Ross doesn’t trust Microsoft’s approach to Web
SocialText’s founder, Ross Mayfield, nails why a bunch of my friends don’t trust Microsoft and are finding what Microsoft’s Web offerings quite boring (or, even worse, worthy of derision).
As I’ve been going around the world I’ve been meeting with many people who’ve built their companies on non-Microsoft stuff. Some of whom have companies worth billions of dollars now. Some of whom you’ve never heard about unless you read TechCrunch. Here’s 12 reasons Web 2.0 entrepreneurs like Ross tell me that they aren’t using Microsoft’s stuff:
1) Startup costs. Linux is free. Ruby on Rails is free. MySQL is free.
2) Performance per dollar. They perceive that a Linux server running Apache has more performance than IIS running .NET.
3) Finding tech staff is easier. There are a whole new raft of young, highly skilled people willing to work long hours at startups who can build sites using Ruby on Rails.
4) Perception of scalability. The geeks who run these new businesses perceive that they can scale up their data centers with Linux and not with Windows (the old “Google runs on Linux” argument).
5) That Microsoft doesn’t care about small businesses. After all, Microsoft is an evil borg, but Ruby on Rails comes from a single guy: David Heinemeier Hansson. He has a blog and answers questions fast.
6) That open source makes it easier to fix problems and/or build custom solutions. A variant of the old “Google or Amazon couldn’t be built on Windows” argument.
7) On clients, they want to choose the highest-reach platforms. That doesn’t mean a Windows app. Or even an app that runs only in IE. It must run on every variant of Linux and Macintosh too.
8) They don’t want to take shit from their friends (or, even, their Venture Capitalist). Most of this is just pure cost-control. I can hear the conversation now: “OK, you wanna go with Windows as your platform, but is the extra feature worth the licensing fees for Windows?”
9) No lockin. These new businesses don’t want to be locked into a specific vendor’s problems, er products. Why? Because that way they can’t shop for the best price among tools (or move to something else if the architecture changes).
10) More security. The new businesses perceive Linux, Apache, Firefox, and other open source stuff to have higher security than stuff built on Windows.
11) More agility. I’ve had entrepreneurs tell me they need to be able to buy a server and have it totally up and running in less than 30 minutes and that they say that Linux is better at that.
12) The working set is smaller. Because Linux can be stripped down, the entrepreneurs are telling me that they can make their server-side stuff run faster and with less memory usage.
Now, why am I telling you this stuff? After all, I’ve just given you a list of perceived competitive advantages for Linux, Ruby on Rails, MySQL, and others. Isn’t this yet another example of why Scoble should be fired for being negative on his own company?
No.
See, I don’t want uninformed customers. That doesn’t help me. It doesn’t help Microsoft. It doesn’t help the customers. I want you to ask your Microsoft salesperson the tough questions before you buy into any of our new Web stuff. And, I start with the presumption my readers are smart enough to use Google or MSN or Yahoo to find out this information anyway. If you don’t get the right answers from Microsoft when it comes time to consider new Web technologies/methodologies/tools, er, if we don’t answer these points above, then I want you to run to the competition (and I’ll help you go there, just like I did when I helped run a camera store in the 1980s). And, when we bring services out, or bring new Web strategies out, I want you to trust us because we treated you right and gave you all the information.
Thanks Ross, though, for bringing your distrust out into the open. That’s helpful cause at least we can work on it now. And deal with it openly, without FUD, is what we’re going to do. Or, we’re going to be fired. That’s my cautionary tale to everyone inside Microsoft. Pay attention to this stuff or you’ll find yourself working somewhere else cause the customers went somewhere else.
What do you think? Did I miss anything in my list of 12?
November 1st, 2005 at 2:54 am
I know you touched on this but some people think it is “leet” to use linux. They think they will be in the “in” crowd, and therefore good with the slashdot people.
November 1st, 2005 at 2:58 am
I would have to consider using OpenSource if I were starting a company, after all who wants to hire someone just to understand the Microsoft licensing scheme. For instance, when I deploy a Microsoft Server product, the first thing it asks me is if I want to run it using a per processor license where each client needs a license, or do I wnat to run it in per client where each connecting client needs a CAL? WTF? I want to buy a server, turn it on and have the owrld connect to it for one price, the price I paid for the damn software! Has Microsoft considered the impetus behind the rise of opensource? Hmmm could it be the fact that MS licensing has gotten so restrictive, ala the ACTIVATION scheme? Imagine all the servers running unlicensed out there 10 years ago compared to today. I am guessing there are less unlicensed servers today, and you are squeezing nickels out of your loyal customers, but you aint spreading your software the same way as you used to. Microsoft in someways was the original shareware, much to their chagrin, however consider the HUGE profits made in the early days when software wasn’t so locked down, make weird sense doesn’t it. Microsoft helped create the opensource market, where people like me, well smarter than me are building their own software and hey guess what, it works well enough.
On the other points, I work with opensource vendors, and believe me they make it up on consulting dollars, where with Microsoft products and tools from ISV’s I can implement a solution in a few days, I see my opensource vendors struggle sometimes for weeks rewriting code. They also suffer when trying to scale out PHP applications to handle thousands of simultaneous connections, CHOKE! They often don’t know why, opensource is supposed to be so good? Microsoft, IBM, Oracle these companies have all been down the paths of customers needs, at the levels that I am talking about. Amazon, Google, their solution seems to mostly be throwing single purpose hardware at their scalability needs, which lately has become cheaper than licensing Micrsofts expensive software to manage capacity. Now I have done it, I have talked myself in to open source. Sorry Robert!
November 1st, 2005 at 3:03 am
Robert, of all the reasons you’ve named - whether merely “perceived” or fact - two of them are at the root of what MS has to address.
“That Microsoft doesn’t care about small businesses.”
Well, if they do care, they have a funny way of showing it. I work for a brick and mortar company of 2000. About half of the IT machines - forget about the others - still run on Windows 2000. Why? Licensing. I was able to teach myself VB and COM programming and eventually get my MCSD 7 years ago. At a cost of a few hundred dollars. Today? Please, don’t tell me about the starter kits.
If I wanted to kickstart my career again (like I did by downloading the FREE VB5 CCE) I would easily recommend putting Apache/PHP/MySQL any day over what Microsoft offers. By the way - those packages offer easy/free installs for Windows machines. Does Microsoft offer something similar for Linux/Mac?
This leads into the second reason….
“More security.”
Yeah yeah, I know. A better and more secure server OS, IIS, and MSIE. But what about that half of my IT department who runs on Windows 2000? They are not only “locked in” to Microsoft, they are “locked out” from anything Microsoft releases unless they ante up quite a bit.
We’re a medium-sized company all told. And the actual costs of maintaining a Windows network is something that I’m sure no small company could EVER afford.
November 1st, 2005 at 3:59 am
[...] You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your ownsite. [...]
November 1st, 2005 at 3:59 am
Another one for the list:
Universities use, and hence teach, open-source software.
Pretty much any student going through a Computer Science-type University course is going to be doing Java, using Apache, Tomcat, etc. Universities and the academic community have always been a big user of UNIX systems, and consequently are embracing Linux too. This translates into a skillset and experience for its students.
Many University lecturers also buy into the “Microsoft is the devil” mindset, and this too is transferred to the students. Consequently, an entire generation of Computer Science graduates leaves University with a pro-open-source mindset, and the skills and experience to match.
Mark
November 1st, 2005 at 4:16 am
[...] Thank you Robert, for addressing the issues that affect everybody thinking of starting a new venture, Web 2.0 (yuck) or not. [...]
November 1st, 2005 at 4:16 am
John (Microsoft Weblog): I am on a spiritual journey, but I’m not leaving Microsoft. The world is about to change, though, you’re right! Microsoft is going to need to change with it.
November 1st, 2005 at 4:31 am
12 reasons MS doesn’t cut it for web development
Robert Scoble has an honest and poignant analysis of why the Microsoft suite of tools just doesn’t cut it for…
November 1st, 2005 at 5:06 am
Number 1) is indeed a problem for startups. When I started my company, I really wanted to use Microsoft products because I felt comfortable using them. But I couldn’t afford them (I do use MS products now).
