On 2/18/2023 8:14 PM, John Dallman wrote:
> In article <tsrrf3$5qhq$6@dont-email.me>, arne@vajhoej.dk (Arne Vajhøj)
> wrote:
>> On 2/18/2023 6:17 PM, Single Stage to Orbit wrote:
>>> On Sat, 2023-02-18 at 14:30 -0500, Arne Vajhøj wrote:
>>>> And honestly I don't see Ada make the cut.
>>> I don't think that will be an issue, as the GNU compiler suite has
>>> that, and once it's been ported, that will be good enough.
>> But will GCC be ported to VMS (again)?
>
> Not apparently necessary. AdaCore are working on their LLVM backend for
> GNAT <https://github.com/AdaCore/gnat-llvm>. Using this on VMS would
> require VMS to be using an up-to-date LLVM, but that's presumably on the
> agenda.

That would avoid the need for GCC.

But I don't expect ACT to reinstate VMS support.

So someone will need to take the GNAT LLVM stuff and
the VSI LLVM stuff and make it all work on VMS.

Maybe it will happen. Maybe not.

Do you know if the GNAT LLVM thingy will generated binaries
with dependencies on GPL'ed stuff without linking exception
like normal GNAT GPL?

Arne

On 2/18/2023 8:44 PM, Arne Vajhøj wrote:
> On 2/18/2023 8:14 PM, John Dallman wrote:
>> In article <tsrrf3$5qhq$6@dont-email.me>, arne@vajhoej.dk (Arne Vajhøj)
>> wrote:
>>> On 2/18/2023 6:17 PM, Single Stage to Orbit wrote:
>>>> On Sat, 2023-02-18 at 14:30 -0500, Arne Vajhøj wrote:
>>>>> And honestly I don't see Ada make the cut.
>>>> I don't think that will be an issue, as the GNU compiler suite has
>>>> that, and once it's been ported, that will be good enough.
>>> But will GCC be ported to VMS (again)?
>>
>> Not apparently necessary. AdaCore are working on their LLVM backend for
>> GNAT <https://github.com/AdaCore/gnat-llvm>. Using this on VMS would
>> require VMS to be using an up-to-date LLVM, but that's presumably on the
>> agenda.
>
> That would avoid the need for GCC.
>
> But I don't expect ACT to reinstate VMS support.
>
> So someone will need to take the GNAT LLVM stuff and
> the VSI LLVM stuff and make it all work on VMS.
>
> Maybe it will happen. Maybe not.
>
> Do you know if the GNAT LLVM thingy will generated binaries
> with dependencies on GPL'ed stuff without linking exception
> like normal GNAT GPL?
>

Sounds nice to have an Ada compiler on VMS but don't most of the
people using Ada other than hobbyists require it to be a validated
compiler?

bill

On Sat, 2023-02-18 at 19:45 -0500, Arne Vajhøj wrote:
> On 2/18/2023 6:17 PM, Single Stage to Orbit wrote:
> > On Sat, 2023-02-18 at 14:30 -0500, Arne Vajhøj wrote:
> > > And honestly I don't see Ada make the cut.
> >
> > I don't think that will be an issue, as the GNU compiler suite has
> > that, and once it's been ported, that will be good enough.
>
> But will GCC be ported to VMS (again)?
>
> The last GCC version I have seen run on VMS is 2.8.0
> (on VMS Alpha).
>
> And that is a let us call it "not latest and greatest".

Wow, jsut wow. 2.8.0? I remember my first time with Linux and GCC
2.7.2.3! And that was the late 90s!
--
Tactical Nuclear Kittens

On 2/18/2023 8:55 PM, Single Stage to Orbit wrote:
> On Sat, 2023-02-18 at 19:45 -0500, Arne Vajhøj wrote:
>> On 2/18/2023 6:17 PM, Single Stage to Orbit wrote:
>>> On Sat, 2023-02-18 at 14:30 -0500, Arne Vajhøj wrote:
>>>> And honestly I don't see Ada make the cut.
>>>
>>> I don't think that will be an issue, as the GNU compiler suite has
>>> that, and once it's been ported, that will be good enough.
>>
>> But will GCC be ported to VMS (again)?
>>
>> The last GCC version I have seen run on VMS is 2.8.0
>> (on VMS Alpha).
>>
>> And that is a let us call it "not latest and greatest".
>
> Wow, jsut wow. 2.8.0? I remember my first time with Linux and GCC
> 2.7.2.3! And that was the late 90s!

2.8.0 is indeed from around that time (I don't remember exact year,
but some of the files are timestamped 1998).

$ typ test.c
#include <stdio.h>

int main()
{ printf("Hello world from C!\n");
return 0;
}

$ gcc test.c
$ gcclink test
$ r test
Hello world from C!

If anybody has something newer then I am interested.

C++ seems to be broken. Not sure if it always has been
so or some update did that.

