus on using tested parts? Because it was so big. Debugging a monster like that would have been impossible; therefore it was assembled slowly out of proven working parts. My approach was based in Waterfall ideas from 1980s with my own refinement - defect forcing. Always bring new/unknown/risky things to the front of a project; as discovering something deeply unhappy at the end leaves no time. So I raked through all the necessary elements and brought things new or untried to the front. The first year was essentially writing and proving libraries / modules which supported the overall application; the last 3 months was assembling the parts together. They fit. It worked. Hooray! :) and I loved the DEC products. Very happy with writing LSI-11 assembler. Had a job doing soak testing of newly built workstations in the early 80's - my LSI-11 had to monitor up to 15 machines for 14 days 24/7; a software/hardware combo that had to be more reliable then newbuilt hardware. Which was being swapped in/out in realtime (whilst the main system kept going). And do all the logging of faults found etc. I was proud of that system - in service testing newbuilds for 14 years with no changes to the code base, reportedly operating for several years without reboot. That was DEC kit for you.

[@joranmulderij]

If you were writing such a large system where correctness was so important, why did you choose to use a dynamically typed language? Just really curious because I would have gone with a language that at least does a small part of correctness proving for you.

[@xlerb_again_to_music7908]

@@joranmulderij Supported and recommended by my Uni, with training courses etc. What language would you have chosen? BTW all my background was with procedural stuff; plus this was not a computer science course - I was simulating electrical system dynamics and wanted something easy to develop in. So python 2.7 it was - ouch. This was c. 2015. Easy was necessary; I'd not done serious coding since um '89 (started back in '67). If I was starting over (no! please, no!) I'd use nim, a compiled language which is C like with python-esc syntax. Plays nice with python; it knows about python data structures. A few weeks back wrote some test code in nim, called it from python and read / printed results in python. Nothing complex, but it was x10 faster than python equivalent. The doctoral work resulted in c. 5 billion tracked EV trips and c. 200TB of output, including all the electrical network responses. A single run on a top-spec PC took 16 hours; I used 4 machines over 6 months to do all the test cases. All about "real-world" dynamics of EV charging loads being modulated by V2G, DR, FR and all the good stuff the smart grid guys get excited about. Except some are not practical (which? why? what can be done to fix it? what happens if there is a comms fail? etc) - so everything was plastic / set in external .csv files which I could ring the changes with. Static languages would have been faster - but one of the speedups was to use _slots_ aka static typing within python classes. I leveraged the C routines I could directly call by python - got it so my bespoke, accurate EV SOC model could spit out 10,000 solutions per thread per second; python it might have been but it was tightly coded calls to C. And found the memory leaks python / Windows have together. Got to the point I was internally batching stuff to get the python optimiser to crank faster. And using 6 of my 8 threads (called from DOS, Windows gives each python instance a new thread). It was great when I'd finished (like banging your head against a wall)

[@rhymereason3449]

Before the craze of "Rapid development" virtually every large scale software project was developed that way... but it's not a guarantee there will be no bugs...

[@xlerb_again_to_music7908]

@@rhymereason3449 Absolutely! I was sweating bullets hoping that this colossus would work OK - so much riding on it! The project was large-scale, accurate simulation of EVs and charging loads. There was one outcome /scenario which looked wrong - I pulled that section apart and crawled over it for a week. Could not find anything. This was to do with the way the simulation had EVs performing much more destination charging in winter. EV range drops when cold, so use of chargers lifts. Cold effects were in the simulation; I was getting 62% winter uplift in daily demand. Two years later real-world results from 180,000 monitored EVs showed the same effect - a 61% uplift. 1% out; oh my. I was more concerned that the whole thing was fantasizing rubbish results... _(cough ...AI..)_ Anyhows charge point use in winter can be x8 up vs. summer, peaking to x20 worst case. Why does that matter? Well, how many charge points do malls, attractions, cities need? Why fit an excess over summers need, when winter is so short? Tip: always charge your EV at home when you can; there may be no free destination chargers when you arrive somewhere. It's unlikely c. 2035 there'll be enough free at malls etc. As 9 or 10 months of the year they get no use, so they don't pay for themselves :( who's going to fork out the $$$ to install known unprofitable kit? Alas, EV stranding rates might rise

[@philsbbs]

I worked in IT for a huge retailer worldwide and they used VMS for there inhouse software with orcale databases

[@timdion9527]

The first system I ever programmed on was a VAX-11/730. Twenty students all worked at the same time on VT100 terminals.

[@fadichamieh]

Amazing interviews, Dave. Keep' em coming!

[@kittel-dev]

Another invisible Hero of my childhood. Thank you sir.

[@rdubb77]

One of the two Gods of OSes, the other of course being Ken Thompson. Will be interesting to hear his take on the evolution of OSes since 1970.