So I guess many companies are starting out with free software purely for financial reasons. If they manage to be successful, it’s likely that they stick to the technology even if they had problems with it (never change a running system). Afterwards it might turn out that a MS solution would have been cheaper overall, but that doesn’t matter anymore. What counts is that you don’t have to spend any money upfront.
As for the other points, I think that a lot of them are purely based on perceived advantages of Open Source and a generally negative attitude towards MS. Almost anyone that I talk to who has a negative attitude towards MS products turns out to have only very limited knowledge about the products and technologies.
And even otherwise reasonable people like David Heinemeier Hansson sometimes dismiss MS products based purely on speculation and misinformation. I remember him posting a dismissal of ATLAS which was just ridiculous. Actually that’s the reason why I stopped reading his blog, although Ruby on Rails as a techology is interesting.
November 1st, 2005 at 5:53 am
Well, my 2 cents on why I don’t use ASP.NET in my next project: because it get’s in my way… i want to have full control on the output of the page and this seems unfeasible. ASP.NET puts viewstate on your page, unwanted javascript which are not easy (nor apparently possible) to disable completely without giving up the functionality that makes ASP.NET “productive”. Also ASP.NET lacks a Publisher Handler like mod_python has.
November 1st, 2005 at 5:55 am
So I’m sitting here wearing my nice Microsoft/Adidas t-shirt that I got off the ecompanystore site. Some people wear a t-shirt of their football or hockey team, so why shouldn’t I wear one of a software company occasionally - it’s comfortable for $45. But the funny thing is that people then think I work for Microsoft… and it almost makes me wish I did.
But there’s another reason I feel a surrogate pride for Microsoft, and that’s because I’m proud of what I’ve learned. As a CTO I make it my job to know as much about every possible technology and benefit that Microsoft can bring.
Also, there just isn’t enough bandwidth in my organisations to try both ways - we are all in on Microsoft.
There’s a downside here - drowning. Robert has mentioned that some of things he’s seen recently have made him happier than ever about Microsoft (paraphrasing). They excite me too - what I’ve been shown (and I’ve seen very C9 video in the last six months because of the value per minute, thanks Robert) - but they also make me realise what an intense and (happily) tiring 12(, 24, 48) months are ahead of me, given all the things that are coming out in one go. That’s a lot of competitive attack surface because there’s a lot of platform to leverage.
I think of Web 2.0 as an HTTP-based platform capable of convergent technology solutions. Take ATLAS, it’s a huge thing, but it’s a last minute tack on. Even Microsoft is playing catch up, and because I’m along for the Microsoft ride, I have to ride the coat tails, but can go only as fast as Microsoft can catch up if I am to wait for the productivity benefits such frameworks will provide.
So I’m saying that Microsoft is a real investment in your life, not just cost, and it’s an intense one to make right now. How many Win32 API gurus are there? Right and how many WinFX gurus are there (and that’s much much smaller)?
I did Unix for 3 years, 13 years ago. It was a blast, but with Microsoft I do know what I’m getting myself in to - it’s an intense ride, but right now, it’s exciting, and I don’t expect it to stop - I could ride all decade…
November 1st, 2005 at 6:46 am
Web Development on a Microsoft Platform
So thats why I hesitate to chose to develop for Microsoft platforms these days. Lack of clarity and flexibility in the licensing structures, and no light-weight scripting language to turn web apps around quickly and easily.
November 1st, 2005 at 6:46 am
Where I work, we are a Microsoft house, which is good and bad. From the web side of things (I’m the webmaster), most of everything has been good. Our SQL Server has been the most reliable server ever. The webservers have been fairly good too, although we have had some problems (like FrontPage extensions allowing anonymous access by default instead of being secured by default), but they have been fairly reliable once we figured out the best way to handle them.
ASP was good but ASP.NET has been great. It’s pretty powerful and I’m able to develop apps so much quicker than I used to be able to. I still refuse to use Visual Studio though, unless I’m building a Windows App, because it’s always been too intrusive and changes too much code as well as add a bunch of extra crap that I don’t need. However, it looks like they’ve finally fixed that with the new versions and I’m seriously thinking of switching to Visual Web Developer, although it’s missing a few important things that I’m not sure I can live without. My biggest complaint has been the lack of XHTML compliance as well as cross-browser JavaScript issues, but it appears that both of these things have been fixed in the 2.0 version as well.
So, from the web side of things, I’ve been pretty happy with Microsoft. Now for the rest, it seems like we are always having problems with our Exchange server and there are always random network issues because of some of the other internal servers. We had a massive Active Directory failure a while back that kept me from being able to do work for two days because Active Directory somehow deleted a bunch of our accounts randomly by itself. From Windows 2000 on, the desktops and laptops have gotten more reliable, but they still are about half as reliable as OSX is. I see our company wasting a lot of money trying to fix these things and it’s frustrating.
So, in my opinion…Microsoft for the web = good, Microsoft for everything else = not so good.
November 1st, 2005 at 6:50 am
Wow, not even one minute after I posted that, the Exchange server went down for one of our other facilities. I think we probably have more problems with Exchange more than any other single Microsoft product.
November 1st, 2005 at 7:02 am
1. This isn’t the most important thing to embrace for Microsoft as this window of VC funding won’t last for ever.
2. Microsoft should offer some type of entry level price for start-ups and such where what the start up pays, grows with how much business they do.
3. Microsoft has been working on modularizing Windows Server, and I am sure they will get better at it. They are doing to best when it comes down to monitoring systems, and ease of use. It seems like there is so much FUD floating around to even carry on a noraml coversation with some people.
4. Two things Microsoft needs to work on, making Microsoft’s tools accessible to low budget business, like a startup, and making their tools and servers more modular.
November 1st, 2005 at 7:04 am
Why I use Linux instead of Windows
First, this is mostly in relation to a post by Scoble here earlier. Before I begin in my rant, let me say the only places I really use Microsoft products: On the desktop as a gaming platform, and on a fileserver/print server/active directory system …
November 1st, 2005 at 7:09 am
“Did I miss anything in my list of 12?”
Yeah, the very fact that in 2005, just a couple of weeks ago, MSFT was AGAIN caught abusing its monopoly power by trying to exclude its competitors in digital media bundling, despite the consent decree. Transparent, kindler, gentler, more emphatic, better listening MSFT? My foot.
November 1st, 2005 at 7:20 am
[...] Robert Scoble: “Here’s 12 reasons Web 2.0 entrepreneurs like Ross (Mayfield) tell me that they aren’t using Microsoft’s stuff” No comments yet. [...]
November 1st, 2005 at 7:37 am
Anona - what are you talking about? Are you talking abuot Media Player, which has been included in the OS since the dawn of time? The fact that Windows, despite including WMP, did nothing else to favor it over competitors? (ie. Many OEMs shipped with different default media players).
Personally, I think that’s rubbish.
November 1st, 2005 at 7:39 am
I am all for start-up pricing. It is a big issue I am dealing with right now with a start-up in the printing industry. If you want to test it out in a pilot program, please contact me.
November 1st, 2005 at 7:39 am
As the founder of a software startup, I faced this very decision (and picked Microsoft). Though the points made here are very strong and many of them hit home with me (as I am living with the consequences).
One point that was not mentioned (that I think is worthy of being on the list):
Startups often believe that they’re going to sell to Google or Yahoo! and that their chances of being acquired are lower if they build on the Microsoft platform.
November 1st, 2005 at 7:45 am
For someone like me that is using technology in a non-profit (a church to be exact), there is no way that I could recommend using Microsoft products to start up. Unless you have a large budget or a grant, it just cost to much money to use the MS solutions. Why would I use a technology that I constantly have to baby sit and worry about as apposed to a technology that I can install and just get work done. While I realize that every solution has problems, I have used Microsoft, Apple, Sun, Linux (many flavors), and over the long run, even though I can set up some microsoft things faster they do not run as well!
November 1st, 2005 at 7:51 am
You left off what I would think would be the biggest one:
If I am starting a Web company, why on earth would I want to be dependent on a competitor for critical infrastructure for that business?
Microsoft competes with all these startups. If any of them find a good business model, Microsoft is likely to jump into that market. It’s done so a million times before.
