Been wanting to wade into this for a few days but needed to think about how. I think that we're all aware that RMS has atrocious personal habits. But I don't think that this mailing list is the place to discuss them unless it's somehow in the context of UNIX. Many seem to excuse RMS's revisionist view of the history of technology on the grounds that RMS claims that his memory isn't very good. I think that if he knows that he doesn't remember things then he should refrain from talking about them as if he does. As others have said, I don't conflate coding prowess with the ability to design. I've had many an argument with John Gilmore (one of the people who doesn't mind footing the cleaning and repair bill after allowing RMS to stay at his place) where he begins with "When I wrote GNU tar..." I've always responded by saying that writing tar is no big deal; the specification was the hard part. One place where I completely disagree with RMS that I think is in context for this list is his claim that Linux should be called GNU/Linux. I've written tons of software in my life, and I don't preface the name of each one with the parts list. Even if one believed that such an attribution scheme made sense, I would claim that it should be called internet/Linux. I would argue that Linux would not have happened without the internet making it possible for folks around the world to participate. And I think that there's a good chance that the tools would have been created anyway. Of course, I acknowledge that the GNU tools have been ported to Linux. Big deal. I haven't seen RMS arguing for GNU/Windows now that Microsoft has seen the light. Like many of you, Linux is not where I first started using GNU tools; I started using them on my Sun machines after Sun started charging extra for the compiler and included a licensing system that was broken and often interfered with getting work done. Jon
I think the RMS stuff should go away. It's not because I love the guy, I don't. It's because we have people like Ken and Rob and other heavy hitters and my hunch is they have little patience for this sort of thing (they might correct me if I'm wrong). I'd love to call out RMS on his BS but this isn't the place. This is the place for people who actually did real work on Unix to share those stories. Or so I think, it's up to Warren, not me.
[-- Attachment #1: Type: text/plain, Size: 763 bytes --] On Sun, Jan 6, 2019, 7:43 PM Jon Steinhart <jon@fourwinds.com wrote <snip> > I would argue that Linux would not have happened without the internet > making it possible for folks around the world to participate. And I think > that there's a good chance that the tools would have been created anyway. > That's more or less how I look at it. Back in the day there was comp.sources.unix for example. In Unix itself, there was /usr/ where tools developed by users other than the core developers belonged, and there was /usr/ucb/ where they put stuff from Berkeley. The culture surrounding Unix has always seemed to encourage outside participation, going back to the lenient licensing of Research Unix, and even before that, when it just existed at Murray Hill. > [-- Attachment #2: Type: text/html, Size: 1527 bytes --]
On 2019-01-06 6:41 PM, Jon Steinhart wrote:
> ...
> As others have said, I don't conflate coding prowess with the ability to
> design. I've had many an argument with John Gilmore (one of the people
> who doesn't mind footing the cleaning and repair bill after allowing RMS
> to stay at his place) where he begins with "When I wrote GNU tar..." I've
> always responded by saying that writing tar is no big deal; the specification
> was the hard part.
>
Hear, hear. I'd aver this is very much the case in any typical
software-related day job, and _definitely_ mine.
--Toby
On Sun, 6 Jan 2019, A. P. Garcia wrote:
> On Sun, Jan 6, 2019, 7:43 PM Jon Steinhart <jon@fourwinds.com wrote
>
> <snip>
>
>>
> I would argue that Linux would not have happened without the internet
>> making it possible for folks around the world to participate. And I think
>> that there's a good chance that the tools would have been created anyway.
>>
>
> That's more or less how I look at it. Back in the day there was
> comp.sources.unix for example. In Unix itself, there was /usr/ where tools
> developed by users other than the core developers belonged, and there was
> /usr/ucb/ where they put stuff from Berkeley. The culture surrounding Unix
> has always seemed to encourage outside participation, going back to the
> lenient licensing of Research Unix, and even before that, when it just
> existed at Murray Hill.
>
>>
>
If not for GNU, Unix would still have been cloned. Net/2 happened in
parallel, did it not?
-uso.
On Sun, Jan 06, 2019 at 08:38:45PM -0500, Toby Thain wrote:
> On 2019-01-06 6:41 PM, Jon Steinhart wrote:
> > ...
> > As others have said, I don't conflate coding prowess with the ability to
> > design. I've had many an argument with John Gilmore (one of the people
> > who doesn't mind footing the cleaning and repair bill after allowing RMS
> > to stay at his place) where he begins with "When I wrote GNU tar..." I've
> > always responded by saying that writing tar is no big deal; the specification
> > was the hard part.
> >
>
> Hear, hear. I'd aver this is very much the case in any typical
> software-related day job, and _definitely_ mine.
Yep. The spec is hard, the code is easy. That is a pattern.
I've been the guy behind decent sized projects, BitKeeper is 2673371
lines of code. Getting to a spec was hard, writing the code was easy.
We were a tiny distributed team of about 10 engineers, the hard part was
agreeing on a design. Which we did by getting on the phone and talking
about what we talked about yesterday. We passed the idea between people
and when we could do the pass back and forth and nothing had changed
from the previous pass, we had a design. Coding that was just typing.
