5) Smash your pager, claim it was "killed in the line of duty".
Please, this is like every sysadmin's wet dream. Show me a sysadmin that doesn't sprint for the phone as soon as the electronic leash starts jerking at 2am, offering an apologetic " is down" to the rest of the party.
I don't know, I am with the other guy a couple of posts up, who wants to be someone's on-call 24/7 lackey. I saw the sysadmin referred to once on Slashdot as "the janitor of the internet" and that is essentially what the position is. The best day of my career was the day I moved from an admin position to a software development one.
I will say that I have seen a great many admins who couldn't cut it as programmers. I have also seen many programmers who could barely log onto a linux box. I do, however, suspect that any competent programmer could easily pick up the "skills" necesssary to admin whereas the converse is far from true.
Is your install from scratch? The performance increase could just be from having a fresh win installation. For whatever reason (probably the registry) Windows seems to get slower and slower the longer the installation has been on a given machine.
The PDF is at http://gigue.peabody.jhu.edu/~mdboom/latency-icmc2 001.pdf . It really doesn't discuss what platform is the "best of the group", only what platform has the lowest latency.
It is noteworthy that I have been unable to view page 3 (the page that actually has the results) of the PDF on every Win box I have tried.
I did notice that the Windows latency that their testing revealed indicated a much higher value than most people experience in reality. I also noticed that some type of multiplatform sound library was used to do the testing. Some of the stuff was curious in that respect. To counter that, however, a coworker knows Karl MacMillan personally and swears that he is a "mad genius" and that the report must be accurate. So...
Some of you who followed it to the end saw what happened to most of the BBS packages. Clark Development went belly up. Searchlight and MSI sold their software to small companies who squeezed every last dime from the software (and still try to market it!). I don't know what Galacticomm ended up doing with MajorBBS/Worldgroup. Lesser-used packaged like TAG and WWIV dropped off the face of the planet.
In all this, there is a neat story, involving Rob Swindell and his Synchronet BBS software. His company, Digital Dynamics, sold Synchronet for a noteable price "back in the day". They had full page spreads in Boardwatch along with Clark, Galacticomm, MSI, and the other big players. However, when the bottom fell out of the market, instead of squeezing every last dime from the product, Rob Swindell cleaned up his code and released everything into the public domain at which time he himself ceased all development.
It gets even cooler than that. About a year ago, Rob picks up the project again and turns it into open source with the release of a Linux version. Synchronet now supports Windows, OS/2, and Linux versions, all free and all GPLd. You can check it out at www.synchro.net .
If anyone here used the ZChat chat door, that was my "child".
In missle silos the presence of water typically means that the silo door leaks when it rains. The sealing of the door is typically the first step in turning it into a residence.
This article completely ignores musicnet, which is BMG, EMI, Warner Music Group and Zomba using Real's format and DRM technology... it's weird to see "music industry" this and "music industry" that in the article without any mention of the musicnet versus pressplay battle that is pending... and the fact that a noteable portion of the music industry is in bed with Real...
What are you talking about? The music industry has eaten essentially every internet-related music distrubution channel over the last year. They haven't lost grips at all. You act as if the MP3 format is alien to the music industry, did you forget who just shelled out the bucks for mp3.com (and its underlying my.mp3 technology)?
Exactly. Just like we have to SSH onto our linux boxes to do an apt-get update & upgrade we also have to VNC into our MS boxes to do a hfcheck. There's no magic tool.
I didn't blame the lack of central administration tools (like HFNetchk?) on a Windows sysadmin, not sure from where in my message you gleaned that tidbit.
What is this standardized tool that allows one to centrally administer non-Windows boxes?
Our organization didn't do squat because we spent five minutes researching commonly accepted practices for securing IIS and NT boxes before we ever put our first box on the net. We do the same for every piece of hardware and software, exploits are not an MS-exclusive thing. The simple act of unmapping unused extensions in IIS has saved us countless hours (or days) of agony on many occasions. I suspect your organization may not contain the level of security-conciousness necessary to properly maintain systems connected to the internet since such security-awareness would have included remedial research into the securest method of presenting a piece of hardware or software to the internet. In other words, if your organization knew what they were doing, the issue you experienced would not have occurred. It's not an apache/IIS issue, it's a poor administration issue that will plague your organization, unless corrected, regardless of what OS and web server software they choose to deploy.
