I have searched for "msdn magazine" on google and the first result is this:

 

  Do they use PHP and use mod_rewrite to make it look like .aspx :)

Well, obviously even Microsoft isn't stupid enough to use ASP.  Of course, it's still M$ so they didn't go for the best web language (ColdFusion) but PHP is a close second. 

huseyint:

Do they use PHP and use mod_rewrite to make it look like .aspx :)

German Wikipedia page for ASP.net:

ASP.NET ist keine Programmiersprache, sondern eine Technologie. Somit kann jede ASP.NET Webseite in allen .Net Sprachen (z. B. C#, VB.NET oder auch PHP) programmiert werden

"ASP.NET is not a programming language, but a technology. Thus, every ASP.NET web site can be coded in any .Net language (e.g. C#, VB.NET or also PHP)"

I'm not familiar with ASP.NET, but assuming this statement is correct, then M$ wouldn't have to use mod_rewrite.

class Paula_Bean
  PAULA = "Brillant"

  def get_paula
    return PAULA
  end
end

What's mod_rewrite got to do with it anyway? You AddHandler with mod_mime.

Interesting... 75,800 results in a search for "site:microsoft.com inurl:PHPSESSID" in Google :P

http://www.google.com/search?num=100&q=site%3Amicrosoft.com+inurl%3APHPSESSID

What's mod_rewrite got to do with it anyway? You AddHandler with mod_mime.

huseyint meant that they could be doing something like RewriteRule ^(.+).aspx$ $1.php or similar

Yeah you're right, mod_mime is a better solution then mod_rewrite. Daniel15 explained what I was thinking with mod_rewrite.

 Or they outsource the site to india and someone there decided to reinvent the wheel and recreate php sessions in asp.net.

derula:

huseyint:

Do they use PHP and use mod_rewrite to make it look like .aspx :)

German Wikipedia page for ASP.net:

ASP.NET ist keine Programmiersprache, sondern eine Technologie. Somit kann jede ASP.NET Webseite in allen .Net Sprachen (z. B. C#, VB.NET oder auch PHP) programmiert werden

"ASP.NET is not a programming language, but a technology. Thus, every ASP.NET web site can be coded in any .Net language (e.g. C#, VB.NET or also PHP)"

I'm not familiar with ASP.NET, but assuming this statement is correct, then M$ wouldn't have to use mod_rewrite.

afaik that statement is theoratically true. C# and VB.NET tools for ASP.NET are supplied by MS. I believe there is some company that supports Perl and Phyton.
I guess Ruby is also possible trough ironRuby or something like that.
However, I've never heard of PHP working on .NET. I know of ironPHP, but that one has been on pre-0.0.1 since 2 years and I don't believe there's any development going on anymore. It also lacks support on almost anything except hello world.

Daniel15:

Interesting... 75,800 results in a search for "site:microsoft.com inurl:PHPSESSID" in Google :P

It seems that msdn, msdn2 and technet use it. Don't exaggerate ;-)
(If the basic technique of those subdomains/systems is PHP all the underlying pages could have a PHPSESSID too)

there's always http://php-compiler.net

WTF? Think about this:

1. Where do links in Google come from?
   Right, from crawlers that crawl other websites.
2. What does crappy forum software do to all links in postings?
   Right, add a PHPSESSID just in case the link was an internal one (or if they use urlrewriter).
3. What does a website do if it does not understand a parameter (in most cases)?
   Right, it ignores it and delivers the site anyway.
4. What happens if enough people link to http://thedailywtf.com/Articles/random.ashx?PHPSESSID=fedd7a5ad9fba819da4bf977aa6fea45?
   Right, it ends up in Google and you all believe that this website uses PHP :)
   

mihi 

WTF. Seems that Hyperlinks do not work in Firefox as expected :(

Trying again:  http://thedailywtf.com/Articles/random.ashx?PHPSESSID=fedd7a5ad9fba819da4bf977aa6fea45

Let's say there are 100 incoming links to MSDN Magazine Homepage with anchor texts "msdn magazine" and let 10 of these come from those crappy forum software with PHPSESSID appended to them. If Google is so intelligent, it will not promote those 10 links, it will use the other 90 correct links in its SERPs. And btw the situation is not only valid for MSDN Magazine Homepage, but also for a bunch of other page with urls starting msdn. msdn2. technet.

huseyint:

Let's say there are 100 incoming links to MSDN Magazine Homepage with anchor texts "msdn magazine" and let 10 of these come from those crappy forum software with PHPSESSID appended to them. If Google is so intelligent, it will not promote those 10 links, it will use the other 90 correct links in its SERPs. And btw the situation is not only valid for MSDN Magazine Homepage, but also for a bunch of other page with urls starting msdn. msdn2. technet.

