Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
Apache Software

August Netcraft Results - Apache up 6%, MS IIS down 6% 51

An Onimous Cow Herd writes "Augusts Netcraft Web survey results are now out. This month's results show a dramatic upsurge of nearly 6% for Apache and a corresponding drop for MS IIS! At this point, Apache's decline has dramatically reversed, regaining the ground it lost to IIS starting mid-2001 and currently Apache's market share stands the highest since Netscraft started their monthly surveys."
This discussion has been archived. No new comments can be posted.

August Netcraft Results - Apache up 6%, MS IIS down 6%

Comments Filter:
  • Looking at the graphs and data, I'd hardly call it dramatic and wave it around yet -- ooh, apache's growing, and microsoft isn't, and soone they'll switch and we'll all be right back here again.

    If it can stay like that long enough, then maybe that's something to talk about.
  • Hard to believe (Score:4, Insightful)

    by bwt ( 68845 ) on Thursday August 22, 2002 @02:08PM (#4120887) Homepage
    It seems to me a hard thing to believe that a swing that big could happen in a month. Anybody have any explaination for how it happened.

    Maybe people are finally deploying the new version of Apache? Or is this just sampling error?
    • Re:Hard to believe (Score:2, Informative)

      by highcaffeine ( 83298 )
      Domain name parking.

      It happens every once in a while. Places like Register.com, Verisign, et. al. that put up those "site coming soon," "buy this domain for $x," and other messages switch their web server software every so often. Don't ask me why they do it exactly, but when it happens it can create a very sudden change in the numbers.

      This report, I believe, had a note that one of those domain name parkers switched to Apache from IIS, which probably accounted for at least a noticable chunk of the 6% swing.
    • Re:Hard to believe (Score:2, Informative)

      by jkramar ( 583118 )
      If you read the article, one of the first sentences you would read would be:
      Microsoft loses around 6% share this month, as register.com continues to fluctuate between using a Windows and Linux front end, and homestead.com, which originally based its business model on support from advertising, cleared away over a million sites.
    • Read a little further into the article:

      "Microsoft loses around 6% share this month, as register.com continues to fluctuate between using a Windows and Linux front end, and homestead.com , which originally based its business model on support from advertising, cleared away over a million sites. Homestead recently raised $5M from its investors to assist its transition to a paid for serivces model. To complete a bad news month for Microsoft's share of the survey, Reuters reports that the Federal Trade Commission will investigate the relationship between Verisign and Interland with respect to marketing domain names. The NSI domain parking system hosted at Interland is the other large repository of parked sites on the Microsoft platform. Earlier in the year large numbers of sites were reaped at Namezero, which had a controversial relationship with NSI regarding reselling domain names."
    • Re:Hard to believe (Score:5, Informative)

      by sab39 ( 10510 ) on Thursday August 22, 2002 @02:18PM (#4120956) Homepage
      Try reading the actual survey. The Ominous Cow Herd and the /. editor that posted this should have done the same thing, and realised that it's not so significant, but that would be asking way too much for /. :)

      What it boils down to is that a couple of mega-hosting-companies that host thousands upon thousands of domains have made changes that have a disproportionate effect due to the sheer number of sites hosted on their systems.

      The survey also shows results for "active domains" which are much more realistic, and while these still show a gain for Apache, it's much more modest and therefore more believable. But it doesn't make for such good headlines...

      Stuart.
      • Re:Hard to believe (Score:2, Interesting)

        by DustMagnet ( 453493 )
        You are right that "active domains" is a more accurate pictures, but what I found most interesting is that Microsoft actually lost sites under active domains, while Apache picked up almost twice that loss.

        I hate misleading catchy titles too. Slashdot is almost as bad as the tech section on abcnews.com.

