The only way to test totally different operating systems like you want is by examing the kernel code itself. With Windows being closed source that is impossible. Until then we are left with imperfect ways of observing behavior. In the same vein _you_ can say "how could you ever compare a micro vs. monolethic kernel implementation unless the test were specific to each paradyn", yet people want/crave a common measurement of each despite what may be under the hood. I believe you will find my methods baised firmly on the papers referenced. I wasn't taking a walk in uncharted space. I'm not trying to change the way you think, but maybe you need to look at my paper as a Linux vs BSD and then Win 2000 PE vs SE, as those can be looked at more fairly the way you are wanting.
Before the flames, by no cross platform optimizing compiler I mean no commercial compiler in which one would hope to expect the most optimized ASM for each host operating system.
Two things, first..No cross platform optimizing compiler exists for each OS. Secondly, if you actually look at the tests used, most would be directly translated into system calls provided by the each kernel's API. While it is hard to argue that there isn't some penalty from the libraries that might be loaded at the load of a program 2 things come into play. Each test time is based on a warm cache, and secondly the extra library support would be roughly equivalent as each environment is the same. (ie, all executables are loading the GNU runtime library, regardless of the OS)
You misread me, the Server edition's installer would not allow an installation to occur. The Professional version's installer, however, did allow installation. That's the reason I had to give results in the tables for both the Professional and Server editions. As for the system that was used, no I had no real choice. The system was one that was borrowed. As in the conclusions I clearly stated that this paper did not include statistics or conclusions about network performance of any operating system. However, I did put the data results from Hbench:OS into the tables of those networks tests that did work.
X-RDate: Fri, 04 Feb 2000 19:09:36 -0500 (EST) Received: from hubcap.clemson.edu (hubcap [130.127.28.32]) by cs.clemson.edu (8.9.3/8.9.3) with ESMTP id TAA18663 for ; Fri, 4 Feb 2000 19:14:01 -0500 (EST) Received: from localhost (willys@localhost) by hubcap.clemson.edu (8.9.3/8.9.3) with ESMTP id TAA16018 for ; Fri, 4 Feb 2000 19:14:01 -0500 (EST) X-Authentication-Warning: hubcap.clemson.edu: willys owned process doing -bs Date: Fri, 4 Feb 2000 19:14:01 -0500 (EST) In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-UIDL: b67b722feb9fbeccfd32605c38f8a208 XFMstatus: 0000 From: Dan Schmiedt To: Chris Freeze Subject: Re: Post
On Thu, 3 Feb 2000, Chris Freeze wrote:
> Little came of this meeting other than one individual who actually > communicated to me in a real fashion his thoughts on my post. As I
I'm sorry that we didn't make much of an impression on you.
> reaction of bringing 15+ people with you, I believe that I have acted > as a catalyst in this change. As technology professionals we must not > only embrace change, but be the tools of it.
You acted only as a catalyst in that you slandered us in front of the world. We work hard everyday to better support the campus' IT needs, and we will continue to do so -- with or without people like you insulting us.
Thank you for clarifying that you did not intend to slander us individually -- criticize DCIT all you want, but do not criticise me unless you know specifically what I do and how I do it.
I can only hope that one day you will find yourself in a support position in a large IT department.
Well here is an update, for those in DCIT that may or may not being monitoring this thread still. Tuesday, Dan as promised showed up at my office door. He also decided to bring ~15 other individuals with him from DCIT, so that I might explain to them my comments. A roughly 30 minute meeting pursued in the hallway outside my office. Little came of this meeting other than one individual who actually communicated to me in a real fashion his thoughts on my post. As I promised him I would, I would like to clarify my previous post. When I labeled DCIT incompetent, I in no way targeted it toward any individual, rather their performance as a collective. In evaluating this division at Clemson, I stand by my original conclusions. I tried to communicate to them why so many at Clemson have this growing belief, but I could not get them to understand why I and others have made this decision. If they are still reading this, which I hope they are, change within your organization needs, must, and will come about. If this change does not come about, your reputation professionally will continue to suffer bad publicity like this again. By your reaction of bringing 15+ people with you, I believe that I have acted as a catalyst in this change. As technology professionals we must not only embrace change, but be the tools of it.