Arne

In article <tsrr9q$5qhq$4@dont-email.me>,
Arne Vajhøj <arne@vajhoej.dk> wrote:
>On 2/18/2023 4:47 PM, Dan Cross wrote:
>> In article <tsrfpl$4bfn$2@dont-email.me>,
>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>> On 2/18/2023 4:01 PM, Dan Cross wrote:
>>>> In article <tsrdl6$4bfn$1@dont-email.me>,
>>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>> But there are a few things to remember before
>>> considering VSI going that path.
>>>
>>> 1) Redhat is doing fine delivering support service. But
>>> they may have done even better if they could also have
>>> charged real license fees, but they cannot because
>>> they mostly did not create the products and the products
>>> are typical under GPL or LGPL. VSI can and do sell
>>> licenses.
>>
>> RedHat got started when the commercial Unix vendors, who did
>> charge for software, were still in their prime. Which among
>> them are still selling licenses?
>
>Most of them are still selling. Oracle is selling Solaris.
>IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>Tru64.

Literally every single one of those has been EOL'ed.
Every. Single. One.

>They are definitely not selling as well as they did back then.
>
>But that does not change that Redhat did not chose
>to open source RHEL, JBoss etc. - it was already open
>source and they did not have any way to close source it.

No, but no one could manage to do what RedHat did with a
commercial Unix version.

Why do you think that is?

>[snip commit stats]
>> Moreoever, this sort of ecosystem doesn't exist around VMS
>> right now because it simply cannot.
>
>It cannot exist for VMS itself.

Of course not. Because VMS is closed and proprietary, and
tightly coupled to a handful of platforms that are dead or
dying, and the port to the most important server platform
currently isn't done yet.

Want to move VMS to ARM? Too bad; you can't do it. RISC-V?
Oh well.

>It can exist for all sorts of applications and tools.
>
>It just don't.

See above.

>>> - the products are widely used products, so even
>>> relative low prices generate a lot of of revenue
>>> Neither will be the case for VSI.
>>
>> Yes. Because insistence on an outdated licensing
>> and revenue model is strangling adoption.
>
>There is a huge server market that are very cost
>sensitive (the people that prefer Ubuntu Server
>or RockyLinux over RHEL).
>
>There is also a huge market where the cost of VMS is
>not a problem - there are still sold a lot
>of expensive software.
>
>VSI probably find the second market more attractive
>than the first.

Right now, VSI doesn't seem to have any real market.

I'll be sad when VMS dies because people couldn't see beyond the
way it's always been done.

- Dan C.

In article <tsrpf1$5qhq$1@dont-email.me>,
Arne Vajhøj <arne@vajhoej.dk> wrote:
>On 2/18/2023 4:01 PM, Dan Cross wrote:
>> In article <tsrdl6$4bfn$1@dont-email.me>,
>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>> On 2/18/2023 3:20 PM, Dan Cross wrote:
>>>> In article <memo.20230218104100.11588B@jgd.cix.co.uk>,
>>>> John Dallman <jgd@cix.co.uk> wrote:
>>>>> In article <tsq2vo$3utev$1@dont-email.me>, jan-erik.soderholm@telia.com
>>>>> (Jan-Erik Söderholm) wrote:
>>>>>> English version of the meeting notes:
>>>>>
>>>>> The license news is good. [snip]
>>>>
>>>> Meh.
>>>>
>>>> I'll be blunt: the only reasonable path for VMS to survive
>>>> is to open source it under an OSI-approved license. VSI
>>>> should dedicated itself to finishing the x86_64 port and
>>>> doing the necessary legal work to make that happen,
>>>
>>> The general assumption is that VSI can't do that as
>>> they don't own VMS - HPE does.
>>
>> Which is why they should start working with HPE now
>> to make it happen. Sun didn't own SVR4; AT&T did.
>> Yet somehow OpenSolaris happened.
>
>At that time the SCO group owned it.

You might want to check your priors on that and the exact
timelines.

>But yes somehow they managed to make it happen.
>
>Maybe SUN had some extra rights because they were
>part of the development of SVR4. Maybe the fact that
>SUN was way bigger than the SCO group made it easier. Or
>something else.
>
>But I suspect that VSI would pay more attention to the
>business impact of the move than the IP aspects.
>
>It did not work out business wise.
>
>The open sourcing did not stop the decline of Solaris.

Yes, it was too little too late. I know a number of former Sun
employees; they almost universally agree that Sun killed itself
by not seeing the economics of the x86 platform and not
embracing open source until it was too late.

Sound familar?

>After a few years Oracle (that had bought SUN) moved back
>to a closed source model.

Almost immediately, in fact. Remind me, where's Solaris now?

>And the open source version is something that practically
>nobody use and very few can even remember the name of (I will
>save people the wikipedia search - it is "illumos").

Now I'm quite sure you're not very familiar with that ecosystem.

Lots more people are using illumos in one form or another than
are using Solaris. SmarOS, Joyent, MNX, Oxide and others are
all using illumos. Who's using Solaris, again?

>Open sourcing Solaris did not solve Solaris'es
>problems.
>
>Maybe it even made them worse.

Oh? How do you figure? Please be specific. Or is that just
idle speculation?

>Not a good example to provide to VSI.

No, Linux is the example here. It's also VSI's primary
competition. Indeed, the tragedy of Solaris reinforces the
thesis that open sourcing is really the only way to go; pointing
out the failure Solaris shows what happens if you _don't_
embrace open source in a timely manner. Or are you simply
saying that VMS can't compete against Linux and is doomed to
failure anyway?

Regardless of all that, however, I'm afraid you missed the point
of the example. The Solaris example is about how an
organization _can_ wrangle the legalities to get something open
sourced, despite similar challenges to VSIs vis ownership of the
code. Solaris is not a model of how to manage open source per
se; simply a roadmap to doing it at all.