    • Well, Apache did just release v 2.0. Now, it is debatable whether some of these are test/redundant machines running essentially mirrors which could account for the increase. But this doesn't account for M$'s decrease. It is possible that the new functionality in 2.0 was enough to convince some poor sysadmins tied to IIS.
  • ..what this survey won't tell you is what products are being used for truly mission critical, complex, back-end processing (high, medium, or low volume). That's where companies will spend their money on licensing (not as important for OS obviously), support, training, etc.

    So, yes, this is a victory for Apache, but it's a bit hollow. What's the point of having all the market share if very few people make not very much money on it?

    One could say that I'm expecting too much of these numbers, and that might appear to be true. But let's not forget how many times the Netcraft numbers have appeared here and how often people will gloat about Apache's figures.
    • What's the point of having all the market share if very few people make not very much money on it?

      Nothing, if your view of importance involves only money.

      But some people play this game only to build the best web server possible. The Apache project includes a lot of these types of people, and right now (according to Netcraft) they're winning.

      But don't worry too much. If you select web servers based on how many people make how much money off you, then you can still choose (what I presume to be) the leader by that criteria.

      • by Da VinMan ( 7669 ) on Thursday August 22, 2002 @02:26PM (#4121065)
        "But some people play this game only to build the best web server possible."

        I suppose it's debatable, but I'll argue that market share has nothing to do with having the best product.

        I believe that many people on /. look at the Apache numbers as some sort of "is Microsoft dead?" meter. They seem to think that if Apache can maintain a big enough market share for long enough, that it will either kill Microsoft or make Microsoft irrelevant. There's no doubt that it could hurt Microsoft's a teeny bit, but that's about it.

        But don't worry too much. If you select web servers based on how many people make how much money off you, then you can still choose (what I presume to be) the leader by that criteria.

        Exactly! And what I'm saying here is that Apache may or may not be a valuable part of many companies' mission critical applications. To me, that's valuable. Just knowing how many functioning instances of Apache are out there isn't too valuable for most businesses in my opinion.
        • No, market share doesn't necessarily mean best product, it's not like netcraft publishes a pro/con review anyway.

          There's no doubt that it could hurt Microsoft's a teeny bit, but that's about it.
          More than a teeny bit, considering that they make a lot more money off of business apps than home apps. You don't think Bill would love the server market? That's where they are currently devoting the greatest share of their attention.

          And what I'm saying here is that Apache may or may not be a valuable part of many companies' mission critical applications.
          Maybe maybe maybe...for people who have used both, they will almost unanimously tell you that apache is MUCH more stable and secure. If you have something mission-critical, you run it on apache. End of story. Throw some stats otherwise if you have them, but you likely won't find them.

          No way in hell I'd use IIS. I could show you server logs of how many CodeRed-style viruses have hit my server (to no avail of course - many before Norton can patch them. Had I been using IIS - kablooie!
        • I believe that many people on /. look at the Apache numbers as some sort of "is Microsoft dead?" meter.

          Actually, I think of it as more of a "has Microsoft taken over everything in the known universe?" meter... or maybe a "is the idea of Free Software in the business world actually tenable" meter.

          Apache, much like the old BSD TCP/IP, is one of those shining success stories in the Free Software world, so I think it's all right if people cheer when it does well in a fluffy survey with pretty graphs. Of course I guess it would be pretty damn silly to have a Netcraft-style survey of networking protocol suites, but I would still cheer to see that TCP went from 99.87% to 99.88% this month :)

        • by Anonymous Coward
          I would agree. If the /. crowd insists that Apache's market share makes it the best web server, then Windows market share would make it the best OS.

          You nutjobs can't have it both ways. Saying market share is irrelevant when talking about OS "quality", then saying it is relevant when it suits you is fscked up.
    • I can tell you right now that my multi-billion dollard ecommerce site uses strictly Apache.
    • The point is this:

      If IIS's market share ever reaches a critical point (probably 80-85%), you can guarantee that the next version of IIS and IE will only work with each other.