[@Akkbar21]

I ran windows 2000 forever. What a gem.

[@ssmith5048]

Don´t forget John Backus....

[@rdubb77]

@@ssmith5048 I believe Backus was more involved in the creation of Fortran and BNF form, not OSes per se. Did he work on an OS?

[@nathansmith3401]

You can sure see Dave C's mind is laser sharp.

[@readytext]

These interviews are great. Even as a relatively low-level software engineer, the culture shock of joining Microsoft after Digital was real!

[@nufosmatic]

2:46 - Amongst the tools Schlumberger put "down hole" was a particle accelerator and return receiver which was a yard long and 1-1/2" in diameter which illuminated the surrounding rock and read the return for petroleum content. The accelerator was powered and controlled through a single coax cable. That cable connected to the sensor head and eventually connected to the VAX...

[@Ausare911]

The company I work for just retired its vax vms system a year ago. It just worked.

[@rw-xf4cb]

VT100 the standard of all terminals to come (VT220 green screen was my favorite) ESC[2J clear screen, from memory good times!

[@burnedoils]

good work dave, ive watched this one many times cuz its so true, it never ends, everytime u think u got it, bang here more bugs

[@iAPX432]

I had a program that ran 3 and half year, to switch communication packets on Turbo-C and MS/DOS, 24/7 and served a lot of users and systems simultaneously. I am absolutely certain there was bugs, because the stats, but some architecture choices helped to mitigate that to make it very solid.

[@Mikey-gs1dx]

These are good. I love watching these.

[@ricksharpe6895]

Nice to see he’s running RSX on his PiDP-11!

[@kaatlev]

really enjoying these interviews

[@dsuess]

Omg these are great!! Thank you, Dave & Dave! I love this kind of living history. Can you get one with the creators or MS C-compiler and Visual Basic?

[@jonathanjacobs6964]

PDS 7.1...!!! What an awesome compiler. The absolute pinnacle of classic Basic.

[@InconspicuousChap]

Basic is a personal sacred cow of mr Gates. He still loves this language most of all, because he cannot comprehend any other language. So this area will always be sponsored within the company, whether it's profitable or not. And when mr. Gates (like king Midas) touches any other language, it turns into Basic. E.g. he had tried to convert Java into Basic in late 1990-s, but get kicked by Sun, so he spawned his own hybrid of Java and Basic and called it C#. Then he took Haskell and injected Basic into it, producing F#. So you should request interview with the big boss.

[@skak3000]

Thanks for making this video. I want to hear more story from the old days of software dev.

[@thecalvinprice]

When I inherited the codebase I'm currently working on, there were just try-catch blocks everywhere as the only way bugs were dealt with. Worse still, the Catch blocks always just threw another, more ambiguous exception without any useful data, and then nothing else. I still get mad thinking about it, even though I've dealt with all of them now.

[@rodhoutx]

Ha, our team had inherited quite a few applications from other departments (non-technical depts. or shadow IT) where everything was wrapped in try-catch, or worse the entire page of code was a huge try-catch. The catch block was..... just ignore it/eat the error with no logging whatsoever! Not sure if they added them because the code was generating so many errors or because they had been taught everything has to have a try-catch block so that the user never knows there was an error (even if it means the system failed to actually perform the transaction)!

[@dmzar]

Yep. I was often asked on weekly status calls: "Any bugs?" Sometimes I could say "no known bugs!" My engineers knew what that meant, but the sales guy didn't. Sigh...

[@anon_y_mousse]

Wow. I wonder when the last time was that a software company shipped a product with *ZERO* *KNOWN* bugs. That might have been the last time, when VMS shipped.

[@captsorghum]

We used to ship with zero known bugs all the time. But we didn't have an SQA department. Customers usually found a few though.

[@anon_y_mousse]

@@captsorghum Was that back in the 90's?

[@captsorghum]

@@anon_y_mousse 80s, early 90s

[@cheako91155]

How many bugs were discovered later, not counting bugs that were introduced when adding features?

[@johnnypeck]

This is so amazing. Thank you for this.

[@havelsand]

Great interview! I like this "Debugging" coffee pot at the right 👍

[@GHHodges]

Nice catch on the mug!

[@TomCee53]

I remember finding a real compiler bug early in my career. There was bug in PL/1 that converted a specific number to the wrong binary value. I don’t remember the actual number but the scenario was if you wrote X=12345.67 Print X You got 6438.5 I did file a bug report with IBM, but I moved to another assignment before I ever heard the result.

[@LTVoyager]

Ah, yes, I remember well the VT1xx family of terminals. This interview is a walk down memory lane. I still remember my visits to The Mill back in the early 80s.