Well, it could be a Google WTF then. And it could be that msdn, msdn2 and technet are the systems most linked to in fora or blogs.
Still: why not any non-msdn|technet|msdn2 page then?
I don't know how PHP url-sessid behavior can be invoked, but I guess it's by disabling cookies. I'm not on my own computer right now (and I'm guaranteed to forget to switch it back on) so I rather not disable them. Anyone willing to try?

Anything that can be compiled into .Net MSIL (MS Intermediate Language), can theoritically be used in ASP. 

 Including:

Cobol.Net (That's right, COBOL web pages!)

I've also ASP done with assembly, just for fun, although a quick google search didn't get my any meaningful results.

 

And finally, an open source project for lolCode.net

mihi:

WTF? Think about this:

1. Where do links in Google come from?
   Right, from crawlers that crawl other websites.
2. What does crappy forum software do to all links in postings?
   Right, add a PHPSESSID just in case the link was an internal one (or if they use urlrewriter).
3. What does a website do if it does not understand a parameter (in most cases)?
   Right, it ignores it and delivers the site anyway.
4. What happens if enough people link to http://thedailywtf.com/Articles/random.ashx?PHPSESSID=fedd7a5ad9fba819da4bf977aa6fea45?
   Right, it ends up in Google and you all believe that this website uses PHP :)
   

mihi 

Seems reasonable enough to me. Especially since I cannot find an internal link in MSDN magacine containing a PHPSESSID. Google could filter those links, but maybe they come from a popular enough blog.

 

Stubb063:

Cobol.Net (That's right, COBOL web pages!)

Someone needs to be shot.  Why would anyone use anything but ColdFusion?

ammoQ:

Seems reasonable enough to me. Especially since I cannot find an internal link in MSDN magacine containing a PHPSESSID. Google could filter those links, but maybe they come from a popular enough blog.

Ignoring my joke post above, this is obviously what is happening.  Did someone really think MS uses PHP for the MSDN site?  *facepalm*  Anyway, when a user agent will not accept cookies PHP will add the session ID to the query string of all links in the output HTML.  Of course, googlebot does not accept cookies so it ends up with PHPSESSID tacked on to all of the links of PHP-based sites it is crawling.  So this doesn't even require crappy forum software, just any site that runs PHP and doesn't force sessions to be cookie-only.

 

What is actually far more interesting to me (and has not yet been commented on) is that the main link for MSDN magazine did not come from a Microsoft site, but from an external site using PHP.  With what public information is available about PageRank, we know that a page's relation to a particular set of keywords is determined through links from other pages.  It is only fitting that the actual MSDN Magazine site be the top result for the query, but obviously not everyone who linked to the page appended the exactly same PHPSESSID.  So PageRank must either ignore all GET params or must be aware of common session ID names like PHPSESSID when correlating URLs from disparate sites.  Regardless, the PHPSESSID param was left in the URL delivered with the search results, so apparently even if Google does strip common session IDs when correlating URLs they don't when providing search results.  From this one can draw to hypotheses: 1) the first URL for MSDN Magazine that googlebot encountered was hosted on a PHP-based site and 2) Google does not use a different indexing strategy for a "known content provider" like MSDN.  Both of these interest me as someone who is drawn to the dark art of SEO as well as to the arcane methods developers at Google use to "index the web".

Stubb063:

Anything that can be compiled into .Net MSIL (MS Intermediate Language), can theoritically be used in ASP. 

 Including:

Cobol.Net (That's right, COBOL web pages!)

I've also ASP done with assembly, just for fun, although a quick google search didn't get my any meaningful results.

