Coffeehouse Thread

62 posts

Why is Channel9 fast?

Back to Forum: Coffeehouse
  • Shining Arcanine

    I am used to Channel9 being a slow, unresponsive site, but today I visit it for the first time in a month and it is as fast as Google. What happened?

  • spivonious

    New server perhaps? It's fast for me too.

  • Shining Arcanine

    spivonious said:

    New server perhaps? It's fast for me too.

    I would ask if the server was upgraded to Linux, but it seems that it is running IIS 7.0, which suggests that it is running either Windows Server 2008 or Windows Vista (unless IIS 7.0 runs under WINE). Whatever Microsoft did to the server should have been done years ago.

  • turrican

    Shining Arcanine said:
    spivonious said:
    *snip*

    I would ask if the server was upgraded to Linux, but it seems that it is running IIS 7.0, which suggests that it is running either Windows Server 2008 or Windows Vista (unless IIS 7.0 runs under WINE). Whatever Microsoft did to the server should have been done years ago.

    "which suggests that it is running either Windows Server 2008 or Windows Vista (unless IIS 7.0 runs under WINE)." ...are you serious?... I mean really? ...are you insane or just uneducated?

  • Sven Groot

    Seems to me that C9 is just going through one of its random fast(er) periods. It'll be back to slow tomorrow (EDIT: actually, it's not even particularly fast for me right now).

     

    And really, if you think C9's slowness is due to the operating system its servers run on, you really know nothing about computers and software development whatsoever.

  • turrican

    Sven Groot said:

    Seems to me that C9 is just going through one of its random fast(er) periods. It'll be back to slow tomorrow (EDIT: actually, it's not even particularly fast for me right now).

     

    And really, if you think C9's slowness is due to the operating system its servers run on, you really know nothing about computers and software development whatsoever.

    +1

     

    ... and regarding the fastness of C9... yeah I agree... it goes up and down all the time. I have been clicking around the site for the past hm... 4 hours and it is fast sometimes and sometime it gets "stuck" and slooooooooooooow as glue. Basically, nothing has changed. hehhe

     

  • vesuvius

    Shining Arcanine said:
    spivonious said:
    *snip*

    I would ask if the server was upgraded to Linux, but it seems that it is running IIS 7.0, which suggests that it is running either Windows Server 2008 or Windows Vista (unless IIS 7.0 runs under WINE). Whatever Microsoft did to the server should have been done years ago.

    I think you are doing very well as a replacement for coronacoder.

     

    I don't mean to gang up or pick on you either, but I was initially impressed by a number of your posts and arguments, but you seem to have completely lost the plot recently.

     

    I think you really need to step out of the bubble you are living in.

     

    Probably the most important thing you can do is to go out and get some real world experience at this stage, before you get to a PhD. I am working with a number of PhD's and the lack of real world exposure sometimes completely negates their academic distinction. If you are going to research real world problems, and fix them, then you must live in the real world, which you clearly are not at the moment, and like some smarter than I, I lack the time or patience to educate you on the error of your ways.

  • stun

    vesuvius said:
    Shining Arcanine said:
    *snip*

    I think you are doing very well as a replacement for coronacoder.

     

    I don't mean to gang up or pick on you either, but I was initially impressed by a number of your posts and arguments, but you seem to have completely lost the plot recently.

     

    I think you really need to step out of the bubble you are living in.

     

    Probably the most important thing you can do is to go out and get some real world experience at this stage, before you get to a PhD. I am working with a number of PhD's and the lack of real world exposure sometimes completely negates their academic distinction. If you are going to research real world problems, and fix them, then you must live in the real world, which you clearly are not at the moment, and like some smarter than I, I lack the time or patience to educate you on the error of your ways.

    Oh yeah what happened to Corona Coder?

    Haven't seen that troll around at all.

    Just to be clear, I'm NOT wishing for him to come back Devil .

     

    Btw, people like CoronaCoder won't listen to reason and live in a bubble.

    Your words will be lost on them because of their extremist views and ridiculous talking points.

    Don't waste your time on them.

  • Bass

    Sven Groot said:

    Seems to me that C9 is just going through one of its random fast(er) periods. It'll be back to slow tomorrow (EDIT: actually, it's not even particularly fast for me right now).

     

    And really, if you think C9's slowness is due to the operating system its servers run on, you really know nothing about computers and software development whatsoever.

    Actually, I really think you are wrong on this. The way an operating system is configured can in fact have a real measurable effect on the performance of an entire system. Google wouldn't be going out of their way to maintain Linux's obsolete O(1) scheduler infrastructure if this wasn't the case.

