free solaris licenses; benchmark restriction

free solaris licenses; benchmark restriction

Post by Neil Corlet » Tue, 18 Aug 1998 04:00:00



I am saddened to notice that Sun have encumbered the licenses for the
no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
the following:

2. Restrictions.
[... deleted ...] You may not  publish or provide the results of any
benchmark or comparison tests run on
Software to any third party without the prior written consent of Sun.
[... deleted ...]

This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
of FCS for Solaris 2.6. This type of licensing is restriction is, in my
opinion, utterly unacceptable. Put simply, if a product is any good,
then it does not require this type of protection - so perhaps they are
implying that Solaris is horribly slow or something. I have certainly
never seen this restriction applied before to a Sun product. The wording
implies that ALL third party standard benchmark reporting becomes
impossible (SPEC , web tests ... anything depending upon the OS) without
the permission of Sun.

I would be really happy if I was wrong about this. And I would be even
happier if Sun would confirm that all future releases of  Solaris (and
definitly the for-fee releases) will not be subject to such a license
restriction.

Cheers,

Neil

readers note that the opinions expressed above are purely personal and
may not reflect those of my employer.

 
 
 

free solaris licenses; benchmark restriction

Post by Pat Wils » Tue, 18 Aug 1998 04:00:00



>I am saddened to notice that Sun have encumbered the licenses for the
>no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
>the following:
>2. Restrictions.
>[... deleted ...] You may not  publish or provide the results of any
>benchmark or comparison tests run on
>Software to any third party without the prior written consent of Sun.
>[... deleted ...]
>This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
>of FCS for Solaris 2.6. This type of licensing is restriction is, in my
>opinion, utterly unacceptable.

Oh, please.  You're looking a gift horse in the mouth here.  Sun
is just trying to stop people from getting a virtually free copy of
their OS, putting it on some trash hardware, and then claiming
"Slowaris SUX" based on the result of some ill-conceived so-called
benchmark.

You might as well complain about the fact that you can't (legally)
use the no-fee license if you're going to conduct business with it -
Sun still needs to make money, and doesn't want to look bad.

If you don't like the license restrictions, stick to Linux.

--
Pat Wilson


 
 
 

free solaris licenses; benchmark restriction

Post by Dan Stromber » Tue, 18 Aug 1998 04:00:00



> I am saddened to notice that Sun have encumbered the licenses for the
> no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
> the following:

> 2. Restrictions.
> [... deleted ...] You may not  publish or provide the results of any
> benchmark or comparison tests run on
> Software to any third party without the prior written consent of Sun.
> [... deleted ...]

> This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
> of FCS for Solaris 2.6. This type of licensing is restriction is, in my
> opinion, utterly unacceptable. Put simply, if a product is any good,
> then it does not require this type of protection - so perhaps they are
> implying that Solaris is horribly slow or something. I have certainly
> never seen this restriction applied before to a Sun product. The wording
> implies that ALL third party standard benchmark reporting becomes
> impossible (SPEC , web tests ... anything depending upon the OS) without
> the permission of Sun.

> I would be really happy if I was wrong about this. And I would be even
> happier if Sun would confirm that all future releases of  Solaris (and
> definitly the for-fee releases) will not be subject to such a license
> restriction.

I've got to agree with the original poster, assuming he's reading the
license correctly.

This kind of restriction is unacceptable.

It reminds me of a story I heard on the radio recently about medical
ethics and pharmaceuticals, where a pharmaceutical company was trying to
prevent a researcher from publishing vital results...  The
pharmaceutical co required approval before such publication, and they
denied permission.

To her credit, the researcher published anyway.

And sun, if you don't plan to deny permission to publish benchmarks in
important situations, why bother having the restriction?  That is, if
you never disallow publication for an unethical reason, isn't it just so
much bureaucracy?

In other words, can there exist an ethical reason to deny publication of
benchmarks?  I don't believe so.  That kind of information is important
to my job.

Finally, I don't particularly trust benchmarks published by vendors, in
fact I tend to ignore them.  I much prefer benchmarks done by third
parties.  However, I'd trust third party benchmarks far less than I do
now, if they're all subject to approval by various vendors.  So I'd
probably end up doing all my own benchmarking to get trustworthy
benchmarks (very expensive), and I wouldn't even be able to give away my
results, to save someone else from going down that same expensive path.