Given that, why would you want to put yourself in the position of having this huge potential competitor deciding what your licensing fees are for your critical software? Or, for that matter, having to call them when things go wrong?
Your point #1 (”Linux is free. Ruby on Rails is free. MySQL is free.”) gets close to this issue, but not close enough. It’s not that Linux, RoR et al are FREE that makes them attractive — it’s that _nobody owns them_. If I’m running a startup based on Linux and Red Hat decides to compete with me, I can switch to Debian or SuSE or Ubuntu. If I’m running a startup based on Windows and MS decides to compete with me, where do I go?
November 1st, 2005 at 8:04 am
I don’t have the time to go over all twelve, but I’ll touch on one: Security.
Security? You’ve got to be joking! Take OpenBSD, for instance. There has been only one remote hole in the default install, in more than 8 years, and that’s more than Windows could EVER claim. Beat that, and *maybe* I’ll think about using a Windows Server, but until then, I care moe about my security than about how great Microsoft and Ballmer’s ego is.
November 1st, 2005 at 8:06 am
I just recently switched from ASP to .NET 2.0, deciding to completely leapfrog 1.1 because it looked too cumbersome and bloated.
Maybe I’m in the minority here, but I really like what MS has done with .NET 2.0. Using Visual Web Developer I was able to whip out a wedding website for some friends in minutes. MINUTES! This included a database backend for content and photo storage. Master pages for managing everything around the page, navigation, sitemap, etc.
This type of development is definitely targeted for the little guy. $50 development tool, free SQL Express. They can tool around with it on their own XP box without even installing IIS.
Then the little guy looks online and runs into the problem of hosting. This is where I think MS loses future MS-based developers. Hosting sites advertises next-to-nothing PHP/MySQL hosting, but the .net hosting is usually lumped into the co-located/expensive packages. Nobody’s going to put a little wedding site on a $100/month host, so the little guy switches gears and learns how to do the site in PHP.
I’ve seen plenty of people do this, and it has everything to do with hosting. After a few months of tinkering with PHP, they’re using that exclusively. Granted, these aren’t developers creating groundbreaking websites, but they will be in a couple of years.
November 1st, 2005 at 8:12 am
>> Many University lecturers also buy into the
>> “Microsoft is the devil” mindset, and this too
>> is transferred to the students.
I’m new to OSS. Only been using it about 3 months now. You wouldn’t believe how much this has changed my views on MS.
It’s really hard to see why the OSS community seems to gripe about MS so much when you’re working under the MS umbrella of technologies. Then you step out for a bit, and you get a taste of what they’ve been complaining about.
Excuse my language, but why, oh why does IIS6 even exist? On the MS side of the fence I think “More security, and a nicer MMC”. I must’ve had pretty low expectations I suppose. On the OSS side of the fence I wonder what PHB decided that the most popular web-servers in the world use modular capabilities, but IIS6 is staying with ISAPI and only ISAPI.
See any FCGI or SCGI ASP.NET modules out of MS?
Fine. How about languages?
Ruby says: “You have the power to do very bad things (that might be useful, so be careful)! Like extending *any* object you want, including Object itself.”.
c# says: “You can’t be trusted to inherit from Form. You can’t be trusted to override method behaviour unless the original author intended you to be able to.”
And Databases? Where’s the Linux MSSQL driver? I have to install FreeTDS and connect with ODBC because MS can’t be troubled to allow customers to use their software on other vendor’s systems?
How about IE? Do I use Firefox because of security? No. As a user, I don’t really care much. I care much more about tabbed browsing, a spiffy interface, and (indirectly) standards compliance because it enhances my web experience.
In the past year I’ve gone from being a blind supporter of MS, to ambivilent, to a pretty negative opinion of MS. If MS wants to change that, they’ll need to be more open. That doesn’t mean creating an IE blog. It means responding to developer questions on that blog in an open and straight-forward manner without being handcuffed by the lawyers. It means realizing MS doesn’t own the world, and a little flexibility with what their products will interoperate with is not a bad thing. It means instead of taking the stance that developers can’t be trusted not to do _very bad things_ in a language (bug-related, not malicious-use related), they need to take the stance that the .NET Framework would’ve been much farther along in 3rd party tools if they hadn’t crippled it with a bunch of sealed classes and default “final” methods.
MS needs to learn how to be enabling again, instead of trying to force bundled products and platforms on me.
November 1st, 2005 at 8:19 am
What do you think? Did I miss anything in my list of 12?
Right off, you missed the entire section titled “And here’s what Microsoft is going to do about it.”
But bigger than that is your bad attitude. You keep writing these are “perceived” negatives. As a technical professional, my studied and informed analysis of the pros and cons of my options is not just my perception. We’re not pulling these reasons out of our asses or basing them on what we read in ad banners, we’re measuring you up and you’re failing. And then you have the gall to write that we’re just seeing things wrong?
So you should also add attitude to your list. This is more than your “Microsoft doesn’t care about you” point, it’s “Microsoft thinks you suck”. When I go drop in on the developers or users of open source software I’ll get folks who want to give me a hand and see me succeed with their tools, not jerks who tell me I can’t do my job.
November 1st, 2005 at 8:38 am
Hi Robert
Great article I just love what your doing there at Microsoft, I just hope they don’t let you down. Heres my reasons for using non-Microsoft technology on projects over recent years.
1) Agility, Our clients require evolutionary products that develop with the user, the old version 1 and version 2 mentality doesn’t cut it anymore. We have to turn on a dime building web apps and expect our tools underneath to do so likewise. (No waitinmg for vista to ship!)
2) Opensource being able to tweeak and fix stuff really helps, sometoimes it’s much faster to fix it yourself than wait for a vendor. (particularly so given point 1). It is also great to have access to teh code sometimes as the codes itself is always the best documentation. Also ones opinions and requirements tend to be taken into account with opensource products, microcoft seeem to have lost that and focused more on the bigger picture that we are meant to follow (Like a dictatorship rather than a community, I can see you are trying to change that however.)
3) Competition and choice we have moved away from single vendor lockin, we cannot afford to put all of our eggs in one basket and prefer to mix and match.
4) Platform agnostic, it must run on all platforms, we should not have to worry about that, our clients use different platforms serevr side and client side.
This is coming from a business who used to use a lot of Microsoft technology. But now we use very little apart from the odd desktop.
I’m sure Microsoft’s anouncements today will fix all of these things (they will have to to compete) and I am looking forward to it, as I love competition in the market palce and will use the best products to do the best job. It would be nice to see Microsoft products coming back into our toolset.
PS it’s much harder for you guys now also because you have to be able to work with the stuff we already have, just like the others, I guess operability should have been point (5) on my list
Looking forward to the announcements ;)
regards
Al
November 1st, 2005 at 8:39 am
When you call your Microsoft office to discuss moving elements of an Internet service infrastructure and the first questions they ask is how many employees are in your company and how many of their desktops are in the US, then I think you have a problem.
November 1st, 2005 at 8:46 am
Reason number 13: when I tell Red Hat what my business needs are, they don’t call them my “perceived” needs, or that I’m trying to sound “cool” or “leet” by talking about my needs, or that I’m a “zealot” because I place my strategic needs first.
November 1st, 2005 at 8:51 am
You may want to add Python and PHP to your list, while Ruby on Rails is great for prototyping and small use solutions (great for small companies), the Python and PHP can scale to greater degrees. Python may be the best development language out there regardless of cost for tools that must scale, but that is my take.
Additionally, the tools you mention also tend to have more graceful upgrades than seem to take place with Microsoft products. There is testing and some tweaking that is needed, but I have yet to see a products upgrades wreak havoc like Microsoft products do.
The fear of competing with Microsoft. Choosing an open source solution will likely never mean you are competing with the creator of the product directly. I have seen Microsoft many times jump the market with a new product that directly competes with many of its “partners” and third-party developers (yes, Apple is notorious for this too). Many entrepreneurs make the explicit decision to use open source for their product for just this reason, they may not mind competing with Microsoft (not anybody’s dream as a start-up), but they don’t want to have their marketplace taken away apples-for-apples.
