If the wiring adheres to local codes in the U.S., then it does not matter if a pro has done it or not...
Well, at least in Farifax county, Virginia, not only do you have to be a pro, you have to have a county-issued permit to do cabling.
Since house wall spaces do not play a part in the HVAC system, you can use PVC.
Also not completely true. There is another category of calbe called "riser", which is fire resistant, but can still be toxic. In wall spaces and such chances are the regulations require it to be at least riser rated.
I think the idea that there is no future for software development is utterly wrong. And software is not tulips or spices (how has the tulip evolved over the past 200 years?).
We have not even scratched the surface of what software can do for us, and becuase of some self-proclaimed "entrepreneurs" who pissed away VC money on corporate yachts, we are now depressed, we're ready to discard software business alltogether.
I predict that software will come back as business, and much sooner than we think. And no offshore programming or GNU is going to be a factor in it - there is going to be thriving market for people capable of developing solutions for businesses, and it's going to be primarily in the US and Europe (and not in India or China).
From the article: I propose that we as a community insist that all distros make the default screensaver be one that randomly displays a different detailed credit for one of the authors of Linux software every 60 seconds.
Erm.. Is Python or Perl or Apache or Emacs - "Linux software"? What about FreeBSD or OpenBSD - that's hardly "Linux software"...
I'm surprised to see someone as knowledgeable as Riser make such a blunder - or is it intentional?
Wiring is the kind of thing best left to people who deal with it every day. I'd get a cabling company to give you a quote on running CAT5 to every unit and instaling a patch panel in some closet. They'll have the right tools to pull the cable, will be able to test it and will be responsible for fixing if anything isn't working.
I really wouldn't recommend pulling the cable yourself unless you really know what you're doing. BTW, depending on where it's pulled, it might need to be plenum or riser rated, and there may additional fire/code regulations for your area. You may need a license for cabling - but the cabling people would know all that.
It's not like wiretapping is not possible right now, it's just that there is no standard way of doing it. I assure you, every reasonable ISP has taps in place, if not for the government, then for its own internal use - to be able to diagnose problems, track DOS's, etc.
The solution is what it always has been - you are responsible for your privacy. Use encryption - SSH, PGP, etc to protect yourself from eavesdropping.
Now if the government passed a law forbidding encryption, that would be a whole different story.
Hmm.. Neat. Is this documented somewhere? I haven't been reading hubs list lately. I'm curious how it determines that a server is loaded, since aside from load, the network location of the client plays a significant role in how performance is perceived by the client.
Don't know where you got this from. Ftp2.freebsd.org resolves (and has for the past couple of years) to 130.94.149.162, which is located at Verio. It's got a GE connection to their backbone and its a beefy machine, which is why it is so fast.
I've thought for a long time that in this sea of OSS projects and various hype surrounding them (which is only bound to get bigger and messier), the experience and knowledge of which ones are good is a very valuable skill. Rather than paying a lot for software, companies should (and many do) pay for people who know software.
It used to be that you start writing a program from scratch (I'm tallking commercial/IT here), but in the last few years it turned into looking for suitable OSS to start with (e.g. lets use Apache as the web server, Postgres as db, now I need to look for some graphics libs, oh and is there an OSS spell-checker out there?... Let's see what license is this under... Looks like this this piece of functionality I will need to write myself... etc..). And the success of the end result seems to be largely depndend on how thorough this decision-making process is. In fact a good real-life example of this is the Safari browser - someone did their homework and chose KDE's engine even thoug Gecko is widely known as *the* engine out there.
As a sidenote, this is also one of the things that might differentiate experienced old-timers from young new-comers. You may be a brilliant 20-year old programmer, but the 35-year old dude in the next office can put a better solution together faster because he knows where to go to look for good source code, what's hype and what really works.
I've worked on AS/400's, albeit a while ago - early/mid 90's. As a matter of fact my first paid programming experience was in RPG IV lanugage. I would not say it was mind-numbing - the documentation provided by IBM was fantastic, and the AS/400 administration is very much automatable using their CL language. The machine was incredibly stable and even most hardware upgrades did not require a reboot (IPL). And even though it doesn't resemble anything UNIX or most other systems I know, and all files are database tables (no such thing as "file" really), I sould say it was very consistent and even intuitive once you get a hang of it.