     

    I'm not saying that is the specific reason for C9's slowness, but it would be incredibly foolish to rule it out.

  • Shining Arcanine

    vesuvius said:
    Shining Arcanine said:
    *snip*

    I think you are doing very well as a replacement for coronacoder.

     

    I don't mean to gang up or pick on you either, but I was initially impressed by a number of your posts and arguments, but you seem to have completely lost the plot recently.

     

    I think you really need to step out of the bubble you are living in.

     

    Probably the most important thing you can do is to go out and get some real world experience at this stage, before you get to a PhD. I am working with a number of PhD's and the lack of real world exposure sometimes completely negates their academic distinction. If you are going to research real world problems, and fix them, then you must live in the real world, which you clearly are not at the moment, and like some smarter than I, I lack the time or patience to educate you on the error of your ways.

    You probably were more fond of my posts until January of this year, when I switched from Windows 7 Professional to Gentoo Linux on my laptop. I use Gentoo Linux as a platform to learn more about how things work than I could in my college classes, so everything I think about has generally revolved around Linux since then.

     

    You will likely not see me write posts like I used to post, because a combination of Microsoft's behavior and my experiences with Gentoo Linux have given me a new perspective on things. While I used to think Microsoft focused on having better products than their competitors, I now think that such an opinion was wrong, for several reasons:

    • Microsoft's claim to have spent 5 years developing Windows Vista when in fact, they had only spent 2.5 years on it because they had to start from scratch mid-way. Paul Thurrot's site documented this, long before they started making the 5 year claims, to my disbelief.
    • Microsoft's outright refusal to port newer versions of DirectX to older platforms after Windows Vista when they had made it a point of policy to backport newer versions of DirectX to their previous platforms. DirectX 9.0c was not only made available by Microsoft on Windows XP, but also on Windows 98, Windows 2000 and Windows ME.
    • Microsoft's coercion of its customers to essentially force them to move to newer products by refusing to patch products that they supposedly support and also by refusing to release service packs/update rollups to the same products.
    • Microsoft's discontinuation of mainstream support for Windows Media Center 2005 Edition 4 to 5 months before it was scheduled to enter extended support. I am still waiting for a class action lawsuit on this, because I purchased a product with a specified period of support and Microsoft abandoned its obligations simply because it was inconvenient for them.
    • Microsoft's sacrifice of sound user interface design principles in the pursuit of psychological effects. In specific, Microsoft abandoned their own User Interface design guidelines with Windows Vista and not only did they do that, but they refused to modify their software (specifically IE) to adhere to such guidelines on operating systems they designed to adhere to them (specifically Windows XP).
    • Microsoft's countless attempts to sabotage standardization, where they continually change the specifications to which their software adheres when such things could have been done either without breaking existing software or in a less severe manner. Their butchering of HTML/CSS is a good example of this. Their handling of Open Document Format is another.
    • Microsoft's continued exhibition of Not Invented Here syndrome, to which it approaches all competition, when the more sensible thing would be to focus on what it is good at doing and in other areas, rely and help others who are better at things than they are. Bing is an excellent example of this.
    • Microsoft's requirement that OEMs purchase Windows licenses for all systems they sell, that companies purchase Windows licenses for all the computers they own and that governments license Windows for all computers that they operate, without regard to whether or not they sell, operate or use Windows on all of their systems in the first place. At this time, it is impossible to buy a computer from a major OEM without also paying a small fee for Windows.
    • Microsoft's circumvention of first-sale doctrine with not only its software (e.g. Windows), but also its hardware (e.g. XBox 360), by using licensing as a means to bypass property ownership on Windows and its XBox Live service as a means to kill any homebrew development on the XBox 360 that is incompatible with its vision, even going so far as to disable all, but the most minimal functionality on such systems (e.g. media center extender support is disabled) is another issue.
    • Microsoft's CEO, who handles its companies' competition, not with a response of competition, but with a response of hurling chairs across his office.
    Any real commitment by Microsoft to adhering at least to its own standards (e.g. by focusing on making things that are better designed and not simply things that look better in appearance,  by adhering to support commitments, etcetera) ended after Windows XP. It is obvious that money was not coming as easily to Microsoft after it delayed Windows Vista, but breaking the few standards it held for itself was inappropriate and it has eliminated any good things that I was able to say. Rather than seeing Microsoft as leading the way to the future as I used to see it, I now see Microsoft as standing in the way to the future, which is why I post the occasional joke that they should be doing more with Linux.
    With Gentoo Linux, I no longer suffer from the fallout of Microsoft's short sighted decisions. I can use the software and technologies that I want, as long as I want and not only can do I that, but I can customize them to my heart's desire, even in terms of the assembly code used in the generation of the binaries. I am no longer locked into a single platform and improvements in technology (such as Link Time Optimization and Stack Smashing Protection) can be back-ported readily. These are things that Microsoft simply does not provide and likely will never provide.
    Is that a step far enough outside of my bubble for you? Tongue Out