Lastly, I many entrepreneurs want to get the best developers and designers possible, which often means letting developers and designers work on their preferred platform. In the top-talent developer and design community Microsoft is not dominant as LINUX and Mac are many times the preferred OS. Like Ross stated with the VCs about taking shit from their friends this very much the case with the top developers and designers, credibility does not come from using Microsoft.
November 1st, 2005 at 8:57 am
[...] Scobleizer - 12 reasons Web 2.0 entrepreneurs tell me that they aren’t using Microsoft’s stuff. via Loud Thinking. [...]
November 1st, 2005 at 9:23 am
Well…if you’re gonna force me into a migration (to .NET (which I’ve read you didn’t didnt’ even use for all of Vista)), thanks for giving me all the reasons why I shouldn’t consider it.
November 1st, 2005 at 9:36 am
Every time I use Word I remember why I avoid Microsoft products. Every time I talk to a computer user who asks me how to get rid of that annoying dancing paper clip, I remember why I avoid Microsoft products. Every time I can’t find simple information about your products using your Web site’s search engine, I remember why I avoid Microsoft products. Most of all every day when I’m forced to switch to IE in order to send replies via my hotmail account because using Mozilla sends gibberish, I remember why I avoid Microsoft products.
November 1st, 2005 at 9:37 am
It is my opinion that my open source software (e.g. Fedora & Firefox) is from a technical viewpoint much more secure than bill’s stuff, but of course, I could be wrong. What I do KNOW is that bill has all kinds of enemies, and if I use his stuff, his enemies become my enemies. Using his stuff would cause me to loss sleep, become very insecure, and perhaps develop (more?) psycological problems. (Sorry, bill)
November 1st, 2005 at 9:43 am
What has happened with ROR is important, not because of how productive it is, but for how they have simplified stuff for the casual user.
In both VB and C# things tend to take longer than you would like when you don’t need all the power and you don’t have the background in programming .net. Database connections are a good example, every book you check has a different method of wiring up connections, commands, adapters etc which can be great when you understand why, but trying to learn by doing is really hard.
What we really need is for an intermediate .net namespace with single line actions, i.e. Array = Go get stuff from this database, and your done.
its really tempting, when the dev tools are cheap and php / mysql is free to jump ship real quickly.
November 1st, 2005 at 9:56 am
I don’t see a lot of talk about Macromedia when these discussions come up Robert. Macromedia’s Flex allows Web 2.0 application development at a whole new level and does so on the Flash platform, which runs on Windows, Linux, Mac and Mobile and which almost everyone has.
It’s such a win-win situation for anyone looking to get into Web 2.0 but yet it wasn’t brought up in this conversation at all.
November 1st, 2005 at 10:00 am
Sam - these are my sentiments on the issue exactly. Microsoft does have some nice products (I hear great things about MSSQL) but I can’t even use them unless I commit to a 100% Microsoft platform. Why can’t I deploy Linux webservers talking to a MSSQL database behind OpenBSD balancer/firewalls? As a self-taught developer I started playing with the Microsoft offerings that were on my W2K box first. Then when I looked into publishing my work, there were no options - it was either pay out the nose or switch. I switched. Microsoft tools suck for the small shop, and like it or not thats where most of the development comes from. Don’t lock me in, don’t lock me in, don’t lock me in. Its MY BUISNESS let me do with it what I want, don’t require me to upgrade just to keep functionality. Don’t require me to buy something else to get one feature that I want. I should be able to purchase directly the exact software solution to the problem I want to solve. Right now I solve my own problems, it doesn’t cost me money - it costs me time. There currently is no way to flip that relationship around, and thats where I think Microsoft’s opportunity in the “Web 2.0″ is. Don’t lock me in, don’t lock me in, don’t lock me in, don’t lock me in….
vanderwal - any proof that Python/PHP scales better than Ruby?
November 1st, 2005 at 10:02 am
#36 (Ryan) - how much is it going to cost me to deploy it?
November 1st, 2005 at 10:26 am
Shame that Mr. Allinthehead (comment 12) has dropped a non-UTF-8 character into your blog.
All I see in my browser is a question mark.
What’s the betting it is an MS Windows character?
http://www.cs.tut.fi/~jkorpela/www/windows-chars.html
He may “hesitate … to develop for Microsoft platforms” but what’s the betting that got there because he composed his post in (spit) Microsoft Word?
November 1st, 2005 at 10:31 am
That’s a good list, Robert. When it came time to building my web venture, I left a two-year MSDN license behind and started learning open-source, moving through PHP to Ruby on Rails. Near-term cost was an issue, but I was also thinking about the wealth of code that’s available in the open source community and the costs of Microsoft databases and servers.
Eclipse is challenging Visual Studio, although VS was probably the sole reason my decision to leave Microsoft was difficult. Ruby is a nicer language to develop in than C++ or C#. MySQL and PostgreSQL does what I need and I don’t have to worry about SQL Server costs escalating in the future as I grow my business. And sharing code within the open source community is front and center, not an ugly stepchild. Compared to open source avenues, how many web frameworks and message board systems are available for free under .NET? The fact that Microsoft hasn’t ruled out legal action against Mono nullified that possible out.
November 1st, 2005 at 10:55 am
attractive start up technology
An interesting article on why ‘Web 2.0′ entrepreneurs scurry away from M$ products. Hasn’t told me much I don’t know, instead intent on informing potential customers of what else is out there.
For my small tech ventures, I’m hungry and bootst…
November 1st, 2005 at 10:55 am
[...] An interesting article on why ‘Web 2.0’ entrepreneurs scurry away from M$ products. Hasn’t told me much I don’t know, instead intent on informing potential customers of what else is out there. [...]
November 1st, 2005 at 10:55 am
Ah, you forgot the most important: Fun. Good programmers like things that they feel are fun to hack around with, because good programmers do their jobs for fun as well as for food and lodging. Good programmers who are having fun write better code than those who are pounding it out because the boss tells them to.
OSS has a culture of fun.
November 1st, 2005 at 10:59 am
Microsoft reminds me of Nero playing the fiddle while all around his empire burns. As an ex-microsoftee from the very early days when there was only DOS, Windows 2.1 and Excel. It saddens me to watch the end. But the end it is. Thank you Bill for all you have given me. Microsoft will continue exist for another xxx years ($50bn guarenttee that) just as IBM do today but only after they morph into their next corporate skin. Microsoft announce their new “live” services strategy today and maybe the license issues mentioned will disappear. Maybe the cost issues will be nullified. Maybe then Microsoft will have once again pulled the rabbit out of the hat. But if not everyone of the 40 comments will go unanswered and withit will go the customers.
November 1st, 2005 at 11:01 am
[...] More… (link on Ross added by me) [...]
November 1st, 2005 at 11:11 am
Brandon: I am talking about this that just happened:
“Microsoft has been trying to find ways to make portable music players that run its software compete better in a market dominated by Apple Computer’s iPod player and companion iTunes software. Under the program Microsoft had proposed, device makers that included a CD with Windows Media Player and other software would have had to agree not to include any other software, including rival media players.”
http://tinyurl.com/7lyf5
November 1st, 2005 at 11:13 am
Damian (comment 39) — that was a trackback, not a comment, and trackbacks are inherently broken w.r.t. internationalization. Every web application that consumes them has to make guesses and/or assumptions when it comes to consuming them. This isn’t just Microsoft’s problem, or WordPress’ or Typepad’s…
November 1st, 2005 at 11:18 am
I think Scoble maybe cleverer than you think. Tonight Microsoft announce their new “live” services strategy - think hosted services. Maybe Scoble is putting out all the reasons why “Opensource is better than Microsoft” so that after the announcement he can post the 12 reasons why Microsoft is now better than Opensource (or at least as cost-effective). In addition he has a host of comments here telling him how stupid/wrong he is in doing so. I think he will love nothing better than to post a very simple I told you so post. Lets wait and see. Over to you Robert ;-)
On the otherhand if Microsoft/Scoble do not make an immediate repost then I think Scoble has shot himslef by publically opening up this Pandora’s box.