It's very similar to what Udi Manber told me as an under grad, he said
writing papers is easy. Not for me. But I came to understand that
papers are two things: a big base of knowledge and an outline. If you
have those two then the paper is just typing. He was right.
[-- Attachment #1: Type: text/plain, Size: 1990 bytes --] On Sun, Jan 6, 2019, 9:12 PM Larry McVoy <lm@mcvoy.com wrote: > On Sun, Jan 06, 2019 at 08:38:45PM -0500, Toby Thain wrote: > > On 2019-01-06 6:41 PM, Jon Steinhart wrote: > > > ... > > > As others have said, I don't conflate coding prowess with the ability > to > > > design. I've had many an argument with John Gilmore (one of the people > > > who doesn't mind footing the cleaning and repair bill after allowing > RMS > > > to stay at his place) where he begins with "When I wrote GNU tar..." > I've > > > always responded by saying that writing tar is no big deal; the > specification > > > was the hard part. > > > > > > > Hear, hear. I'd aver this is very much the case in any typical > > software-related day job, and _definitely_ mine. > > Yep. The spec is hard, the code is easy. That is a pattern. > > I've been the guy behind decent sized projects, BitKeeper is 2673371 > lines of code. Getting to a spec was hard, writing the code was easy. > We were a tiny distributed team of about 10 engineers, the hard part was > agreeing on a design. Which we did by getting on the phone and talking > about what we talked about yesterday. We passed the idea between people > and when we could do the pass back and forth and nothing had changed > from the previous pass, we had a design. Coding that was just typing. > > It's very similar to what Udi Manber told me as an under grad, he said > writing papers is easy. Not for me. But I came to understand that > papers are two things: a big base of knowledge and an outline. If you > have those two then the paper is just typing. He was right. > Coming back full circle, maybe the person with the right viewpoint was actually RMS. It's not about the technology. It's not about the code or the spec. Maybe it really is about the freedoms that he talks about. Maybe if bit keeper were FOSS from the start, the world would be using that instead of git. But where would be the value proposition in that? It's a tough question. > [-- Attachment #2: Type: text/html, Size: 2652 bytes --]
[-- Attachment #1: Type: text/plain, Size: 1586 bytes --] On Sun, Jan 6, 2019, 7:06 PM Steve Nickolas <usotsuki@buric.co wrote: > On Sun, 6 Jan 2019, A. P. Garcia wrote: > > > On Sun, Jan 6, 2019, 7:43 PM Jon Steinhart <jon@fourwinds.com wrote > > > > <snip> > > > >> > > I would argue that Linux would not have happened without the internet > >> making it possible for folks around the world to participate. And I > think > >> that there's a good chance that the tools would have been created > anyway. > >> > > > > That's more or less how I look at it. Back in the day there was > > comp.sources.unix for example. In Unix itself, there was /usr/ where > tools > > developed by users other than the core developers belonged, and there was > > /usr/ucb/ where they put stuff from Berkeley. The culture surrounding > Unix > > has always seemed to encourage outside participation, going back to the > > lenient licensing of Research Unix, and even before that, when it just > > existed at Murray Hill. > > > >> > > > > If not for GNU, Unix would still have been cloned. Net/2 happened in > parallel, did it not? > Berkeley actively rewrote most of unix yes. Net/1 was released about the same time GNU was getting started. Net/2 and later 4.4 BSD continued this trend, where 4.4 was finally a complete system. BSD386 only lagged Linux by about a year and had much stronger networking support, but supported fewer obscure devices than linux... Warner Ps I know this glosses over a lot, and isn't intended to be pedantic as to who got where first. Only they were about the same time... and I'm especially glossing over the AT&T suits, etc. > [-- Attachment #2: Type: text/html, Size: 2389 bytes --]
[-- Attachment #1: Type: text/plain, Size: 967 bytes --] On Sun, Jan 6, 2019, 9:39 PM Warner Losh <imp@bsdimp.com wrote: > > > On Sun, Jan 6, 2019, 7:06 PM Steve Nickolas <usotsuki@buric.co wrote: > >> On Sun, 6 Jan 2019, A. P. Garcia wrote: >> >> If not for GNU, Unix would still have been cloned. Net/2 happened in >> parallel, did it not? >> > > Berkeley actively rewrote most of unix yes. Net/1 was released about the > same time GNU was getting started. Net/2 and later 4.4 BSD continued this > trend, where 4.4 was finally a complete system. BSD386 only lagged Linux by > about a year and had much stronger networking support, but supported fewer > obscure devices than linux... > > Warner > > Ps I know this glosses over a lot, and isn't intended to be pedantic as to > who got where first. Only they were about the same time... and I'm > especially glossing over the AT&T suits, etc. > It's really hard to say. How would you compile it? Clang didn't come along until 2007. The Amsterdam Compiler Kit, perhaps? > [-- Attachment #2: Type: text/html, Size: 2040 bytes --]