- Dan C.

In article <tsrpoc$5qhq$2@dont-email.me>,
Arne Vajhøj <arne@vajhoej.dk> wrote:
>On 2/18/2023 5:08 PM, Dan Cross wrote:
>> In article <tsrf3b$4krc$1@dont-email.me>,
>> Dave Froble <davef@tsoft-inc.com> wrote:
>>> What benefits do you imagine for VSI, for customers,
>>> if VSI were to do what you suggest. Talking about
>>> the "open source" issue.
>>
>> Establishment of a developer ecosystem, crowd-sourced fixes for
>> bugs, security auditing,
>
>That all sounds very nice.
>
>But how does it work for the VMS stuff that are already
>open source?
>
>I can tell you: two handful of people are doing all the
>work.

Yes. Because there is no incentive for anyone else, because VMS
is clinging to an antiquated closed model and maintainers see it
as a dead platform. Do you not see that as a problem?

>It is problematic to find people to maintain the ifdefs
>and build scripts of for VMS in many open source projects.

Have you ever stopped to wonder why that is, and how one might
go about changing it?

>Expecting the community to do serious work on 25 MLOC
>is unrealistic.

It is unrealistic to expect VMS to be able to compete against
Linux with the current model. It's a shame that so many VMS
users are perfectly ok with that.

- Dan C.

In article <tsrqg0$5qhq$3@dont-email.me>,
Arne Vajhøj <arne@vajhoej.dk> wrote:
>On 2/18/2023 6:20 PM, Dan Cross wrote:
>> Interesting and superior technology always loses relative to
>> simple economics. Linux is available gratis; VMS is not. Ergo,
>> VMS cannot compete. 30 years ago, Linux was strictly worse than
>> VMS in every measurable way; now the inverse is true. This is
>> not an accident; economics dictate that proprietary will always
>> lose going forward.
>
>Everybody likes free.
>
>But even though creating an additional copy of software
>is free/effortless, then creating the software is not
>free/effortless.

Non-sequitur.

>Fundamentally it cost the same to create 1 line of
>open source code as 1 line of closed source code.

False. It is amortized over every contributor versus having one
organization shoulder the entire cost.

>There are some open source projects that can make
>it work. Linux is a good example of a huge success.
>
>But there are also some that can't make it work.
>My last post mentioned OpenSolaris. But products
>like ElasticSearch and Akka are also switching
>from open source licenses to different licensing.

You fundamentally missed the point of the Solaris reference in
my earlier post.

Let's turn this around: what do you think that VMS's future
prospects look like?

Dan Cross wrote:
> In article <k5crhdFroqeU1@mid.individual.net>,
> Michael Kraemer @ home <M.Kraemer@gsi.de> wrote:
>
>>Dan Cross wrote:
>>
>>>I'll be blunt: the only reasonable path for VMS to survive
>>>is to open source it under an OSI-approved license. VSI
>>>should dedicated itself to finishing the x86_64 port and
>>>doing the necessary legal work to make that happen, and
>>>then pivot to consulting and services (honestly: this is
>>>what DEC should have done, and it's largely what IBM did
>>>in order to survive in the 00's).
>>
>>what exactly did IBM opensource in the 00's?
>>MVS? AIX? i? the Tivoli stuff?
>
>
> JFS,

most certainly not.
Maybe GPFS, but not JFS[2].
Wouldn't make sense anyway,
ripping an integral part off the OS,
in the hope some community will take care of it.

a bunch of stuff that went into Linux, etc, etc,
> etc.

Would be interesting to learn about the etc, etc, etc.

Dan Cross wrote:
> In article <tsrr9q$5qhq$4@dont-email.me>,
> Arne Vajhøj <arne@vajhoej.dk> wrote:
>
>>On 2/18/2023 4:47 PM, Dan Cross wrote:
>>
>>>In article <tsrfpl$4bfn$2@dont-email.me>,
>>>Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>
>>>>On 2/18/2023 4:01 PM, Dan Cross wrote:
>>>>
>>>>>In article <tsrdl6$4bfn$1@dont-email.me>,
>>>>>Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>
>>>>But there are a few things to remember before
>>>>considering VSI going that path.
>>>>
>>>>1) Redhat is doing fine delivering support service. But
>>>> they may have done even better if they could also have
>>>> charged real license fees, but they cannot because
>>>> they mostly did not create the products and the products
>>>> are typical under GPL or LGPL. VSI can and do sell
>>>> licenses.
>>>
>>>RedHat got started when the commercial Unix vendors, who did
>>>charge for software, were still in their prime. Which among
>>>them are still selling licenses?
>>
>>Most of them are still selling. Oracle is selling Solaris.
>>IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>>Tru64.
>
>
> Literally every single one of those has been EOL'ed.
> Every. Single. One.

Latest release of AIX (7.3) was end 2021,
about one year ago.
And two years *after* IBM acquired RH.
Doesn't sound like EOL to me.

John Dallman <jgd@cix.co.uk> wrote:
>In article <tsrmfk$b8i$1@reader2.panix.com>,
>cross@spitfire.i.gajendra.net (Dan Cross) wrote:
>
>> Establishment of a developer ecosystem, crowd-sourced fixes
>> for bugs, security auditing . . .
>
>Can you be confident these things would happen?