X-RDate: Mon, 31 Jan 2000 09:55:56 -0500 (EST) Received: from CLEMSON.EDU (mail.clemson.edu [130.127.28.87] (may be forged)) by cs.clemson.edu (8.9.3/8.9.3) with ESMTP id JAA07930; Mon, 31 Jan 2000 09:58:29 -0500 (EST) Received: from mail (mail.clemson.edu [130.127.28.87]) by CLEMSON.EDU (8.9.3/8.9.3) with ESMTP id JAA13182; Mon, 31 Jan 2000 09:47:32 -0500 (EST) Received: from CLEMSON.EDU by CLEMSON.EDU (LISTSERV-TCP/IP release 1.8d) with spool id 1709068 for CLEMSON_STUDENTS-L@CLEMSON.EDU; Mon, 31 Jan 2000 09:47:27 -0500 Approved-By: CJDCK@CLEMSON.EDU Received: from ira (ira.clemson.edu [130.127.4.101]) by CLEMSON.EDU (8.9.3/8.9.3) with ESMTP id JAA03197; Mon, 31 Jan 2000 09:37:25 -0500 (EST) X-Sender: (Unverified) X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.2 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Message-ID: Date: Mon, 31 Jan 2000 09:37:22 -0500 Reply-To: DCIT Publications X-UIDL: 126d73787d80a7195d4c1a172a0cd983 XFMstatus: 0000 Sender: CU STUDENTS - DCIT LIST From: DCIT Publications To: CLEMSON_STUDENTS-L@mail.clemson.edu Subject: DCIT Bulletin
DCIT Bulletin Vol. 1 No. 7
Access to voice over internet services is back!
The group of students, faculty, and staff that is looking into how the university should incorporate internet phone service into its infrastructure will be meeting next Thursday. The study group will recommend policies and procedures related to internet phone service and suggest ways in which the university can optimize its use.
------------------------------------------------ -- Seems like the bad press the University has received has made them rethink their policy of evaluating dialpad services.
I was wondering when you're usually in your office. I would like very much to stop by so you can "tell me to my face how truely [sic] incompetent I really am."
So...I'm going to be using 2 different computers at the same time? And yes, I do have both a student and an employee account. And if you had read my post, I said "intranet" not Internet. As for my skills....I'll put them up against most of the people in my fellow grad students in the CS department. I came to Clemson due to the reputation of my department in today's technology businesses.. And as such, they are paying me to go to school here, I'm not paying...
After reading this in disgust once again, a few things to point out...
We are a University not a College
40,000 users? Clemson reports about 18,000 students, so we have 22,000 employees here at Clemson? How about canning a few and improving some of the buildings on campus.
All dorms have had network service for 2+ years? Mine has had ether jacks for less than a year, Thornhill Village.
Free access? What about the new printer quota you have started this year? Not totally free wouldn't you say...
All the 'student' you know, represent such a small fraction of the student population. Consider the intake of new freshmen that happen every year, your talking mostly about students that have had no access to an intranet.
And another thing, if students are leaving Clemson to go to other Universities that do not provide the latest in technology, what is that saying about Clemson's undergrads? Couldn't get into a better school? What authority did you have to post on behalf of DCIT, you make Clemson sound like a second rate University..
And why should this be integrated into Clemson's existing long distance service? What does a teleco have to do with a tcp/ip stream running across our ATM link?
How about trashing the LD service, any one can get cheaper rates through a national teleco. But I guess for you like alot of the "adminstration" that I have meet look for the buck.
After reading this in disgust once again, a few things to point out... We are a University not a College 40,000 users? Clemson reports about 18,000 students, so we have 22,000 employees here at Clemson? How about canning a few and improving some of the buildings on campus. All dorms have had network service for 2+ years? Mine has had ether jacks for less than a year, Thornhill Village. Free access? What about the new printer quota you have started this year? Not totally free wouldn't you say... All the 'student' you know, represent such a small fraction of the student population. Consider the intake of new freshmen that happen every year, your talking mostly about students that have had no access to an intranet. And another thing, if students are leaving Clemson to go to other Universities that do not provide the latest in technology, what is that saying about Clemson's undergrads? Couldn't get into a better school? What authority did you have to post on behalf of DCIT, you make Clemson sound like a second rate University.. And why should this be integrated into Clemson's existing long distance service? What does a teleco have to do with a tcp/ip stream running across our ATM link? How about trashing the LD service, any one can get cheaper rates through a national teleco. But I guess for you like alot of the "adminstration" that I have meet look for the buck.
