I think Microsoft has a more hidden purpose to it's "shared source" initiative. Everyone who gets to look at their source code undoubtedly must sign an NDA saying that they'll never reveal the source to the public, nor use the code or ideas for any other project.
Every developer that agrees to it is now "tainted" when it comes to working on an open source software project. MS now has grounds for going after individual developers working on open source software, and has an indimidation tool for those just thinking about it. And as we know with legal stuff, it doesn't necessarily have to hold up in a trial, legal arguments just have to sound plausible enough for threats.
Ideally (from MS's point of view), everyone in the world would sign such NDAs, and thereby stifle open source software development.
Other roadblocks to PVRs might save the day for networks and advertisers. For example, people are beginning to reevaluate spending high annual fees for unimportant services. Do you want to spend $120 a year to operate a TiVo unit just to skip a few commercials?
The value of my time varies depending on what I do, but the top end is $120 / hour when I'm doing consulting work.
With TiVo, I can skip the 15 minutes of commercials in a typical 1 hour program. If watching commercials is as onerous as consulting work, then I've recouped my investment in just 4 hours of use.
Check out that cost-benefit ratio Mr. PC Expert Dvorak.
Did anyone else besides me first think this was about that '70s era Japanese animation about the five kids that flew in a blue spaceship which could turn into a phoenix?
Yeah, I read that paragraph a couple times trying to make sense of it. The first two sentences make sense together. Then after that it can best be described as incoherent.
Maybe his pet monkey got to his computer while he was taking a coffee break...
What you say may be true, but you're not really listing all the options. You can also:
A) Restructure the application so that it doesn't need rental software.
B) Restructure the application so that you can write your own stuff and not infringe on the hypothetical patents.
C) Restructure the business so that it doesn't need the application, or needs a different application.
D) Move the business into some other field.
All of these have been done. Me I'm kinda the spiteful type. If I believe someone is extorting me (especially in regards to software) then I'll go way out of my way to stop the situation. Not everyone is like that though.
Hee hee.
Of course, what would be really funny is to have religious websites required to link to pr0n sites for the "opposing viewpoint".
And the MPAA would have to put links to DeCSS sites... I'm sure that would go over real well.
I just love it when some professor (who's supposed to be smarter than average) doesn't stop to ask someone else if his ideas are completely stupid and unworkable.
Classic, totally classic.
Free speech is a precious thing. People have made a good case for limiting it in certain situations, like yelling "fire" in a crowded theater. Generally though, any suggestions to further limit speech should be view with great suspicion.
Well, maybe I am a dumbass, but I didn't know that. I always thought I'd have to find the specific download package, which is not that easy to find from their website.
What's funny about the situation with washirv (the original poster) is that, OK, he's got a copy of the public key. But what good is that going to do him? Without the originally generated secret key, the server can't verify itself to incoming SSL connections.
The information he got from Verisign was almost useless, and his company will have to shell out another $500 for a new certificate (which as someone else pointed out isn't a bad idea anyway).
For example, I can put security.debian.org in my/etc/apt/sources.list, and set cron to run apt-get upgrade nightly. This will automagically install any security patches with no user intervention.
I hope you only do that for your desktop machine, and not any production servers.
An example for why this is a bad idea: I tried upgrading Zope today in response to a security alert for Debian. I install the package as normal, however, when I try to access the web server, it asks for a password, and doesn't accept any valid ones. This is for the front page!
If that had happened during the night, our server would have been unavailable for hours. As it is, I just re-installed the old version, so our downtime was limited to just a few minutes.
I usually try to test out upgrades first, but since the last Zope upgrade went very smoothly, I started to get cocky (and thereby less careful).
Well, there's at least one project that has a better possibility of reducing launch costs to LEO. Check out Scorpius by Microcosm.
Sure, it's not as sexy as new, untested wizz-bang technologies (like aerospike engines, composite H2 tanks, scramjets, etc.). It pretty much follows LEO on the Cheap by Lt. Col John R. London. After reading this book, you'll start to really wonder why projects like the X-33 and X-34 were funded in the first place. You'll start to see the massive politics behind the space industry.
There was only one show of the last season that didn't suck: "The Last Flight of Starbuck".
Stuck on a deserted planet, with only some destroyed Cylons for company... I think I cried at the end (gimme a break, I was like 12 at the time).
As for the rest of the season, the messiah kid (forshadowing of Wesley Crusher), the invisibility schtick, the two guys as replacements of Starbuck and Apollo... bah!
I also don't remember why the Glactica crew didn't just land and say "Honey, I'm home!" What was the reason for hiding?