There are many, many open source operating systems where that never happened.
There are two very popular ones where it did. Maybe three if you want to
count ReactOS. Statistically speaking, your chances are not very good.

>People from the wider open source community would be facing an alien
>environment that uses weird programming languages. They'd also have to
>work with plenty of people who think things were better when DEC was
>around.
>
>It seems likely to me that an open source VMS would not develop a large
>enough community to keep it going. Plenty of open source projects fail.

The vast majority of open source projects fail. The good news is that
they are still usable and maintainable when they do, becase there is source.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On 2/18/2023 10:06 PM, Dan Cross wrote:
> In article <tsrpoc$5qhq$2@dont-email.me>,
> Arne Vajhøj <arne@vajhoej.dk> wrote:
>> On 2/18/2023 5:08 PM, Dan Cross wrote:
>>> In article <tsrf3b$4krc$1@dont-email.me>,
>>> Dave Froble <davef@tsoft-inc.com> wrote:
>>>> What benefits do you imagine for VSI, for customers,
>>>> if VSI were to do what you suggest. Talking about
>>>> the "open source" issue.
>>>
>>> Establishment of a developer ecosystem, crowd-sourced fixes for
>>> bugs, security auditing,
>>
>> That all sounds very nice.
>>
>> But how does it work for the VMS stuff that are already
>> open source?
>>
>> I can tell you: two handful of people are doing all the
>> work.
>
> Yes. Because there is no incentive for anyone else, because VMS
> is clinging to an antiquated closed model and maintainers see it
> as a dead platform. Do you not see that as a problem?
>
>> It is problematic to find people to maintain the ifdefs
>> and build scripts of for VMS in many open source projects.
>
> Have you ever stopped to wonder why that is, and how one might
> go about changing it?

It is not obvious to me why VMS being open source should
make it more attractive to develop open source on VMS.

There is no (non-religious) reason for an open source developer
to not develop open source on a closed source OS.

Back in the days the commercial Unix'es did not seem
to have a problem attracting open source developers.

Back in the same days a lot of free stuff (which today
would have gotten an open source license slapped on)
was available for VMS - VMS SIG tapes and L&T SIG tapes
were full of such stuff. VMS not being free did not
prevent that.

The Java world (here I am going with Bjarne's "Java isn't
platform independent; it is a platform") saw a flood
of open source before OpenJDK.

Open source simply requires people developing
open source.

A couple of well known quotes:

Benjamin Franklin - Well done is better than well said

John F Kennedy - Ask not what your country can do for you – ask what you
can do for your country

VMS does not need people that say:
- VSI please open source VMS
- someone please port GNAT to VMS
- someone please port Rust to VMS
- someone please port XYZ to VMS

VMS need people that say:
- I have ported XYZ to VMS
- I have created ABC on VMS

Arne

On 2/18/2023 9:49 PM, Dan Cross wrote:
> In article <tsrr9q$5qhq$4@dont-email.me>,
> Arne Vajhøj <arne@vajhoej.dk> wrote:
>> On 2/18/2023 4:47 PM, Dan Cross wrote:
>>> In article <tsrfpl$4bfn$2@dont-email.me>,
>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>> On 2/18/2023 4:01 PM, Dan Cross wrote:
>>>>> In article <tsrdl6$4bfn$1@dont-email.me>,
>>>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>> But there are a few things to remember before
>>>> considering VSI going that path.
>>>>
>>>> 1) Redhat is doing fine delivering support service. But
>>>> they may have done even better if they could also have
>>>> charged real license fees, but they cannot because
>>>> they mostly did not create the products and the products
>>>> are typical under GPL or LGPL. VSI can and do sell
>>>> licenses.
>>>
>>> RedHat got started when the commercial Unix vendors, who did
>>> charge for software, were still in their prime. Which among
>>> them are still selling licenses?
>>
>> Most of them are still selling. Oracle is selling Solaris.
>> IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>> Tru64.
>
> Literally every single one of those has been EOL'ed.
> Every. Single. One.

You can state that and sound totally convincing.

The problem is that everybody that know how to use
basic search on the internet can detect that it is
a lie.

AIX:

AIX 7.3. TL1 was released in December 2022.

https://www.ibm.com/support/pages/aix-support-lifecycle-information
states that IBM supports AIX 7.3 TL1 until end of 2025.

HP-UX:

There was an update to HP-UX 11iv3 in May 2022.

https://www.hpe.com/psnow/doc/4AA4-7673ENW states that HPE supports
HP-UX 11iv3 (on Integrity) until at least end of 2025.

Solaris:

Solaris 11.4 SRU53 was release in January 2023.

https://www.oracle.com/us/support/library/lifetime-support-hardware-301321.pdf
says that Oracle will support Solaris 11.4 until
November 2021 / November 2034.

Do they have a future? No or probably not.

(HP-UX will die with Itanium, Oracle has clearly indicated that they
are putting much effort into Solaris, IBM did not invest in Redhat
to push AIX)

But they are not EOL today. And will not be for several years.

>> But that does not change that Redhat did not chose
>> to open source RHEL, JBoss etc. - it was already open
>> source and they did not have any way to close source it.
>
> No, but no one could manage to do what RedHat did with a
> commercial Unix version.

In their days the commercial unixes did pretty well.

