In the current system, a person/company has some fixed amount of time (1 year? 6 months? I don't recall) to file a patent after the invention has been mentioned publicly. Some companies rely on this by shipping the product first, then worrying about filing the patent applications. "First to file" will likely delay many product releases, as the inventor will be required to get the patent application process started before release.
You joke, but that was one of the (many, excellent) points of Van Jacobson's recent net channels presentation. From his slides:
Networking gear has gotten fast enough (10Gb/s) and cheap enough ($10 for an 8port Gb switch) that it's changing from a communications technology to a backplane technology.
This reminds me of an experience I had many years ago at a camera shop. They had one of those new (at the time) self-service kiosks for printing digital pictures. I asked the sales drone about the media the kiosk accepted. The conversation went something like this:
Her: It takes compact flash and floppies, but you'll get the best results with compact flash. Me: Wha?? Her: Oh yes, you lose resolution when you copy your digital photos to floppies. Me: Uhh... no. They use the same filesystem (FAT), so the photos will be identical regardless of media. Her: (annoyed, rolling eyes) I'm not going to *argue* about this! Me: Well, it's best that you don't argue about it, because you're 100% wrong. Nevermind, I'll go elsewhere.
As several people have already pointed out, Microsoft is just responding to VMware's free-as-in-beer release of its VMPlayer and VMServer (beta 2) products. It will be interesting to see how VMware responds.
I hope part of their response includes a port of VMWorkstation, VMPlayer, and VMServer to *BSD, *especially* OSX. An Intel-based Mac running VMware would make a damn fine host for Windows/Linux/etc development and testing.
So far, the official responses on VMware's forums to "OSX hosting" requests has been luke warm at best. That said, it's hard to tell if they're really not that interested in doing the work, or if they're just "playing their cards close to their chest". Time will tell.
I agree with un1x10ser. I've used various versions of VMware Workstation starting with 3.2 up to the current (5.5.1) with many different Linux distributions (Redhat 8 through Fedora Core 5 beta, SuSe 8 - 10, plus various other distributions). I've never had a problem with Grub in a VM.
I leased a truck for 5 years. I had to pay property taxes on it. I tried the same "why should I pay property taxes on something I don't own" argument, but it fell on deaf ears.
I have the same problem with this layout as I have with every other alternative keyboard layout (including Dvorak): I want to be able to sit down at any computer, anywhere, and touch type. If I commit the COLMAK layout to memory, I'll have big problems the next time I go to a friend's house, an internet cafe, whatever.
Spend one day giving them (anonymously, if you wish) to people you see every day who make your life easier and who could really use the money. The "single mother" waitress who keeps your coffee cup filled at your favorite breakfast place. The anonymous immigrant who cleans the bathroom at your office. The teenage student who sacks your groceries when not studying for an exam. The elderly person working at the fast-food joint because they cannot survive on their retirement benefits.
These are the faceless people who keep our society running. We all depend on them.
No, you cannot deduct these donations from your taxes, but fuck it. Do it anyway.
Hungarian Notation is the tactical nuclear weapon of source code obfuscation techniques; use it!
Hungarian is a great idea, but it doesn't go far enough. Function names should encode the return type and the type of each parameter. For example, a simple:
int main( int argc, char * argv[] );
Becomes the much more readable and instantly understandable:
int imaincapsz( int cArgs, char * apszArgs[] );
Of course, this is C, and arguments are typically pushed on the stack in reverse order. If you use a low-level symbolic debugger (not one of these new-fangled IDEs), you can save yourself a few debugging headaches by reversing the order of argument types. In other words:
int imainapszc( int cArgs, char * apszArgs[] );
Now, when you see imainapszc in a stack trace, it will be trivial to decode the functions parameters.
Slashdot Mirror
In the current system, a person/company has some fixed amount of time (1 year? 6 months? I don't recall) to file a patent after the invention has been mentioned publicly. Some companies rely on this by shipping the product first, then worrying about filing the patent applications. "First to file" will likely delay many product releases, as the inventor will be required to get the patent application process started before release.