I'm not sure you really understand Clemson. Clemson by definition of it's founding is a "land grant" University. Which may or may not follow some of what you have said.
Can I say truely what a wind bag you are, we have 5 T1's and a 10 bursting to 25 megabit ATM link out of this University... So you are going to say that our IT department is going to follow the policy of let's cut the the arm off to see if the red spot is a rash, and if it isn't we'll sew it back on later...if it doesn't get swept under the rug and left disconnected...Sadly disappointing from DCIT, but most students when accessing the campus intranet have become used to this type of poor IT service...
Incompentency is not the word, as a grad student in the Computer Science department at Clemson I can quickly say that the level of skill possessed by the employees of the IT department is deplorable. Constant network outages, slow service, constantly crashing web servers, in some of the classrooms I teach in getting logged into this "World's largest Novell Network" is an impossible task. I can only hope that the idoits within this department use this as a wake up call. Btw, if any of you want to talk to me, I'll be in my office in Jordan G-06. I would love to tell you to your face how truely incompetent you really are...
Slashdot Mirror
www.attrition.org should have the hacked page.
Since about 1.4.2 of XFMail we have supported sending email's to different SMTP gateways based on the address you are sending from.
The only way to test totally different operating systems like you want is by examing the kernel code itself. With Windows being closed source that is impossible. Until then we are left with imperfect ways of observing behavior. In the same vein _you_ can say "how could you ever compare a micro vs. monolethic kernel implementation unless the test were specific to each paradyn", yet people want/crave a common measurement of each despite what may be under the hood. I believe you will find my methods baised firmly on the papers referenced. I wasn't taking a walk in uncharted space. I'm not trying to change the way you think, but maybe you need to look at my paper as a Linux vs BSD and then Win 2000 PE vs SE, as those can be looked at more fairly the way you are wanting.
Is there one that does target the tested operating systems? I did not know of one that did.
The paper was geared at testing each operating system in a low and high memory environment. I hate it when people scan instead of read.
Before the flames, by no cross platform optimizing compiler I mean no commercial compiler in which one would hope to expect the most optimized ASM for each host operating system.
Two things, first..No cross platform optimizing compiler exists for each OS. Secondly, if you actually look at the tests used, most would be directly translated into system calls provided by the each kernel's API. While it is hard to argue that there isn't some penalty from the libraries that might be loaded at the load of a program 2 things come into play. Each test time is based on a warm cache, and secondly the extra library support would be roughly equivalent as each environment is the same. (ie, all executables are loading the GNU runtime library, regardless of the OS)
You misread me, the Server edition's installer would not allow an installation to occur. The Professional version's installer, however, did allow installation. That's the reason I had to give results in the tables for both the Professional and Server editions. As for the system that was used, no I had no real choice. The system was one that was borrowed. As in the conclusions I clearly stated that this paper did not include statistics or conclusions about network performance of any operating system. However, I did put the data results from Hbench:OS into the tables of those networks tests that did work.
I actually did my master's research on a comparison of multi-user operating systems. http://www.cs.clemson.edu/~cfreeze/research.html . I had fun doing it, and was suprised.