November 1st, 2005 at 11:32 am
Mr. Scoble, A couple of weeks ago I emailed our government sales rep with a licensing question. We’re putting up an AVL solution that only runs on Windows. I wanted to know what I had to buy from MS so that I would not overpay nor be illegal. The sales rep referred me to the TAM. He asked what software we were going to run and then (in email!) started into a pitch about 3 different server products. The AVL will have 400 vehicles reporting through a cell network receiver. The only clients logging in will be admins, the AVL data will be delivered to 10-20 users via a browser. I don’t even have to ask this question of RedHat. The best answer that I received came from Software Spectrum. I still don’t know what license to buy, how many CAL’s, etc. I bought the cheapest license and sent my emails to the person on my staff in charge of licensing, in case the MS thugs from the BSA come around.
November 1st, 2005 at 11:42 am
d.w.: two days later I coded up a fix which was adoped as a part of the spec on August 1, 2004.
November 1st, 2005 at 11:45 am
You already got most of my reasons, but here’s one more. I choose Linux for my servers because it’s my development environment. I choose linux as my development environment because of several reasons, but mostly because the Windows command line is a joke (yes, yes, Monad, but that’s still in beta). Where’s grep? where’s locate? Where are my pipes, io redirection? It really grates on you when you’re used a powerful shell.
November 1st, 2005 at 11:45 am
The BIG issue is Trust. Who trusts Microsoft? Do you trust that they won’t abandon the path they led you down? Do you trust they won’t abuse your personal information? Can you trust a company that’s coming out with an operating system that lets Hollywood control your computer? That lets publishers disable features on your system? Why isn’t Microsoft suing Sony for installing evil DRM rootkit technology on people’s computers? Or releasing a patch to prevent it?
When someone buys a computer, they want to own it. They want the ability to control it. You can do that with open source. You’ll yield control if you trust the hardware and software. You’ll scream if you’re not able to control what gets installed, and what works, and what you can uninstall.
Microsoft hasn’t shown that it’s trustworthy.
November 1st, 2005 at 11:52 am
Sam,
We will be answering all these someday. Maybe not today. But, I wanted to get everything out on the table. This is how we are being judged on the street. As we try to get you to adopt new things, this list is helpful to look back at.
November 1st, 2005 at 11:53 am
I would agree strongly with comment #2 (and many of the following comments) - Complexity and unclearness in the licensing makes it very dificult to recommend MSFT solutions to startups, or even to large businesses looking at external facing applications.
I have yet to see an easy to obtain quote (and explanation) for how to build an unlimitted number of users web 2.0 application based on MSFT servers in a legal and fully licensed manner.
With open source systems in contrast I know already what the costs would be to scale up the number of servers - ZERO for the software, just the cost of the raw servers and rack space.
In working with my clients - both startups, investors and very large firms we rarely focus on the cost of a single server (perhaps a set of servers) but we do look at the overall cost and impact of a technology decision. With an open source solution we know that we’ll primarily be paying for specifically the software and services we need. With Microsoft based solutions we know that a noticible (and generally higher than the cost of the raw hardware) portion of our cost will be going to pay what we term “the Microsoft Tax” and that just gets us in the game, it is not typically the services or features we need, those cost addition amounts.
But then as the issues around licensing crop up, we find that we either have to invest a lot of time (and thus money) into understanding the nuances of MSFT’s licensing/versioning schemes or we can avoid all that by going with much cleaner and more scalable licenses from open source applications.
Consider planning an application that, if successful, will have 100,000 users. Assuming that it is more complex than a “simple” web app - do you really want to try to figure out how to buy 100,000 access licenses from MSFT? (yes, I know MSFT probably has some options that handle this case but even as a fairly sophisticated consultant, I’m not at all sure how to look for them)
- just take a look at http://selectug.mslicense.com/L1033/default.aspx
Please tell me that there is something better?
WHERE do I get a simple, clear, concise answer to “how much will it cost my startup to build, deploy and scale a MSFT based solution?” How much will it cost for our production systems? What about for our three tier deployment systems (i.e. development, testbed, production) along with backup/scale up capacity?
Shannon
November 1st, 2005 at 11:55 am
Sam — in finest Emily Litella mode — “nevermind”… I remembered the bloodletting but not the cure. :) Good to hear that the spec was revved and clarified. The rest of you, talk amongst yourselves.
November 1st, 2005 at 12:02 pm
Very interesting post and comments…
Quite a few people have commented on startup costs and licensing so I wanted to comment on both of those from an ISV perspective.
Startup Costs
- Of course Microsoft would want to have a means of providing startups with an easy and cost effective way to build their product (and run their infrastructure) on our platform and tools. That’s the motive behind the Empower program for Independent Software Vendors. Basically, for $350 you get an MSDN subscription that provides access to server products, development tools and SDK’s, desktop applications and operating systems etc. for development purposes as well as internal licenses.
Check it out:
http://members.microsoft.com/partner/asia/isv/readytomarket/understandproducts/licensingResources_empower.aspx
Licensing
- One pain point for companies building a product on top of something like SQL Server/BizTalk/MBS etc. is that the ISV would like to somehow embed the license into their solution. This would allow the ISV to sell their product to the end customer without having to require the end customer to somehow acquire a SQL/BTS/MBS license from a reseller (and therefore greatly simply the procurement of their product). Well, that’s exactly what the ISV Royalty program does. It allows an ISV to embed quite a few of our products right into their solution from a licensing perspective.
https://partner.microsoft.com/global/40010439?nav=ln
Competing with Microsoft
- One comment I thought I’d make on this. In the majority of cases you will find that you aren’t competing with Microsoft, you are competing with a company that has built a vertical solution on the Microsoft platform. I find this quite often comes up in the MBS (aka Dynamics) world. So you could go and write your own CRM system targeted towards a specific vertical, and then find you’re up against “Microsoft CRM” - when in fact, you’re up against another company who has built upon CRM by customizing it for that vertical.
SQL 2005 Express and the Visual Studio Express SKU’s
You guys know you can use SQL 2005 Express for free right? (and it even includes support for reporting). The fact that Visual Studio pricing was a barrier of entry for quite a few people was pretty much the motive for the Visual Studio Express line of SKU’s. Check both out here:
VS2005 Express
http://lab.msdn.microsoft.com/express/
SQL2005 Express
http://lab.msdn.microsoft.com/express/sql/
One final tangent
Another important factor to consider if you are planning on building a product, and ultimately take it to a wide market (which is kind of the whole point isn’t it?), is to also consider who you will partner with in order to make that happen. Meaning, look beyond just development, consider the sales side of things as well as that is just as critical to the success of your product. Ideally you want to work with a company that will help drive sales of your product as it’s in their best interest (since your product drives their product). I can honestly say, as I do this every day with all shapes and sizes of ISV’s, that Microsoft has a very effective and progamatic way of allowing an ISV to get started with our platform and tools, develop a product, and then take that to a wide market. This is all driven by the Microsoft Partner Program. I’ll be first to admit that there’s a very large surface area to that program but I work with ISV’s every day who leverage this program with huge success. I mean, why not let Microsoft help drive sales of your product? Let’s be honest here, if you built a Microsoft solution of course we want to help you be as successful as possible selling that solution as we’re also driving Microsoft product in the sale.
cheers,
Ryan Storgaard
http://blogs.msdn.com/stoey
November 1st, 2005 at 12:03 pm
Great post Robert and I agree that this is exactly the dialog needed to expose the issues and FUD. In looking at the points you listed however, I have to wonder what this has to do with the “web” specifically? My complaints about Microsoft’s approach to the web are more about a lack of stewardship and compliance with standards. That I still have separate CSS for browsers and many, many compatability issues makes web development very hard. Second, why did it take Firefox to light a fire over there in Redmond to do something about IE? With the focus on smart clients and .net, it looked like the Microsoft browser was purposely put on ice. Just think where Microsoft would be if the development of the browser was a priority. And lastly, the current ASP.NET implementations drive very poor application design. Not directly MS’s fault there, but if I have to change or debug another “lets put everything in the code behind file” web app I’m going to scream. Yes, stupidity is hard to design for, but the tools and infrastructure should encourage good behavior and not require a ton of certification to be competent about not screwing up.
