SPEC CPU2006 announced

[Scott Dorsey]

I did note that elsethread, and it's a valid point. It's unfortunate
that comp.lang.c does not have an official charter, the existence of
which would render these periodic topicality debates irrelevant.

I don't think compiler authors should be allowed to even look at
benchmarks. Too much looking at benchmarks results in writing compilers
that compile benchmarks really well, at the expense of actual running code.

I am reminded of a verified Ada compiler from a major manufacturer in
the eighties. It could readily compile anything in the verification suite
but when presented with actual code it was unable to handle even a Hello
World. Our suspicion was that it detected the various programs of the
verification suite and spit out hand-tailored machine code for each one.
--scott

 

[Eugene Miya]

Yes, but this understates the issue. A serious proposal follows.

"System-specific features are nontopical in these groups."
Wow. What have Ellis and Truscott wought?

I'm touched.

2) People like Andy Glew, Eugene Miya not only have contributed to
Usenet, but actually have a raft of real-world *contributions* to the
discipline included within the posted newsgroups, as have several
posters. Interestingly, people that I either know directly, or by
real-world reputation/contributions (like Richard Maine: standards
efforts are *really* hard work and under-appreciated), and/or by long
history of sensible posts on Usenet, seem to think the SPEC
announcement was fine.

5) Unfortunately, I don't know any of the people who seem upset about ....
6) I've done a quick rummage, so I've found:
a) Added up, according to Google profiles, they have written about
50-60,000 posts over the last few years. Presumably, given email
address changes, this is a lower limit, although I'm not exactly sure
how they count.

Well I hope mere numbers aren't a factor. Most of those are
over 200 automated FAQs, but the bulk of those are *.test group daily
and hourly beacons.

b) I'm trying to understand the weight I should assign to these
opinions, as compared to opinions held by people who I know have
actually contributed a lot to the relevant disciplines.

I have considered a demographic weighting business.
But I think Google is also already mining posts.

This is a serious question; I'm *not* trying to give anybody a hard
time, I'm just trying to understand whether this is a real issue or
not, and given the joint posting rate of ~10,000/year, or
1600/person/year, one more post apiece is just a drop in the ocean.

8) Here are things that would be plusses:
a) Work on either official or adhoc language/OS standards committees.
b) Work on relevant conference committees.
c) Relevant refereeing work.

Refereeing in some cases must remain anonymous.
I run NSF review proposal review panels. I am also not above
watching people's posting patterns and asking them to serve
on such panels.

d) Relevant refereed papers, invited talks, published books
e) Contribution of benchmarks to any of the benchmarking groups
f)* Especially insightful postings on relevant topics, i.e., ones that
are often back-referenced by others, copied onto others' websites,
edited into journals, etc.
g)* Creation of relevant software that gets widespread external use;

Also work prohibits my release of code w/o extensive, Cold war era,
and in some cases, rrelevant review.

h) Usenet longevity is relevant (and is hard to determine
mechanically, given email address changes). Also, SPEC familiarity is
somewhat relevant, i.e., it is a real plus to be familiar with the
relevant benchmarking in the 1980s or earlier, before SPEC started. (A
few of the complaints appeared, on the surface, to be from people who
didn't know what SPEC was.)

Architectural changes need to be made before I will ever be perfectly
pleased with SPEC, but that should not stop them from trying.

--

 

[Default User]

Anton Ertl wrote:

One other thing: The complaints seem to come from the comp.lang.c
camp. My impression of this newsgroup is that the only on-topic
postings there are postings explaining why some other posting was
off-topic.

You were doing pretty well until you came up with this lie. CLC is a
high-traffic group with a number of very knowledgable contributors, who
do a great job of answering on-topic questions.

You should be ashamed of yourself and should apologize to the group.





Brian

 

[Default User]

Thanks, man, for making my morning! Thank God the coffee wasn't ready
yet, otherwise I would now need a new keyboard.

To keep with your apodictical style, you're a disgrace to whatever
groups you do post in.

As you didn't follow usenet standard posting methods by leaving out the
attribution, it's very difficult to tell who you have directed this
insult towards.




Brian

 

[Default User]

Hardly seriously upset. I simply posed the question. I would have
been quite satisfied if the original announced "there is a charge,
to cover expenses".

I'm certainly not all that upset by the original post. I consider it to
be marginal in topicality, exacerbated by the commercial aspect.

What I am becoming increasingly irritated by are the unfair
characterizations of comp.lang.c, a newsgroup that has been a huge
benefit to me over the years. It's a newgroup that defends its
topicality fiercely, and in my mind that's one of the things that kept
it so useful.

It has one of the highest concentrations of expert-level contributors
of any technical newsgroup I've ever used. Questions are answered
quickly and thoroughly (and yes, sometimes the answer is "not here").



Brian

 

[Greg Lindahl]

What I am becoming increasingly irritated by are the unfair
characterizations of comp.lang.c,

First you make comp.lang.c look bad by whining about a harmless
cross-post about something which a subset of people think is
incredibly important.