 
 
 

free solaris licenses; benchmark restriction

Post by Philip Bro » Tue, 18 Aug 1998 04:00:00



>...
>I've got to agree with the original poster, assuming he's reading the
>license correctly.
>..

>It reminds me of a story I heard on the radio recently about medical
>ethics and pharmaceuticals, where a pharmaceutical company was trying to
>prevent a researcher from publishing vital results...  The
>pharmaceutical co required approval before such publication, and they
>denied permission.

Well, sun isn't requireing "permission", it's just requiring you to BUY
a copy of solaris at full price, if you're going to go around publishing
benchmarks.
This encourages a pattern of having benchmarks published by those who
are serious about it, and know something about running benchmarks properly.

--
[trim the no-bots from my address to reply to me by email!]
 --------------------------------------------------
"initiating.. 'getting the hell out of here' maneouver" - Lennier, babylon5

 
 
 

free solaris licenses; benchmark restriction

Post by Thad Flory » Tue, 18 Aug 1998 04:00:00


| >
| > [...]
| > 2. Restrictions.
| > [... deleted ...] You may not  publish or provide the results of any
| > benchmark or comparison tests run on
| > Software to any third party without the prior written consent of Sun.
| > [... deleted ...]
| >
| > This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
| > of FCS for Solaris 2.6. This type of licensing is restriction is, in my
| > opinion, utterly unacceptable. Put simply, if a product is any good,
| > [...]
|
| I've got to agree with the original poster, assuming he's reading the
| license correctly.
|
| This kind of restriction is unacceptable.
|
| It reminds me of a story I heard on the radio recently about medical
| ethics and pharmaceuticals, where a pharmaceutical company was trying to
| prevent a researcher from publishing vital results...  The
| pharmaceutical co required approval before such publication, and they
| denied permission.
|
| To her credit, the researcher published anyway.

Was she sued?  The implication of what she did re: "the pharmaceutical co
required approval before such publication" sounds like a tort.

| And sun, if you don't plan to deny permission to publish benchmarks in
| important situations, why bother having the restriction?  That is, if
| you never disallow publication for an unethical reason, isn't it just so
| much bureaucracy?
|
| In other words, can there exist an ethical reason to deny publication of
| benchmarks?  I don't believe so.  That kind of information is important
| to my job.

Several comments:

1. someone else already mentioned the possibility of an end-user installing
   the software on some *hardware, thus biasing benchmark "results".
   Given my observations of people installing MS SW on x86 platforms without
   the installers reading release notes, README files, installing patches,
   etc., Sun is justified examining the platform and/or tests before granting
   publication approval.

2. Your "That kind of information is important to my job" smacks of a business
   use of the SW in violation of the offer's restrictions, but please accept
   my apologies if I misinterpreted your statement's intent.

3. FWIW, my own observations of Solaris 2.6's (and 2.5.1's) performance vs.
   SunOS 4.1.{34] on indentically-configured systems show Solaris 2.[56] to
   be far more efficient [than SunOS 4.1.[34]] beyond any doubt.   Granted
   there are other OS possibilities (e.g. Linux, FreeBSD, etc.), but you're
   going to have to spend TIME futzing with those whereas Solaris 2.* is a
   Plug'N'Play solution (from my own experience).

4. Ignoring the present "free" offer from Sun for a moment, even the 2-year
   Sunsoft Subscriptions (Get Current, Stay Current (or whatever its name is
   presently)) at approx. US$400+ is one heck of a good deal when you consider
   it bundles support. updates, etc.  FOr comparison, look at the price (from
   PC Connection, a discounter) for NT 4.0 at $274.95 + a one-year Microsoft
   TechNet subscription at $274.95 for a total of $549.90.   Sun wins big.

| Finally, I don't particularly trust benchmarks published by vendors, in
| fact I tend to ignore them.  I much prefer benchmarks done by third
| parties.  However, I'd trust third party benchmarks far less than I do
| now, if they're all subject to approval by various vendors.  So I'd
| probably end up doing all my own benchmarking to get trustworthy
| benchmarks (very expensive), and I wouldn't even be able to give away my
| results, to save someone else from going down that same expensive path.