      And then all the Windows desktop users will upgrade, and then Microsoft will have won, because anyone still running a non-IIS web server will be screwed.
  • the second graph doesn't show a sharp decline in IIS usage...so are IIS users switching to apache, or are new sites choosing apache over IIS...or both?
    • It's all about the Switch.

      I'm pulling this out of my ass, but every new Mac OS X install comes with Apache as the webserver.

      Put a few thousand of those personal/small biz websites up there and you might add something to the numbers. Not 6% or anything crazy like that but I'm sure it's had some effect.

      E
  • Read the Fine Print (Score:3, Informative)

    by SQL Error ( 16383 ) on Thursday August 22, 2002 @02:16PM (#4120936)
    From Netcraft's site:

    Around the Net

    Microsoft loses around 6% share this month, as register.com continues to fluctuate between using a Windows and Linux front end, and homestead.com , which originally based its business model on support from advertising, cleared away over a million sites. Homestead recently raised $5M from its investors to assist its transition to a paid for serivces model. To complete a bad news month for Microsoft's share of the survey, Reuters reports that the Federal Trade Commission will investigate the relationship between Verisign and Interland with respect to marketing domain names. The NSI domain parking system hosted at Interland is the other large repository of parked sites on the Microsoft platform. Earlier in the year large numbers of sites were reaped at Namezero, which had a controversial relationship with NSI regarding reselling domain names.
  • Hits/Server Type (Score:2, Interesting)

    by zmalone ( 542264 )

    It would be interesting to know what kind of traffic each web server tends to take. It just seems to me that the end user perspective would be important here. If 90% of the machines out there run IIS, but the average person mostly accesses Apache machines, which would you say is the dominant server software?

    Unfortunatly, I cannot think of an easy way to do any such overview.

    • Unfortunatly, I cannot think of an easy way to do any such overview.

      How about porn sites as a good indicator?
      I once heard that 10% of the sites on the internet got 90% of the hits. I had also heard that the majority of these were porn sites. Disclaimer: These are obviously hearsay statistics.

      So what would be really interesting is to see what the major porn sites are running, the major news sites, and the major search engines. Google must account for a huge amounts of hits.

      Streching the porn argument further, it would be interesting to see a report on which webservers push the most bandwidth. I work for a small web host and had a customer tell me that she talked with a guy that had once pushed 1.5 GB of data transfer in a month on an NT server. I then replied that I had a Linux server that had pushed out 1600 GB in a single month. (This was a plain white box PC with mild hardware by today's standards.) She had no idea such things were possible. Again, these kinds of numbers would be difficult to get, but very cool to view.

      • Well, really.. thats a silly example. Obvioulsy the customer has a fuzzy view of nadwidth and file sizes: My brother uploads 1.5 GB of Anime on a Win98 server every day or two. I mean, 1.5 GB... thats less than 3 CDs full of data. I don't think it matter what OS you are using for that, so all it proves is shes confused.
        Still, 1600 GB is a good amount and i bet that pathetic NT server would have had to reboot at least 6 times to manage that.
  • by dpilot ( 134227 ) on Thursday August 22, 2002 @02:42PM (#4121286) Homepage Journal
    While one may argue that a web composed of Apache servers is more secure than a web composed of IIS, it's better yet to have a plurality of servers. I know we're not down to one or even two servers, yet. But we should keep in mind that "World Domination" by a web server (or even OS) is not a particularly good outcome.

    It's unfortunate that the computing industry has this completely stupid "One Leader" mindset. First it was IBM, then Microsoft, and even if Linux is a more benign "One Leader" than the other two, it's the "One" in there that's flawed.

    We need to re-orient our heads a bit, and IMHO the Linux camp is almost as guilty as the Windows camp. Mitigating factor, at least the Linux camp begrudges *BSD the right to exist, but is it any better than the Windows/Mac relationship?
    • The Virtuous Cycle (Score:4, Insightful)

      by cpeterso ( 19082 ) on Thursday August 22, 2002 @03:34PM (#4121823) Homepage

      The emergence of a de facto standard is inevitable. It's called the Virtuous Cycle. Users and vendors are more likely to use and invest in the market leader, which increases is market lead. Rinse and repeat. For better or worse, that's why Windows rules the desktop today. It's easier and cheaper to just support one standard.

