Blah, blah, Microsoft sucks, blah ... "AZURE IS GOING TO WIN"..
Wait, what? Either this is off the mark or I'm missing a major initiative that could change the internet and save Microsoft. Is there a great article lingering about explaining why Azure is going to win?
Oh I think Azure will do well. And win in the Enterprise market. Look, outside the bubble of the bay area startup land - the world is different. MS Office is everywhere; SharePoint is everywhere; Windows Server is everywhere.
MS is positioned very well. In the Enterprise world. Not consumer world.
Outside of the xbox (nd that has it's issues as well) MS has already lost the consumer market.
Consumers don't buy PC cause of Windows. The get Widows becuase it's the default OS on the PC they are buying. Most care-less what OS is running as long as they can check their email, play a few game and look at look/listen/create media.
Yep the consumers buy a laptop because it is pink or cheap which usually ends up running windows as its the only thing at the price point. When they buy a tablet, they buy a shit no brand one running an ancient android release, realise it sucks and just go and buy a shit windows laptop.
Less than ten percent of consumers are not moronic enough to buy something better the first time.
The businesses buy windows because its good enough and already works for what they need and they don't want to learn it all again.
I live in an expensive bit of London where people can afford i-this and that and everyone still had cheap pink laptops running windows. Phones are different but there is a definite shift towards blackberry and windows phone around me from iPhones. Most people whinge about their iPhone these days.
Lost me at "Let me make this clear, AZURE IS GOING TO WIN the cloud computing infrastructure and platform battle." without providing any further justification for this bold claim.
Does anyone who actually uses Azure want to chime in? I admit that I may live in a bubble, but if Azure is going to win, and I have heard almost nothing about it that gives it the edge over the Heroku/AWS/GAE triumvirate, I would certainly like to know.
I haven't used it myself, but I think in the enterprise market it actually is being used quite a bit. I have heard good things from people who used it, but am at a loss at giving you a concrete example. Maybe it's worth giving it a go, just to form an opinion...
It's tidier than AWS and GAE, better performing and slightly more expensive. Tooling is good, support is excellent (I can get someone who knows arse from elbow on the phone in 2 minutes flat). Some of it is a little odd, but it makes sense eventually. API is a little clunky though.
We have some kit in it (Linux and windows which might sound odd), but it seems to be pretty reliable. Our AD forest root servers are in it and two OpenLDAP roots for clients as is our web site (which is 3rd party PHP).
It basically does everything. It's sort of GAE's app model and AWS's server + network model in one.
We did have a network/vlan issue but they sorted that on the phone in literally 10 minutes.
Microsoft has a very dedicated set of devs that like .Net, and for them Azure is lovely. I doubt if any of them are on Hacker News though. These will be the people this chap is talking to, so he won't see the existential threat of Cloud to microsoft.
If you're using any Microsoft technology people around here assume that you are a boring programmer that's a slave to the evil Redmond overloards and are not in anyway interested in fancy tech news that are posted in Hacker News.
Keep up the good work, you guys in the ASP.NET team are doing great :)
There are also official SDKs for Windows Azure in Java, Python, JavaScript and PHP. You don't need to be a .NET developer to think it's a swell platform to develop for.
I work in a smallish company, with about 30 coders. I would not say that "Microsoft has" us, but C# and .Net is our primary software development environment.
We all read hacker news.
If you divide coders into two distinct categories - ".Net devs" and "hacker news readers", then you are mixing up at least 2 different things that aren't that strongly related.
I think a lot of people will be surprised how good the Microsoft web stack have become. They have copied a lot from java and ruby and added great tooling. I think the .Net stack from 4.0 and later is one of the best development environments, and without a doubt the one with the best commercial support. The .Net team have really done a good job and are listening to their customers (developers).
I own a mac and am generally a web developer and open source fan boy.
That being said i develop games in C# under Windows with Visual Studio and its hands down the best development experience i had yet.
I get that feeling that alot of the hip webdevs these days think they are something special by using ruby/node/vim etc as if a experienced .NET programmer isnt worth their salt.
Tell you what, John Carmack and Notch, two of the most respected programmers in the world program on Windows with (yikes!) IDEs like Visual Studio and Eclipse.