----- The following addresses had transient non-fatal errors -----
----- Transcript of session follows -----
... while talking to smtp02.egghead.com.:
>>> RCPT To:
... Deferred: 452 4.2.2 Mailbox full
Warning: message still undelivered after 4 hours
Will keep trying until message is 5 days old
Most of the commercial insurance policies in the US do in fact cover terrorism and shares of insurance and reinsurance carriers are already suffering today. The World Trade Center is insured for 1.5 billion. Swiss Re and Munich Re probably have the most exposure in this incident. As of this morning, Swiss Re shares are down more than 17% and Munich Re shares had fallen more than 15%. Lloyds of London also has exposure, most likely through the aircraft. This is not good for Lloyds, who is just getting back on its feet after liabilities from asbestos, Exxon Valdez, and some big storms in 1989
They must already have at least chess in the pen. A friend of mine has a tattoo from prison and it was made from ink created by grinding up plastic chess pieces.
Probably wasn't too bright to use your development bugzilla server as a message forum. Someone made a cutesy-ass entry in bugzilla that has nothing to do with any sort of bug and now people are going there to check it out. Does this surprise you?
The music industry doesn't work that way. Music is a business, the musicians who are commercially successful are generally more businesspeople and less musicians. It's all about who the record labels promote, which certainly is not directly related to (maybe inversely, actually) the quality of the artists' works.
The technical issue you describe isn't really the limitation (the payment mechanism needn't be built into the player). What is the limitation is the lack of digital rights management in the current media players. That will be resolved in the next couple of months, when Real starts rolling out en-masse the new version of Real player. The player supports digital rights management, where a content provider can provide exactly the service you describe. The player keeps track of the purchased licenses. There's obviously more to it but I can't really talk about it.
From an artist's perspective, for me, 128k is an irritation. We produce our music au gratis exclusively for our mp3.com audience. I mix something down and the vocals (our vocalist is extremely good) have a particular sound and level. After I make the mp3, the sound is just slightly different, just not as clear, just not as good. Very often the dynamics will change such that a particular "track" in the song that was mixed to sit at a particular volume level relative to other parts will end up just a tiny bit too loud or too soft in the mp3. I find myself encoding the mp3, then going back and adjusting the original mix, then reencoding the mp3, and repeating this process 4 or 5 times. The problem is always with the vocals. They just lose their "smoothness" at 128K.
I would like to be able to have 192K content available on mp3.com . I believe this would make the sound of the mp3 significantly closer to my wave mixdown without adding too much additional size to the files.
The reason it isn't trivial is because there is the issue of when/where to inject the ads. The program material (we are referring to live sources only, ads in on-demand content is trivial) would have spaces where the ad content goes.
One method would be to list in the metadata associated with the media on the server side the exact start time and length of each of the "commercial-ready" areas. There are a couple of issues with this method, the client and server would have to time synchronize (no big deal) but more importantly the broadcaster would have to perform additional steps not necessary for their RF broadcasts.
Another method that is used injects tones into the broadcast in the location where commercial insertion is supposed to occur. Equipment at the stations detects the tones and inserts commercials. For this to work on a net stream, some point in the encoding process would need to scan the stream and look for the tone.
None of it is rocket science but, with the exception of the codecs themselves (and perhaps the RTSP transport), the Real streaming system is relatively unimpressive. I do not know how much resources codec development would take, but that must be where all the development efforts occur, because the server-side stuff Real sells really couldn't get much more simplistic and feature-lacking. A great example is the method used to prevent stream theft from a standard Real server (this doesn't include the new Digital Rights Management stuff). Without using a hacked-together plugin Real only seems to offer to "special" customers and whose method of theft prevention is pretty crude, you are stuck with the situation where anyone can connect directly to your real servers and download any of your content. It's an expensive streaming system that doesn't even have a way to prevent unauthorized access to those streams out-of-the-box. If they can't handle an issue like security in a elegant way, I have no problem understanding why the issue of ad insertion in live streams is apparently such a challenge.
Slashdot Mirror
Of course Slash, the ultimate BOFH, mangled part of the message. "is down" actually read "(insert daemon or geekily-named machine) is down".
maru
5) Smash your pager, claim it was "killed in the line of duty".
Please, this is like every sysadmin's wet dream. Show me a sysadmin that doesn't sprint for the phone as soon as the electronic leash starts jerking at 2am, offering an apologetic " is down" to the rest of the party.
maru
I don't know, I am with the other guy a couple of posts up, who wants to be someone's on-call 24/7 lackey. I saw the sysadmin referred to once on Slashdot as "the janitor of the internet" and that is essentially what the position is. The best day of my career was the day I moved from an admin position to a software development one.