Arne

On Sat, 2023-02-18 at 21:24 -0500, Arne Vajhøj wrote:
> > Wow, jsut wow. 2.8.0? I remember my first time with Linux and GCC
> > 2.7.2.3! And that was the late 90s!
>
> 2.8.0 is indeed from around that time (I don't remember exact year,
> but some of the files are timestamped 1998).

2.7.2.3 came out in August 1997, whilst 2.8.0 was released in Jan 1997.

I was just looking at my current GCC compiler (12.2.1) documentation
and it looks like it still has support for Alpha, VAX but not Itanium
(couldn't find it!)
--
Tactical Nuclear Kittens

On 2/19/2023 9:09 AM, Arne Vajhøj wrote:
> On 2/18/2023 9:49 PM, Dan Cross wrote:
>> In article <tsrr9q$5qhq$4@dont-email.me>,
>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>> On 2/18/2023 4:47 PM, Dan Cross wrote:
>>>> In article <tsrfpl$4bfn$2@dont-email.me>,
>>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>> On 2/18/2023 4:01 PM, Dan Cross wrote:
>>>>>> In article <tsrdl6$4bfn$1@dont-email.me>,
>>>>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>> But there are a few things to remember before
>>>>> considering VSI going that path.
>>>>>
>>>>> 1) Redhat is doing fine delivering support service. But
>>>>> they may have done even better if they could also have
>>>>> charged real license fees, but they cannot because
>>>>> they mostly did not create the products and the products
>>>>> are typical under GPL or LGPL. VSI can and do sell
>>>>> licenses.
>>>>
>>>> RedHat got started when the commercial Unix vendors, who did
>>>> charge for software, were still in their prime. Which among
>>>> them are still selling licenses?
>>>
>>> Most of them are still selling. Oracle is selling Solaris.
>>> IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>>> Tru64.
>>
>> Literally every single one of those has been EOL'ed.
>> Every. Single. One.
>
> You can state that and sound totally convincing.
>
> The problem is that everybody that know how to use
> basic search on the internet can detect that it is
> a lie.
>
> AIX:
>
> AIX 7.3. TL1 was released in December 2022.
>
> https://www.ibm.com/support/pages/aix-support-lifecycle-information states that
> IBM supports AIX 7.3 TL1 until end of 2025.
>
> HP-UX:
>
> There was an update to HP-UX 11iv3 in May 2022.
>
> https://www.hpe.com/psnow/doc/4AA4-7673ENW states that HPE supports HP-UX 11iv3
> (on Integrity) until at least end of 2025.
>
> Solaris:
>
> Solaris 11.4 SRU53 was release in January 2023.
>
> https://www.oracle.com/us/support/library/lifetime-support-hardware-301321.pdf
> says that Oracle will support Solaris 11.4 until
> November 2021 / November 2034.
>
> Do they have a future? No or probably not.
>
> (HP-UX will die with Itanium, Oracle has clearly indicated that they
> are putting much effort into Solaris, IBM did not invest in Redhat
> to push AIX)
>
> But they are not EOL today. And will not be for several years.
>
>>> But that does not change that Redhat did not chose
>>> to open source RHEL, JBoss etc. - it was already open
>>> source and they did not have any way to close source it.
>>
>> No, but no one could manage to do what RedHat did with a
>> commercial Unix version.
>
> In their days the commercial unixes did pretty well.
>
> Arne
>

A question I'd ask, is, would Linux have done so well, if it was nothing more
than "free Unix"?

--
David Froble Tel: 724-529-0450
Dave Froble Enterprises, Inc. E-Mail: davef@tsoft-inc.com
DFE Ultralights, Inc.
170 Grimplin Road
Vanderbilt, PA 15486

Dave Froble <davef@tsoft-inc.com> wrote:
>
>A question I'd ask, is, would Linux have done so well, if it was nothing more
>than "free Unix"?

When Linux was new, that was all it was. There were some competitors like
xinu and minix, but Linux actually provided a full functional system. Because
it was unixlike, there was plenty of existing software for it (including the
whole gnu back catalogue). Because it was free, the barriers to entry were
very low.

One might ask if VMS would have done so well if it hadn't come with the Vax.
Lots of folks bought VMS because of the hardware, not because of the software.
Those folks were the first to abandon ship when the cheap and fast workstations
came out in the eighties. (I'm talking here of mostly development folks and
scientific computing folks who weren't so tied to architecture and who were
very sensitive to price/performance.)
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On 2/19/2023 1:12 PM, Dave Froble wrote:
> On 2/19/2023 9:09 AM, Arne Vajhøj wrote:
>> On 2/18/2023 9:49 PM, Dan Cross wrote:
>>> No, but no one could manage to do what RedHat did with a
>>> commercial Unix version.
>>
>> In their days the commercial unixes did pretty well.
>
> A question I'd ask, is, would Linux have done so well, if it was nothing
> more than "free Unix"?

I am pretty sure that it was not the technical differences
between Linux and Unix that made Linux beat the
commercial Unixes.

License cost must have been a huge factor. Directly: companies
liked to avoid the license fee. But also indirectly: students
learned Linux because it was free and companies liked an
OS where their new hires had skill.

But there was also hardware cost. Solaris, AIX, HP-UX and
Tru64 was running on expensive hardware while Linux ran
on cheaper x84-64 hardware (Solaris did also support x86-64
besides the primary platform SPARC).

