If you really wanted to demonstrate such a high bandwidth link AND really impress the clients, why not show them something really practical... Create an iSCSI based mini-SAN.
What I would do is use a software based iSCSI target like the one from Ardis Technologies and use it to share out a ram-drive. Obviously, you would need a machine with a fair amount of horsepower and quite a bit of RAM, but when you will be able to demonstrate transfer speeds of 250 Megabytes a second, that should be able to adequately show just how fast this link is.
You MIGHT be able to get that speed in a burst from SATA Raid or SCSI Raid, but I doubt you would be able to sustain it - this is why I am recommending a ramdrive.
What I would like to see is a way to route my calls to someone who actually speaks English as a FIRST language.
My Review of KDE 3.2
on
Review: KDE 3.2
·
· Score: 1, Interesting
Ok, so I am not exactly someone who is well known in the Linux community. I am also not exactly the best writer in the world, so this "review" will be a bit terse...
First of all, getting the new KDE was troublesome. Obviously, the prime site - ftp.kde.org - was a touch busy. However, NONE of the mirror sites (in the US anyhow) had it at all. One site had the directory structure for 3.2, but NO CONTENT. The other US sites didn't even have the 3.2 directory structure. In other words, the KDE group was woefully unprepared to actually release the software.
I was finally able to download from ftp.kde.org - at a whopping 7.6k per second. Yep, that means that they were REALLY BUSY. I was on an unloaded T1 so it probably wasn't congestion at my end.
I finally got the whole package set downloaded and installed. That part, at least, was fairly painless. The fact that I'm running Slackware on an old P3-700 laptop notwithstanding...
On starting up the new KDE, one thing that becomes clear almost immediately is that it is MUCH SLOWER at the initial loadup. Probably a 20% time increase. Not painful though.
Things that I like about the new KDE - yes, application startup feels faster. I also liked that KDE recognized an app (Pan) locked up and wasn't responding and offered to kill it for me (when I tried to close the app window).
Some things just simply don't work though. The RDP protocol implementation inside Konqueror doesn't like Windows Server 2003 at all.
When you install the new KDE, it does you a "favor" and resets your media associations to use Noatun and Kaboodle - inspite of the fact that neither of these apps play modern video files very well if at all.
Overall, it's not too awful, but it's not a compelling upgrade either.
(mythical) 5 points to the first person who can identify both the processor (by make/model) and the machine that the processor resided in for the following snippet of code. Bonus if you can identify WHAT this code does...
ld hl,3c00 ld de,3c01 ld bc,03ff ld (hl),20 ldir
Ok, so it's been a few (tens) of years since I have had to do this, so I may have the declaration of the memory address wrong. Just for clarification, the values are all in hex.
Hopefully, they will work out a SINGLE standard for getting copy/cut and paste working correctly.
I can't tell you how infuriating it is when you go to copy a page of text from, say, openoffice.org, and paste it into a webform in Mozilla - only to find that perhaps the first half a paragraph out of 6 made it over.
The absolute worst job I had was as an audio technician at a public theatre in Saginaw. While getting to see the shows was cool, it was absolute hell trying to stay awake through 3 days of technical conference talking about the different uses of silicone (and not, they didn't mention "medical products" like breast implants once). The conferences were 12 hours a day for 3 days.
The Law of the Land, however, does grant you the right to sell, lease or otherwise transfer or dispose of your own property
Dude: this is why software in this country is licensed and not sold. The software manufacturers here in the states want to maximize their profits, so they license the software to you *but not sell it*. This allows them to impose all sorts of additional conditions and terms on the use of said software.
Unfortunately, there are several court cases here in the states that establish this as the way things are - regardless of the "looks like a sale, acts like a sale" bit that occurs when you purchase a retail box at your local store.
Ok, I'll bite - why not just insert a "sleep (10);" line into the connection response of sendmail (or qmail, or whatever MTA you are using)? By making the sender wait 10 seconds before delivery can begin, you get the same effect as a tar-pit...
I thought this bit interesting - from Microsoft Visual Studio.Net Enterprise Architect...
/*** *errno.h - system wide error numbers (set by system calls) * * Copyright (c) Microsoft Corporation. All rights reserved. * *Purpose: * This file defines the system-wide error numbers (set by * system calls). Conforms to the XENIX standard. Extended * for compatibility with Uniforum standard. * [System V] * * [Public] * ****/
Better call the SCO police on their licensee Microsoft - releasing "Unix" proprietary methods without authorization....