I will say that I have seen a great many admins who couldn't cut it as programmers. I have also seen many programmers who could barely log onto a linux box. I do, however, suspect that any competent programmer could easily pick up the "skills" necesssary to admin whereas the converse is far from true.
maru
I would suspect that the sig is a joke and the misspelling is the "punch line"...
maru
www.mp3.com/pixal
Is your install from scratch? The performance increase could just be from having a fresh win installation. For whatever reason (probably the registry) Windows seems to get slower and slower the longer the installation has been on a given machine.
maru
www.mp3.com/pixal
The PDF is at http://gigue.peabody.jhu.edu/~mdboom/latency-icmc
It is noteworthy that I have been unable to view page 3 (the page that actually has the results) of the PDF on every Win box I have tried.
I did notice that the Windows latency that their testing revealed indicated a much higher value than most people experience in reality. I also noticed that some type of multiplatform sound library was used to do the testing. Some of the stuff was curious in that respect. To counter that, however, a coworker knows Karl MacMillan personally and swears that he is a "mad genius" and that the report must be accurate. So...
maru
www.mp3.com/pixal
Some of you who followed it to the end saw what happened to most of the BBS packages. Clark Development went belly up. Searchlight and MSI sold their software to small companies who squeezed every last dime from the software (and still try to market it!). I don't know what Galacticomm ended up doing with MajorBBS/Worldgroup. Lesser-used packaged like TAG and WWIV dropped off the face of the planet.
In all this, there is a neat story, involving Rob Swindell and his Synchronet BBS software. His company, Digital Dynamics, sold Synchronet for a noteable price "back in the day". They had full page spreads in Boardwatch along with Clark, Galacticomm, MSI, and the other big players. However, when the bottom fell out of the market, instead of squeezing every last dime from the product, Rob Swindell cleaned up his code and released everything into the public domain at which time he himself ceased all development.
It gets even cooler than that. About a year ago, Rob picks up the project again and turns it into open source with the release of a Linux version. Synchronet now supports Windows, OS/2, and Linux versions, all free and all GPLd. You can check it out at www.synchro.net .
If anyone here used the ZChat chat door, that was my "child".
maru
www.mp3.com/pixal
In missle silos the presence of water typically means that the silo door leaks when it rains. The sealing of the door is typically the first step in turning it into a residence.
maru
www.mp3.com/pixal
This article completely ignores musicnet, which is BMG, EMI, Warner Music Group and Zomba using Real's format and DRM technology... it's weird to see "music industry" this and "music industry" that in the article without any mention of the musicnet versus pressplay battle that is pending... and the fact that a noteable portion of the music industry is in bed with Real...
maru
www.mp3.com/pixal
What are you talking about? The music industry has eaten essentially every internet-related music distrubution channel over the last year. They haven't lost grips at all. You act as if the MP3 format is alien to the music industry, did you forget who just shelled out the bucks for mp3.com (and its underlying my.mp3 technology)?
maru
www.mp3.com/pixal
Exactly. Just like we have to SSH onto our linux boxes to do an apt-get update & upgrade we also have to VNC into our MS boxes to do a hfcheck. There's no magic tool.
maru
www.mp3.com/pixal
I didn't blame the lack of central administration tools (like HFNetchk?) on a Windows sysadmin, not sure from where in my message you gleaned that tidbit.
What is this standardized tool that allows one to centrally administer non-Windows boxes?
maru
www.mp3.com/pixal
Our organization didn't do squat because we spent five minutes researching commonly accepted practices for securing IIS and NT boxes before we ever put our first box on the net. We do the same for every piece of hardware and software, exploits are not an MS-exclusive thing. The simple act of unmapping unused extensions in IIS has saved us countless hours (or days) of agony on many occasions. I suspect your organization may not contain the level of security-conciousness necessary to properly maintain systems connected to the internet since such security-awareness would have included remedial research into the securest method of presenting a piece of hardware or software to the internet. In other words, if your organization knew what they were doing, the issue you experienced would not have occurred. It's not an apache/IIS issue, it's a poor administration issue that will plague your organization, unless corrected, regardless of what OS and web server software they choose to deploy.
Hope this helps,
maru
www.mp3.com/pixal
Don't forget, the VCs made their killing as soon as the IPO hit...
maru
www.mp3.com/pixal
Did you not know that the customers at a colocation facility pay for bandwidth?
maru
www.mp3.com/pixal
Of course slashcode stripped half of the text out of the message...