Having said that, I wouldn't want to do any AS/400 work now - but probably more because I don't like a typical AS/400 user (brick and mortar businesses not focused on technology) than the system itself.
I can't say I'm thrilled with Red Hat's distribution - it's hardly innovative, arcane, their RPM system leaves a lot to be desired and the hole company IMHO has nothing but the exploitation of Linux popularity going for it. Somehow they've managed to become synonymous with Linux (someone already mentioned Linux 9) which I think only does the community a disservice while RH's owners are getting richer. Slackware or Debian are far superior to RH, though I've become so fed up with Linux hype lately that I primarily use FreeBSD.
So I don't know if it's just me, but this is hardly front-page news.
I doubt you can patent such a fundumental principal - how do you think transformers work? Also my Braun electric toothbrush charges its battery wirelessly and I've had it for years.
I tried Apple's X11, but couldn't figure out how to run it full-screen with WindowMaker. (I wouldn't need this had Apple provided virtual desktops that could be switched as fast as WindowMaker does it).... So I gave up and went back to XDarwin.
Before XML there was (and still is) RFC822 which describes how headers are formatted in e-mail, HTTP and a slew of other protocols.
I've been down the route where I tried to use XML where something as simple as "key: value" would do, and before I knew it, my program became a bloat relying on third-party XML libs, the config files were only marginally human-readable and a lot of time was wasted thinking about virtues of DOM vs SAX. In the end I learned that using XML for sake of XML isn't worth it.
I think XML is OK if used appropriately - for example I think XML is perfect for something like storing word processing documents. But the idea that every config file and every bit of network traffic should be XML is stupid IMHO.
Anything else is an art. In my experience best contributors range from those that come completely out of the blue to people you know professionally in person.
Generally, lack of contributors shouldn't be a concern - if it is, then the scope of your project is too big. Keep your scope narrow and goals simple (work on the assumption that you will never get any contributors other than yourself), focus on perfection rather than size or time, and your project will be successful.
Tarkovsky is in the same league with Fellini and Kurosawa - why even bother comparing? 30 years from now I doubt anyone will even know that Soderbergh's Solaris existed.
I use XDarwin full screen with a black background. For whatever reason now the mouse pointer is now always black, which makes it invisible on black background. What a bummer...
Interesting... Would you care to elaborate on what application was this protocol intended to solve, i.e. was it the same kind of stuff as UPnP, or something different altogether? Also, why ICMP? And lastly - was UPnP available at the time of writing and what you think of UPnP in general, if anything? Thanks!
HTTP over UDP is a hack, and the RFC is written by someone at Microsoft.
IMHO, the Rendezvous method of using DNS SRV records for service discovery is much better, because DNS has been around for ages, and we know it works. Only time will show what sorts of problems will appear with UPnP's HTTPU reliance.
First one that comes to mind is reliability - UDP doesn't perform well on busy networks (and its not unlikely that the network over which these packets will travel will indeed be busy - imagine 802.11b maxed out while a video is being streamed over it). UPnP requires devices to resend their messages after an interval because of UDP's unreliable nature - but if I hit "play" on my remote control, I want the thing to play instantly, not a few seconds later.
Another one is scale - large amounts of data can't be sent over UDP. As it looks right now, the packets shouldn't be large, but we don't know what the future holds - as devices get more complicated and support more features, the UPnP service description messages could get fairly large.
Of course the above is a moot point given that there are no other protocols that do this....
Yes, but Rendezvous is merely a protocol for automatic network configuration and discovering services (and a really good one too). UPnP goes a step further by providing means for device interaction.
The point is that if I were to write a piece of UPnP capable software, when the UPnP DVD players (or whatever) come out, chances are this software will be able to work with it. With Rendezvous - the device-specific interface is undefined.
Jini seems to be a Java-specific thing, so I'm not really interested in it.
