...I wonder if the CAD vendors and other vendors are going to consider DirectX in the future, especially with so many vendors shifting more focus towrads Windows in the last couple of years.
I certainly hope not. UNIX is still a more suitable platform for intensive large-scale CAD, since CAD is one domain where the software can take advantage of everything UNIX and high-end workstations have to offer. For example: multiple high-end framebuffers, large memory space, big CPU caches, multiple CPUs, high system bandwidth, and linear scalability in the operating system.
I've used Pro/E on both Sun workstations and on decent PCs, and the PCs just leave something to be desired. On the Suns, the image quality is better, Solaris is more consistent and better-behaved, and the whole package just seems to take the abuse better. And, considering the licensing costs of Pro/E, the extra cost for a good Sun workstation really isn't that bad.
To be honest, I feel that Windows and PCs have cheapened the high-end CAD industry in more ways than just up-front cost. Some good analogies: Replacing a Sun workstation with a Windows PC is somewhat like replacing a BMW M3 with a modded Chevy Cavalier or like replacing a genuine Swiss Army Knife with a cheap Chinese knock-off. Sure the Cavalier will probably get you from point A to point B, but it handles poorly and just feels cheap no matter how much horsepower you think it has. Sure the cheap knife has the same tools, but they break and wear out in no time. Is a PC with Windows supposed to be how we reward our hard-working professionals?
The whole concept, dumb graphics terminals tied to application servers, is obsolete.
Network-transparency in the windowing system is a truly useful feature and is definitely not obselete. People who complain about X are usually people who forget where X's real powers lie. Any adequate replacement for X would have to preserve network-transparency; anything less is simply not an adequate replacement.
The whole concept, dumb graphics terminals tied to application servers, is obsolete.
You have to be careful with phrases like "regression testing". Here on one project, all the managers where spouting it when no one knew what it meant. They just enjoyed saying "regression" (sounds smart) and putting it on the reports to imply quality where there was none.
So, while I still don't understand what regression testing really is, I do know to warn you to learn what it is first before you begin employing it. That way, fewer people on the project will be fooling themselves about the quality of the end product.
Flash peeves will never be redundant, since Flash is simply an awful tool for developing websites. As the original poster said, Flash is best used for value-added things like demonstrations, interactive tidbits, etc. Whenever I see an all-Flash website, I groan as it takes longer to load, sucks my machine's resources, doesn't always work, sometimes crashes my browser, is often counter-intuitive, and is tacky, in general. The result: I try to avoid Flash-based websites as much as possible, and I certainly never bookmark them (they aren't worth my time).
The error message actually says to upgrade to IE or Netscape 6. However, I'm using Netscape 6, and I still got the error message. Perhaps they didn't like the fact my browser blocks all their cookies and isn't running on a Microsoft OS. Well, that's just too bad for them.
As far as your girlfriend's comments about Macs, I've noticed that a lot of people don't know where one piece of software begins and another ends when using a computer. For example, the people testing my software comeback with all sorts of crazy problem reports that are totally unrelated to my code, but with everything in the same context they just don't see the separation.
Thats the big problem MS has at the moment, they have sold their software to almost everybody so now they will have to look for ways to make these people pay once more to be able to keep making profit on windows and office.
One critical problem with mandatory upgrades is that Microsoft will be imposing risk on its customers. Trust me, software upgrades are risky. This is why many people are still using Windows NT or Solaris 2.6 as in the earlier post (i.e., their working business is more important than a software "upgrade").
Any company who knowingly forces its customers to make decisions they otherwise wouldn't make is simply a company that eventually loses its customers.
My reputation? What the hell are you talking about?
Maybe I'm weird, but I don't buy computer components in order to get into online pissing matches...
I was referring to professional reputation. This isn't a matter of getting bragging rights among friends, it is a matter of building systems upon which a business can establish its reputation.
SCSI can help in this regard. For example, it isn't difficult to build a RAID array of SCSI disks, which can take full advantage of the SCSI bus' bandwidth and UNIX's inherent support of concurrent I/O. Going further by adding a second SCSI controller to the host and a second RAID array can allow for a disk system with no single points of failure. You can maintain it on the fly, and the users will never know. With no disk-related downtime, the system might even pay for itself.