  • joechung

    Shining Arcanine said:
    vesuvius said:
    *snip*

    You probably were more fond of my posts until January of this year, when I switched from Windows 7 Professional to Gentoo Linux on my laptop. I use Gentoo Linux as a platform to learn more about how things work than I could in my college classes, so everything I think about has generally revolved around Linux since then.

     

    You will likely not see me write posts like I used to post, because a combination of Microsoft's behavior and my experiences with Gentoo Linux have given me a new perspective on things. While I used to think Microsoft focused on having better products than their competitors, I now think that such an opinion was wrong, for several reasons:

    • Microsoft's claim to have spent 5 years developing Windows Vista when in fact, they had only spent 2.5 years on it because they had to start from scratch mid-way. Paul Thurrot's site documented this, long before they started making the 5 year claims, to my disbelief.
    • Microsoft's outright refusal to port newer versions of DirectX to older platforms after Windows Vista when they had made it a point of policy to backport newer versions of DirectX to their previous platforms. DirectX 9.0c was not only made available by Microsoft on Windows XP, but also on Windows 98, Windows 2000 and Windows ME.
    • Microsoft's coercion of its customers to essentially force them to move to newer products by refusing to patch products that they supposedly support and also by refusing to release service packs/update rollups to the same products.
    • Microsoft's discontinuation of mainstream support for Windows Media Center 2005 Edition 4 to 5 months before it was scheduled to enter extended support. I am still waiting for a class action lawsuit on this, because I purchased a product with a specified period of support and Microsoft abandoned its obligations simply because it was inconvenient for them.
    • Microsoft's sacrifice of sound user interface design principles in the pursuit of psychological effects. In specific, Microsoft abandoned their own User Interface design guidelines with Windows Vista and not only did they do that, but they refused to modify their software (specifically IE) to adhere to such guidelines on operating systems they designed to adhere to them (specifically Windows XP).
    • Microsoft's countless attempts to sabotage standardization, where they continually change the specifications to which their software adheres when such things could have been done either without breaking existing software or in a less severe manner. Their butchering of HTML/CSS is a good example of this. Their handling of Open Document Format is another.
    • Microsoft's continued exhibition of Not Invented Here syndrome, to which it approaches all competition, when the more sensible thing would be to focus on what it is good at doing and in other areas, rely and help others who are better at things than they are. Bing is an excellent example of this.
    • Microsoft's requirement that OEMs purchase Windows licenses for all systems they sell, that companies purchase Windows licenses for all the computers they own and that governments license Windows for all computers that they operate, without regard to whether or not they sell, operate or use Windows on all of their systems in the first place. At this time, it is impossible to buy a computer from a major OEM without also paying a small fee for Windows.
    • Microsoft's circumvention of first-sale doctrine with not only its software (e.g. Windows), but also its hardware (e.g. XBox 360), by using licensing as a means to bypass property ownership on Windows and its XBox Live service as a means to kill any homebrew development on the XBox 360 that is incompatible with its vision, even going so far as to disable all, but the most minimal functionality on such systems (e.g. media center extender support is disabled) is another issue.
    • Microsoft's CEO, who handles its companies' competition, not with a response of competition, but with a response of hurling chairs across his office.
    Any real commitment by Microsoft to adhering at least to its own standards (e.g. by focusing on making things that are better designed and not simply things that look better in appearance,  by adhering to support commitments, etcetera) ended after Windows XP. It is obvious that money was not coming as easily to Microsoft after it delayed Windows Vista, but breaking the few standards it held for itself was inappropriate and it has eliminated any good things that I was able to say. Rather than seeing Microsoft as leading the way to the future as I used to see it, I now see Microsoft as standing in the way to the future, which is why I post the occasional joke that they should be doing more with Linux.
    With Gentoo Linux, I no longer suffer from the fallout of Microsoft's short sighted decisions. I can use the software and technologies that I want, as long as I want and not only can do I that, but I can customize them to my heart's desire, even in terms of the assembly code used in the generation of the binaries. I am no longer locked into a single platform and improvements in technology (such as Link Time Optimization and Stack Smashing Protection) can be back-ported readily. These are things that Microsoft simply does not provide and likely will never provide.
    Is that a step far enough outside of my bubble for you? Tongue Out

    Sorry, there's not much new about your "new" perspective.  People have switched from Windows to Linux before.