Slashdot Mirror
I believe it surfaced a while back on /. - can't find any links at the moment, but AFAIK the entire Google index is stored in RAM.
Well, at least in Farifax county, Virginia, not only do you have to be a pro, you have to have a county-issued permit to do cabling.
Since house wall spaces do not play a part in the HVAC system, you can use PVC.
Also not completely true. There is another category of calbe called "riser", which is fire resistant, but can still be toxic. In wall spaces and such chances are the regulations require it to be at least riser rated.
I think the idea that there is no future for software development is utterly wrong. And software is not tulips or spices (how has the tulip evolved over the past 200 years?).
:-)
We have not even scratched the surface of what software can do for us, and becuase of some self-proclaimed "entrepreneurs" who pissed away VC money on corporate yachts, we are now depressed, we're ready to discard software business alltogether.
I predict that software will come back as business, and much sooner than we think. And no offshore programming or GNU is going to be a factor in it - there is going to be thriving market for people capable of developing solutions for businesses, and it's going to be primarily in the US and Europe (and not in India or China).
Stop wining please!
From the article: I propose that we as a community insist that all distros make the default screensaver be one that randomly displays a different detailed credit for one of the authors of Linux software every 60 seconds.
Erm.. Is Python or Perl or Apache or Emacs - "Linux software"? What about FreeBSD or OpenBSD - that's hardly "Linux software"...
I'm surprised to see someone as knowledgeable as Riser make such a blunder - or is it intentional?
Wiring is the kind of thing best left to people who deal with it every day. I'd get a cabling company to give you a quote on running CAT5 to every unit and instaling a patch panel in some closet. They'll have the right tools to pull the cable, will be able to test it and will be responsible for fixing if anything isn't working.
I really wouldn't recommend pulling the cable yourself unless you really know what you're doing. BTW, depending on where it's pulled, it might need to be plenum or riser rated, and there may additional fire/code regulations for your area. You may need a license for cabling - but the cabling people would know all that.
It's not like wiretapping is not possible right now, it's just that there is no standard way of doing it. I assure you, every reasonable ISP has taps in place, if not for the government, then for its own internal use - to be able to diagnose problems, track DOS's, etc.
The solution is what it always has been - you are responsible for your privacy. Use encryption - SSH, PGP, etc to protect yourself from eavesdropping.
Now if the government passed a law forbidding encryption, that would be a whole different story.
Hmm.. Neat. Is this documented somewhere? I haven't been reading hubs list lately. I'm curious how it determines that a server is loaded, since aside from load, the network location of the client plays a significant role in how performance is perceived by the client.
Don't know where you got this from. Ftp2.freebsd.org resolves (and has for the past couple of years) to 130.94.149.162, which is located at Verio. It's got a GE connection to their backbone and its a beefy machine, which is why it is so fast.
Check out the traffic graph for ftp2. Now slashdot that!
I've thought for a long time that in this sea of OSS projects and various hype surrounding them (which is only bound to get bigger and messier), the experience and knowledge of which ones are good is a very valuable skill. Rather than paying a lot for software, companies should (and many do) pay for people who know software.
It used to be that you start writing a program from scratch (I'm tallking commercial/IT here), but in the last few years it turned into looking for suitable OSS to start with (e.g. lets use Apache as the web server, Postgres as db, now I need to look for some graphics libs, oh and is there an OSS spell-checker out there?... Let's see what license is this under... Looks like this this piece of functionality I will need to write myself... etc..). And the success of the end result seems to be largely depndend on how thorough this decision-making process is. In fact a good real-life example of this is the Safari browser - someone did their homework and chose KDE's engine even thoug Gecko is widely known as *the* engine out there.
As a sidenote, this is also one of the things that might differentiate experienced old-timers from young new-comers. You may be a brilliant 20-year old programmer, but the 35-year old dude in the next office can put a better solution together faster because he knows where to go to look for good source code, what's hype and what really works.