For desktops, SCSI has fewer applications, but I have seen cases where a user's productivity is limited by disk bandwidth. An array of two or three SCSI disks in the workstation could solve this problem, and, again, it might pay for itself.
Another point about SCSI is that it allows computers to have longer useful lives. Being able to tack on more devices is really useful in the long-haul, and good SCSI compatibility allows adding big drives to old computers (there is a 36GB Fast SCSI drive available for under $400--not bad for a computer built in 1992 that is still going strong). Fewer whole-server upgrades allows controlling costs on a larger scale, where there is less disruption of users' work and less bereaucratic overhead aquiring new hardware.
In conclusion, I really don't percieve SCSI to be very expensive.
Because Scsi drives can get throughput up to 360MB/sec in certain modes.
Correction: Arrays of SCSI drives can get that sort of throughput. Each drive itself is crippled to the speed of the platter, an average of 20 to 30 MB/sec or so.
The only advantage to ATA is monetary cost. If that is all you care about, then that is fine. If you value your time and your reputation, then use SCSI or Fibre Channel. It's pretty simple.
media are market-driven, not substance-or-content driven
This is very evident in the evolution of news media over the years. TV news brodcasts, such as local news and now even CNN Headline News, are little more than a glorified cross between Cops and Entertainment Tonight. Local newspapers often run poor national news and sorely biased local news dotting the pages amid huge advertisements and coupon sections. Now, I watch almost nothing on TV, I've stopped reading the newspaper, and I'm finding that public TV, public radio, and the WWW are the last refuges of worthwhile content.
One other thing to look our for is HTML email (gack!) with loads an image from the spammer's site. There's usually some id tag sent with the image request so that the spammer gets confirmation on your email address just by reading the email.
Ugh. This is the absolutely worst aspect of HTML e-mail. Just by sending you an unsolicited HTML e-mail, a company can get your browser model and version, whether JavaScript and Java are enabled, your IP address and hostname, the operating system, and roughly where you are located in the world (sometimes down to the city). First, they know you check your e-mail; second, in one click, you just provided a wealth of marketing information; and, third, they can tailor their future e-mail to your system's security vulnerabilities.
Does anyone know of cases, where e-mail was used to install spyware on the client?
Whoever first concieved of HTML-based e-mail should cower in shame for lack of foresight. And all those who chose to implement HTML-based e-mail clients should also cower in shame. HTML-based e-mail is simply irresponsible. I'm simply tired of people who insist in making their e-mail pretty, while unknowingly sacrificing their privacy and security.
Tell that to the secondary market. Hardware is a physical tangible hard-to-copy thing that is owned. If I had an X-Box, I could take little wires and a soldering iron and do anything I want to its innards, risking only voiding the warranty. If I just wanted to use the CPU to keep a little cup of tea warm, I could, and you couldn't stop me.
If you ever tried that with my gaming console (yes, I'm a developer for a major game company, not MS), I'll send my lawyers after you so fast you'll be in jail getting assfucked by Bubba and his friends.
How long will all this last after one disgruntled engineer posts the critical Palladium specs to an anonymous forum (electronic or otherwise). Or, when foriegn intelligence agencies steal Palladium's secrets for their own nation's defense interests. Palladium's weaknesses will leak, it is just a matter of time before it is an expensive useless chip in our then-new computers.
There are huge companies who dispise Microsoft and have their own fabrication plants. Texas Instruments is a long-time partner with Sun Microsystems, for example. IBM has mixed feelings about Microsoft and has their own factories. Taiwan is home to many chip makers, and they are considering open source software more and more. Many of these companies currently build whole computer systems without even a hint of Microsoft software of x86 hardware (I'm using such a computer right now, and it's wonderfully useful).
World-wide, Microsoft's plans will probably be an up-hill battle the entire way.
Learning about the diamond industry pissed me off enough that I will never buy another diamond. Microsoft is no different. If they get their way, where I can not choose how to develop and distribute software, then I will probably find a new career. It'll be hard to find one not dependent on Microsoft software, but I'll try.
What if you, a copyright holder, choose to use P2P as one of your distribution mechanisms. This is a legitimite use of P2P.