Only the "free" version of Solaris 2.6 is encumbered by the publication
restriction; there is nothing preventing you (or any independent reviewer or
test lab) buying a copy "off the shelf", running whatever tests you deem
necessary, and publishing the results.  I don't see ANY restrictions on one's
ability to publish test results stated in the licenses for the desktop and
server editions of Solaris 2.6 or 2.5.1 (which are the only copies I have
immediately at hand) and I don't recall any such restrictions for earlier
versions of Solaris 2.* or Solaris 1.* (SunOS 4.1.3 and 4.1.4) though I could
check if necessary.

Thad

 
 
 

free solaris licenses; benchmark restriction

Post by Atif Ahmad Kh » Tue, 18 Aug 1998 04:00:00


I agree with the original poster.  Benchmarking should be allowed regardless
of how much you paid for the software.  I paid for my copy of 2.6 few months
ago.  I am just as much capable of putting it on a trashy machine and telling
people that it sucks as anybody else.  The fear is baseless.  People do use
Linux and love it.  I dont think that the freedom for people to put it on
trashy machines and benchmark it has helped stop the spreading of it any.



>>I am saddened to notice that Sun have encumbered the licenses for the
>>no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
>>the following:
>>2. Restrictions.
>>[... deleted ...] You may not  publish or provide the results of any
>>benchmark or comparison tests run on
>>Software to any third party without the prior written consent of Sun.
>>[... deleted ...]
>>This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
>>of FCS for Solaris 2.6. This type of licensing is restriction is, in my
>>opinion, utterly unacceptable.
>Oh, please.  You're looking a gift horse in the mouth here.  Sun
>is just trying to stop people from getting a virtually free copy of
>their OS, putting it on some trash hardware, and then claiming
>"Slowaris SUX" based on the result of some ill-conceived so-called
>benchmark.
>You might as well complain about the fact that you can't (legally)
>use the no-fee license if you're going to conduct business with it -
>Sun still needs to make money, and doesn't want to look bad.
>If you don't like the license restrictions, stick to Linux.
>--
>Pat Wilson


 
 
 

free solaris licenses; benchmark restriction

Post by Jens Ha » Wed, 19 Aug 1998 04:00:00



Quote:

> I agree with the original poster.  Benchmarking should be allowed regardless
> of how much you paid for the software.  I paid for my copy of 2.6 few months
> ago.  I am just as much capable of putting it on a trashy machine and telling
> people that it sucks as anybody else.

No, you don't. If you had ever seen me fiddling around with a good running system
turning it into some trash no wouldn't say that ;-).
To be serious: I think the free offer is for people like me: I have just started
to work with solaris systems and want to get to know it in more detail, play around
with it and maybe write some nice little tools. A benchmark made by someone like
me is a benchmark of the admin's abilities, not of the OS.
People who know what they're doing with Solaris probably learned it because they're
real professionals, and they have a commercial license somewhere (at the office
or so), so they can publish their results.

Quote:> The fear is baseless.  People do use
> Linux and love it.  I dont think that the freedom for people to put it on
> trashy machines and benchmark it has helped stop the spreading of it any.

You can't really compare it with linux. Linux is cult. It's a party. Someone who's
happy having a good OS on his PC for free and first of all not by M$ doesn't really
care about benchmarks. With very few exceptions maybe. Anyway you don't have much
choice if you want a free OS that you can use even for commercial purposes.

Jens

 
 
 

free solaris licenses; benchmark restriction

Post by Achim Grat » Wed, 19 Aug 1998 04:00:00



> This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
> of FCS for Solaris 2.6. This type of licensing is restriction is, in my
> opinion, utterly unacceptable. Put simply, if a product is any good,
> then it does not require this type of protection - so perhaps they are
> implying that Solaris is horribly slow or something. I have certainly
> never seen this restriction applied before to a Sun product. The wording
> implies that ALL third party standard benchmark reporting becomes
> impossible (SPEC , web tests ... anything depending upon the OS) without
> the permission of Sun.

