Have you ever seen an old apple tree that's so old and knarly and overgrown that it has to be propped up, and even though its almost dead, it still keeps making apples?
That's kind of what I think of apple.
Why not sell a Mac environment for Linux and yield to a better OS? Why make a career out of jerking around customers with bad, confusing, and costly ideas? Everything that apple wants to do can be done under Linux w/ add-ons. Isn't that painfully obvious?
Let's compare pricewatch prices for roughly equivalent drives:
IDE: Deskstar 34.2GB, 34GXP, 9ms, ULTRA DMA/66 2MB 7200rpm MODEL: 2 or 3 year warrantee, $322 or $9.41/GB
ATLAS IV 36.4GB, 7200RPM ULTRA-2 WIDE 68PIN 3.5 SCSI, $740 on pricewatch, typically 5yr warrantee, $20.32 Seagate 23.2 GB, SCSI-3 UW, , $385 or $17.11/GB
Of course for $100 bucks more, you can get a 50GB barracuda, SCSI-3, 5yr warrantee, which you can't get at all in an ide form running at any speed. This works out to $16.80/GB
FACT 1: So, SCSI-3 is ~2x as expensive, but comes with a longer expected life.
FACT 2: NCR-based devices were selling for $100 several years ago and have worked great since Debian 1.x and RedHat 5.x. This is not nearly as bad as IDE people think. You can also boot from various ID numbers, CD, and have both narrow and wide on the same card. Not sure where to get a price on one right now.
So what do we get in return for this extra cost?
1. Faster spindle speed and seek times are definitely the result of marketing, as surely as Bill Gates is planning to increase the CPU requirements of the free world.
2. Reordering of reads and writes by the SCSI drives themselves is one fantastic benefit. However, MS-Windows users benefit from this more than Linux users because of the way Microsoft has created the registry system. Sure file servers benefit from this also, but they benefit even more from point #1.
3. Peripherals and hot swappable drives work great on SCSI and not so well on IDE.
4. SCSI machines idle faster than IDE machines.:) But the same can be said for Athlons vs. Pentium IIIs.
But the biggest hit to performance is running a perl script to update a list of comments every time someone hits the back button on their browser. Isn't that what the reload button is for? You must need a quad bus SCSI/160 with maxed out 10000RPM redundant raids to handle that load....*ahem*.... gentlemen.
Perhaps my anticorporate paranoia is running away with me here, but doesn't it seem like major news outlets like CNN, NASA, foxnews, etc. are driving the need for plugins?
For instance, why do I need a shockwave plugin when the sorts of lame graphics we see could have easily been done with an animated GIF? Why would someone produce a video clip that could ONLY be viewed with the latest shockwave or quicktime when it is clear that the same video clip could be viewed in a format readable by all shockwave or quicktime plugins? Why are some sites giving you a "slideshow", which is nothing more than a few pictures displayed sequentially, using javascript and special plugins? Couldn't this more easily be done using a standard link to a page with a few.jpg files?
I personally believe there are two reasons: 1) Incentives offered by the plugin companies to use their "latest". 2) Stupid pointy haired bosses (PHBs) who demand the "latest" software.
If this is really going on as I believe, it's wasting bandwidth and complicating the development of the web by making things inconvenient for the casual user with a slow connection.
Shouldn't there be some sort of guidelines for web sites and producers of plugins on backward compatability with previous versions of plugins? At least that way once you get a Linux plugin to work, programmers who are donating their time for a Linux port wouldn't have to hit such a moving target.
Slashdot Mirror
Have you ever seen an old apple tree that's so old and knarly and overgrown that it has to be propped up, and even though its almost dead, it still keeps making apples?
That's kind of what I think of apple.
Why not sell a Mac environment for Linux and yield to a better OS? Why make a career out of jerking around customers with bad, confusing, and costly ideas? Everything that apple wants to do can be done under Linux w/ add-ons. Isn't that painfully obvious?
Let's compare pricewatch prices for roughly equivalent drives:
IDE: Deskstar 34.2GB, 34GXP, 9ms, ULTRA DMA/66
2MB 7200rpm MODEL: 2 or 3 year warrantee, $322 or $9.41/GB
ATLAS IV 36.4GB, 7200RPM ULTRA-2
WIDE 68PIN 3.5 SCSI, $740 on pricewatch, typically 5yr warrantee, $20.32
Seagate 23.2 GB, SCSI-3 UW, , $385 or $17.11/GB
Of course for $100 bucks more, you can get a 50GB barracuda, SCSI-3, 5yr warrantee, which you can't get at all in an ide form running at any speed. This works out to $16.80/GB
FACT 1: So, SCSI-3 is ~2x as expensive, but comes with a longer expected life.
FACT 2: NCR-based devices were selling for $100 several years ago and have worked great since Debian 1.x and RedHat 5.x. This is not nearly as bad as IDE people think. You can also boot from various ID numbers, CD, and have both narrow and wide on the same card. Not sure where to get a price on one right now.
So what do we get in return for this extra cost?
:) But the same can be said for Athlons vs. Pentium IIIs.
1. Faster spindle speed and seek times are definitely the result of marketing, as surely as Bill Gates is planning to increase the CPU requirements of the free world.
2. Reordering of reads and writes by the SCSI drives themselves is one fantastic benefit. However, MS-Windows users benefit from this more than Linux users because of the way Microsoft has created the registry system. Sure file servers benefit from this also, but they benefit even more from point #1.
3. Peripherals and hot swappable drives work great on SCSI and not so well on IDE.
4. SCSI machines idle faster than IDE machines.
But the biggest hit to performance is running a perl script to update a list of comments every time someone hits the back button on their browser. Isn't that what the reload button is for? You must need a quad bus SCSI/160 with maxed out 10000RPM redundant raids to handle that load....*ahem*.... gentlemen.
Lurking behind every Red Hat site is a Woody user.
:)
HEADLINE: Woody penetrates server market, Microsoft falls flat.
For the first time ever, Finnish Wood is better than Norwegian Wood. (With apologies to Linus)
[Hint for those under 30: Look at http://rmb.simplenet.com/publi c/files/faqs/said.html]
Perhaps my anticorporate paranoia is running away with me here, but doesn't it seem like major news outlets like CNN, NASA, foxnews, etc. are driving the need for plugins?
For instance, why do I need a shockwave plugin when the sorts of lame graphics we see could have easily been done with an animated GIF? Why would someone produce a video clip that could ONLY be viewed with the latest shockwave or quicktime when it is clear that the same video clip could be viewed in a format readable by all shockwave or quicktime plugins? Why are some sites giving you a "slideshow", which is nothing more than a few pictures displayed sequentially, using javascript and special plugins? Couldn't this more easily be done using a standard link to a page with a few
I personally believe there are two reasons:
1) Incentives offered by the plugin companies to use their "latest".
2) Stupid pointy haired bosses (PHBs) who demand the "latest" software.
If this is really going on as I believe, it's wasting bandwidth and complicating the development of the web by making things inconvenient for the casual user with a slow connection.
Shouldn't there be some sort of guidelines for web sites and producers of plugins on backward compatability with previous versions of plugins? At least that way once you get a Linux plugin to work, programmers who are donating their time for a Linux port wouldn't have to hit such a moving target.