I'm not entirely agreeing with everything here - there was a lot of communication going on through the Windows developer blogs, with (at times ridiculously long) postings on the state of Win8 development. The problem was however that this was a total one-way communication. The hundreds of comments (some angry, some reasonable) below were ignored. I don't think there was a single positive response to ditching the start button over the metro interface. Yet still they did it (and it's shit, as expected (imho)).
Granted, there was a lot of whining about dropping the Glass look as well, and I have to admit that the new win8 theme looks quite nice now that I see it in real life.
The fact however that the first thing I do is put a shortcut to a shutdown batch file on my desktop, so I don't have to jump through hoops to find that menu, says it all. I ditched win8 within days.
The piece isn't really about that though. Despite being an extreme claim, it's really just a tangential opinion. I think the piece can stand pretty well even if you disagree with this, but I would certainly be interested in a separate article on the topic.
Agreed with a lot of the article up until: "AZURE IS GOING TO WIN".
This is likely the result of spending too much time inside a Microsoft cocoon and has no idea that most of the outside world isn't built on Azure, or even windows servers. The only thing Azure is going to win at is being the default/preferred cloud provider for Microsoft MVPs and other VS.NET devs primarily based on a Microsoft tech stack.
Outside bay area and startups world, windows servers are gold standard. Particularly, large enterprises. They have created nice lock-ins too. Majority of large enterprises (not just in US but around world) have Windows stack meaning majority of tech team consists of .net folks with all the .net tools like visual studio, MSDN licenses etc. And with Azure, you add cloud computing too. So it's likely that all these big enterprises are going to use Azure with their fat checquebooks.
P.S. I haven't used Azure platform yet so I can't comment on product quality of Azure
In the general purpose business IT world. Just about every small to medium business that isn't heavily tech focused is built around Microsoft software.
There is a ton of business IT that is entirely based around the Microsoft platform, and they are making good inroads there with their cloud services. Why run your own Sharepoint? Why buy everyone a desktop license for Office, when you can run it in the cloud and easily sync?
Azure is quite scary good but the main problem is that people really don't understand what it is. If they actually work through that, I fully believe it will win.
We have a couple of Linux instances on azure to give you an idea how odd it is.
Yep it is expensive, bit hetzner are unreliable, their kit performs poorly, their support sucks and you can't deploy a sql server instance quickly (the latter is quite useful if you use sql server).
Hetzner has never failed me, it processes 3k+ C# redis commands in <1s which was 8x faster than my previous old Linux server of Leaseweb (costing 4x more). Although my old Linux server did quite well with over 480 days uptime and processed more than 10M commands.
http://www.servicestack.net/mythz_blog/?p=838
In cloud-speak, that's more than 99.99999% uptime.
It seems to be what's most important to you. Uptime and Value are other important metrics cherished by sites that want to scale reliably and efficiently.
"Stuff breaks. Being able to contact someone, get a root cause of an issue and get help to mitigate it is what is important."
This.
This reminds me of the article about Heroku's "intelligent routing" and a discussion about scaling/scalability and performance.
We pay more for mitigating and reducing risks. Cloud services, particularly Iaas and Paas, such as Azure, AWS, GAE reduce (or are at least suppose to) the downtime when crap hits the fan and also enable more flexibility, speed, and agility in building out solutions.
The interesting part about the Heroku discussion pertained to the following:
Assume we have a straight forward, minimum layered architecture and that the average response time is 1.45 seconds. Using average responses times is a bad idea for determining true performance but let's continue for purposes of illustration. To get this average lets say 19 out of 20 requests return in 1 second, 1 out of 20 takes 10 seconds (Please correct me if my math is off/wrong). If we could add a layer that would shorten our long tailed curve that would be an improvement. However, adding a layer will incur some overheard. Lets assume for illustration purposes this layer will add 500ms to every request but will reduce the worst case from 10 seconds to 5 secords. So, 1 out of 19 take 1.5 sec, 1 our of 20 takes 5 for an average of 1.675 seconds. The average time is worse! However, the worst case scenario is much better. We've mitigated the costs of the worst case.