I think you're referring to what is known as a class-action lawsuit.
Um.... no. A class action is meant to consolidate an action into one setting. I am suggesting just the opposite. Thousands of smaller actions (bee-stings).
Besides, it's very rare to get any sort of meaningful win out of a class action suit. It's more likely to get a weasel-phrase like "We are sorry it happened but we don't admit to causing it" and possibly a bunch of $5.00 off coupons for their products. Generally, the only people who win in a Class Action are the lawyers.
Unless and until we all get off of our collective lazy butts and do something about this, there is absolutely nothing else that is going to stop it.
What this means is that US, the authors, movers and shakers in the Free Software community (not just us Linux folks, but you BSDers, too), have to attack back lest we risk loosing our hard work.
If we can not defend what is ours, then we will loose all rights to our work. How can this be put any more simpler. SCO is attacking our property and we are all waiting for "someone else" to defend it. This includes people like Linus and Andrew Tridgell (Samba) as well as the little folks like myself.
People, get this straight... if we do not attack back (using the courts), then we WILL loose our property and our community. We can't afford to wait for the IBM and/or RedHat cases to play out. We can't afford to let "someone else" deal with it. We MUST act today, now!
My proposed attack method is akin to a bee-sting. Except in rare cases (allergies), a bee-sting is not fatal, but 1000 bee-stings almost certainly will be. What we need to do is file 1000's of small claims against SCO in your local courts, alleging copyright violation. Seek the maximum allowed for your jurisdiction. Be prepared to show that your work was submitted to the Linux Kernel (or any other project that SCO is distributing like Samba) and that said project is indeed licensed under the GPL. Yes, you may need to consult a lawyer to be effective, but this is the price we MUST pay to attack SCO back, and be effective.
As long as our dislike of the legal system keeps us from using it, we WILL loose because of it.
While it does sound a bit like Windows Roaming Profiles, it sounds a lot more like USMT - User Settings Migration Tool - a utility included with Windows XP for migrating user settings from one machine to another.
The bigger question is: does "using the world wide web" mean using TCP ports 80 & 443, or does it mean the more broad using TCP/IP connectivity between computers/servers.
An electronic election machine solution
on
CNN Reports on Diebold
·
· Score: 4, Insightful
If everyone if so concerned about verifibility of the election results, then give them what they want. This seems so obvious to me...
Use a machine (e.g. a touchscreen based computer) to generate a paper ballot. This paper ballot should both contain a human readable printout of what you (the voter) just voted for, as well as a mag-stripe encoding of the same. Use the same basic technology as used in the airline industry - human readable on the front, machine readable on the back. These printed, mag-stripe coded ballots are then given to the kind people at the balloting place where it is deposited into a locked steel box for counting later. No electronic counting on-premesis. No "internet connections". Just consistent, countable, checkable, permanent and persistent results.
The results are electronically counted thanks to the mag-stripe encoding. If someone or some organization wishes to contest the count results, there is the printed version on the front of each card to give an actual, unmistakable account of that vote.
The ballot generating machines would be there strictly to generate a "valid" ballot. Valid in this sense meaning checking that someone isn't exceeding the number of votes per race allowed (e.g. not voting for more than 1 person for the presidential election). The machine would also generate a "review" screen before the ballot is actually printed to allow the voter to make sure that all their votes were properly tabulated.
The whole point of this mindless exercise is to produce consistent, unmistakable results, right? No more "hanging chads" or partial punch-thru's, right? No more presidential election decisions by the Supreme Court, right?
This is because SCSI, by defination, is an intelligent bus. In very simple terms, with SCSI, you basically tell the drive to go to sector x, read y sectors, and let me know when the data is available in it's entirety. While waiting for the process to complete, your OS can go about doing something else as it only has to wait for the data to become available (DMA transfered so it just shows up in memory).
In IDE, the OS has to position the head, wait x sectors, read a sector, save it into memory, go to the next sector, read again, store again, and so on. In other words, there is no offloading of the actual data retrieval process, so the OS has to do it all, causing significant overhead in data transfers. Because of the timing sensitive nature of reading several megabytes worth of data in a few milliseconds, letting the OS do other things while reading the data is not a really good idea.
This sounds like "session cookies" to me. A Unique IDentifier(UID) is generated on first connect to a site and that UID is used to track your movement and configuration through the website.