Also, you are paying for your internet connection. Would the RIAA be liable for real monetary damage done to you, since a DoS attack robs you of paid-for bandwidth that is really being used for legitimate purposes?
I'm bothered that this article focused so much on testing, when the real solution lies so much further up the development "pipeline".
Testing is very important, but it really is of limited use when the software has chronic design problems. If the nature of the software lends itself to hack-n-patch type fixes that build up into a painful thicket of code, then no real progress is made through testing.
Testing is intended to shake out the last few bugs in a well-thought out system, and the users delight in using the final product....<wakes up>...oh, I must have dozed off there...where was I?...oh...<hack-hack-hack punch-punch compile damn-it! hack-hack...>
Everything's always "only $10 a month!" or whatever, and before you know it - you're talking thousands a year in these non-essential services, chipping away at your income.
Companies simply love this. It's sad when I see friends who are still struggling to pay school loans fall for the "inexpensiveness" of cell phones, short-term financing plans, extended warranties, etc. These sorts of people are causing the rise of the easy-bankruptcy we'll-bail-you-out debt-consolodation industry. It's possible that the amount of average individual expenses and debt will reach a threshold where the subscription service business model crumbles due to no one being able to afford anything but their monthly triple-mortgage payment.
And in fact, the network admins here would likely tell you the *nix boxes they administer go down more often.
This means the people administering your UNIX boxes really don't know what they are doing. The Solaris kernel, for example, can easily go under high load for a year or more. There are mini-outages, such as restarting daemons, but these are peripheral to the operating system, which simply doesn't need restarting. This modularity of UNIX is one reason some people can claim very long uptimes--they maintain the running system. The lack of modularity in Windows has always disturbed me (what does a person do when the GUI goes flaky? In Solaris, I can just resurrect CDE with no one else noticing.)
Slashdot Mirror
...I wonder if the CAD vendors and other vendors are going to consider DirectX in the future, especially with so many vendors shifting more focus towrads Windows in the last couple of years.
I certainly hope not. UNIX is still a more suitable platform for intensive large-scale CAD, since CAD is one domain where the software can take advantage of everything UNIX and high-end workstations have to offer. For example: multiple high-end framebuffers, large memory space, big CPU caches, multiple CPUs, high system bandwidth, and linear scalability in the operating system.
I've used Pro/E on both Sun workstations and on decent PCs, and the PCs just leave something to be desired. On the Suns, the image quality is better, Solaris is more consistent and better-behaved, and the whole package just seems to take the abuse better. And, considering the licensing costs of Pro/E, the extra cost for a good Sun workstation really isn't that bad.
To be honest, I feel that Windows and PCs have cheapened the high-end CAD industry in more ways than just up-front cost. Some good analogies: Replacing a Sun workstation with a Windows PC is somewhat like replacing a BMW M3 with a modded Chevy Cavalier or like replacing a genuine Swiss Army Knife with a cheap Chinese knock-off. Sure the Cavalier will probably get you from point A to point B, but it handles poorly and just feels cheap no matter how much horsepower you think it has. Sure the cheap knife has the same tools, but they break and wear out in no time. Is a PC with Windows supposed to be how we reward our hard-working professionals?
The whole concept, dumb graphics terminals tied to application servers, is obsolete.
Network-transparency in the windowing system is a truly useful feature and is definitely not obselete. People who complain about X are usually people who forget where X's real powers lie. Any adequate replacement for X would have to preserve network-transparency; anything less is simply not an adequate replacement.
The whole concept, dumb graphics terminals tied to application servers, is obsolete.
Slightly off-topic: Have you ever used the WWW?
You have to be careful with phrases like "regression testing". Here on one project, all the managers where spouting it when no one knew what it meant. They just enjoyed saying "regression" (sounds smart) and putting it on the reports to imply quality where there was none.
So, while I still don't understand what regression testing really is, I do know to warn you to learn what it is first before you begin employing it. That way, fewer people on the project will be fooling themselves about the quality of the end product.
Flash peeves will never be redundant, since Flash is simply an awful tool for developing websites. As the original poster said, Flash is best used for value-added things like demonstrations, interactive tidbits, etc. Whenever I see an all-Flash website, I groan as it takes longer to load, sucks my machine's resources, doesn't always work, sometimes crashes my browser, is often counter-intuitive, and is tacky, in general. The result: I try to avoid Flash-based websites as much as possible, and I certainly never bookmark them (they aren't worth my time).