And then there is the vendor independence. With Linux you
were not tied to SUN/IBM/HP/DEC. Some companies liked that.
Especially after all the Unix chaos with OSF vs UI and
the nasty interaction between usage of different code bases
and corporate business strategies.

But there were also free Unixes running on x86-64 available
back then. Why Linux and not them? My best guess is that
Linux had better "marketing" - not traditional marketing
aka slick sales people selling to big companies, but the
internet developer to developer talk type of marketing -
Linux was cool while *BSD was old.

Arne

In article <k5e9reF3sueU1@mid.individual.net>,
Michael Kraemer @ home <M.Kraemer@gsi.de> wrote:
>Dan Cross wrote:
>> In article <k5crhdFroqeU1@mid.individual.net>,
>> Michael Kraemer @ home <M.Kraemer@gsi.de> wrote:
>>
>>>Dan Cross wrote:
>>>
>>>>I'll be blunt: the only reasonable path for VMS to survive
>>>>is to open source it under an OSI-approved license. VSI
>>>>should dedicated itself to finishing the x86_64 port and
>>>>doing the necessary legal work to make that happen, and
>>>>then pivot to consulting and services (honestly: this is
>>>>what DEC should have done, and it's largely what IBM did
>>>>in order to survive in the 00's).
>>>
>>>what exactly did IBM opensource in the 00's?
>>>MVS? AIX? i? the Tivoli stuff?
>>
>> JFS,
>
>most certainly not.
>Maybe GPFS, but not JFS[2].

https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/tree/fs/jfs?h=next-20230217

>Wouldn't make sense anyway,
>ripping an integral part off the OS,
>in the hope some community will take care of it.

See above.

>a bunch of stuff that went into Linux, etc, etc,
>> etc.
>
>Would be interesting to learn about the etc, etc, etc.

: spitfire; pwd
/home/cross/reference/linux
: spitfire; git log | grep '^Author: .*ibm.com' | wc -l
24615
: spitfire;

Feel free to browse the commits. Evidently some
will be very surprising to you.

- Dan C.

In article <k5ea4nF3u82U1@mid.individual.net>,
Michael Kraemer @ home <M.Kraemer@gsi.de> wrote:
>Dan Cross wrote:
>> In article <tsrr9q$5qhq$4@dont-email.me>,
>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>
>>>On 2/18/2023 4:47 PM, Dan Cross wrote:
>>>
>>>>In article <tsrfpl$4bfn$2@dont-email.me>,
>>>>Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>
>>>>>On 2/18/2023 4:01 PM, Dan Cross wrote:
>>>>>
>>>>>>In article <tsrdl6$4bfn$1@dont-email.me>,
>>>>>>Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>>
>>>>>But there are a few things to remember before
>>>>>considering VSI going that path.
>>>>>
>>>>>1) Redhat is doing fine delivering support service. But
>>>>> they may have done even better if they could also have
>>>>> charged real license fees, but they cannot because
>>>>> they mostly did not create the products and the products
>>>>> are typical under GPL or LGPL. VSI can and do sell
>>>>> licenses.
>>>>
>>>>RedHat got started when the commercial Unix vendors, who did
>>>>charge for software, were still in their prime. Which among
>>>>them are still selling licenses?
>>>
>>>Most of them are still selling. Oracle is selling Solaris.
>>>IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>>>Tru64.
>>
>>
>> Literally every single one of those has been EOL'ed.
>> Every. Single. One.
>
>Latest release of AIX (7.3) was end 2021,
>about one year ago.
>And two years *after* IBM acquired RH.
>Doesn't sound like EOL to me.

https://www.theregister.com/2023/01/17/unix_is_dead/

- Dan C.

On 2/19/2023 3:17 PM, Scott Dorsey wrote:
>
> One might ask if VMS would have done so well if it hadn't come with the Vax.

Time to trot out one of my favorite anecdotes.

Working on a RFP for a mini-computer. My employer was offering
A Prime 50-Series. Only competitor was DEC offering an as yet
non-existent VAX Model. RFP required a benchmark. We showed up
with several boxes of green-bar with results. DEC showed up with
and envelope containing a letter that stated: "Here is what it
would do if we actually had one to test." The person driving the
bus was an Astronomy Professor. He closed the meeting with the
statement I don't care who wins as long as it says VAX on the front.
We withdrew. He got his VAX. Well, a smaller slower VAX than what
he wanted but they promised him the better one in a year or so.
The best part of it was he got all the Astronomy programs he wanted
to run from Kitt's Peak Observatory. They were all for BSD. :-)

But the point is it was a VAX he wanted, no necessarily VMS. I
expect al lot of the early VMS sites started that way and then just
grew to be stuck on VMS.

bill

In article <tstaid$dhc4$2@dont-email.me>,
Arne Vajhøj <arne@vajhoej.dk> wrote:
>On 2/18/2023 9:49 PM, Dan Cross wrote:
>> In article <tsrr9q$5qhq$4@dont-email.me>,
>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>> On 2/18/2023 4:47 PM, Dan Cross wrote:
>>>> In article <tsrfpl$4bfn$2@dont-email.me>,
>>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>> On 2/18/2023 4:01 PM, Dan Cross wrote:
>>>>>> In article <tsrdl6$4bfn$1@dont-email.me>,
>>>>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>>> But there are a few things to remember before
>>>>> considering VSI going that path.
>>>>>
>>>>> 1) Redhat is doing fine delivering support service. But
>>>>> they may have done even better if they could also have
>>>>> charged real license fees, but they cannot because
>>>>> they mostly did not create the products and the products
>>>>> are typical under GPL or LGPL. VSI can and do sell
>>>>> licenses.
>>>>
>>>> RedHat got started when the commercial Unix vendors, who did
>>>> charge for software, were still in their prime. Which among
>>>> them are still selling licenses?
>>>
>>> Most of them are still selling. Oracle is selling Solaris.
>>> IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>>> Tru64.
>>
>> Literally every single one of those has been EOL'ed.
>> Every. Single. One.
>
>You can state that and sound totally convincing.
>
>The problem is that everybody that know how to use
>basic search on the internet can detect that it is
>a lie.

