June Netcraft Survey

Andy Cheung writes "http://www.netcraft.com/survey/ The Netcraft Web Server Survey for June is out. Apache market share rises 3.46%; MS down -2.72". Scroll down past the graph on servers and check out the information on current exploits. It makes you wonder why "immediate death of the internet" has not happened.

The time is ripe to take over the Internet...

[hackwrench]

Come on, somebody, get your code together that takes advantage of all the exploits and eforces your view of the way things should be. There's hardly a better time!

What would be more interesting...

[JMZero]

Is a breakdown of how each company's base is growing.

How much comes from switching from another product?

How much comes from new domains starting up?

How much comes from existing domains adding servers?

Re:What would be more interesting...

[ThePilgrim]

Net Craft will provide that info if you are willing to pay them.

Thats where they make most of their income.

Re:Here we go again...

[rodolfo.borges]

On Mozilla it's not wide.

you are wrong!

[BigBir3d]

what is important is how many are active:

Apache -0.69

Wrong?

[OldMiner]

Although the amount of active sites [netcraft.com] is certainly important, it should be noted that the drop you state is a (miniscule) drop in the total market share of internet sites, not a drop in usage. There was still a 5% increase (10.4 million to about 11 million) in active sites served by Apache. Also worth noting is the -0.85 drop by MS (also with about 5% increase in total active sites). What is somewhat perplexing is that of the four major developers listed, only iPlanet showed a growth in the market share of active sites, and its growth is not great enough to account for the drop in Apache and MS. Who are the tiny developers that are taking over relatively large amounts of market share, then?

Re:Wrong?

[pete-classic]

it should be noted that the drop you state is a (miniscule) drop in the total market share of internet sites, not a drop in usage.

That isn't right, either.

That number is the change (drop) in percent of publicly accessable Apache servers that are acutally hosting a page.

So the original posters point is actually in favor of Apache, since over 64% of Apache servers (acording to the survey) serve an actual site (as opposed to an "It Works!" type page), where IIS has just under 25%.

OTOH these numbers are skewed by the fact that there are oodles of "branded" Apache packages out there with custom "It Works!" pages, and undoubtably Netcraft gives Apache more credit than it deserves on this account. But we can't really be sure of anything without more info on their methodology.

-Peter

Re:Wrong?

[pete-classic]

To be clearer, Apache's share of total installs went up. Apache's number of active sites went up. The rate of increase of total installs went up by more than the rate of active sites, so the only drop was in percent active sites. The number in question is orthoginal to number of active sites.

-Peter

Hmmm

[Treeluvinhippy]

Scroll down past the graph on servers and check out the information on current exploits. It makes you wonder why "immediate death of the internet" has not happened.

Well since the only people capable of bringing the world wide web to its knees are l33t Ub3rh4x0rs, it will be with us for quite some time.

Odd

[LinuxGeek8]

Of course you can't put real value into statistics without finding out what the reasons are from people changing their webserver.

But I find it odd that when an expoit, worm or virus comes out for a certain webserver, that webserver rises in marketshare.
Like the rise of Microsoft IIS, after Code red, and now the rise of Apache.
Would it then be true that bad news is even better marketing then no news?
Is that why Microsoft Windows is even so successfull?

Re:Odd

[sporktoast]

I'll risk a guess that the exploit may wake up some percentage of admins or site managers, reminding them that they have let their site sit for months/years with just an "It works!" type page.

They update/upgrade to try and patch against the exploit and maybe spend a short while putting in some more personalized placeholder pages (Coming soon: Our Calendar!), before they back burner it again.

Just a guess, anyways.

-Sporktoast

Their exploitable server guess is a bit flawed...

[weave]

They are assuming that any apache site out there under 1.3.26 is vulnerable. But that's not true. Redhat back-ports patches into their supported revs when needed. RH 7.3 patched Apache is at 1.3.23 and RH 7.2 patched Apache is at 1.3.22. I'm sure a lot of other distros back port patches as well.