The error message actually says to upgrade to IE or Netscape 6. However, I'm using Netscape 6, and I still got the error message. Perhaps they didn't like the fact my browser blocks all their cookies and isn't running on a Microsoft OS. Well, that's just too bad for them.
As far as your girlfriend's comments about Macs, I've noticed that a lot of people don't know where one piece of software begins and another ends when using a computer. For example, the people testing my software comeback with all sorts of crazy problem reports that are totally unrelated to my code, but with everything in the same context they just don't see the separation.
Thats the big problem MS has at the moment, they have sold their software to almost everybody so now they will have to look for ways to make these people pay once more to be able to keep making profit on windows and office.
One critical problem with mandatory upgrades is that Microsoft will be imposing risk on its customers. Trust me, software upgrades are risky. This is why many people are still using Windows NT or Solaris 2.6 as in the earlier post (i.e., their working business is more important than a software "upgrade").
Any company who knowingly forces its customers to make decisions they otherwise wouldn't make is simply a company that eventually loses its customers.
My reputation? What the hell are you talking about?
Maybe I'm weird, but I don't buy computer components in order to get into online pissing matches...
I was referring to professional reputation. This isn't a matter of getting bragging rights among friends, it is a matter of building systems upon which a business can establish its reputation.
SCSI can help in this regard. For example, it isn't difficult to build a RAID array of SCSI disks, which can take full advantage of the SCSI bus' bandwidth and UNIX's inherent support of concurrent I/O. Going further by adding a second SCSI controller to the host and a second RAID array can allow for a disk system with no single points of failure. You can maintain it on the fly, and the users will never know. With no disk-related downtime, the system might even pay for itself.
For desktops, SCSI has fewer applications, but I have seen cases where a user's productivity is limited by disk bandwidth. An array of two or three SCSI disks in the workstation could solve this problem, and, again, it might pay for itself.
Another point about SCSI is that it allows computers to have longer useful lives. Being able to tack on more devices is really useful in the long-haul, and good SCSI compatibility allows adding big drives to old computers (there is a 36GB Fast SCSI drive available for under $400--not bad for a computer built in 1992 that is still going strong). Fewer whole-server upgrades allows controlling costs on a larger scale, where there is less disruption of users' work and less bereaucratic overhead aquiring new hardware.
In conclusion, I really don't percieve SCSI to be very expensive.
Because Scsi drives can get throughput up to 360MB/sec in certain modes.
Correction: Arrays of SCSI drives can get that sort of throughput. Each drive itself is crippled to the speed of the platter, an average of 20 to 30 MB/sec or so.
Your post is flamebait.
The only advantage to ATA is monetary cost. If that is all you care about, then that is fine. If you value your time and your reputation, then use SCSI or Fibre Channel. It's pretty simple.
media are market-driven, not substance-or-content driven
This is very evident in the evolution of news media over the years. TV news brodcasts, such as local news and now even CNN Headline News, are little more than a glorified cross between Cops and Entertainment Tonight. Local newspapers often run poor national news and sorely biased local news dotting the pages amid huge advertisements and coupon sections. Now, I watch almost nothing on TV, I've stopped reading the newspaper, and I'm finding that public TV, public radio, and the WWW are the last refuges of worthwhile content.
With 30 billion planets to visit, Star Trek Enterprise will be the longest running series ever!
One other thing to look our for is HTML email (gack!) with loads an image from the spammer's site. There's usually some id tag sent with the image request so that the spammer gets confirmation on your email address just by reading the email.
Ugh. This is the absolutely worst aspect of HTML e-mail. Just by sending you an unsolicited HTML e-mail, a company can get your browser model and version, whether JavaScript and Java are enabled, your IP address and hostname, the operating system, and roughly where you are located in the world (sometimes down to the city). First, they know you check your e-mail; second, in one click, you just provided a wealth of marketing information; and, third, they can tailor their future e-mail to your system's security vulnerabilities.
Does anyone know of cases, where e-mail was used to install spyware on the client?