November 1st, 2005 at 12:06 pm
Google vs. Microsoft
With Microsoft announcing hosted services to reel in small businesses, Rick Segal and Robert Scoble had posted how start ups view Microsoft compared to Google and open source alternatives. Scoble has 12 reasons why start ups don’t use MS product…
November 1st, 2005 at 12:27 pm
12 Gründe, warum Web 2.0 fast ohne Microsoft stattfindet
Microsoft’s Robert Scoble hat 12 Punkte aufgeschrieben, die zeigen, warum viele Web 2.0-Firmen und -Start-Ups auf OpenSource-Software wie Linux, Apache, MySQL, PHP, Ruby etc. anstatt auf die Produkte seiner Firma setzen. Anfangskosten Preis/Leistungsv…
November 1st, 2005 at 12:28 pm
The thing is, a lot of those reasons aren’t true:
Linux really isn’t free. Do you want support for that OS? Have fun asking 13 year old l33t h4×0rz.
No lockin is a myth. Once you’ve started using a platform, you’re locked into it, because migrating from any platform is a royal PITA. Unix to Linux might be easy, since you don’t have to rewrite scripts or whatever as much.
More security… neither platform is really 100% secure. I’m going to say that security is more a function of proper administration at this point. If I put an unpatched Linux box on the net and it had no firewall, the thing would get hacked in less than a few hours; the same applies to Windows. If you’re a sloppy moron of an IT admin, you’re going to get hacked. Of course, how easy it is to secure the thing is a big factor.
Anyhow, there is a lot of FUD and untruths being spewed by Linux Zealots. (Not that Microsoft hasn’t been guilty of misinformation)
November 1st, 2005 at 12:29 pm
To expand on my comment - I just did the following experiment:
- go to http://www.microsoft.com
- follow links to Server systems
- try to walk through all the links to get to pricing (and figure out products) that might be involved in a Startup deploying a MSFT based solution
- get lost in MANY different, sales orientated (but overly text/jargon filled) sites each linking back and forth and ALL focused mostly on desktop systems. See http://www.microsoft.com/licensing/resources/default.mspx for information on the apparently FIVE different “volume” licensing programs MSFT offers.
As a startup where would I start?
I would not have a lot of desktops, certainly less than 250, but I would want to try to just get a handle on costs (to plug into my business plan say). But each specific application I might look at building on top of - say MSFT Content Management Server, in turn requires other systems:
- see http://www.microsoft.com/CMServer/howtobuy/default.mspx
It requires Windows Server 2003 service pack 2 and SQL Server 2000.
I would then need to figure out the licensing costs for each.
Looking at SQL Server - http://www.microsoft.com/sql/howtobuy/default.mspx I see that there are THREE different versions, THREE different pricing plans (unlimitted users, client CAL, user CAL) but I also further see that the “example” pricing differs fairly radically (50% discount for some plans) but to get that, I have to figure out the volume license - and once again, I may be planning on only 10-20 desktops for my startup….
I further find that I have to be clear at this point in terms of how many processors on my server, how many servers, etc (and I’m not clear yet whether I need TWO or ONE or more servers just to deploy this solution - i.e. does SQL Server get deployed on its own server - I would generally recommend that but how does that impact my licensing needs?
And that’s just thinking about production. I need to replicate this environment in development and ideally being a company that plans for the future also have a test bed that is a staging area between development and product (i.e. a clean test environment) and probably some backup system for failover purposes.
So I look at “developer edition” - http://www.microsoft.com/sql/howtobuy/development.mspx
Now I get pricing that is dependant on a “per developer” license. Again, I’m planning a startup so this is a bit unknown yet (and I’m not clear who is a “developer” - are testers develoers - probably. What about the CEO/management of the company? (for a startup probably as well)
This seems like a subscription (though pricing is somewhat unclear on this point) so I now have to build into my business plan about $3k per developer - probably workable for developers I hire full time, but unclear how I handle contract developers or other project work, or developers who might build on top of my systems but not work for me (i.e. I want to open up my api and deliver some of my stuff as open source, if we put up a development server and open it up for these developers to contribute to it what licensing covers that circumstance?)
And I don’t quite know what licenses cover my test bed.
Further - from just looking here it isn’t entirely clear if these cover EXACTLY the same systems as I would deploy in production. i.e. I don’t want developers building against a non-production ready product - I want them building (or at least capable of fixing bugs) against development environments which are EXACTLY the same as their code will see in production. (though possibly different hardware, I want the software versions to be precisely the same, down to the patch order and level).
Now I go back to the SQL server page, I see SEVEN different subjects under “licensing considerations” as well as SIX different components which could be licensed.
At this point I figure that understanding the licensing options NOW will take me 1-2 weeks, possibly more. That’s just to figure it out sufficiently for my business plan (and remember, this scenario is at the business plan writing stage, we don’t yet have our code written or fully understand our customers, usage levels or scaling needs - I just want enough information to put numbers on paper which I and my financial backers could live with)
So I try reading the downloadable white papers/FAQs (we’ll leave aside that these are downloadable only in .doc format - I’m being generous and trust that MSFT of all people won’t have a virus attached…).
I’m still not at all clear - and further the note that there are volume discounts as well as negotiated discounts from ISPs, ASPs and OEMs - not sure eactly what those are in this context but in any case it seems to mean that I can probably toss the retail price out….
Frustrating!
In contrast, to do this same exercise for a LAMP stack solution I can buy (or download) all the software I need on a few cd’s/dvds from RedHat (or IBM or many others), deploy it as many times as I need on interchangable systems (with how ever many processors as I want/care to buy). I can then get a fairly simple quote from one or more vendors for support and some advanced features (developer licenses from ZEND for example for PHP) but for planning purposes I can get all this pricing in a few days, feel very comfortable about it and see how and where my costs would grow as my business takes off.
Further, I know that if I need to quickly add systems in such a solution, I do not also have to quickly provision licenses (say I’m recovering from a disaster, building out a server farm in a new country, scaling up from being Slashdotted etc). Nothing in what I’ve read on microsoft.com makes me confident that “quick” and “licensing” belong in the same sentence.
Shannon
November 1st, 2005 at 12:31 pm
“Do you want support for that OS? Have fun asking 13 year old l33t h4×0rz.”
Like, say, IBM?
November 1st, 2005 at 1:30 pm
“Like, say, IBM?”
Why would IBM be interested in supporting the countless linux distros out there? How are they affiliated with them in anyway?
Like, say, duh?
But really, I can tell Matt has never used Linux or just isn’t a people person because support for many linux distros can be obtained through their respective websites. Most of these sites have forums FILLED with people who are anxious to help, it’s not uncommon to post a problem and get a response within minutes. The open source community is much more just that, a community than he would have you believe. After all, it is MADE by us, and we are all eagar to share.
November 1st, 2005 at 1:31 pm
[...] Robert Scoble sparked an active discussion with his post commenting on Ross Mayfield’s list of 12 reasons why Ross does not trust Microsoft. [...]
November 1st, 2005 at 1:44 pm
Rich — he seemed to be implying that Linux support consisted of teenaged forum trolls. I think everyone realizes that there are now many organizations out there that are happy to offer consulting services for all manner of Linux installations. IBM simply happens to be the biggest / most well known. My point was simply that open source support = m4dly sk1ll3d teenagers is a pretty specious argument. But of course you realize this too. :)
But I agree, for a small organization, there are really few configuration questions that can’t be answered by an open-minded individual with an inquisitive nature and a search engine or two.
November 1st, 2005 at 1:50 pm
My newest web app (GreedyMe.com - wishlist site) is, like so many others, running on LAMP. I can get awesome Linux hosting for $5/month, as opposed to much more expensive Windows hosting. For us small ISV’s, nickels are carefully counted!
November 1st, 2005 at 1:57 pm
I like Microsoft’s products, and I like their development tools.
I think that Smart Clients are far superior to the AJAX hype.
However, I’m quite hesitant to develop for the Windows client because I’ve seen all of the examples of ISVs becoming too successful on the Windows platform. Through bundling and privileged insight to the Windows API, it’s almost trivial for Microsoft to take a successful software model for itself. It’s happened with development tools, the web browser, IM, media software, ad infinitum. And now it looks like Microsoft is focused on wiping out Adobe and Quicken. The ISVs who have made Windows the most popular platform ultimately have given Microsoft the power to cut their throats.