This is a basic feature in PHP - the "register_session("$var")" function . I'm sure there are numerous other examples of this being used in other web languages (Perl, ASP, Java, etc...). This has obvious practical use in things like weblogs and even in e-commerce sites. Including such sites as our patent-hungry friends at Amazon.com.
The entire one-click thing they have is built on the premise of being able to remember the user and what that user has ordered while they are online - in other words, tracking their session. Session cookies by any other name.
Either Amazon is using this newly patented technology without a license, or Microsoft has a bogus patent (surprise surprise) based on already patented prior art.
I smell a big patent infringment party coming on here - real fast....
Typically, especially in a civil suit, a case can be sealed on request of both parties involved. This is typically done when one of the parties (usually some company) doesn't want "damning evidence or testimony" to be made public. This company would most likely propose a settlement offer that is contingent on the case being sealed and that the other party be prohibited from discussing the case publically.
In short, it's to prevent bad publicity from getting out.
Remember the battle-cry of the Slashdot'er - IANAL!!!
Considering that their most basic entry - a 2x15" system - costs $2399, I would have to say that they obviously don't plan to sell many of these...
I can go on eBay and buy 15" LCD monitors for $200 a piece (or less) and that for another $200 I can get a custom aluminum monitor chassis built, these things are nothing but very overpriced toys...
Well, if you go with a completely modular kernel, then you already are there. You can add and remove kernel level device drivers at will, while the system is running.
Of course, this takes up a huge amount of hard drive space (bloat) but it get's you where you say you want to go. Our setup program is called "modprobe" and it doesn't require any clicks.
If you want even more fancy stuff going on, then check out the hotplug system (look on http://freshmeat.net - search for hotplug). When setup correctly, hotplug takes care of the modprobe stuff for you!
Slashdot Mirror
If you really wanted to demonstrate such a high bandwidth link AND really impress the clients, why not show them something really practical... Create an iSCSI based mini-SAN.
What I would do is use a software based iSCSI target like the one from Ardis Technologies and use it to share out a ram-drive. Obviously, you would need a machine with a fair amount of horsepower and quite a bit of RAM, but when you will be able to demonstrate transfer speeds of 250 Megabytes a second, that should be able to adequately show just how fast this link is.
You MIGHT be able to get that speed in a burst from SATA Raid or SCSI Raid, but I doubt you would be able to sustain it - this is why I am recommending a ramdrive.
What I would like to see is a way to route my calls to someone who actually speaks English as a FIRST language.
Ok, so I am not exactly someone who is well known in the Linux community. I am also not exactly the best writer in the world, so this "review" will be a bit terse...
First of all, getting the new KDE was troublesome. Obviously, the prime site - ftp.kde.org - was a touch busy. However, NONE of the mirror sites (in the US anyhow) had it at all. One site had the directory structure for 3.2, but NO CONTENT. The other US sites didn't even have the 3.2 directory structure. In other words, the KDE group was woefully unprepared to actually release the software.
I was finally able to download from ftp.kde.org - at a whopping 7.6k per second. Yep, that means that they were REALLY BUSY. I was on an unloaded T1 so it probably wasn't congestion at my end.
I finally got the whole package set downloaded and installed. That part, at least, was fairly painless. The fact that I'm running Slackware on an old P3-700 laptop notwithstanding...
On starting up the new KDE, one thing that becomes clear almost immediately is that it is MUCH SLOWER at the initial loadup. Probably a 20% time increase. Not painful though.
Things that I like about the new KDE - yes, application startup feels faster. I also liked that KDE recognized an app (Pan) locked up and wasn't responding and offered to kill it for me (when I tried to close the app window).
Some things just simply don't work though. The RDP protocol implementation inside Konqueror doesn't like Windows Server 2003 at all.
When you install the new KDE, it does you a "favor" and resets your media associations to use Noatun and Kaboodle - inspite of the fact that neither of these apps play modern video files very well if at all.
Overall, it's not too awful, but it's not a compelling upgrade either.
Absolutely correct - you win!!
:)
What you win, I am not sure, but you DO win!
for whatever it's worth!
(mythical) 5 points to the first person who can identify both the processor (by make/model) and the machine that the processor resided in for the following snippet of code. Bonus if you can identify WHAT this code does...
ld hl,3c00
ld de,3c01
ld bc,03ff
ld (hl),20
ldir
Ok, so it's been a few (tens) of years since I have had to do this, so I may have the declaration of the memory address wrong. Just for clarification, the values are all in hex.