There's one other significant difference between monochrome and color monitors.
If you look at two monitors that were produced, say, 10 years ago, you'll notice some significant differences. The color VGA monitor, even at the default 640x480 resolution, still has somewhat fuzzy text. In contrast (pun intended), the monochrome monitor has very sharp text.
This is because of the dot pitch used on those older monitors, and the whole issue of having a screen mask (dots commonly, or wires with a Trinitron display). Because monochrome monitors didn't have a mask, there's nothing that diffuses the electron beam striking the phosphor. There are also no focus and alignment issues with multiple electron guns.
Back at school, we had 19" monochrome monitors for our Xterminals, they rocked! Too bad you can't buy that kind of stuff anymore.
Of course, now I'm completely dependent on my LCD screens (SGI 1600SWs of course). I am so spoiled, I don't know how I managed without colorization of program code and color-ls.
Um, sorry, but I think that socialism in the U.S.A. started with the New Deal back in the 1930s.
It could also be argued that creeping socialism started with the creation of the federal income tax, which, if I recall correctly, was around 1900.
Republicanism: We'll tell you what to do because you're morally weak, but we won't take care of you.
Democratism: We won't tell you what to do, but we'll take care of you anyway because you don't know any better.
Communism: We'll tell you what to do because you're stupid, and we take care of you... minimally. Don't cause problems, though, or we'll have to kill you for the good of society.
Libertarianism: We won't tell you what to do, and we won't take care of you either. You'd better buy a gun to defend yourself.
Yah, that's more what I had in mind, I should have made myself clear.
You'd definitely need to keep the core interpreter written in C plus assembly. All the device drivers would be Python extensions written similarly.
But at the user level, all you'd have is Python code (source and/or compiled). There'd be no C library or other stuff.
Hopefully, it would be a fun development environment, that would make it easy to plug together Python components into applications. Kinda like a Smalltalk system, where everything lives inside one big virtual machine. Context switches wouldn't be a problem, because the interpreter enforces security at a higher level.
Dunno how fast it would run, though the Squeak guys seem to be somewhat satisfied with their project in that regard.
DVI and OpenLDI are not directly compatable. You can get the multi-link adapter from SGI that will convert DVI to OpenLDI. However, you can't run the wide resolution (1600x1024) because of limitations with DVI. I believe the best you can get is 1280x1024, which loses the advantage of the SGI 1600SW in the first place.
There's been some previous discussion of the pros and cons of digital flat panels on Slashdot, try searching.
If you can live with a system that has a video card designed by a company that has gone bankrupt (so no manufacturer driver updates), then the SGI + RevIV is a good setup. I use it every day. Be warned that the RevIV is only good for 2D applications, 3D ain't so hot.
Hmmm... For good efficiency, you'd need a quarter wave dipole antenna. If you want the antenna to fit inside the longest dimension, you're now talking 10GHz or so. Extremely high frequency microwave radar and such. If you wanna flood your workplace with that, that's fine, but don't do it while I'm around.
Yup. That's part of it. Another part is thermal stress, which is one of the big killers for electromechanical systems.
They are made of different materials, which expand and contract at different rates as the heat and cool down. This causes microscopic flexing every time the device is power cycled. Some components, that have wide engineering margins, can handle lots of power cycles. Others can't. Which one do you have? Only one way to find out, and you don't want to...
This combines the best of both worlds. This also means that while it's easy to corrupt your database with a single bug in your code, you can always re-build it from the on-disk messages.
Yes, it is great until the two get out of sync. If you can limit access to the raw filesystem, then that'll eliminate most of the problems, and most of the advantages.
Besides, databases are a lot better (these days) at storing large hunks of arbitrary data, so I'd just stick everything in the database.
That or use a future version of reiserfs, which could give you a database-like view of your filesystem.
While it's true that it is usually 10 bits per byte on asynchronous links (1 start bit, 8 data bits, 1 stop bit, no parity), that's not true for things like leased lines. They use synchronous serial communication, so it is 8 bits per byte. IIRC, ISDN is also synchronous, but I don't know about some of the newer signaling schemes.
1) Confusing configuration. Get rid of/etc and replace it with something sane.
Oh man I hope you're kidding. It's having all the configuration files for the system in one directory that helps keep things sane when doing a lot of system administration.
And having them be plain text files is just as important. So easy to copy configurations and edit, without special tools. Try to adminster more than a few NT machines on a network (even with remote control software such as VNC or PCAnywhere) for a month, and come back and tell me you don't like text configuration files.