I think the problem here is that you lack the sophistication to
understand that small security releases (likely as part of
ongoing contractual obligations) do not mean that something
isn't EOL'ed.

>AIX:
>
>AIX 7.3. TL1 was released in December 2022.
>
>https://www.ibm.com/support/pages/aix-support-lifecycle-information
>states that IBM supports AIX 7.3 TL1 until end of 2025.

Such long lead times! A little under three years! What a smart
time to invest in AIX.

>HP-UX:
>
>There was an update to HP-UX 11iv3 in May 2022.
>
>https://www.hpe.com/psnow/doc/4AA4-7673ENW states that HPE supports
>HP-UX 11iv3 (on Integrity) until at least end of 2025.

WOW, so alive....

>Solaris:
>
>Solaris 11.4 SRU53 was release in January 2023.
>
>https://www.oracle.com/us/support/library/lifetime-support-hardware-301321.pdf
>says that Oracle will support Solaris 11.4 until
>November 2021 / November 2034.

You'll noticed that Solaris 12 has disappeared form Oracle's
roadmap.

>Do they have a future? No or probably not.
>
>(HP-UX will die with Itanium, Oracle has clearly indicated that they
>are putting much effort into Solaris, IBM did not invest in Redhat
>to push AIX)
>
>But they are not EOL today. And will not be for several years.

I can still buy licenses for VMS for VAX from HPE; does that
mean that OpenVMS/VAX hasn't been EOL'ed?

>>> But that does not change that Redhat did not chose
>>> to open source RHEL, JBoss etc. - it was already open
>>> source and they did not have any way to close source it.
>>
>> No, but no one could manage to do what RedHat did with a
>> commercial Unix version.
>
>In their days the commercial unixes did pretty well.

Times change.

- Dan C.

On 2/19/2023 3:56 PM, Dan Cross wrote:
> In article <k5ea4nF3u82U1@mid.individual.net>,
> Michael Kraemer @ home <M.Kraemer@gsi.de> wrote:
>> Dan Cross wrote:
>>> In article <tsrr9q$5qhq$4@dont-email.me>,
>>> Arne Vajhøj <arne@vajhoej.dk> wrote:
>>>> Most of them are still selling. Oracle is selling Solaris.
>>>> IBM is selling AIX. HPE is selling HP-UX. HPE is not selling
>>>> Tru64.
>>>
>>>
>>> Literally every single one of those has been EOL'ed.
>>> Every. Single. One.
>>
>> Latest release of AIX (7.3) was end 2021,
>> about one year ago.
>> And two years *after* IBM acquired RH.
>> Doesn't sound like EOL to me.
>
> https://www.theregister.com/2023/01/17/unix_is_dead/

The headline says that Unix is dead.

Those that only read the headline may think that means
all Unix is EOL.

Those that read the content below the headline will
see that it just says that AIX, Solaris and HP-UX
are in "maintenance mode", which by definition is
not EOL.

Arne

Dave Froble <davef@tsoft-inc.com> wrote:
>A question I'd ask, is, would Linux have done so well, if it was nothing more
>than "free Unix"?

Linux was in the right place at the right time. The industry as
a whole was trying to move away from Unix, in no small part due
to AT&T having been freed from the shackles of consent decrees
and attempting to enter the computer market. They had Unix, and
Unix had become the "open system" of record, so they wanted to
exercise control over the considerable financial pie that had
baked around Unix. Simultaneously, 20 years of research into
systems had been (at best) haphazardly retrofitted onto Unix,
which had become a bloated mess and people legitimately thought
that moving systems in new technical directions was the right
way to go. Remember the hype around microkernels?

Meanwhile, the whole industry missed three critical factors: the
power of installed base and source-level compatibility across
Unix variatns; Moore's law the in particular the implications
this had for making cheap x86 machines performance competitive
against much larger workstation and server-class machines at a
fraction of the cost; the emotional attachment people had to
systems beyond any rational basis for favor.

So people wanted Unix, and Unix was getting hard to get. UC
Berkeley had produced a port that removed almost all of the
AT&T code from the base system, but of course AT&T sued
Berkeley over that. Enter Linux: a from-scratch
reimplementation of the basic system interface, but unencumbered
in any way by AT&T/USL copyright. Oh, and it ran on cheap PCs,
not expensive workstations.

I remember at one point in the mid-90s someone showed me a PC
they had bought. It was maybe half as good as a mid-range Sun
workstation, but at a quarter of the cost. Moreover, PCs were
getting better for the cost faster than workstations were; it
was therefore axiomatic that at some point the price/performance
curves would cross, and that the future was not in the
workstation market. Of course, the same thing had happened to
the minicomputer market a decade prior.