Whoever first concieved of HTML-based e-mail should cower in shame for lack of foresight. And all those who chose to implement HTML-based e-mail clients should also cower in shame. HTML-based e-mail is simply irresponsible. I'm simply tired of people who insist in making their e-mail pretty, while unknowingly sacrificing their privacy and security.
Oh please, when was the last time you actually bought a microsoft product?
Oh, about five years ago.
Congratulations, you have been trolled.
Thanks!
You're replying to the parent of my post? Perhaps I should have been clearer with the quoted text.
You license the hardware, you don't own it.
Tell that to the secondary market. Hardware is a physical tangible hard-to-copy thing that is owned. If I had an X-Box, I could take little wires and a soldering iron and do anything I want to its innards, risking only voiding the warranty. If I just wanted to use the CPU to keep a little cup of tea warm, I could, and you couldn't stop me.
If you ever tried that with my gaming console (yes, I'm a developer for a major game company, not MS), I'll send my lawyers after you so fast you'll be in jail getting assfucked by Bubba and his friends.
Whatever.
How long will all this last after one disgruntled engineer posts the critical Palladium specs to an anonymous forum (electronic or otherwise). Or, when foriegn intelligence agencies steal Palladium's secrets for their own nation's defense interests. Palladium's weaknesses will leak, it is just a matter of time before it is an expensive useless chip in our then-new computers.
There are huge companies who dispise Microsoft and have their own fabrication plants. Texas Instruments is a long-time partner with Sun Microsystems, for example. IBM has mixed feelings about Microsoft and has their own factories. Taiwan is home to many chip makers, and they are considering open source software more and more. Many of these companies currently build whole computer systems without even a hint of Microsoft software of x86 hardware (I'm using such a computer right now, and it's wonderfully useful).
World-wide, Microsoft's plans will probably be an up-hill battle the entire way.
Cartels like the diamond industry?
Learning about the diamond industry pissed me off enough that I will never buy another diamond. Microsoft is no different. If they get their way, where I can not choose how to develop and distribute software, then I will probably find a new career. It'll be hard to find one not dependent on Microsoft software, but I'll try.
This plan is so large that they must have thought this through.
Fallacy. There is no reason to believe they aren't just blinded by their egos and a huge marketing budget.
Uh, I don't know. The answer escapes me.
Ha ha.
What if you, a copyright holder, choose to use P2P as one of your distribution mechanisms. This is a legitimite use of P2P.
Also, you are paying for your internet connection. Would the RIAA be liable for real monetary damage done to you, since a DoS attack robs you of paid-for bandwidth that is really being used for legitimate purposes?
I'm bothered that this article focused so much on testing, when the real solution lies so much further up the development "pipeline".
Testing is very important, but it really is of limited use when the software has chronic design problems. If the nature of the software lends itself to hack-n-patch type fixes that build up into a painful thicket of code, then no real progress is made through testing.
Testing is intended to shake out the last few bugs in a well-thought out system, and the users delight in using the final product....<wakes up>...oh, I must have dozed off there...where was I?...oh...<hack-hack-hack punch-punch compile damn-it! hack-hack...>
Everything's always "only $10 a month!" or whatever, and before you know it - you're talking thousands a year in these non-essential services, chipping away at your income.
Companies simply love this. It's sad when I see friends who are still struggling to pay school loans fall for the "inexpensiveness" of cell phones, short-term financing plans, extended warranties, etc. These sorts of people are causing the rise of the easy-bankruptcy we'll-bail-you-out debt-consolodation industry. It's possible that the amount of average individual expenses and debt will reach a threshold where the subscription service business model crumbles due to no one being able to afford anything but their monthly triple-mortgage payment.
And in fact, the network admins here would likely tell you the *nix boxes they administer go down more often.
This means the people administering your UNIX boxes really don't know what they are doing. The Solaris kernel, for example, can easily go under high load for a year or more. There are mini-outages, such as restarting daemons, but these are peripheral to the operating system, which simply doesn't need restarting. This modularity of UNIX is one reason some people can claim very long uptimes--they maintain the running system. The lack of modularity in Windows has always disturbed me (what does a person do when the GUI goes flaky? In Solaris, I can just resurrect CDE with no one else noticing.)