This is why so many new developers have opted to develop on a web server. Microsoft is still a competitive threat, but they have to play fair because they don’t own the platform (web).
November 1st, 2005 at 1:59 pm
That’s an interesting list, but it misses out what is in my opinion the most important benefit of all: building on open source lets you take charge of your own destiny. If you’re developing software why settle for a platform that doesn’t give you total control?
November 1st, 2005 at 2:07 pm
“No lockin is a myth. Once you’ve started using a platform, you’re locked into it, because migrating from any platform is a royal PITA.”
What naked sophistry! The virtue of “no lock-in” is that you don’t have to upgrade your platform merely on Microsoft’s say-so. “By the way, we’ve decided you’ll upgrade all your Win98 boxes this year because we’re stopping all support.” With Unix and Linux you can reliably find support for even the earliest distros.
November 1st, 2005 at 2:34 pm
You missed: 13. Tux is cute!
November 1st, 2005 at 3:55 pm
Just look at Live.com to see it’s the same old, same old, zero or flakey cross platform support. 1. Site doesn’t support Firefox or Safari. 2. Chat app is also still Windows only.
And we’re supposed to take this seriously? The Web Isn’t Windows. Repeat. The Web Isn’t Windows.
November 1st, 2005 at 3:56 pm
#13: Microsoft licensing is hostile to open source business strategies. If part of my strategy (or potential ones I want to consider) depends on releasing software under a dual license (say, GPL/commercial) then that pretty much rules out using Microsoft platforms as infrastructure.
#14: Hardware deployment flexibility. Shoestring operations need to be able to swap hardware and software around between boxes to deploy in different configurations and squeeze more use out of existing components. Frankenservers are everywhere, and Windows licensing reauthorization is a royal pain in the ass.
It’s also important to realize that several of the reasons you cited ’stack’ or reinforce each other to result in larger overall savings. In particular, when you can plow the money you save due to #1 into more hardware that inherently performs better due to #2 and #12, you get a pretty dramatic cost-effectiveness improvement. Microsoft’s solutions would have to be 10x better along some other axis in order to overcome this, and ‘Windows compatibility’ doesn’t count.
November 1st, 2005 at 5:14 pm
Charles Garry makes a good point over at Eweek about ‘world view’ shaping markets.
http://www.eweek.com/article2/0,1895,1879884,00.asp
I am only interested in technology as a tool to accomplish an end, both OSS and proprietary software have specific areas in which they excel. But the biggest ‘trust’ issue in my opinion, which has been mentioned here, is that Microsoft continues to create a difficult environment for small to medium businesses with a complex licensing model. The other points are mostly false. Linux is by no means “free”, and if you want IBM support instead of the leet kid, and you certainly should have it if you are doing business, then it is still expensive and more difficult to implement than with MS.
November 1st, 2005 at 7:15 pm
Reason #42 If you aren’t a 100% Windows Shop, MS hates you, and wants you to die. Try using Windows Servers in a heterogeneous environment with multi-platform clients.
November 1st, 2005 at 7:16 pm
“But really, I can tell Matt has never used Linux”
-untrue. I’ve used plenty of distros. Is there anything else you would like to divine from the background cosmic radiation about me?
“or just isn’t a people person”
-not in the slightest. I like making small children cry.
November 1st, 2005 at 7:17 pm
“think everyone realizes that there are now many organizations out there that are happy to offer consulting services for all manner of Linux installations.”
Yes, but for a price. That was my point.
November 1st, 2005 at 7:26 pm
“Why would IBM be interested in supporting the countless linux distros out there? How are they affiliated with them in anyway?”
Um, IBM will happily sell you support services for Red Hat and Novell/SuSE, both of which they have business alliances with.
http://www-1.ibm.com/services/us/index.wss/offering/its/a1000030
Nice try, though.
November 1st, 2005 at 7:56 pm
Umm, dude…
Uninformed customers are why the Evil Empire is in business. If they were fully cognizant of the alternatives, MS would collapse like the Roman Empire.
November 1st, 2005 at 8:22 pm
You left out several reasons, so let me try to pick some off the top of my head. I’ll also toss in examples from personal experience to back up my point, when I can:
Morals and ethics: There’s a large perception that these are missing at the top, and quite a bit of the way down. That’s not to say that everyone there is morally bankrupt, but the perception of many is that the leadership is. I share that perception, and current events only reinforce that perception.
Trust: As you yourself pointed out, there’s a lot of distrust of MS. Here’s a hint wrt to that distrust: MS’ own actions aren’t helping, not even a little. This goes hand-in-hand with the above. The PR people may be working over time trying to convince people that MS actually cares, but what we see is a behemoth that is bent on the single minded mission of wiping out any and all other companies which have even the potential to compete with it, no matter the cost, and no matter who gets hurt by this behavior.
Support for standards: That means real support, not the well-known “embrace, extend, extinguish” support that we’ve all seen in the past (and present). The MA. mixup is a perfect example. What does MS care if ODF isn’t as “rich” as MSXML? It’s what the customer wants. Try listening to your customers, instead of some paid focus group or where ever you’re currently getting your information. Or worse, telling the customer, as you did with MA., that they don’t know what they’re talking about.
Trouble shooting: Who decides what to put in your error log messages? Is writing “error 234213613246123451672759078345 has occurred” supposed to help me somehow? Because it isn’t. When I have a problem on other platforms, I recompile with –enable-debug=yes, and I have my answer in short order, even if I have to go digging in the code to find it. That means the problem generally gets resolved in less than a day (that the messages are mostly plain english with some memory addresses helps a lot). This is in contrast to when I have to open up a ticket with MS. If I actually get answer, it takes days or weeks, unless it’s a “known issue”.
Stability: You think that’s a “perception”? Bull pucky. No matter how well I maintain my MS servers, and I do maintain several and support a lot more, they always end up deteriorating, just like my workstations. The servers don’t get it as bad, because they tend to be dedicated boxes running very limited and locked down installs, so bitrot takes longer to set in. A perfect example would be one of my Win2k DCs… ran fine for a couple of years. Then, for no obvious reason, it stopped authenticating network connections after about 2-3 weeks, requiring a reboot. After several weeks of trying to trouble shoot (it would be the domain infrastructure master, wouldn’t you know?), I finally transferred the role and rebuilt it. On the other hand, I have *NIX boxes that have been around since the ’90s without a rebuild, just patches and upgrades. And those are boxes I built when I was first starting with *NIX. I have my MCSE and 15 years of windows experience, and a windows box going more than a year without a complete rebuild is the exception, not the rule. And pretty much every windows admin I talk to has this same problem. Why is that?
So there’s a few more for you. I do, however, have one comment on your statement that the scalability issue is a “perception”. Sorry, that’s fact. I support an application that requires the the TCP/IP stack to make routing and forwarding decisions very quickly. Windows tops out at around 18 MBit, no matter how much hardware we throw at it. I’ve seen a linux box running the same app do over *80* MBit on nothing but a P2 w/ 384 MB RAM, and even then the processor was mostly idle. I also see, time and again, a windows deployment for hosting, for applications, for what have you, require about 2-4 times the hardware that the same app requires on a Linux or BSD deployment (synthetic benchmarks, no withstanding). Windows scales. It just takes ridiculous amounts of hardware to get it there. Maybe when you get a bit closer to a 1-1 processor parity, we’ll talk again.
And that’s why I deploy anything but a MS solution unless there’s no other choice. Or, to put it another way, the more I work with Unix, the more I loathe Windows. I do really like Exchange though, so go figure…
November 1st, 2005 at 8:23 pm
Another one you missed.
Microsoft has a reputation for walking away from the platforms it develops.
VB.NET is VB in name only, and all the training and infrastructure that people invested in it got yanked out from under them. It’s as easy to migrate from VB6 to Python as it is to VB.NET; and it’s easy to fear that VB.NET will go away too as C# quickly moves on to 3.0 and beyond and VB seems to be stagnating.
Even Perl 3->4->5 had more graceful backwards compatability support than VB is gettting.