[@franksimon6617]

I was a computer consultant on many IBM 370 models. A friend of mine was the system programmer. He told me that the OS operating system always had the same number of bugs. IBM would fix some and break as many (at least) as they fixed.

[@VolkerBaier]

Thanks a lot for this interesting interview!

[@ebridgewater]

Please boost the volume for the final video.

[@larrycleeton]

You can't emphasize enough that in the past (and even the present) a computer company sells hardware. The software was/is an essential accessory.

[@RichardSkokowski]

I once had a programmer swear that the software was absolutely perfect and bug-free. "How can you be sure?" -- "The compiler says 'No Errors Found'." (Face palm)

[@Akkbar21]

Back when you had to go physical gold, you couldn’t ship the messes you see as v1.0 these days. Games for example: wing commander 3 and 4 shipped with zero bugs I ever saw. Nowadays games ship outright broken. Better company standards and devs with better commitment to their projects. Now it’s MORE FOR LESS! 😰

[@Moon-Raven]

Mr. Cutler has a lot of nice looking things in the background, does anyone know what they are?

[@RichardSkokowski]

@4:32 the CC is wrong: Dave says "ASR 33" which is the old chunka-chunka teletypewriter (the OG "TTY"). Those keys must have needed 1/4 pound of force. I could touch type pretty quickly, but never on a 33.

[@BL-ob9fn]

I didn't get a notification for these last two videos, only discovered that they were posted by a coincidence. 🤔 Were they posted with notification disabled or something? Or is KZbin hiding them because it doesn't like too many notifications from one channel within a short span of time? If so, maybe splitting videos up into shorter segments isn't a good idea...

[@ronm6585]

Thanks for sharing. 👍🏻😊

[@starra0787]

added the 21st to my watch calendar!

[@mercster]

I used VMS on a VAX in community college (although by that time I was an early adopter of Linux and much preferred that system)... could never get the hang of VMS's idiosyncrasies, but admire its reputation and pedigree. Although I will say... no non-trivial software has zero bugs. ;-) But releasing with no known bugs in the hopper is pretty cool.

[@mercster]

"Cutler is known for his disdain for Unix. Said one team member who worked with Cutler: Unix is like Cutler's lifelong foe. It's like his Moriarty. He thinks Unix is a junk operating program designed by a committee of PhDs. There's never been one mind behind the whole thing, and it shows. So he's always been out to get Unix." 😡

[@InconspicuousChap]

@@mercsterDave is not alone. Many engineers hated Unix in 1990-s for being a junk OS. But by now the choice of operating systems has been reduced to 2: Unix family and David Cutler's OS. Unix has a nightmare architecture covered by attractive instruments and illisional simplicity. Windows has a properly designed VMS core built by Dave's team and sloppy application level built by accountants. Only Dave could still discuss these matters - nobody else is capable of building operating systems these days. It's a forgotten handicraft.

[@garyclouse4164]

We had mostly Hazeltinr 1500s and a few vt220s. I think I stills have my old VAX assembler textbook

[@HansOtten]

VAX/VMS I remember the days of early VAX 750 versions, it did crash in internal beta's ... Later versions, never seen such a stable OS with uptime of years in clusters with major OS upgrades and no downtine/reboots!

[@tvdan1043]

My college had a VAX 8350. I'm not sure anybody really knew how to use it other than a couple of professors who constantly bogged it down doing "look how smart I am" stuff. By the time I graduated it had basically been relegated to email server.

[@aaron___6014]

How do you fix let allow find 200 bugs a day? Sounds way too high.

[@Zuklaak]

It's not a bug, it's a feature!

[@Dereks06]

Where do I go for the full interview? Can't find it anywhere.

[@richardwernst]

Interesting, VMS, eh? Working for DEC, I assume? I worked at a DEC (early DECSystem10 4 KLs!), then VAXs at Copley Computer Services (later Copley Information Services( CCS/CIS) for many years, lots of fun. We had a lot of I/O, not a lot of number crunching so moving from the super fast KL/DecSystem10 to VAXs was a bear, and only had to do that because the DEC Jupiter project which was to be super fast/faster than the 4 KL cpus we had, apparently didn't live up to the specs expected and was canceled so we had to go with VAX.

[@hye181]

dave cutler was the primary architect of VMS

[@DankyMankey]

Semantic error, am I right?

[@RaminOhebshalom]

Thank You

[@jchavins]

I'd be happy just to have a local network that was actually consistent so that I could read the desktops and laptops consistently. This isn't happening. One day they can read each other and then the next day they can't. Next week it will be different....

[@barrywalker1971]

I would have killed to work with Cutler in his heyday. Dude is a legend.

[@user-tc2ky6fg2o]

I had a brainwave while I was watching this video: bug is a habit.