      • The virtuous cycle is only reality because binaries from one platform don't run on other platforms...

        Whether the solution is some common runtime like IL or the JRE or some of the smaller OS backers swallowing their pride and running binaries from larger platforms, there is a solution. I'm not sure the 'winner takes all' phenomenon is a fundamental part of the computing landscape.

        One more beer-laden post....
    • An important item to note, is that in the Linux sector, there isn't just ONE item. Linux has many similiar but distinct distributions. There are even several variations of the kernel. Mix and match to your hearts content. To say that there is ONE Linux is to be just as blind as saying there is one Human. While they are all somewhat alike, they are all still quite different.
  • Ah... look again (Score:5, Insightful)

    by stonewolf ( 234392 ) on Thursday August 22, 2002 @04:19PM (#4122249) Homepage
    If you look down the page for the number of active websites instead of the number of total websites you see that Apache gained 1.43% and IIS lost 0.51%

    There is a lot of difference between active sites and "parked" inactive sites. On the netcraft site you can find a good discussion of the difference between total sites and active sites.

    The 6% swing seems to be the result of some major web parking sites switching to Apache.

    I love to see open source advance, but we need to be as honest as possible in selecting statistics to present.

    Stonewolf
    • I love to see open source advance, but we need to be as honest as possible in selecting statistics to present.

      It's hard to be honest with statistics when everyone around you is being dishonest with them. As soon as you put a percent sign or something next to your figures, people seem to believe it without question.

      Anytime there's a survey with a figure that even marginally supports Microsoft, it will be played up as much as possible by them, which makes it difficult to compete if you stick to honest stats.

      Along the same lines, take a look at the megahertz ratings on PCs and see how the public treat them - the figures are almost completely meaningless, but people still use them as one of the major determining factors in which PC to buy. So what has AMD done? They've tried to rig their side of the equation.

      All is fair in love and war... and the computer business definitely counts as a war these days.
    • by g4dget ( 579145 )
      Microsoft cites the same statistics when the wind blows their way. And it is meaningful when a large site that parks domains decides that IIS doesn't fit the bill.
  • Other news (Score:5, Funny)

    by Snafoo ( 38566 ) on Thursday August 22, 2002 @07:32PM (#4123609) Homepage
    Other holds steady at 12%.

    Although not as popular as Apache and IIS, Other offers an incredibly scalable and diverse development platform; and although its share may wax and wane over time, it never drops to zero: In Other surveys of Other products and markets, not once has any Other competitor been successful at completely eliminating this tenacious brand. (Some Independent checks by Some Independent regulatory body ensured that the numbers remained accurate, although that Some-body could not be reached for comment.)
  • It looks like Apache's share hasn't had 2 months of consecutive growth or shrinkage since early 2001. We'll have to want until next month to see if this is the start of a trend, or just one more blip.
  • This survey also doesn't cover large corporate internal networks. I've found that a surprising amount of corporations will use web-based interfaces for transmitting information over the internet. The last large corp I worked for used some small IIS servers (which incidentally, all of us in the web-development branch secretly hated) for the INTRAnet, and apache webservers for both the INTRAnet and INTERnet services.
    • This survey also doesn't cover large corporate internal networks.

      It doesn't cover small corporate intranets either!

      I had a contrating gig last year with a large pharma - they were using Netscape Enterprise on HP-UX. For some reason Netscape Enterprise seems to be very popular with VERY large companies; General Electric etc.

  • yes, apache rocks. I run it on my win2k server and it is stable as a rock, go apache!

Physician: One upon whom we set our hopes when ill and our dogs when well. -- Ambrose Bierce

Working...