Then buy Solaris instead of whining about a (reasonable, IMHO)
restriction in a free version.  If you want to publish a benchmark
result, it is quite likely you have a professional interest in it.
Oh, by the way - did you complain about this very same restriction
when it came with other, much more expensive software?

Also, there are far too many idiots out there that don't know how to
spell benchmark, let alone run one properly in a reproducable fashion.
Even reputable magazines (e.g. the german iX) have been goofing up on
benchmarks in the past and that is not counting the occasions were the
selection of benchmarks was skewed to ascertain the outcome.  One
other possible reason that comes immediately to mind is that someone
who didn't want to pay for Solaris is unlikely to pay for a good
compiler.

Achim Gratz.

--+<[ It's the small pleasures that make life so miserable. ]>+--
WWW:    http://www.inf.tu-dresden.de/~ag7/{english/}

Phone:  +49 351 463 - 8325

 
 
 

free solaris licenses; benchmark restriction

Post by e.. » Wed, 19 Aug 1998 04:00:00



Quote:> I am saddened to notice that Sun have encumbered the licenses for the
> no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
> the following:

> 2. Restrictions.
> [... deleted ...] You may not  publish or provide the results of any
> benchmark or comparison tests run on
> Software to any third party without the prior written consent of Sun.
> [... deleted ...]

It's obvious why.  They know that they will get beaten on lmbench
by Linux on both x86 and Sparc, and they are tired of seeing
people publishing comparisons on the web.  Not to say that Linux
is better for everything, it probably isn't, but in OS benchmarks
it is making Sun look silly.

Not letting people benchmark makes them look silly too of course.

Lmbench homepage
        http://www.bitmover.com/lmbench/lmbench.html
Some rather outdated benchmarks
        http://www.caip.rutgers.edu/~davem/scoreboard.html

--
Erik Corry

 
 
 

free solaris licenses; benchmark restriction

Post by Neil Corlet » Wed, 19 Aug 1998 04:00:00



> I am saddened to notice that Sun have encumbered the licenses for the
> no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
> the following:

> 2. Restrictions.
> [... deleted ...] You may not  publish or provide the results of any
> benchmark or comparison tests run on
> Software to any third party without the prior written consent of Sun.
> [... deleted ...]

Probably better reply to my own posting here ;-)

Basically, I dislike the restriction on Oracle. Finding that Sun are
employing it is bad news. For personal release, perhaps it is not such a
major restriction. A lot of benchmarking would be rather too ad-hoc.

But for *educational site licenses*, it's a bummer. How, for example is
anyone going to quote any results obtained on Solaris in an academic paper
under such terms? This type of license term is not something that should be
encouraged for any software product, regardless of complexity.  It is easy
to say now  'go buy a license if you want to benchmark', but what happens
if ALL 2.7 licenses are encumbered this way?

Why do this now? Sun have lived a long time against systems that are faster
in absolute terms (e.g. UnixWare 2/x86/SPARC ; Digital Unix/Alpha) with
similar price/performance. So I rather think Solaris can cope with Linux...
Whilst it Linux might be better on one specific benchmark, it does not win
on all of them.

Anyway, in the meantime I'll hassle Sun directly and see what they have to
say.

Cheers,

Neil Corlett

 
 
 

free solaris licenses; benchmark restriction

Post by Bob Lamoth » Wed, 19 Aug 1998 04:00:00



> I am saddened to notice that Sun have encumbered the licenses for the
> no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
> the following:

> 2. Restrictions.
> [... deleted ...] You may not  publish or provide the results of any
> benchmark or comparison tests run on
> Software to any third party without the prior written consent of Sun.
> [... deleted ...]