[-- Attachment #1: Type: text/plain, Size: 1892 bytes --] On Sun, Jan 6, 2019, 7:59 PM A. P. Garcia <a.phillip.garcia@gmail.com wrote: > > > On Sun, Jan 6, 2019, 9:39 PM Warner Losh <imp@bsdimp.com wrote: > >> >> >> On Sun, Jan 6, 2019, 7:06 PM Steve Nickolas <usotsuki@buric.co wrote: >> >>> On Sun, 6 Jan 2019, A. P. Garcia wrote: >>> >>> If not for GNU, Unix would still have been cloned. Net/2 happened in >>> parallel, did it not? >>> >> >> Berkeley actively rewrote most of unix yes. Net/1 was released about the >> same time GNU was getting started. Net/2 and later 4.4 BSD continued this >> trend, where 4.4 was finally a complete system. BSD386 only lagged Linux by >> about a year and had much stronger networking support, but supported fewer >> obscure devices than linux... >> >> Warner >> >> Ps I know this glosses over a lot, and isn't intended to be pedantic as >> to who got where first. Only they were about the same time... and I'm >> especially glossing over the AT&T suits, etc. >> > > It's really hard to say. How would you compile it? Clang didn't come along > until 2007. The Amsterdam Compiler Kit, perhaps? > The portable c compiler PCC was used to bootstrap a lot of this. It kinda sucked, but was decent enough. Early unix vendors used it on a variety of platforms. Here different universities produced different back ends. But there was no central clearing house. Gcc was a bit innovative in that it provided that, which allowed people to cooperate enough to make it better than PCC, at first. Then better or comparable to vendor compilers. Competition with gcc in large measure drove Sun to unbundle its compilers so there was a revenue stream that could be pointed at technology improvements. Somewhere between 4.3 and 4.4 BSD started using gcc over pcc since it was easier to distribute. The gnu project was important, but not because it rewrote the kernel. It provided the enabling compilers for that... Warner > [-- Attachment #2: Type: text/html, Size: 3518 bytes --]
On 2019-01-06 9:59 PM, A. P. Garcia wrote: > > > On Sun, Jan 6, 2019, 9:39 PM Warner Losh <imp@bsdimp.com > <mailto:imp@bsdimp.com> wrote: > > > > On Sun, Jan 6, 2019, 7:06 PM Steve Nickolas <usotsuki@buric.co > <mailto:usotsuki@buric.co> wrote: > > On Sun, 6 Jan 2019, A. P. Garcia wrote: > > If not for GNU, Unix would still have been cloned. Net/2 > happened in > parallel, did it not? > > > Berkeley actively rewrote most of unix yes. Net/1 was released about > the same time GNU was getting started. Net/2 and later 4.4 BSD > continued this trend, where 4.4 was finally a complete system. > BSD386 only lagged Linux by about a year and had much stronger > networking support, but supported fewer obscure devices than linux... > > Warner > > Ps I know this glosses over a lot, and isn't intended to be pedantic > as to who got where first. Only they were about the same time... and > I'm especially glossing over the AT&T suits, etc. > > > It's really hard to say. How would you compile it? Clang didn't come > along until 2007. The Amsterdam Compiler Kit, perhaps? > If you're asking about non-gcc ANSI C compilers? There were dozens; sometimes several per platform. One random example was lcc* (1994). And of course the vendor compilers at which gcc specifically took aim. There were also quite a number of non-GNU C++ compilers. --Toby * - https://en.wikipedia.org/wiki/LCC_(compiler)
On Sun, Jan 6, 2019 at 9:01 PM A. P. Garcia <a.phillip.garcia@gmail.com> wrote: > > > > On Sun, Jan 6, 2019, 9:39 PM Warner Losh <imp@bsdimp.com wrote: >> >> >> >> On Sun, Jan 6, 2019, 7:06 PM Steve Nickolas <usotsuki@buric.co wrote: >>> >>> On Sun, 6 Jan 2019, A. P. Garcia wrote: >>> >>> If not for GNU, Unix would still have been cloned. Net/2 happened in >>> parallel, did it not? >> >> >> Berkeley actively rewrote most of unix yes. Net/1 was released about the same time GNU was getting started. Net/2 and later 4.4 BSD continued this trend, where 4.4 was finally a complete system. BSD386 only lagged Linux by about a year and had much stronger networking support, but supported fewer obscure devices than linux... >> >> Warner >> >> Ps I know this glosses over a lot, and isn't intended to be pedantic as to who got where first. Only they were about the same time... and I'm especially glossing over the AT&T suits, etc. > > > It's really hard to say. How would you compile it? Clang didn't come along until 2007. The Amsterdam Compiler Kit, perhaps? I find it ironic that BSD people are so quick to bash RMS, yet they have been using his tools (gcc and various utils) for years... By reading this thread it appears there are more people that have personal issues with RMS than technical ones. I find usually there are two sides to each story though. One side is eloquently presented in 2001 movie "Revolution OS"[1] starring RMS amongst others. [1] https://www.youtube.com/watch?v=4vW62KqKJ5A --Andy --Andy