  • Shining Arcanine

    joechung said:
    Shining Arcanine said:
    *snip*

    Sorry, there's not much new about your "new" perspective.  People have switched from Windows to Linux before.

    I went from adoring Microsoft's actions to despising Microsoft's actions. While that might not be as new of a perspective as say Fascism was in the 1930s, it is a new perspective for me, because I have never had it until now.

     

    That is why I say "a combination of Microsoft's behavior and my experiences with Gentoo Linux have given me a new perspective on things". The use of "me" as an indirect object was intended to state that it [a new perspective] is new in the context of my mind.

  • Shining Arcanine

    Bass said:
    Sven Groot said:
    *snip*

    Actually, I really think you are wrong on this. The way an operating system is configured can in fact have a real measurable effect on the performance of an entire system. Google wouldn't be going out of their way to maintain Linux's obsolete O(1) scheduler infrastructure if this wasn't the case.

     

    I'm not saying that is the specific reason for C9's slowness, but it would be incredibly foolish to rule it out.

    Do you have any references on Google's use of the O(1) scheduler? I would be interested in seeing them.

  • AndyC

    Shining Arcanine said:
    Bass said:
    *snip*

    Do you have any references on Google's use of the O(1) scheduler? I would be interested in seeing them.

    I'm rather more curious as to what you think is obsolete about the O(1) scheduler. It was introduced in the 2.6 kernel to replace the rather hideous O(n) scheduler they had previously, finally catching up with NT in being able to schedule threads without getting bogged down when the system was heavily loaded.

     

    In fact, I seem to recall Google were spending rather a lot of effort backporting it to the 2.4 kernel because they badly needed it and couldn't move to 2.6 without either major reworking of their internals or actually releasing a bunch of their proprietary code back to the FOSS comunity.

  • Shining Arcanine

    AndyC said:
    Shining Arcanine said:
    *snip*

    I'm rather more curious as to what you think is obsolete about the O(1) scheduler. It was introduced in the 2.6 kernel to replace the rather hideous O(n) scheduler they had previously, finally catching up with NT in being able to schedule threads without getting bogged down when the system was heavily loaded.

     

    In fact, I seem to recall Google were spending rather a lot of effort backporting it to the 2.4 kernel because they badly needed it and couldn't move to 2.6 without either major reworking of their internals or actually releasing a bunch of their proprietary code back to the FOSS comunity.

    I never called it obsolete.

     

    Anyway, as far as what I know is concerned, the Completely Fair Scheduler (an O(log(n)) scheduler) replaced the O(1) scheduler in Linux. Wikipedia has an article on it where it says that the reason behind it was to make systems more interactive:

     

    http://en.wikipedia.org/wiki/Completely_Fair_Scheduler

     

    It seems that the only reason why Linux does not support both schedulers is Linus Torvalds' desire that there only be one CPU scheduler in Linux.

     

    The difference between O(1) and O(log(n)) is not particularly large, especially considering that it is an operation done only a hundred to a thousand times a second on Linux systems.

     

    Lastly, I never heard about this regarding Google and the O(1) scheduler. Do you have any references? How did the creation of the Completely Fair Scheduler change things for them?

  • PaoloM

    Shining Arcanine said:
    vesuvius said:
    *snip*

    You probably were more fond of my posts until January of this year, when I switched from Windows 7 Professional to Gentoo Linux on my laptop. I use Gentoo Linux as a platform to learn more about how things work than I could in my college classes, so everything I think about has generally revolved around Linux since then.