X-RDate: Fri, 04 Feb 2000 19:09:36 -0500 (EST)
Received: from hubcap.clemson.edu (hubcap [130.127.28.32]) by cs.clemson.edu
(8.9.3/8.9.3) with ESMTP id TAA18663 for ;
Fri, 4 Feb 2000 19:14:01 -0500 (EST)
Received: from localhost (willys@localhost) by hubcap.clemson.edu (8.9.3/8.9.3)
with ESMTP id TAA16018 for ;
Fri, 4 Feb 2000 19:14:01 -0500 (EST)
X-Authentication-Warning: hubcap.clemson.edu: willys owned process doing -bs
Date: Fri, 4 Feb 2000 19:14:01 -0500 (EST)
In-Reply-To:
Message-ID:
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-UIDL: b67b722feb9fbeccfd32605c38f8a208
XFMstatus: 0000
From: Dan Schmiedt
To: Chris Freeze
Subject: Re: Post
On Thu, 3 Feb 2000, Chris Freeze wrote:
> Little came of this meeting other than one individual who actually
> communicated to me in a real fashion his thoughts on my post. As I
I'm sorry that we didn't make much of an impression on you.
> reaction of bringing 15+ people with you, I believe that I have acted
> as a catalyst in this change. As technology professionals we must not
> only embrace change, but be the tools of it.
You acted only as a catalyst in that you slandered us in front of the
world. We work hard everyday to better support the campus' IT needs, and
we will continue to do so -- with or without people like you insulting us.
Thank you for clarifying that you did not intend to slander us
individually -- criticize DCIT all you want, but do not criticise me
unless you know specifically what I do and how I do it.
I can only hope that one day you will find yourself in a support position
in a large IT department.
Take care, and have a good weekend.
Dan
:1,1s/being/be
Well here is an update, for those in DCIT that may or may not being monitoring this thread still. Tuesday, Dan as promised showed up at my office door. He also decided to bring ~15 other individuals with him from DCIT, so that I might explain to them my comments. A roughly 30 minute meeting pursued in the hallway outside my office. Little came of this meeting other than one individual who actually communicated to me in a real fashion his thoughts on my post. As I promised him I would, I would like to clarify my previous post. When I labeled DCIT incompetent, I in no way targeted it toward any individual, rather their performance as a collective. In evaluating this division at Clemson, I stand by my original conclusions. I tried to communicate to them why so many at Clemson have this growing belief, but I could not get them to understand why I and others have made this decision. If they are still reading this, which I hope they are, change within your organization needs, must, and will come about. If this change does not come about, your reputation professionally will continue to suffer bad publicity like this again. By your reaction of bringing 15+ people with you, I believe that I have acted as a catalyst in this change. As technology professionals we must not only embrace change, but be the tools of it.
X-RDate: Mon, 31 Jan 2000 09:55:56 -0500 (EST)
- --
Received: from CLEMSON.EDU (mail.clemson.edu [130.127.28.87] (may be forged))
by cs.clemson.edu (8.9.3/8.9.3) with ESMTP id JAA07930;
Mon, 31 Jan 2000 09:58:29 -0500 (EST)
Received: from mail (mail.clemson.edu [130.127.28.87]) by CLEMSON.EDU
(8.9.3/8.9.3) with ESMTP id JAA13182; Mon, 31 Jan 2000 09:47:32 -0500 (EST)
Received: from CLEMSON.EDU by CLEMSON.EDU (LISTSERV-TCP/IP release 1.8d) with
spool id 1709068 for CLEMSON_STUDENTS-L@CLEMSON.EDU;
Mon, 31 Jan 2000 09:47:27 -0500
Approved-By: CJDCK@CLEMSON.EDU
Received: from ira (ira.clemson.edu [130.127.4.101]) by CLEMSON.EDU
(8.9.3/8.9.3) with ESMTP id JAA03197; Mon, 31 Jan 2000 09:37:25 -0500 (EST)
X-Sender: (Unverified)
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.2
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Message-ID:
Date: Mon, 31 Jan 2000 09:37:22 -0500
Reply-To: DCIT Publications
X-UIDL: 126d73787d80a7195d4c1a172a0cd983
XFMstatus: 0000
Sender: CU STUDENTS - DCIT LIST
From: DCIT Publications
To: CLEMSON_STUDENTS-L@mail.clemson.edu
Subject: DCIT Bulletin
DCIT Bulletin
Vol. 1 No. 7
Access to voice over internet services is back!
The group of students, faculty, and staff that is looking into how the
university should incorporate internet phone service into its
infrastructure will be meeting next Thursday. The study group will
recommend policies and procedures related to internet phone service and
suggest ways in which the university can optimize its use.