[@jensBendig]

Windows NT was the only Windows, I could accept. Until today.

[@jimfoye1055]

I can't give someone software knowing it has "known bugs". I just can't ever bring myself to do it.

[@leetucker9938]

I like your lamp

[@FlyingPhilUK]

To be honest, I often ship Zero bug software... and Zero bug h/w too.... 😅

[@mathiaskeulen]

You will always have bugs, but if you have bugs in your happy flow then there are pretty much 0 excuses imo.

[@wojtekdudek2374]

Look. It's an interview, and the guy is giving his 100%, it is just rude to randomly say things that don't introduce anything but just annoy the person like: "ok", "right", etc. when the person is talking. And it interrupts the whole guy's story. Just don't do it.

[@Solarbonite]

Tbh when you see behind the scenes of interviews I have seen interviewers do that. It normally is cut out, but it's hard to get someone to talk to a camera for 45 minutes without having a conversation.

[@stephenfwadsworth9565]

As always thanks for sharing. Bugs are like self generating code, expected, but not predictable always. Had this in a website my brother and I worked on recently, but experienced way back in the 8 bit days on the c64. Bugs, fix current problems, but like any system, expect new bugs. The perfect system, has no unclosed loops. Funny how when you translate this out to Project Management/Structured Systems closed loops can be just as dangerous, creating information silo's. Why just because a computer is binary do we think it will yield only binary results, when a human programs it? Why we have PI, but do like the Mesopotamian system for programming. :) Ran(). ? ?? Pi, E=Mc2 :)

[@max.r4]

Big fan of you. How are you today :)

[@ifstatementifstatement2704]

Tell that to video game developers today

[@kainx99]

I totally agree, we need people that just do what needs to be done. I do not care if you have a nice title, or are just called Code Janitor... Succesfull software is done by people who act. However, problem in modern dev environment, people with big title take decisions, and if they are not _doers_, they usually stupid ass decisions, and we the actual doers are left with the messes...

[@TheSkunkyMonk]

KZbin may of blocked my ad blocker but I can still download and watch, f u Google.

[@joe3276865536]

KZbin premium is $14 per month (US) and helps pay for servers and staff hosting content like this. Nothing in life is free.

[@TheSkunkyMonk]

@@joe3276865536 Erm Google has plenty money, they want us wasting our time and engaging on there platforms it all adds to the share price. Oh and the Air we breath, the water and fish in the river, the joy of making someone smile, half the aplications we have today wouldn't be here without the open source community, plenty of free things in life you just need to open your eyes and see them. But it is easier to keep em shut and just pay the piper.

[@Vectorh]

I guess PMs have always been this way eh?

[@adokapo]

It sounds like working at mine. 😅

[@custardtart1312]

Zero known bugs is I’m afraid a meaningless statement. There’s a continuum here, the harder you look, the more you find. One could influence the point you decide there are zero known bugs by simply not looking so hard. More properly, one could say there were zero known bugs based on the testing resource we were able to invest.

[@InconspicuousChap]

That's an excellent example of today's mindset. Everyone is lying to each other and to themselves. And that's why it's important to listen to people from older generations who haven't adopted it.

[@TommyParnell]

I ship with no known bugs, by not looking for them 🤣

[@channelzero2252]

No known bugs? Obviously not the philosophy of whoever OK'ed Windows Vista. I'm not having a go at the coders, I'm having a go at management for releasing a product that clearly wasn't ready. Looking forward to the full interview!

[@InconspicuousChap]

There is an interview about Vista on this channel.

[@astrahcat1212]

You fix 200, then 200 more crop up. Yes, exactly, 20% making the thing, 80% of your time debugging.

[@robertcromwell9736]

Zero Bugs.... LMAO there is no OS like that. I worked software development and testing for quite a few years.

[@liamhotspur9182]

Build bridges like the software people do and we have hundreds people dying every day...

[@dzalejandro]

we need this people at ms again, win 11 and 10 are 💩

[@gawelek1980]

🌀Windows 14

[@lucidmoses]

Microsoft has never shipped a product with a bug in it. Well.... In a "it's not a bug it's an undocumented feature" kind of way. :p

[@brianmorgan9811]

VMS had bugs! I remember writing a Digital VT340 software oscilliscope program using Digital Regis graphics on a VT340 graphics terminal and the system would fall over if the keys were pressed in quick succession to move the oscilliscope cursors...The O/S couldn't cope...

[@playitlouder451]

I was using DECplan on a VMS system, and I remember I executed some operation and the screen went dead. OK - I crashed the app. But then the whole office of 70 developers was out in the aisles wondering why their screens also stopped. Sysop says give me 20 min to reboot and then we go back to work. I try my edit again ...... So now I knew how to give the office a 20 min coffee break any time things were getting a bit stressed