Hopefully, they will work out a SINGLE standard for getting copy/cut and paste working correctly.
I can't tell you how infuriating it is when you go to copy a page of text from, say, openoffice.org, and paste it into a webform in Mozilla - only to find that perhaps the first half a paragraph out of 6 made it over.
The absolute worst job I had was as an audio technician at a public theatre in Saginaw. While getting to see the shows was cool, it was absolute hell trying to stay awake through 3 days of technical conference talking about the different uses of silicone (and not, they didn't mention "medical products" like breast implants once). The conferences were 12 hours a day for 3 days.
Click on [start], then click on [run]. Type in the following in the run dialog box:
notepad c:\windows\system32\drivers\etc\hosts
Edit to your hearts content - then click on file, then save.
To close off notepad, click on the "X" in the upper right hand corner of the notepad window.
Obviously, anyone who jumps to this conclusion has NEVER ran the Eclipse IDE before.
I like Eclipse, don't get me wrong, but ANY editor that keeps the CPU running at around a 70% load while just editing a file or 2 is just plain wrong.
The Law of the Land, however, does grant you the right to sell, lease or otherwise transfer or dispose of your own property
Dude: this is why software in this country is licensed and not sold. The software manufacturers here in the states want to maximize their profits, so they license the software to you *but not sell it*. This allows them to impose all sorts of additional conditions and terms on the use of said software.
Unfortunately, there are several court cases here in the states that establish this as the way things are - regardless of the "looks like a sale, acts like a sale" bit that occurs when you purchase a retail box at your local store.
Ok, I'll bite - why not just insert a "sleep (10);" line into the connection response of sendmail (or qmail, or whatever MTA you are using)? By making the sender wait 10 seconds before delivery can begin, you get the same effect as a tar-pit...
I thought this bit interesting - from Microsoft Visual Studio .Net Enterprise Architect...
Better call the SCO police on their licensee Microsoft - releasing "Unix" proprietary methods without authorization....
Windows Server 2003 does have some "Unix" type technologies in it, including NUMA and LPR/LPD. Their event scheduler is based on AT.
Then, if you want even more Unix type technologies, there is always the addon package called "Unix Services for Windows".
I think you're referring to what is known as a class-action lawsuit.
Um.... no. A class action is meant to consolidate an action into one setting. I am suggesting just the opposite. Thousands of smaller actions (bee-stings).
Besides, it's very rare to get any sort of meaningful win out of a class action suit. It's more likely to get a weasel-phrase like "We are sorry it happened but we don't admit to causing it" and possibly a bunch of $5.00 off coupons for their products. Generally, the only people who win in a Class Action are the lawyers.
I've said this before, and I'll say it again...
Unless and until we all get off of our collective lazy butts and do something about this, there is absolutely nothing else that is going to stop it.
What this means is that US, the authors, movers and shakers in the Free Software community (not just us Linux folks, but you BSDers, too), have to attack back lest we risk loosing our hard work.
If we can not defend what is ours, then we will loose all rights to our work. How can this be put any more simpler. SCO is attacking our property and we are all waiting for "someone else" to defend it. This includes people like Linus and Andrew Tridgell (Samba) as well as the little folks like myself.
People, get this straight... if we do not attack back (using the courts), then we WILL loose our property and our community. We can't afford to wait for the IBM and/or RedHat cases to play out. We can't afford to let "someone else" deal with it. We MUST act today, now!
My proposed attack method is akin to a bee-sting. Except in rare cases (allergies), a bee-sting is not fatal, but 1000 bee-stings almost certainly will be. What we need to do is file 1000's of small claims against SCO in your local courts, alleging copyright violation. Seek the maximum allowed for your jurisdiction. Be prepared to show that your work was submitted to the Linux Kernel (or any other project that SCO is distributing like Samba) and that said project is indeed licensed under the GPL. Yes, you may need to consult a lawyer to be effective, but this is the price we MUST pay to attack SCO back, and be effective.
As long as our dislike of the legal system keeps us from using it, we WILL loose because of it.
Remember too, that rebates, in their current form, are nothing more than an INTEREST FREE LOAN from the consumer (YOU) to the manufacturer.
