Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

1972 Unix V2 "Beta" Resurrected

tuhs.org

Post reply on HN
Print view View original

141–150 of 153 posts

Re: 1972 Unix V2 "Beta" Resurrected

#141

Post by pjmlp »

kragen wrote: post #131
Cthulhu_ wrote: post #53

Earlier quoted context omitted.

A newer joke is that Ken Thompson (along with Rob Pike and Robert Griesemer) designed Go while waiting for C / C++ to compile.

Not C/C++. Specifically C++. C compiles pretty fast. And it's not really a joke, though obviously it wasn't a single build they were waiting for.

Back in 2000, our builds, C modules to be consumed by Tcl, Apache and IIS modules, DB drivers took about one hour per OS for a full platform release.

OS being Windows NT/2000, Aix, HP-UX, Solaris, Red-Hat Linux.

Re: 1972 Unix V2 "Beta" Resurrected

#142

Post by kragen »

pjmlp wrote: post #141
kragen wrote: post #131

Earlier quoted context omitted.

Not C/C++. Specifically C++. C compiles pretty fast. And it's not really a joke, though obviously it wasn't a single build they were waiting for.

Back in 2000, our builds, C modules to be consumed by Tcl, Apache and IIS modules, DB drivers took about one hour per OS for a full platform release. OS being Windows NT/2000, Aix, HP-UX, Solaris, Red-Hat Linux.

Yeah, that's similar to my experience. C++ projects commonly took a week to build in the 90s.

Re: 1972 Unix V2 "Beta" Resurrected

#143

Post by pjmlp »

kragen wrote: post #142
pjmlp wrote: post #141

Earlier quoted context omitted.

Back in 2000, our builds, C modules to be consumed by Tcl, Apache and IIS modules, DB drivers took about one hour per OS for a full platform release. OS being Windows NT/2000, Aix, HP-UX, Solaris, Red-Hat Linux.

Yeah, that's similar to my experience. C++ projects commonly took a week to build in the 90s.

I can assure that wasn't the case in our desktop software for Windows, and that already feels like trolling.

Not even Nokia NetAct took half as much, given its complexity of distributed CORBA services written in C++, nor the CERN TDAQ/HLT builds I was responsible for.

Re: 1972 Unix V2 "Beta" Resurrected

#144

Post by BrendanEich »

wbl wrote: post #122
mr_toad wrote: post #99

Earlier quoted context omitted.

I’ve heard that Torvalds build Git in 5 (or 10) days and that Brendan Eich created JavaScript in 10 days. Maybe the average programmer is less efficient, but the distribution is probably heavily skewed these days.

Brendan Eich would say "10 days" whenever one of the big warts from the that are unfixable came up.

The Mocha prototype was better than what I did after the ten days. The biggest wart, == implicitly converting operand types, came after the ten days and was me being an idiot by agreeing with early inside-Netscape adopter requests for slop.

Ryan Dahl gave a speech decades later saying "don't do them" when you are tempted to add little features that might be "cute": https://youtu.be/M3BM9TB-8yA?t=900.

Unfixability is a property of the Web and applies to CSS and HTML as well as JS.

Re: 1972 Unix V2 "Beta" Resurrected

#145

Post by kragen »

pjmlp wrote: post #143
kragen wrote: post #142

Earlier quoted context omitted.

Yeah, that's similar to my experience. C++ projects commonly took a week to build in the 90s.

I can assure that wasn't the case in our desktop software for Windows, and that already feels like trolling. Not even Nokia NetAct took half as much, given its complexity of distributed CORBA services written in C++, nor the CERN TDAQ/HLT builds I was responsible for.

You were probably working on more competently designed software than I was, but sure, desktop software for Windows usually didn't take that long.

Re: 1972 Unix V2 "Beta" Resurrected

#146

Post by api »

kragen wrote: post #140
api wrote: post #61

Earlier quoted context omitted.