I've worked on AS/400's, albeit a while ago - early/mid 90's. As a matter of fact my first paid programming experience was in RPG IV lanugage. I would not say it was mind-numbing - the documentation provided by IBM was fantastic, and the AS/400 administration is very much automatable using their CL language. The machine was incredibly stable and even most hardware upgrades did not require a reboot (IPL). And even though it doesn't resemble anything UNIX or most other systems I know, and all files are database tables (no such thing as "file" really), I sould say it was very consistent and even intuitive once you get a hang of it.
Having said that, I wouldn't want to do any AS/400 work now - but probably more because I don't like a typical AS/400 user (brick and mortar businesses not focused on technology) than the system itself.
So I don't know if it's just me, but this is hardly front-page news.
I doubt you can patent such a fundumental principal - how do you think transformers work? Also my Braun electric toothbrush charges its battery wirelessly and I've had it for years.
I tried Apple's X11, but couldn't figure out how to run it full-screen with WindowMaker. (I wouldn't need this had Apple provided virtual desktops that could be switched as fast as WindowMaker does it).... So I gave up and went back to XDarwin.
Before XML there was (and still is) RFC822 which describes how headers are formatted in e-mail, HTTP and a slew of other protocols.
I've been down the route where I tried to use XML where something as simple as "key: value" would do, and before I knew it, my program became a bloat relying on third-party XML libs, the config files were only marginally human-readable and a lot of time was wasted thinking about virtues of DOM vs SAX. In the end I learned that using XML for sake of XML isn't worth it.
I think XML is OK if used appropriately - for example I think XML is perfect for something like storing word processing documents. But the idea that every config file and every bit of network traffic should be XML is stupid IMHO.
Anything else is an art. In my experience best contributors range from those that come completely out of the blue to people you know professionally in person.
Generally, lack of contributors shouldn't be a concern - if it is, then the scope of your project is too big. Keep your scope narrow and goals simple (work on the assumption that you will never get any contributors other than yourself), focus on perfection rather than size or time, and your project will be successful.
Tarkovsky is in the same league with Fellini and Kurosawa - why even bother comparing? 30 years from now I doubt anyone will even know that Soderbergh's Solaris existed.
No, there is a workaround. Read the mod_python README, "OS Hints" section.
Since both Python and Apache default to using threads (except for FreeBSD, where Apache defaults to not using threads), this is no longer an issue.
Whose great idea was it I wonder?
I use XDarwin full screen with a black background. For whatever reason now the mouse pointer is now always black, which makes it invisible on black background. What a bummer...
Yahoo maps used mod_python according to the slides.
Interesting... Would you care to elaborate on what application was this protocol intended to solve, i.e. was it the same kind of stuff as UPnP, or something different altogether? Also, why ICMP? And lastly - was UPnP available at the time of writing and what you think of UPnP in general, if anything? Thanks!
HTTP over UDP is a hack, and the RFC is written by someone at Microsoft.
IMHO, the Rendezvous method of using DNS SRV records for service discovery is much better, because DNS has been around for ages, and we know it works. Only time will show what sorts of problems will appear with UPnP's HTTPU reliance.
First one that comes to mind is reliability - UDP doesn't perform well on busy networks (and its not unlikely that the network over which these packets will travel will indeed be busy - imagine 802.11b maxed out while a video is being streamed over it). UPnP requires devices to resend their messages after an interval because of UDP's unreliable nature - but if I hit "play" on my remote control, I want the thing to play instantly, not a few seconds later.
Another one is scale - large amounts of data can't be sent over UDP. As it looks right now, the packets shouldn't be large, but we don't know what the future holds - as devices get more complicated and support more features, the UPnP service description messages could get fairly large.
Of course the above is a moot point given that there are no other protocols that do this....
Yes, but Rendezvous is merely a protocol for automatic network configuration and discovering services (and a really good one too). UPnP goes a step further by providing means for device interaction.
The point is that if I were to write a piece of UPnP capable software, when the UPnP DVD players (or whatever) come out, chances are this software will be able to work with it. With Rendezvous - the device-specific interface is undefined.
Jini seems to be a Java-specific thing, so I'm not really interested in it.