Then you get irritated because it's "unfair" that all the comp.lang.c
whiners about off-topic cross-posters have caused the rest of us to
think bad things about comp.lang.c. Hint: that's because that's all we
see from comp.lang.c denizens.

Try to not be part of the problem.

-- greg

 

[glen herrmannsfeldt]

First you make comp.lang.c look bad by whining about a harmless
cross-post about something which a subset of people think is
incredibly important.

I think both are right. c.l.c does whine a lot about
off topic posts, but also has many good posts. I don't
have time to read it so much lately, but it really is a
good group. Maybe an analogy:

Some years ago I would read comp.arch.powerpc, a group for
discussing the architecture of the PowerPC processor. It seems,
though, that anyone with a PowerPC mac thought that it was the
appropriate group to discuss Macintosh hardware or software
questions, completely overwhelming any PPC architecture
discussions. The solution was comp.arch.powerpc.tech.

-- glen

 

[Default User]

First you make comp.lang.c look bad by whining about a harmless
cross-post about something which a subset of people think is
incredibly important.

This is nonsense. Being protective of topicality is not "whining". In
particular I was talking about the individuals who falsely maintain
that all clc does is tell posters they are off-topic. It's a false
characterization and unfair to some of the very good and knowledgable
people there.

Then you get irritated because it's "unfair" that all the comp.lang.c
whiners about off-topic cross-posters have caused the rest of us to
think bad things about comp.lang.c. Hint: that's because that's all we
see from comp.lang.c denizens.

That fact that you choose not to actually check out the newsgroup is my
fault? I'll point out that same applies, what I see of comp.arch etc.
are largely in this thread. Yet, somehow, I'm able to understand that a
tiny slice of posts about a hot topic aren't representative of a
newsgroup. Are you saying that it's too tough for you?



Brian

 

[Greg Lindahl]

That fact that you choose not to actually check out the newsgroup is my
fault?

No. The irony that you're the same kind of whining idiot that created
the external impression of comp.lang.c is your fault.

I'll point out that same applies, what I see of comp.arch etc.
are largely in this thread.

It is extremely rare that comp.arch complains about cross-posted
threads. So it would be impossible for you to build up a long-term
impression about comp.arch this way.

FWIW, I'm also a comp.lang.fortran and comp.benchmarks reader.

Yet, somehow, I'm able to understand that a
tiny slice of posts about a hot topic aren't representative of a
newsgroup. Are you saying that it's too tough for you?

I was only talking about the external impression of comp.lang.c. That
you read too much into my statement is your problem, not mine.

-- greg

 

[Default User]

No. The irony that you're the same kind of whining idiot that created
the external impression of comp.lang.c is your fault.

Well, here's the problem. I raised an objection to unfair
characterization, and that makes me a "whining idiot"? But I suppose
your whining about the complaints from clc are perfectly legit?

You aren't a person with a reasonable grasp of the facts. Plain and
simple. As such, you aren't going to have anything worthwhile to say.
As such, the killfile seems like the reasonable solution.

So long.

[remainder disregarded]




Brian

 

[Andy Glew]

As I noted elsethread, that irritation - in that SPEC can not and will not
offer it for anonymous download - will never go away for legal reasons:
There are copyright issues, and SPEC particularly wants to make sure that
the run rules and the publication rules are obeyed. The latter alone make
it imperative that you sign a contract with them.

The FSF copyright restrictions do not apply to freely downloaded software?

This legal argument is bogus. More likely, SPEC doesn't want to be
troubled by a host of tech support questions.

Now, they could lower the fee, sure. They already have done so several
times in the past. I do think the $200 is about adequate to cover their
administrative costs for one licensee, and you sure wouldn't want SPEC
to go down because of a DDoS attack, would you? And frankly, anybody to
whom the $200 rate isn't applicable - the additional $600 is just in noise.
Heck, you could re-use runspec as your benchmark harness and save developing
one yourself - that's substantially more in savings.

All I know is, there was a time when I was a starving student,
developing some of the ideas that I contributed to P6, and hence to
the computers that most of you run on, where 200$US, inflation
adjusted, was a significant barrier. Heck, man, I was eating dogfood
when I wanted a taste of meat, buying moldy cucumbers and cutting out
the bad spots to afford "fresh" vegetables.

I will always be sympathetic to people who don't have much money.

Moreover, I am not so certain that I may not end up that way again.

 

[Bill Todd]

Andy Glew wrote:

....

Moreover, I am not so certain that I may not end up that way again.

So those rumors about Intel having the long knives out for real are true?

- bill

[Newsgroups deliberately not cropped to annoy those who have so recently
earned it.]

 

[Greg Lindahl]

Well, here's the problem. I raised an objection to unfair
characterization, and that makes me a "whining idiot"? But I suppose
your whining about the complaints from clc are perfectly legit?

Yes, and yes. Trust me, after N years of exposure to the moronic objections
of c.l.c to cross-posted questions, any reasonable person would go mad. Either
you're mad, or you're inured to the insanity of c.l.c.

As such, the killfile seems like the reasonable solution.

Thank you - at least this will reduce the amount of whining.