And finally, an open source project for lolCode.net

You could do it in IL Assembler rather easily.  more traditional versions of assembly would be a bit harder though and would require you to code your own compiler.  or you could cheat and do it in c# and just put everything in ASM blocks. 

 I think we need either Fortran.Net or PunchCards.Net.

apparently someone already made Fortran.net.

http://www.codeproject.com/KB/net-languages/intro_fortran.aspx

morbiuswilters:

ammoQ:

Seems reasonable enough to me. Especially since I cannot find an internal link in MSDN magacine containing a PHPSESSID. Google could filter those links, but maybe they come from a popular enough blog.

Ignoring my joke post above, this is obviously what is happening.  Did someone really think MS uses PHP for the MSDN site?  *facepalm*  Anyway, when a user agent will not accept cookies PHP will add the session ID to the query string of all links in the output HTML.  Of course, googlebot does not accept cookies so it ends up with PHPSESSID tacked on to all of the links of PHP-based sites it is crawling.  So this doesn't even require crappy forum software, just any site that runs PHP and doesn't force sessions to be cookie-only.

 

What is actually far more interesting to me (and has not yet been commented on) is that the main link for MSDN magazine did not come from a Microsoft site, but from an external site using PHP.  With what public information is available about PageRank, we know that a page's relation to a particular set of keywords is determined through links from other pages.  It is only fitting that the actual MSDN Magazine site be the top result for the query, but obviously not everyone who linked to the page appended the exactly same PHPSESSID.  So PageRank must either ignore all GET params or must be aware of common session ID names like PHPSESSID when correlating URLs from disparate sites.  Regardless, the PHPSESSID param was left in the URL delivered with the search results, so apparently even if Google does strip common session IDs when correlating URLs they don't when providing search results.  From this one can draw to hypotheses: 1) the first URL for MSDN Magazine that googlebot encountered was hosted on a PHP-based site and 2) Google does not use a different indexing strategy for a "known content provider" like MSDN.  Both of these interest me as someone who is drawn to the dark art of SEO as well as to the arcane methods developers at Google use to "index the web".

I don't think it's obvious at all. If Google kept the PHPSESSID variable from links you would see it everywhere on non-PHP sites in Google, but these MS sites are the first I've seen of it. In fact, in one Google query I was able to find a Microsoft Support page that links to MSDN with PHPSESSID: http://support.microsoft.com/kb/950606 (two links under More Information heading). I think we can therefore conclude that Microsoft is responsible for putting PHPSESSID into their own links. It's possible that Microsoft uses some PHP mixed with ASP.NET and requires the PHP session id variable passed when this is in use. Maybe Microsoft's support site is PHP and the CMS creates links this way. Maybe all of Microsoft's sites are PHP disguised to look like ASP.NET. Maybe the programmers used PHPSESSID to pass around their own .NET data for reasons unknown to us.

shakin:

I don't think it's obvious at all. If Google kept the PHPSESSID variable from links you would see it everywhere on non-PHP sites in Google, but these MS sites are the first I've seen of it. In fact, in one Google query I was able to find a Microsoft Support page that links to MSDN with PHPSESSID: http://support.microsoft.com/kb/950606 (two links under More Information heading). I think we can therefore conclude that Microsoft is responsible for putting PHPSESSID into their own links. It's possible that Microsoft uses some PHP mixed with ASP.NET and requires the PHP session id variable passed when this is in use. Maybe Microsoft's support site is PHP and the CMS creates links this way. Maybe all of Microsoft's sites are PHP disguised to look like ASP.NET. Maybe the programmers used PHPSESSID to pass around their own .NET data for reasons unknown to us.

This does happen all the time with other sites and is a completely sensible explanation, given the promiscuous link-modification of PHP.  What is not sensible is believing that Microsoft is using PHP and not their own development stack for running their own sites.  I suppose they use Eclipse to develop in, as well?  Store their PHP sources in SVN and host it on Apache boxes?  Pfft.

mihi:

WTF. Seems that Hyperlinks do not work in Firefox as expected :(

Yeah, Firefox has this weird behavior where the target of a hyperlink is determined by the contents of the href attribute. I think it's been like that since back when it was still Netscape Navigator.