Make the system cohesive. Get back to the UNIX roots. You know how UNIX treats *everything* as a file?
From a programmer's point of view, I agree with you (check out Plan9 from Bell Labs sometime, you'll probably like it).
But from a user's point of view, that is irrellavant.
This is what the Network Appliance boxen do to speed NFS writes.
All NFS write transactions are commited to NVRAM first, so that they can be acknowledged. Then the writes to disk are sorted and blasted out. Very efficient, very fast.
It is this NVRAM (as well as using a modified RAID-4 on top of the WAFL filesystem) that makes a NetApp much faster (yet still safer) than most other NFS servers. I've often thought about creating just such an NVRAM board for a PC, so that I could do the same thing with my Linux fileservers.
Note that the NetApp implementation caches NFS requests, not filesystem-level data. Say I'm changing 1 byte in a block. If I buffer filesystem data, I have to cache the whole block. If I'm buffering the NFS request, it'll be much smaller.
Buffering (in NVRAM) the log data might work well for something like ext3.
I was talking to a guy that had the multi-link adapter, and was trying to use one of those "superwide savvy" cards that SGI lists.
Though he had a card with a DVI connector (-D or -I... don't remember), he was not able to run the SGI at 1600x1024 through the DVI, only through the analog. From what I recall of the standard, you're only going to get 1280x1024 through DVI. (Correct me if I'm wrong)
Ya see, there really was a good reason why SGI decided to go with OpenLDI instead of DVI or that other VESA DFP standard. Two words: bandwidth limitation.
I'm just glad to be able to still run the old Number Nine RevIV card under XFree86. Those of us that have them will just have to hang tight for another year or so until displays like that expensive monster panel from IBM come down in price.
So, even though I love my SGI's (have two, no dead pixels), I'm not sure I can really recommend anyone buying one for now, unless you're fully aware of the limitations.
Yup. You can look at the Space Shuttle's next launch for confirmation. The main engines (on the orbiter itself) burn H2 +O2, and it is almost invisible.
Slashdot Mirror
An evil thought just occured to me...
I think Microsoft has a more hidden purpose to it's "shared source" initiative. Everyone who gets to look at their source code undoubtedly must sign an NDA saying that they'll never reveal the source to the public, nor use the code or ideas for any other project.
Every developer that agrees to it is now "tainted" when it comes to working on an open source software project. MS now has grounds for going after individual developers working on open source software, and has an indimidation tool for those just thinking about it. And as we know with legal stuff, it doesn't necessarily have to hold up in a trial, legal arguments just have to sound plausible enough for threats.
Ideally (from MS's point of view), everyone in the world would sign such NDAs, and thereby stifle open source software development.
Ol' John sure is a riot:
Other roadblocks to PVRs might save the day for networks and advertisers. For example, people are beginning to reevaluate spending high annual fees for unimportant services. Do you want to spend $120 a year to operate a TiVo unit just to skip a few commercials?
The value of my time varies depending on what I do, but the top end is $120 / hour when I'm doing consulting work.
With TiVo, I can skip the 15 minutes of commercials in a typical 1 hour program. If watching commercials is as onerous as consulting work, then I've recouped my investment in just 4 hours of use.
Check out that cost-benefit ratio Mr. PC Expert Dvorak.
Did anyone else besides me first think this was about that '70s era Japanese animation about the five kids that flew in a blue spaceship which could turn into a phoenix?
Bummer. I wish they'd bring that back.
Yeah, I read that paragraph a couple times trying to make sense of it. The first two sentences make sense together. Then after that it can best be described as incoherent.
Maybe his pet monkey got to his computer while he was taking a coffee break...
What you say may be true, but you're not really listing all the options. You can also:
A) Restructure the application so that it doesn't need rental software.
B) Restructure the application so that you can write your own stuff and not infringe on the hypothetical patents.
C) Restructure the business so that it doesn't need the application, or needs a different application.
D) Move the business into some other field.
All of these have been done. Me I'm kinda the spiteful type. If I believe someone is extorting me (especially in regards to software) then I'll go way out of my way to stop the situation. Not everyone is like that though.
Hee hee. Of course, what would be really funny is to have religious websites required to link to pr0n sites for the "opposing viewpoint".
And the MPAA would have to put links to DeCSS sites... I'm sure that would go over real well.
I just love it when some professor (who's supposed to be smarter than average) doesn't stop to ask someone else if his ideas are completely stupid and unworkable.
Classic, totally classic.