-- greg

 

[Greg Lindahl]

Heck, man, I was eating dogfood
when I wanted a taste of meat, buying moldy cucumbers and cutting out
the bad spots to afford "fresh" vegetables.

Er, that's when you're supposed to become a vegetarian. Really.

-- greg

 

[David Kanter]

I'll make my piece short and to the point:

The people complaining about John Henning's post are at best misguided
and somewhat overzealous denizens of usenet, and at worst, uniformed
idiots.

If you happen to think that SPEC is irrelevant to the future of C, C++
and Fortran, you really need to work at a compiler dev. group...

DK

 

[Chris Thomasson]

I'll make my piece short and to the point:

The people complaining about John Henning's post are at best misguided
and somewhat overzealous denizens of usenet, and at worst, uniformed
idiots.

http://groups.google.com/group/comp.arch/msg/4ed798b8862ab8f9?hl=en
(read about halfway into the first paragraph...)



[...]

 

[Chris Thomasson]

Your wish is granted.

Humm... Well, yes... After spending one second reading through your
postings; you did grant his wish... Indeed...

Stupidity has its own reward.

Why would you killfile Andy or Eugene? IMO , they are very smart, and clever
individuals...

I can even make $ off the instruction sets that one probably created...

Do some research before you mouth off...

You sure do know how to let you ignorance shine brightly and clearly, for
every one to see...

WOW...

Okay... Now, please go ahead and killfile me! I dare you too...

 

[Alexander Terekhov]

Andy Glew wrote:
[...]

The FSF copyright restrictions do not apply to freely downloaded software?

I assume that by "freely" you mean lack of manifestation of assent to
indicate acceptance of moronic GPL restrictions regarding downloaded
stuff.

This legal argument is bogus.

Not at all. 17 USC 109 (distribution) and 117 (private modifications
aka "adaptations"). See

http://www.copyright.gov/reports/studies/dmca/sec-104-report-vol-1.pdf

"There is no dispute that section 109 applies to works in digital
form. Physical copies of works in a digital format, such as CDs or
DVDs, are subject to section 109 in the same way as physical
copies in analog form. Similarly, a lawfully made tangible copy
of a digitally downloaded work, such as a work downloaded to a
floppy disk, Zip™ disk, or CD-RW, is clearly subject to section
109."

More quotes from dmca/sec-104-report-vol-<2|3>.pdf:

Red Hat, Inc.:

Let me just clarify that I don't think anyone today intends to
impact our licensing practices. I haven't seen anything in the
comments, nor have I heard anything today that makes me think
someone does have that intention. What we're concerned about
are unintended consequences of any amendments to Section 109.
The primary difference between digital and nondigital products
with respect to Section 109 is that the former are frequently
licensed. ... product is also available for free downloaded
from the Internet without the printed documentation, without
the box, and without the installation service. Many open source
and free software products also embody the concept of copyleft.
... We are asking that amendments not be recommended that would
jeopardize the ability of open source and free software
licensor to require [blah blah]

Time Warner, Inc.:

We note that the initial downloading of a copy, from an
authorized source to a purchaser's computer, can result in
lawful ownership of a copy stored in a tangible medium.

Library Associations:

First, as conceded by Time Warner, digital transmissions can
result in the fixation of a tangible copy. By intentionally
engaging in digital transmissions with the awareness that a
tangible copy is made on the recipient's computer, copyright
owners are indeed transferring ownership of a copy of the work
to lawful recipients. Second, the position advanced by Time
Warner and the Copyright Industry Organizations is premised
on a formalistic reading of a particular codification of the
first sale doctrine. When technological change renders the
literal meaning of a statutory provision ambiguous, that
provision "must be construed in light of its basic purpose"
and "should not be so narrowly construed as to permit evasion
because of changing habits due to new inventions and
discoveries." Twentieth Century Music Corp. v. Aiken, 422 U.S.
151, 156-158 (1975). The basic purpose of the first sale
doctrine is to facilitate the continued flow of property
throughout society.

See also (Pg 27):

http://www.usfca.edu/law/determann/softwarecombinations060403.pdf

regards,
alexander.

 

[Richard Tobin]

If you happen to think that SPEC is irrelevant to the future of C, C++
and Fortran, you really need to work at a compiler dev. group...

I think the view of the more extreme comp.lang.c complainers is that
that group is not about "things relevant to the future of C", nor is
it about compiler development. Rather it is about the C language itself,
as standardized. I don't agree with this narrow view.

-- Richard

 

[Ketil Malde]

I don't think compiler authors should be allowed to even look at
benchmarks. Too much looking at benchmarks results in writing compilers
that compile benchmarks really well, at the expense of actual running code.

Well, you can try to click your heels together, and see if it helps.

The problem is that compiler writers need to test their compilers, and
if you don't have an official benchmark, you'll get unofficial ones -
at worst, you get a compiler that only can compile itself.

So if you are interested in performance, you really should care about
benchmarks as well.

SPEC isn't perfect, but at least it is an attempt at a large and
comprehensive suite, and my impression is that they try hard to
represent a large set of problem domains and workloads.

-k