----- The following addresses had transient non-fatal errors -----
----- Transcript of session follows -----
... while talking to smtp02.egghead.com.:
>>> RCPT To:
... Deferred: 452 4.2.2 Mailbox full
Warning: message still undelivered after 4 hours
Will keep trying until message is 5 days old
Most of the commercial insurance policies in the US do in fact cover terrorism and shares of insurance and reinsurance carriers are already suffering today. The World Trade Center is insured for 1.5 billion. Swiss Re and Munich Re probably have the most exposure in this incident. As of this morning, Swiss Re shares are down more than 17% and Munich Re shares had fallen more than 15%. Lloyds of London also has exposure, most likely through the aircraft. This is not good for Lloyds, who is just getting back on its feet after liabilities from asbestos, Exxon Valdez, and some big storms in 1989
maru
www.mp3.com/pixal
They must already have at least chess in the pen. A friend of mine has a tattoo from prison and it was made from ink created by grinding up plastic chess pieces.
maru
www.mp3.com/pixal
Probably wasn't too bright to use your development bugzilla server as a message forum. Someone made a cutesy-ass entry in bugzilla that has nothing to do with any sort of bug and now people are going there to check it out. Does this surprise you?
maru
www.mp3.com/pixal
The music industry doesn't work that way. Music is a business, the musicians who are commercially successful are generally more businesspeople and less musicians. It's all about who the record labels promote, which certainly is not directly related to (maybe inversely, actually) the quality of the artists' works.
maru
www.mp3.com/pixal
The technical issue you describe isn't really the limitation (the payment mechanism needn't be built into the player). What is the limitation is the lack of digital rights management in the current media players. That will be resolved in the next couple of months, when Real starts rolling out en-masse the new version of Real player. The player supports digital rights management, where a content provider can provide exactly the service you describe. The player keeps track of the purchased licenses. There's obviously more to it but I can't really talk about it.
maru
www.mp3.com/pixal
=====
them. If MP3.com charges, the bulk of that money better go the the artists that spent their hard earned dough to write and record it.
=====
I suspect that is as likely to occur at mp3.com as it is at the traditional labels owned by Vivendi.
maru
www.mp3.com/pixal
From an artist's perspective, for me, 128k is an irritation. We produce our music au gratis exclusively for our mp3.com audience. I mix something down and the vocals (our vocalist is extremely good) have a particular sound and level. After I make the mp3, the sound is just slightly different, just not as clear, just not as good. Very often the dynamics will change such that a particular "track" in the song that was mixed to sit at a particular volume level relative to other parts will end up just a tiny bit too loud or too soft in the mp3. I find myself encoding the mp3, then going back and adjusting the original mix, then reencoding the mp3, and repeating this process 4 or 5 times. The problem is always with the vocals. They just lose their "smoothness" at 128K.
I would like to be able to have 192K content available on mp3.com . I believe this would make the sound of the mp3 significantly closer to my wave mixdown without adding too much additional size to the files.
maru
www.mp3.com/pixal
The reason it isn't trivial is because there is the issue of when/where to inject the ads. The program material (we are referring to live sources only, ads in on-demand content is trivial) would have spaces where the ad content goes.
One method would be to list in the metadata associated with the media on the server side the exact start time and length of each of the "commercial-ready" areas. There are a couple of issues with this method, the client and server would have to time synchronize (no big deal) but more importantly the broadcaster would have to perform additional steps not necessary for their RF broadcasts.
Another method that is used injects tones into the broadcast in the location where commercial insertion is supposed to occur. Equipment at the stations detects the tones and inserts commercials. For this to work on a net stream, some point in the encoding process would need to scan the stream and look for the tone.
None of it is rocket science but, with the exception of the codecs themselves (and perhaps the RTSP transport), the Real streaming system is relatively unimpressive. I do not know how much resources codec development would take, but that must be where all the development efforts occur, because the server-side stuff Real sells really couldn't get much more simplistic and feature-lacking. A great example is the method used to prevent stream theft from a standard Real server (this doesn't include the new Digital Rights Management stuff). Without using a hacked-together plugin Real only seems to offer to "special" customers and whose method of theft prevention is pretty crude, you are stuck with the situation where anyone can connect directly to your real servers and download any of your content. It's an expensive streaming system that doesn't even have a way to prevent unauthorized access to those streams out-of-the-box. If they can't handle an issue like security in a elegant way, I have no problem understanding why the issue of ad insertion in live streams is apparently such a challenge.
maru