-----------------------------------------------
Seems like the bad press the University has received has made them rethink their policy of evaluating dialpad services.
Hey Chris ...
...
/=] (864)656-7556 WILLYS@clemson.edu
I was wondering when you're usually in your office. I would like very
much to stop by so you can "tell me to my face how truely [sic]
incompetent I really am."
Please do let me know when would be a good time.
Thanks
Dan
\
\
\
__________ _________ Dan Schmiedt
| ( / |) Network Services
| oooo -------- oooo | Clemson University DCIT
[=\ oo oo \------/ oo oo
oooo oooo
So...I'm going to be using 2 different computers at the same time? And yes, I do have both a student and an employee account. And if you had read my post, I said "intranet" not Internet. As for my skills....I'll put them up against most of the people in my fellow grad students in the CS department. I came to Clemson due to the reputation of my department in today's technology businesses.. And as such, they are paying me to go to school here, I'm not paying...
Since when does my technology access fee, become a gift to Clemson. I'm paying for service.
After reading this in disgust once again, a few things to point out...
We are a University not a College
40,000 users? Clemson reports about 18,000 students, so we have 22,000 employees here at Clemson? How about canning a few and improving some of the buildings on campus.
All dorms have had network service for 2+ years? Mine has had ether jacks for less than a year, Thornhill Village.
Free access? What about the new printer quota you have started this year? Not totally free wouldn't you say...
All the 'student' you know, represent such a small fraction of the student population. Consider the intake of new freshmen that happen every year, your talking mostly about students that have had no access to an intranet.
And another thing, if students are leaving Clemson to go to other Universities that do not provide the latest in technology, what is that saying about Clemson's undergrads? Couldn't get into a better school? What authority did you have to post on behalf of DCIT, you make Clemson sound like a second rate University..
And why should this be integrated into Clemson's existing long distance service? What does a teleco have to do with a tcp/ip stream running across our ATM link?
How about trashing the LD service, any one can get cheaper rates through a national teleco. But I guess for you like alot of the "adminstration" that I have meet look for the buck.
After reading this in disgust once again, a few things to point out... We are a University not a College 40,000 users? Clemson reports about 18,000 students, so we have 22,000 employees here at Clemson? How about canning a few and improving some of the buildings on campus. All dorms have had network service for 2+ years? Mine has had ether jacks for less than a year, Thornhill Village. Free access? What about the new printer quota you have started this year? Not totally free wouldn't you say... All the 'student' you know, represent such a small fraction of the student population. Consider the intake of new freshmen that happen every year, your talking mostly about students that have had no access to an intranet. And another thing, if students are leaving Clemson to go to other Universities that do not provide the latest in technology, what is that saying about Clemson's undergrads? Couldn't get into a better school? What authority did you have to post on behalf of DCIT, you make Clemson sound like a second rate University.. And why should this be integrated into Clemson's existing long distance service? What does a teleco have to do with a tcp/ip stream running across our ATM link? How about trashing the LD service, any one can get cheaper rates through a national teleco. But I guess for you like alot of the "adminstration" that I have meet look for the buck.
I'm not sure you really understand Clemson. Clemson by definition of it's founding is a "land grant" University. Which may or may not follow some of what you have said.
Can I say truely what a wind bag you are, we have 5 T1's and a 10 bursting to 25 megabit ATM link out of this University... So you are going to say that our IT department is going to follow the policy of let's cut the the arm off to see if the red spot is a rash, and if it isn't we'll sew it back on later...if it doesn't get swept under the rug and left disconnected...Sadly disappointing from DCIT, but most students when accessing the campus intranet have become used to this type of poor IT service...
Incompentency is not the word, as a grad student in the Computer Science department at Clemson I can quickly say that the level of skill possessed by the employees of the IT department is deplorable. Constant network outages, slow service, constantly crashing web servers, in some of the classrooms I teach in getting logged into this "World's largest Novell Network" is an impossible task. I can only hope that the idoits within this department use this as a wake up call. Btw, if any of you want to talk to me, I'll be in my office in Jordan G-06. I would love to tell you to your face how truely incompetent you really are...