You joke, but that was one of the (many, excellent) points of Van Jacobson's recent net channels presentation. From his slides:
GPE = Guinea Pig Equivalent
When I was a kid, they still roamed the Earth.
This reminds me of an experience I had many years ago at a camera shop. They had one of those new (at the time) self-service kiosks for printing digital pictures. I asked the sales drone about the media the kiosk accepted. The conversation went something like this:
Her: It takes compact flash and floppies, but you'll get the best results with compact flash.
Me: Wha??
Her: Oh yes, you lose resolution when you copy your digital photos to floppies.
Me: Uhh... no. They use the same filesystem (FAT), so the photos will be identical regardless of media.
Her: (annoyed, rolling eyes) I'm not going to *argue* about this!
Me: Well, it's best that you don't argue about it, because you're 100% wrong. Nevermind, I'll go elsewhere.
No, in Longhorn, it's called COW-tipping(tm).
As several people have already pointed out, Microsoft is just responding to VMware's free-as-in-beer release of its VMPlayer and VMServer (beta 2) products. It will be interesting to see how VMware responds.
I hope part of their response includes a port of VMWorkstation, VMPlayer, and VMServer to *BSD, *especially* OSX. An Intel-based Mac running VMware would make a damn fine host for Windows/Linux/etc development and testing.
So far, the official responses on VMware's forums to "OSX hosting" requests has been luke warm at best. That said, it's hard to tell if they're really not that interested in doing the work, or if they're just "playing their cards close to their chest". Time will tell.
Random train of thought...
2 bits = 25 cents
4 bits = 50 cents
8 bits = byte
4 bits = nibble
Therefore:
50 Cent = Nibble
I think everyone should refer to the rapper 50 Cent as "Nibble" from now on.
Sure. That'll happen.
I agree with un1x10ser. I've used various versions of VMware Workstation starting with 3.2 up to the current (5.5.1) with many different Linux distributions (Redhat 8 through Fedora Core 5 beta, SuSe 8 - 10, plus various other distributions). I've never had a problem with Grub in a VM.
I leased a truck for 5 years. I had to pay property taxes on it. I tried the same "why should I pay property taxes on something I don't own" argument, but it fell on deaf ears.
http://www.doonesbury.com/strip/dailydose/index.ht ml?uc_full_date=20051218
I have the same problem with this layout as I have with every other alternative keyboard layout (including Dvorak): I want to be able to sit down at any computer, anywhere, and touch type. If I commit the COLMAK layout to memory, I'll have big problems the next time I go to a friend's house, an internet cafe, whatever.
Not worth the trouble.
These are the faceless people who keep our society running. We all depend on them.
No, you cannot deduct these donations from your taxes, but fuck it. Do it anyway.
Warrants are so "last century".
The first time I read that quote, I thought "most people don't know what malignant melanoma is, so why should they care about it?"
FWIW, it works fine with Mozilla 1.7.12.
Looks like they'll try again tomorrow at 1:00pm EST. I wish them the best of luck.
Of course not! The RIAA has already signed the iceberg to an exclusive 10-year contract.
Get a rope.
Hungarian is a great idea, but it doesn't go far enough. Function names should encode the return type and the type of each parameter. For example, a simple:
Becomes the much more readable and instantly understandable:
Of course, this is C, and arguments are typically pushed on the stack in reverse order. If you use a low-level symbolic debugger (not one of these new-fangled IDEs), you can save yourself a few debugging headaches by reversing the order of argument types. In other words:
Now, when you see imainapszc in a stack trace, it will be trivial to decode the functions parameters.
</sarcasm>
Q: What's purple and commutes?
A: An Abelian grape.
And it rocks absolutely, too.
Would you run anything from a CD you received on "trick-or-treat" night?