Free speech is a precious thing. People have made a good case for limiting it in certain situations, like yelling "fire" in a crowded theater. Generally though, any suggestions to further limit speech should be view with great suspicion.
Well, maybe I am a dumbass, but I didn't know that. I always thought I'd have to find the specific download package, which is not that easy to find from their website.
What's funny about the situation with washirv (the original poster) is that, OK, he's got a copy of the public key. But what good is that going to do him? Without the originally generated secret key, the server can't verify itself to incoming SSL connections.
The information he got from Verisign was almost useless, and his company will have to shell out another $500 for a new certificate (which as someone else pointed out isn't a bad idea anyway).
For example, I can put security.debian.org in my /etc/apt/sources.list, and set cron to run apt-get upgrade nightly. This will automagically install any security patches with no user intervention.
I hope you only do that for your desktop machine, and not any production servers.
An example for why this is a bad idea: I tried upgrading Zope today in response to a security alert for Debian. I install the package as normal, however, when I try to access the web server, it asks for a password, and doesn't accept any valid ones. This is for the front page!
If that had happened during the night, our server would have been unavailable for hours. As it is, I just re-installed the old version, so our downtime was limited to just a few minutes.
I usually try to test out upgrades first, but since the last Zope upgrade went very smoothly, I started to get cocky (and thereby less careful).
Well, there's at least one project that has a better possibility of reducing launch costs to LEO. Check out Scorpius by Microcosm.
Sure, it's not as sexy as new, untested wizz-bang technologies (like aerospike engines, composite H2 tanks, scramjets, etc.). It pretty much follows LEO on the Cheap by Lt. Col John R. London. After reading this book, you'll start to really wonder why projects like the X-33 and X-34 were funded in the first place. You'll start to see the massive politics behind the space industry.
Hrrm... That would explain why these major media companies (like Fox and Sony) have been buying up actual media outlets left and right.
You'd think in that case that the RIAA would be all over electronic distribution as a way of sidestepping the middlemen.
There was only one show of the last season that didn't suck: "The Last Flight of Starbuck".
Stuck on a deserted planet, with only some destroyed Cylons for company... I think I cried at the end (gimme a break, I was like 12 at the time).
As for the rest of the season, the messiah kid (forshadowing of Wesley Crusher), the invisibility schtick, the two guys as replacements of Starbuck and Apollo... bah!
I also don't remember why the Glactica crew didn't just land and say "Honey, I'm home!" What was the reason for hiding?
There's one other significant difference between monochrome and color monitors.
If you look at two monitors that were produced, say, 10 years ago, you'll notice some significant differences. The color VGA monitor, even at the default 640x480 resolution, still has somewhat fuzzy text. In contrast (pun intended), the monochrome monitor has very sharp text.
This is because of the dot pitch used on those older monitors, and the whole issue of having a screen mask (dots commonly, or wires with a Trinitron display). Because monochrome monitors didn't have a mask, there's nothing that diffuses the electron beam striking the phosphor. There are also no focus and alignment issues with multiple electron guns.
Back at school, we had 19" monochrome monitors for our Xterminals, they rocked! Too bad you can't buy that kind of stuff anymore.
Of course, now I'm completely dependent on my LCD screens (SGI 1600SWs of course). I am so spoiled, I don't know how I managed without colorization of program code and color-ls.
Um, sorry, but I think that socialism in the U.S.A. started with the New Deal back in the 1930s.
It could also be argued that creeping socialism started with the creation of the federal income tax, which, if I recall correctly, was around 1900.
Republicanism: We'll tell you what to do because you're morally weak, but we won't take care of you.
Democratism: We won't tell you what to do, but we'll take care of you anyway because you don't know any better.
Communism: We'll tell you what to do because you're stupid, and we take care of you... minimally. Don't cause problems, though, or we'll have to kill you for the good of society.
Libertarianism: We won't tell you what to do, and we won't take care of you either. You'd better buy a gun to defend yourself.
Take your pick!
Yah, that's more what I had in mind, I should have made myself clear.
You'd definitely need to keep the core interpreter written in C plus assembly. All the device drivers would be Python extensions written similarly.
But at the user level, all you'd have is Python code (source and/or compiled). There'd be no C library or other stuff.
Hopefully, it would be a fun development environment, that would make it easy to plug together Python components into applications. Kinda like a Smalltalk system, where everything lives inside one big virtual machine. Context switches wouldn't be a problem, because the interpreter enforces security at a higher level.
Dunno how fast it would run, though the Squeak guys seem to be somewhat satisfied with their project in that regard.
One of these years...
Actually, I really would like to write a Python-based OS.