> This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
> of FCS for Solaris 2.6. This type of licensing is restriction is, in my
> opinion, utterly unacceptable. Put simply, if a product is any good,
> then it does not require this type of protection - so perhaps they are
> implying that Solaris is horribly slow or something. I have certainly
> never seen this restriction applied before to a Sun product. The wording
> implies that ALL third party standard benchmark reporting becomes
> impossible (SPEC , web tests ... anything depending upon the OS) without

        Climb down off your high horse.  Most of us know that benchmarks
don't amount to diddly.  They rarely, if ever, give you an idea of how
a machine will peform for a certain task under the variety of conditions
that the task will be required to face.  Even if benchmarks could be
considered anything more than fluff who can be certain that your
test was desiged or performed correctly.  Why should Sun take the chance
that your poorly designed, badly run test might make it look bad?  How
many benchmarks have you seen where NT outperforms everything else only
to find that when run again with properly built and tuned systems that
NT dogs behind?  The problem is that its harder to retract a false
statement then it is not to make it in the first place.  All Sun is
doing is insuring that you won't go off half*ed and make a bad
statement without being certain that the test was unbiased.

                                                -Bob
--
* Robert L. Lamothe            | Unix/Internet Contract Professional  *
* (603)929-0353                |                                      *

* URL: http://www.veryComputer.com/| Grampa, What was Microsoft?          *
* remove nospam from address to reply                                 *

 
 
 

free solaris licenses; benchmark restriction

Post by Dan Stromber » Wed, 19 Aug 1998 04:00:00




> | It reminds me of a story I heard on the radio recently about medical
> | ethics and pharmaceuticals, where a pharmaceutical company was trying to
> | prevent a researcher from publishing vital results...  The
> | pharmaceutical co required approval before such publication, and they
> | denied permission.
> |
> | To her credit, the researcher published anyway.

> Was she sued?  The implication of what she did re: "the pharmaceutical co
> required approval before such publication" sounds like a tort.

Yes, I believe legal action was taken against her.

Quote:> | And sun, if you don't plan to deny permission to publish benchmarks in
> | important situations, why bother having the restriction?  That is, if
> | you never disallow publication for an unethical reason, isn't it just so
> | much bureaucracy?
> |
> | In other words, can there exist an ethical reason to deny publication of
> | benchmarks?  I don't believe so.  That kind of information is important
> | to my job.

> Several comments:
> 2. Your "That kind of information is important to my job" smacks of a business
>    use of the SW in violation of the offer's restrictions, but please accept
>    my apologies if I misinterpreted your statement's intent.

I'm at a university site.

Quote:> 3. FWIW, my own observations of Solaris 2.6's (and 2.5.1's) performance vs.
>    SunOS 4.1.{34] on indentically-configured systems show Solaris 2.[56] to
>    be far more efficient [than SunOS 4.1.[34]] beyond any doubt.   Granted
>    there are other OS possibilities (e.g. Linux, FreeBSD, etc.), but you're
>    going to have to spend TIME futzing with those whereas Solaris 2.* is a
>    Plug'N'Play solution (from my own experience).

Sure, Solaris 2.x is nice stuff.

Quote:> 4. Ignoring the present "free" offer from Sun for a moment, even the 2-year
>    Sunsoft Subscriptions (Get Current, Stay Current (or whatever its name is
>    presently)) at approx. US$400+ is one heck of a good deal when you consider
>    it bundles support. updates, etc.  FOr comparison, look at the price (from
>    PC Connection, a discounter) for NT 4.0 at $274.95 + a one-year Microsoft
>    TechNet subscription at $274.95 for a total of $549.90.   Sun wins big.

Yup, Solaris 2.x pricing is better than NT's.

Quote:> | Finally, I don't particularly trust benchmarks published by vendors, in
> | fact I tend to ignore them.  I much prefer benchmarks done by third
> | parties.  However, I'd trust third party benchmarks far less than I do
> | now, if they're all subject to approval by various vendors.  So I'd
> | probably end up doing all my own benchmarking to get trustworthy
> | benchmarks (very expensive), and I wouldn't even be able to give away my
> | results, to save someone else from going down that same expensive path.

> Only the "free" version of Solaris 2.6 is encumbered by the publication
> restriction; there is nothing preventing you (or any independent reviewer or
> test lab) buying a copy "off the shelf", running whatever tests you deem
> necessary, and publishing the results.  I don't see ANY restrictions on one's
> ability to publish test results stated in the licenses for the desktop and
> server editions of Solaris 2.6 or 2.5.1 (which are the only copies I have
> immediately at hand) and I don't recall any such restrictions for earlier
> versions of Solaris 2.* or Solaris 1.* (SunOS 4.1.3 and 4.1.4) though I could
> check if necessary.