     

    You will likely not see me write posts like I used to post, because a combination of Microsoft's behavior and my experiences with Gentoo Linux have given me a new perspective on things. While I used to think Microsoft focused on having better products than their competitors, I now think that such an opinion was wrong, for several reasons:

    • Microsoft's claim to have spent 5 years developing Windows Vista when in fact, they had only spent 2.5 years on it because they had to start from scratch mid-way. Paul Thurrot's site documented this, long before they started making the 5 year claims, to my disbelief.
    • Microsoft's outright refusal to port newer versions of DirectX to older platforms after Windows Vista when they had made it a point of policy to backport newer versions of DirectX to their previous platforms. DirectX 9.0c was not only made available by Microsoft on Windows XP, but also on Windows 98, Windows 2000 and Windows ME.
    • Microsoft's coercion of its customers to essentially force them to move to newer products by refusing to patch products that they supposedly support and also by refusing to release service packs/update rollups to the same products.
    • Microsoft's discontinuation of mainstream support for Windows Media Center 2005 Edition 4 to 5 months before it was scheduled to enter extended support. I am still waiting for a class action lawsuit on this, because I purchased a product with a specified period of support and Microsoft abandoned its obligations simply because it was inconvenient for them.
    • Microsoft's sacrifice of sound user interface design principles in the pursuit of psychological effects. In specific, Microsoft abandoned their own User Interface design guidelines with Windows Vista and not only did they do that, but they refused to modify their software (specifically IE) to adhere to such guidelines on operating systems they designed to adhere to them (specifically Windows XP).
    • Microsoft's countless attempts to sabotage standardization, where they continually change the specifications to which their software adheres when such things could have been done either without breaking existing software or in a less severe manner. Their butchering of HTML/CSS is a good example of this. Their handling of Open Document Format is another.
    • Microsoft's continued exhibition of Not Invented Here syndrome, to which it approaches all competition, when the more sensible thing would be to focus on what it is good at doing and in other areas, rely and help others who are better at things than they are. Bing is an excellent example of this.
    • Microsoft's requirement that OEMs purchase Windows licenses for all systems they sell, that companies purchase Windows licenses for all the computers they own and that governments license Windows for all computers that they operate, without regard to whether or not they sell, operate or use Windows on all of their systems in the first place. At this time, it is impossible to buy a computer from a major OEM without also paying a small fee for Windows.
    • Microsoft's circumvention of first-sale doctrine with not only its software (e.g. Windows), but also its hardware (e.g. XBox 360), by using licensing as a means to bypass property ownership on Windows and its XBox Live service as a means to kill any homebrew development on the XBox 360 that is incompatible with its vision, even going so far as to disable all, but the most minimal functionality on such systems (e.g. media center extender support is disabled) is another issue.
    • Microsoft's CEO, who handles its companies' competition, not with a response of competition, but with a response of hurling chairs across his office.
    Any real commitment by Microsoft to adhering at least to its own standards (e.g. by focusing on making things that are better designed and not simply things that look better in appearance,  by adhering to support commitments, etcetera) ended after Windows XP. It is obvious that money was not coming as easily to Microsoft after it delayed Windows Vista, but breaking the few standards it held for itself was inappropriate and it has eliminated any good things that I was able to say. Rather than seeing Microsoft as leading the way to the future as I used to see it, I now see Microsoft as standing in the way to the future, which is why I post the occasional joke that they should be doing more with Linux.
    With Gentoo Linux, I no longer suffer from the fallout of Microsoft's short sighted decisions. I can use the software and technologies that I want, as long as I want and not only can do I that, but I can customize them to my heart's desire, even in terms of the assembly code used in the generation of the binaries. I am no longer locked into a single platform and improvements in technology (such as Link Time Optimization and Stack Smashing Protection) can be back-ported readily. These are things that Microsoft simply does not provide and likely will never provide.
    Is that a step far enough outside of my bubble for you? Tongue Out

    I wish at least something you said was true. Alas, not.

     

    It is sad, indeed. Sad

  • spivonious

    Leave it to Shining to get this thread way off track.

     

    C9 guys - anything upgraded today? It's definitely faster all-around.

  • Bass

    AndyC said:
    Shining Arcanine said:
    *snip*

    I'm rather more curious as to what you think is obsolete about the O(1) scheduler. It was introduced in the 2.6 kernel to replace the rather hideous O(n) scheduler they had previously, finally catching up with NT in being able to schedule threads without getting bogged down when the system was heavily loaded.

     

    In fact, I seem to recall Google were spending rather a lot of effort backporting it to the 2.4 kernel because they badly needed it and couldn't move to 2.6 without either major reworking of their internals or actually releasing a bunch of their proprietary code back to the FOSS comunity.

    I call it obsolete because it is no longer maintained (actually, no longer exists) in mainline, and hasn't been there for some time. Linus dropped in favor of the Completely Fair Scheduler (CFS), which has an O(lg n) context switching complexity.

Comments closed

Comments have been closed since this content was published more than 30 days ago, but if you'd like to continue the conversation, please create a new thread in our Forums, or Contact Us and let us know.