The only problem is I have this day job... I could get by without money, new clothes or a car, but not an Internet connection!
DVI and OpenLDI are not directly compatable. You can get the multi-link adapter from SGI that will convert DVI to OpenLDI. However, you can't run the wide resolution (1600x1024) because of limitations with DVI. I believe the best you can get is 1280x1024, which loses the advantage of the SGI 1600SW in the first place.
There's been some previous discussion of the pros and cons of digital flat panels on Slashdot, try searching.
If you can live with a system that has a video card designed by a company that has gone bankrupt (so no manufacturer driver updates), then the SGI + RevIV is a good setup. I use it every day. Be warned that the RevIV is only good for 2D applications, 3D ain't so hot.
Hmmm... For good efficiency, you'd need a quarter wave dipole antenna. If you want the antenna to fit inside the longest dimension, you're now talking 10GHz or so. Extremely high frequency microwave radar and such. If you wanna flood your workplace with that, that's fine, but don't do it while I'm around.
Yup. That's part of it. Another part is thermal stress, which is one of the big killers for electromechanical systems.
They are made of different materials, which expand and contract at different rates as the heat and cool down. This causes microscopic flexing every time the device is power cycled. Some components, that have wide engineering margins, can handle lots of power cycles. Others can't. Which one do you have? Only one way to find out, and you don't want to...
This combines the best of both worlds. This also means that while it's easy to corrupt your database with a single bug in your code, you can always re-build it from the on-disk messages.
Yes, it is great until the two get out of sync. If you can limit access to the raw filesystem, then that'll eliminate most of the problems, and most of the advantages.
Besides, databases are a lot better (these days) at storing large hunks of arbitrary data, so I'd just stick everything in the database.
That or use a future version of reiserfs, which could give you a database-like view of your filesystem.
While it's true that it is usually 10 bits per byte on asynchronous links (1 start bit, 8 data bits, 1 stop bit, no parity), that's not true for things like leased lines. They use synchronous serial communication, so it is 8 bits per byte. IIRC, ISDN is also synchronous, but I don't know about some of the newer signaling schemes.
Just tryin' to keep the math straight.
1) Confusing configuration. Get rid of /etc and replace it with something sane.
Oh man I hope you're kidding. It's having all the configuration files for the system in one directory that helps keep things sane when doing a lot of system administration.
And having them be plain text files is just as important. So easy to copy configurations and edit, without special tools. Try to adminster more than a few NT machines on a network (even with remote control software such as VNC or PCAnywhere) for a month, and come back and tell me you don't like text configuration files.
Make the system cohesive. Get back to the UNIX roots. You know how UNIX treats *everything* as a file?
From a programmer's point of view, I agree with you (check out Plan9 from Bell Labs sometime, you'll probably like it).
But from a user's point of view, that is irrellavant.
This is what the Network Appliance boxen do to speed NFS writes.
All NFS write transactions are commited to NVRAM first, so that they can be acknowledged. Then the writes to disk are sorted and blasted out. Very efficient, very fast.
It is this NVRAM (as well as using a modified RAID-4 on top of the WAFL filesystem) that makes a NetApp much faster (yet still safer) than most other NFS servers. I've often thought about creating just such an NVRAM board for a PC, so that I could do the same thing with my Linux fileservers.
Note that the NetApp implementation caches NFS requests, not filesystem-level data. Say I'm changing 1 byte in a block. If I buffer filesystem data, I have to cache the whole block. If I'm buffering the NFS request, it'll be much smaller.
Buffering (in NVRAM) the log data might work well for something like ext3.
I was talking to a guy that had the multi-link adapter, and was trying to use one of those "superwide savvy" cards that SGI lists.
Though he had a card with a DVI connector (-D or -I... don't remember), he was not able to run the SGI at 1600x1024 through the DVI, only through the analog. From what I recall of the standard, you're only going to get 1280x1024 through DVI. (Correct me if I'm wrong)
Ya see, there really was a good reason why SGI decided to go with OpenLDI instead of DVI or that other VESA DFP standard. Two words: bandwidth limitation.
I'm just glad to be able to still run the old Number Nine RevIV card under XFree86. Those of us that have them will just have to hang tight for another year or so until displays like that expensive monster panel from IBM come down in price.
So, even though I love my SGI's (have two, no dead pixels), I'm not sure I can really recommend anyone buying one for now, unless you're fully aware of the limitations.
Yup. You can look at the Space Shuttle's next launch for confirmation. The main engines (on the orbiter itself) burn H2 +O2, and it is almost invisible.