This means that educational sites are going to have to go out of their
way to publish benchmarks - but right now, a lot of the good benchmark
data comes from .edu's.  Which means an important source of good
benchmark data may shrivel up and die.

This is the reverse of progress.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
..
.
.
.
.
.
.
.
.
.
.
.
.
.
.

 
 
 

free solaris licenses; benchmark restriction

Post by Thad Flory » Wed, 19 Aug 1998 04:00:00


| > [...]
| > Only the "free" version of Solaris 2.6 is encumbered by the publication
| > restriction; there is nothing preventing you (or any independent reviewer or
| > test lab) buying a copy "off the shelf", running whatever tests you deem
| > necessary, and publishing the results.  I don't see ANY restrictions on one's
| > ability to publish test results stated in the licenses for the desktop and
| > server editions of Solaris 2.6 or 2.5.1 (which are the only copies I have
| > immediately at hand) and I don't recall any such restrictions for earlier
| > versions of Solaris 2.* or Solaris 1.* (SunOS 4.1.3 and 4.1.4) though I could
| > check if necessary.
|
| This means that educational sites are going to have to go out of their
| way to publish benchmarks - but right now, a lot of the good benchmark
| data comes from .edu's.  Which means an important source of good
| benchmark data may shrivel up and die.
|
| This is the reverse of progress.

I don't follow your reasoning.  The present "free" offer is for individuals
and is not (to the best of my recollection) a site license for Universities
which have other means to secure highly-discounted site licenses.

Supporting one of your comments: a LOT of good stuff migrates outwards from
the universities and colleges.

I'm curious, though: what benchmarks have been attributed to universities?

FWIW and IIRC, the SPECmark tests are distributed by a consortium and are
closely-held to prevent alteration and to thwart surreptitious massaging.

Out of curiousity, what types of benchmarks were you considering such that
the present Sun license for the "free" Solaris would prevent publication?

Thad

 
 
 

free solaris licenses; benchmark restriction

Post by Dan Stromber » Wed, 19 Aug 1998 04:00:00




> > I am saddened to notice that Sun have encumbered the licenses for the
> > no-fee educational and 'personal/developer' licenses of Solaris 2.6 with
> > the following:

> > 2. Restrictions.
> > [... deleted ...] You may not  publish or provide the results of any
> > benchmark or comparison tests run on
> > Software to any third party without the prior written consent of Sun.
> > [... deleted ...]

> > This restriction is NOT PRESENT in the license for Hardware 5/98, 3/98
> > of FCS for Solaris 2.6. This type of licensing is restriction is, in my
> > opinion, utterly unacceptable. Put simply, if a product is any good,
> > then it does not require this type of protection - so perhaps they are
> > implying that Solaris is horribly slow or something. I have certainly
> > never seen this restriction applied before to a Sun product. The wording
> > implies that ALL third party standard benchmark reporting becomes
> > impossible (SPEC , web tests ... anything depending upon the OS) without

>         Climb down off your high horse.  Most of us know that benchmarks
> don't amount to diddly.  They rarely, if ever, give you an idea of how
> a machine will peform for a certain task under the variety of conditions
> that the task will be required to face.  Even if benchmarks could be
> considered anything more than fluff who can be certain that your
> test was desiged or performed correctly.  Why should Sun take the chance
> that your poorly designed, badly run test might make it look bad?  How
> many benchmarks have you seen where NT outperforms everything else only
> to find that when run again with properly built and tuned systems that
> NT dogs behind?  The problem is that its harder to retract a false
> statement then it is not to make it in the first place.  All Sun is
> doing is insuring that you won't go off half*ed and make a bad
> statement without being certain that the test was unbiased.

If this is sun's attitude, then why do they publish their own
benchmarks?

If this is sun's attitude, why do they not simply disallow all benchmark
publishing?  Why only allow it if approved by sun?

Sorry, this is tantamount to censoring anything that doesn't make sun
look good - which means benchmark data is less trustworthy.