One of the many things I dislike about the SaaS era is that this will never happen. Nobody in 2075 will boot up an old version of Notion or Figma for research or nostalgia. Like the culture produced and consumed on social media and many other manifestations of Internet culture it is perfectly ephemeral and disposable. No history, no future. SaaS is not just closed but often effectively tied to a literal single instal…

A lot of software in 01972 was also effectively tied to a literal single installation. Most of the software people ran under Unix at the time was only present on one of the ten Unix installations and has consequently been lost. The shrink-wrapped mass-distribution software epoch was still ten years in the future.

That's why software archaeology from that era is hard, but it's still a lot easier than today's cloud software. It's a lot more complex, much more of a moving target, and is more interdependent with other services all of which would have to be either emulated or restored.

My other point was that we've gone back to the mainframe era. The PC revolution has mostly been abandoned.

Re: 1972 Unix V2 "Beta" Resurrected

#147

Post by EfficientDude »

kragen wrote: post #66
ThePowerOfFuet wrote: post #59

Earlier quoted context omitted.

Now we know where SMTP got it, I guess.

That's possible but unlikely. MTP as defined by Suzanne Sluizer and Jon Postel in RFC 772 in September 01980 https://datatracker.ietf.org/doc/html/rfc772 seems to have been where SMTP got that convention for ending the message: > ...and considers all succeeding lines to be the message text. It is terminated by a line containing only a period, upon which a 250 completion reply is returned. But in 01980 Unix had only b…

Why the leading zero in the year dates but no leading zero for the RFC numbers, word length, or bits?

EDIT: I checked your web site and you DON'T use leading zeros in front of dates there. Just trying something out? Enlighten me please...

Re: 1972 Unix V2 "Beta" Resurrected

#148

Post by EfficientDude »

api wrote: post #61
typeofhuman wrote: post #15

Software archeology

One of the many things I dislike about the SaaS era is that this will never happen. Nobody in 2075 will boot up an old version of Notion or Figma for research or nostalgia. Like the culture produced and consumed on social media and many other manifestations of Internet culture it is perfectly ephemeral and disposable. No history, no future. SaaS is not just closed but often effectively tied to a literal single instal…

Nobody will care what Notion or Figma are in the future. I don't know what they are, nor do I care to find out. Saas simply doesn't exist for many people, it's not a software model that all people are comfortable with.

Re: 1972 Unix V2 "Beta" Resurrected

#149

Post by kragen »

api wrote: post #146
kragen wrote: post #140

Earlier quoted context omitted.

A lot of software in 01972 was also effectively tied to a literal single installation. Most of the software people ran under Unix at the time was only present on one of the ten Unix installations and has consequently been lost. The shrink-wrapped mass-distribution software epoch was still ten years in the future.

That's why software archaeology from that era is hard, but it's still a lot easier than today's cloud software. It's a lot more complex, much more of a moving target, and is more interdependent with other services all of which would have to be either emulated or restored. My other point was that we've gone back to the mainframe era. The PC revolution has mostly been abandoned.

Everybody I see on the bus has a personal computer in their hand, and a lot of them also have an additional personal computer in each ear. USB-C chargers typically each contain a personal computer to decide what voltage to output. All this doesn't necessarily result in enhanced user autonomy and agency, though; I wrote this essay about the disturbing trend in the late 90s: https://www.gnu.org/philosophy/kragen-software.html

Re: 1972 Unix V2 "Beta" Resurrected

#150

Post by BrendanEich »

mr_toad wrote: post #99
m4r1k wrote: post #47

I once saw a talk from Brian Kernighan who made a joke about how in three weeks Ken Thompson wrote a text editor, the B compiler, and the skeleton for managing input/output files, which turned out to be UNIX. The joke was that nowadays we're a bit less efficient :-D

I’ve heard that Torvalds build Git in 5 (or 10) days and that Brendan Eich created JavaScript in 10 days. Maybe the average programmer is less efficient, but the distribution is probably heavily skewed these days.

My understanding (direct from Graydon, some 18 years ago) was that Linus admired Monotone except for its slowness. Graydon liked making things correct then fast, but was on a break in Köln when Linus (provoked by loss of BitKeeper) emailed on Friday endorsing Monotone's design and asking for speed. Graydon said "let me buy a laptop and I'll start on Monday".

Monday rolled around and Linus said "too late, I wrote it already" (it => git).

Post reply on HN