While it does sound a bit like Windows Roaming Profiles, it sounds a lot more like USMT - User Settings Migration Tool - a utility included with Windows XP for migrating user settings from one machine to another.
The bigger question is: does "using the world wide web" mean using TCP ports 80 & 443, or does it mean the more broad using TCP/IP connectivity between computers/servers.
If everyone if so concerned about verifibility of the election results, then give them what they want. This seems so obvious to me...
Use a machine (e.g. a touchscreen based computer) to generate a paper ballot. This paper ballot should both contain a human readable printout of what you (the voter) just voted for, as well as a mag-stripe encoding of the same. Use the same basic technology as used in the airline industry - human readable on the front, machine readable on the back. These printed, mag-stripe coded ballots are then given to the kind people at the balloting place where it is deposited into a locked steel box for counting later. No electronic counting on-premesis. No "internet connections". Just consistent, countable, checkable, permanent and persistent results.
The results are electronically counted thanks to the mag-stripe encoding. If someone or some organization wishes to contest the count results, there is the printed version on the front of each card to give an actual, unmistakable account of that vote.
The ballot generating machines would be there strictly to generate a "valid" ballot. Valid in this sense meaning checking that someone isn't exceeding the number of votes per race allowed (e.g. not voting for more than 1 person for the presidential election). The machine would also generate a "review" screen before the ballot is actually printed to allow the voter to make sure that all their votes were properly tabulated.
The whole point of this mindless exercise is to produce consistent, unmistakable results, right? No more "hanging chads" or partial punch-thru's, right? No more presidential election decisions by the Supreme Court, right?
No Linux support. Nothing to see here. Move along. Another irrelevent product that can't be used without spending $$$ more on the base OS...
This is because SCSI, by defination, is an intelligent bus. In very simple terms, with SCSI, you basically tell the drive to go to sector x, read y sectors, and let me know when the data is available in it's entirety. While waiting for the process to complete, your OS can go about doing something else as it only has to wait for the data to become available (DMA transfered so it just shows up in memory).
In IDE, the OS has to position the head, wait x sectors, read a sector, save it into memory, go to the next sector, read again, store again, and so on. In other words, there is no offloading of the actual data retrieval process, so the OS has to do it all, causing significant overhead in data transfers. Because of the timing sensitive nature of reading several megabytes worth of data in a few milliseconds, letting the OS do other things while reading the data is not a really good idea.
Before anyone asks, I did not RTFP...
This sounds like "session cookies" to me. A Unique IDentifier(UID) is generated on first connect to a site and that UID is used to track your movement and configuration through the website.
This is a basic feature in PHP - the "register_session("$var")" function . I'm sure there are numerous other examples of this being used in other web languages (Perl, ASP, Java, etc...). This has obvious practical use in things like weblogs and even in e-commerce sites. Including such sites as our patent-hungry friends at Amazon.com.
The entire one-click thing they have is built on the premise of being able to remember the user and what that user has ordered while they are online - in other words, tracking their session. Session cookies by any other name.
Either Amazon is using this newly patented technology without a license, or Microsoft has a bogus patent (surprise surprise) based on already patented prior art.
I smell a big patent infringment party coming on here - real fast....
Typically, especially in a civil suit, a case can be sealed on request of both parties involved. This is typically done when one of the parties (usually some company) doesn't want "damning evidence or testimony" to be made public. This company would most likely propose a settlement offer that is contingent on the case being sealed and that the other party be prohibited from discussing the case publically.
In short, it's to prevent bad publicity from getting out.
Remember the battle-cry of the Slashdot'er - IANAL!!!
Considering that their most basic entry - a 2x15" system - costs $2399, I would have to say that they obviously don't plan to sell many of these...
I can go on eBay and buy 15" LCD monitors for $200 a piece (or less) and that for another $200 I can get a custom aluminum monitor chassis built, these things are nothing but very overpriced toys...
Well, if you go with a completely modular kernel, then you already are there. You can add and remove kernel level device drivers at will, while the system is running.
Of course, this takes up a huge amount of hard drive space (bloat) but it get's you where you say you want to go. Our setup program is called "modprobe" and it doesn't require any clicks.
If you want even more fancy stuff going on, then check out the hotplug system (look on http://freshmeat.net - search for hotplug). When setup correctly, hotplug takes care of the modprobe stuff for you!
How about throwing Postscript in the garbage - where it belongs....
Oops - not in the kernel.... sorry about that. My Bad.