(And I never have, and never will, buy the argument that benchmarks are
useless.  So long as you keep in mind that benchmarks are positively but
imperfectly correlated with reality, you're fine)

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

 
 
 

free solaris licenses; benchmark restriction

Post by Steffen Klu » Thu, 20 Aug 1998 04:00:00




>Several comments:
>[...]
>3. FWIW, my own observations of Solaris 2.6's (and 2.5.1's) performance vs.
>   SunOS 4.1.{34] on indentically-configured systems show Solaris 2.[56] to
>   be far more efficient [than SunOS 4.1.[34]] beyond any doubt.

And compared to Linux, *BSD, etc.?

Quote:>   Granted
>   there are other OS possibilities (e.g. Linux, FreeBSD, etc.), but you're
>   going to have to spend TIME futzing with those whereas Solaris 2.* is a
>   Plug'N'Play solution (from my own experience).

Have you ask yourself why? Have a look at the list of supported
hardware options of both Solaris x86 and, say, Linux. If you stay
with a plain vanilla off-the-shelf configuration and you get one
of those contemporary Linux distributions there will be NO futzing
at all (from my own experience with both Solaris and Linux).

Honestly, I don't believe Solaris x86 has got any particular
advantages over the free (as in free source) Unices unless
you need to run applications that require Solaris. Those
applications will typically be expensive ones, so what's the
point for the hobbyist? I buy the argument that someone who
uses Solaris at work and wants to get more proficient in it,
(s)he may want to have it at home. But despite the e*ment
around the free (as in no money) Solaris I doubt that there
will be many people willing to swap their Linux or *BSD for it.

If you know of any reason why I should prefer Solaris over a
free Unix I'd be glad to hear it. If you are going to talk
about things like stability, support etc. save your time,
I won't buy those.

Btw, does Solaris x86 come with a free CDE runtime license?

Cheers
Steffen.

--

Fujitsu Australia Ltd
Keywords: photography, Mozart, UNIX, Islay Malt, dark skies
--

 
 
 

1. License restrictions suck. Try Linux, free.

Rather than jerk the whole site from one platform to another, a more
reasonable approach would be to pick one or two functions currently being
performed by your NT server and put up a Linux system to mirror them.
Observe the systems and benchmark them against each other.  Put together a
list of relative pros and cons.  If you do this fairly, you may (and I believe
you will) discover that Linux can be a big win.

Ah, but it's been fixed, and fast (days if not hours).  How about the bug in
NT that lets it generate non-RFC ping packets?  It's a pretty horrible bug in
practice.  Is there a fix for that available?  Will it *ever* be available?
Can I fix it myself?  Can I pay someone else to fix it?  Just how many zillion
dollars of software do I have to buy from M$ each year in order for them to
consider it to be worth their while to fix it for me?

I believe Corel's WordPerfect is available for Linux, and if you believe their
ads, that *is* the most popular word processor available.  You can also run
WABI on Linux, which would give access to further Windows programs.

That notwithstanding, if the main function of a machine is to have a
non-technical user run word processors and spreadsheets, I wouldn't pick Linux
either--I'd choose Windows 95 or MacOS.  (For now.  WINE progress could change
that.)

Cheers,
--Mike

--
Mike Coleman                                  http://ctr.cstp.umkc.edu/~coleman
  The only "intuitive" interface is the nipple.  After that, it's all learned.
                        --Bruce Ediger, on X interfaces.

2. Pascal tools/source/info related to Linux

3. End of "Free" Solaris binary licence program - now it's free.

4. SuSE 7.0: OSLoader?

5. Help needed "benchmarking" solaris 2.6 free memory

6. apache/mod_perl: "Resource temporarily unavailable" errors

7. GPL Licensing Complication (free speech AND free beer?)

8. Installed TightVNC but....

9. A forgetful newbie question:free Education License of Solaris 9

10. Solaris 8 Free Biniary License

11. Solaris 8 Free Binary License Program

12. Why I can run "cfgadm" command on the Free Solaris 8 Binary License system?

13. Do Solaris 8 Free Binary License system not suport "cfgadm" command?