If an open source platform had the installed base of loyal users that VB6 had, you’d have dozens of companies supporting it and keeping it on the leading edge. But what does Microsoft do? Kill it to try to get people to pay for upgrades to something different!?!?! No Thanks.
So I’d think you should add “Open Source has better long term support and doesn’t leave you stranded cold with abandoned technologies” to your list — and even in the rare cases where it doesn’t (X11), other organizations (Xorg) will fill the gaps to support what the original maintainers didn’t handle well.
November 1st, 2005 at 9:01 pm
[...] From Scoble: H’es talking about Web 2.0, but it applies to any deve,lopment / startup / etc. And he forgot to say what Microsoft is going to do about it. 1) Startup costs. Linux is free. Ruby on Rails is free. MySQL is free.2) Performance per dollar. They perceive that a Linux server running Apache has more performance than IIS running .NET.3) Finding tech staff is easier. There are a whole new raft of young, highly skilled people willing to work long hours at startups who can build sites using Ruby on Rails.4) Perception of scalability. The geeks who run these new businesses perceive that they can scale up their data centers with Linux and not with Windows (the old “Google runs on Linux” argument).5) That Microsoft doesn’t care about small businesses. After all, Microsoft is an evil borg, but Ruby on Rails comes from a single guy: David Heinemeier Hansson. He has a blog and answers questions fast.6) That open source makes it easier to fix problems and/or build custom solutions. A variant of the old “Google or Amazon couldn’t be built on Windows” argument.7) On clients, they want to choose the highest-reach platforms. That doesn’t mean a Windows app. Or even an app that runs only in IE. It must run on every variant of Linux and Macintosh too.8) They don’t want to take shit from their friends (or, even, their Venture Capitalist). [...]
November 1st, 2005 at 9:14 pm
Isn´t kind of ironic your blog is running on open-source? I rest my case.
November 1st, 2005 at 9:17 pm
[...] Read this post and tell me that blogging is not already a big thing. Or that it hasn’t already changed the world. [...]
November 1st, 2005 at 9:33 pm
Wow, what a list. For me, it’s MSFT’s tools’ attitude toward HTML. ASP.NET treats HTML as if it’s a burden, an embarassment. It’s very hard to get good valid, predictable HTML out of it.
it’s frankly an attitude of disrespect for the way that web development has gone in the last three or four years. HEre are some suggestions:
- Stop bothering with demo-only featrures like “draging and dropping” controls to build a page, which then of course uses fucking pixel-based positioning, destroying my page layout. You know how often it’s useful to be able to drag a checkbox control (a frigging *control*! For a single HTML tag!) onto a page, rather than writing it? Never ever ever. That how often.
- Everyone buidling web apps that I know wants to render clean, validating HTML, or as close as reasonably possible. Stop treating HTML as merely an output format, to be slapped together as you see fit.
- Everyone buidling web apps in 2005 wants to style the page almost entirely through CSS. That professional MSFT tools automatically insert spacer gifs is an embarassment. Table-row background colors in tr-tags is an insult. Support for older browsers should no longer be a default, it should be an option.
- Everyone buidling web apps wants to do more with DOM manipulation than ever: Ajax, page effects, etc. Existing libraries (like Prototype) do this cross-browser, and are nicely designed. MSFT’s “Atlas” feels like some thick-client geek was assigned by a middle-manager to reinvent the whole damn paradigm cause it wasn’t “flexible enough.”
November 2nd, 2005 at 1:53 am
[...] Drew points out (in a response to an article posted by Scoble a few days ago), why non-windows web development beats windows development hands down. [...]
November 2nd, 2005 at 2:18 am
Web apps and services on all platforms
Scoble has a 12-point list from Ross - the 12 reasons Web 2.0 entrepreneurs aren’t using Microsoft’s stuff: 7) On clients, they want to choose the highest-reach platforms. That doesn’t mean a Windows app. Or even an app that runs…
November 2nd, 2005 at 3:11 am
[...] Scobleizer - Microsoft Geek Blogger » Ross doesn’t trust Microsoft’s approach to Web As I’ve been going around the world I’ve been meeting with many people who’ve built their companies on non-Microsoft stuff. Some of whom have companies worth billions of dollars now. Some of whom you’ve never heard about unless you read TechCrunch. Here’s 12 reasons Web 2.0 entrepreneurs like Ross tell me that they aren’t using Microsoft’s stuff: [...]
November 2nd, 2005 at 5:22 am
[...] Scobleizer: Ross doesn’t trust Microsoft’s approach to Web [...]
November 2nd, 2005 at 8:37 am
One more reason: Microsoft is an American company. Conspiracy theory or not, one has to wonder about MS’s relationship with the US government. Some EU government entities (mainly at local level) are choosing non-MS solutions simply to avoid foreign solutions.
You can also add in general anti-Americanism in the wake of Iraq/Bush/etc. Rightly or wrongly MS, like McDonalds, is perceived as a symbol of America. This brings with it its own set of pros and cons which
Linux/MySQL/Apache/etc don’t have to contend with.
November 2nd, 2005 at 9:12 am
You see the fatal flaw with using linux and other opensource flavors for development is that you don’t have decent tools. I asked if there was any intellisense-like featured ruby on rails text editor, and the response wasn’t just no, but its not possible. That’s the kind of attitude that makes me steer clear of platforms. With MS I have dev studio. And I don’t care what anyone says nothing can beat Dev Studio as an IDE. 2005 is amazing.
Linux is a big no for me. For any server solutions I use FreeBSD. And this isn’t just preference. It makes more sense to use an option that doesn’t tie me down. The GPL isn’t exactly a business friendly licence. On top of that linux is just a kernel. FreeBSD is a complete system with a very robust set of tools. But again just because its my server platform of choice does it mean its my development? No. I still develope in windows and port to unix. I said unix, not linux. Linux has too many variants for me to keep track of. Do I build with 2.4 libc or 2.6 libc will it run on Red Hat or SuSE? I don’t care I compile wiht FreeBSD’s libc and be done with it. And since there’s only one FreeBSD flavor I don’t have to worry about it. NetBSD and OpenBSD generally don’t have a problem running the program natively, at least in my situations.
Next I don’t like projects with just one controller, its a nice beginning. An organized development team is always a good thing. Linux has Linus. Linus in my opinion isn’t the greatest developer ever. He is a smart man however. He took an old idea repackaged it gave it a new licence and its now the hottest buzz word in the tech industry. I used linux many years ago and tried again recently, I’ve seen improvments but its still not FreeBSD. And KDEV is not Dev Studio. Nor is Eclipse. They don’t even come close. Linux just might be turning into that evil demon microsoft used to be. Then again I’m just one lowely developer.
November 2nd, 2005 at 9:28 am
Many of you have pointed out a need for MS software license for start-ups. One of the best kept secrets is Empower for ISVs at https://partner.microsoft.com/global/40010429. The subscription is $375 a year and includes a host of benefits, i.e., product licenses, MSDN universal and support services. If you have other things in mind, please send me your feedback directly at markbar@microsoft.com. We have been analyzing this for some time and would be useful to validate our thinking.
November 2nd, 2005 at 11:10 am
Mark,
While Empower is useful for addressing a startup’s internal software needs - it does not address a couple of very important issues.
1. It assumes that a startup is building packaged software - if you are building a web application/service at least on first, plain English reading you don’t qualify.
(i.e. requirements read: “Commit to developing one (1) packaged and resalable software application that supports at least one of the following Microsoft technologies:”)
I would argue that this program should be extended to companies building web applications as well.
2. It does not address or help simplify the cost of liceneses for PRODUCTION. Both a web service you (the company) run and manage or what you sell and deploy to clients - the cost of the underlying required MSFT licenses is not at all addressed by the Empower program.
So if I were writing a business plan, Empower would address some of my internal costs and budget (at least until I reached over 5 developers) but it does nothing to help me structure my sales prices or deployment costs, nor does it help in the building of a Web 2.0 type service.
And it would assume and require that even as a small company (i.e. less than 5 developers) I have the resources to manage MSFT servers for internal use (Exchange, Sharepoint etc.) This is a big assu