So as PCs got better and better, the price of hardware started
driving towards zero. These days, I can buy a Raspberry Pi for
USD 100 that gives performance comparable to a top-of-the-line
gaming rig from a decade ago.

In that environment, software costs start to dominate, at which
point there's a strong incentive to drive those towards zero, as
well; using a gratis operating system becomes almost a given.

The next big cost center, of course, is on support; but by using
a common offering (like Linux), one can amortize that with
economies of scale.

In article <tsu03v$8td$1@panix2.panix.com>,
Scott Dorsey <kludge@panix.com> wrote:
>When Linux was new, that was all it was. There were some competitors like
>xinu and minix, but Linux actually provided a full functional system. Because
>it was unixlike, there was plenty of existing software for it (including the
>whole gnu back catalogue). Because it was free, the barriers to entry were
>very low.

Neither Xinu nor Linux was a competitor to Linux; both were
pedagogical systems designed for teaching. I do think they
filled a bit of a niche among enthusiasts, but Linux's early
competitors were things like COHERENT (absolutely killed by
Linux) and commercial distributions of System V. I suppose
Arne will now tell us those are doing just fine.

>One might ask if VMS would have done so well if it hadn't come with the Vax.
>Lots of folks bought VMS because of the hardware, not because of the software.
>Those folks were the first to abandon ship when the cheap and fast workstations
>came out in the eighties. (I'm talking here of mostly development folks and
>scientific computing folks who weren't so tied to architecture and who were
>very sensitive to price/performance.)

Just so. Hence, if one wishes to see VMS survive, one must find
a way to make VMS attractive to new developers and so on.

- Dan C.

Dan Cross <cross@spitfire.i.gajendra.net> wrote:
>Scott Dorsey <kludge@panix.com> wrote:
>>When Linux was new, that was all it was. There were some competitors like
>>xinu and minix, but Linux actually provided a full functional system. Because
>>it was unixlike, there was plenty of existing software for it (including the
>>whole gnu back catalogue). Because it was free, the barriers to entry were
>>very low.
>
>Neither Xinu nor Linux was a competitor to Linux; both were
>pedagogical systems designed for teaching. I do think they
>filled a bit of a niche among enthusiasts, but Linux's early
>competitors were things like COHERENT (absolutely killed by
>Linux) and commercial distributions of System V. I suppose
>Arne will now tell us those are doing just fine.

Linux was ALSO a pedagogical system designed for teaching, and came very
stripped down without much more than a kernel at first. But it grew.
And folks added other stuff into the distribution, most of it from gnu.

It took a couple years before Linux was able to compete with something
like COHERENT or QNX.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

In article <tsu14t$g5l7$1@dont-email.me>,
Arne Vajhøj <arne@vajhoej.dk> wrote:
>On 2/19/2023 1:12 PM, Dave Froble wrote:
>> On 2/19/2023 9:09 AM, Arne Vajhøj wrote:
>>> On 2/18/2023 9:49 PM, Dan Cross wrote:
>>>> No, but no one could manage to do what RedHat did with a
>>>> commercial Unix version.
>>>
>>> In their days the commercial unixes did pretty well.
>>
>> A question I'd ask, is, would Linux have done so well, if it was nothing
>> more than "free Unix"?
>
>I am pretty sure that it was not the technical differences
>between Linux and Unix that made Linux beat the
>commercial Unixes.

Not at the time, nope.

>License cost must have been a huge factor. Directly: companies
>liked to avoid the license fee. But also indirectly: students
>learned Linux because it was free and companies liked an
>OS where their new hires had skill.
>
>But there was also hardware cost. Solaris, AIX, HP-UX and
>Tru64 was running on expensive hardware while Linux ran
>on cheaper x84-64 hardware (Solaris did also support x86-64
>besides the primary platform SPARC).

x86_64 didn't exist when this was happening. It came in
2000 or so. There's an ironic DEC tie-in, but that's
another story.

>And then there is the vendor independence. With Linux you
>were not tied to SUN/IBM/HP/DEC. Some companies liked that.

"Sun" not "SUN"

>Especially after all the Unix chaos with OSF vs UI and
>the nasty interaction between usage of different code bases
>and corporate business strategies.
>
>But there were also free Unixes running on x86-64 available
>back then. Why Linux and not them? My best guess is that
>Linux had better "marketing" - not traditional marketing
>aka slick sales people selling to big companies, but the
>internet developer to developer talk type of marketing -
>Linux was cool while *BSD was old.

Nope. It had more to do with the USL/UCB lawsuit and the
presumed effect that that would have on the freely-available
BSD code. Since Linux was a clean-room reimplementation, it
wasn't encumbered by AT&T copyright, and would thus be "safe"
should AT&T win in court.

What's ironic is that the people making these decisions didn't
quite understand the AT&T lawsuit, which was about _trade secret
status_ of Unix, not copyright. AT&T was trying to say that it
was the system interface that was the real intellectual property
that was being violated, not the specific expression of that
interface. This would have affected Linux, too, despite being
an independent reimplementation.

Fortunately, the court sided with Berkeley and rejected USL's
trade secret claims on the basis that the system interface was
well documented and had been in the industry for decades (and
that Ken and Dennis used to mail tapes to universities for the
cost of media and shipping).

- Dan C.