Using cloud services and vendors that have expertise and the ability to quickly address hardware issues is suppose to reduce the risk and cost when things go bad. Yeah you might have hiccups here and there that put you down for a few hours. But that is better than being down for days! It's really all about risk mitigation.
One of our clients has 17,000 windows desktops, 3,500 windows laptops, 500 windows servers and a team of 80 people looking after it all. None of this is connected to the internet or is tongue-wagged about profusely. There are thousands of companies that size hiding that no one hears about here be use they are not clever, vc funded or bleeding edge.
The world is built on lots of things, but a chunk of it certainly is built on windows, possibly more than you can see with the naked eye.
No not silent majority, not even close. Naming a single instance is not representative of the real-world at all.
Unfortunately Amazon/Azure are coy with their numbers but I expect there is likely an order of magnitude more instances deployed on Amazon than Azure.
Some rough numbers to provide some meaningful context of Microsoft's overall server market share:
That's not a meaningful context. That web server market share doesn't include all the on-premises servers for Intranets, mail, shared folders... Neither the Github percentage is representative. In fact, when your parent was referring to with 'silent majority' is that those C# projects are not on GitHub, not open source, just internal tools.
I don't say that .NET is a majority, or minority, or whatever. I sincerely don't have the data. But those numbers you gave don't prove anything.
That's web sites. That's a fraction of the real world, which is my point. Most businesses don't run over HTTP. If most of the top 500 web sites disappeared, the world wouldn't stop.
GitHub is not a good measure of popularity of a proprietary language mainly aimed at internal and closed source software. In fact, if you added our 4.2mloc of total c# code to githib I reckon the stats would be different.
HTTP-based services and websites is going to be the majority of what services that are going to be hosted on Amazon/Azure.
Yes, many enterprises also makes use of MQ's and NoSQL, which due to their weak offerings and poor positioning is relatively non-existent on Microsoft platforms.
I think you are smoking crack. Our AD is on azure. Its not designed for http specifically. The azure web instances bit is but the whole platform is IaaS style.
So hadoop with SQL server (even ESENT if you want a key-value store) and MSMQ/BizTalk are not used? Only the entire health service in the UK runs on their 'poor offerings'...
Deep sigh. I think there are two views of reality.
If we're going to start sharing personal opinions, I think you're naive, insulated inside a Microsoft bubble and unable to comprehend the written word.
I said "majority" of services going on the cloud will be HTTP-based services - feel free to provide any evidence that contradicts this (and no, "But we host our AD on Azure!" doesn't count). That's also where the most of their Azure marketing efforts are going (which has historically had a strong impact in how .NET devs behave)
I said "relatively non-existent", not that they're not used at all. i.e. MQ's are sparingly used in .NET than say compared to the JVM platform which has a metric ton more quality MQ solutions than what's offered from Microsoft. Also NoSQL is effectively thriving on most other platforms which benefit from active communities and solid language bindings, in contrast, most of .NET still uses SQL Server for most things.
I think you are the one insulated inside a startup-hiptech-webdev centric bubble which doesnt represent the real world.
Almost any buisness that isnt strictly IT focused is running on MS Tech. Not for webservers, but for client pcs and network servers. The entire german healthcare (basically any Hospital, Doctor etc) runs on MS infrastructure and Azure also has benefits for these kind of businesses, far beyond any web related services.
Github and the web development world is a very bad metric to measure this.
Your opinion/prediction is pretty far off: I've spent most of my 13+ years experience working for Australian and UK Governments and Large enterprises, mostly as a .NET developer (Java 2 years). I also actively participate in OSS, NoSQL forums and maintain MQ, NoSQL, ORM clients and .NET server frameworks giving me a balanced perspective of Microsoft usage in and outside of the enterprise.
It's strange to see you've got this far and still haven't worked out this thread about hosted services on Amazon/Azure. Thanks for dropping a useless anecdote about the German healthcare windows infrastructure, please enlighten us with what % of the sensitive data and infrastructure will the German healthcare system be moving to an external cloud hosting provider? When did it become commonplace for health institutions to host their sensitive/confidential information on an external hosting provider?
Healthcare was an example for absolut MS dominance. That being said, for most companies that run local IT, secuirty is pretty abmissal and i doubt moving to the cloud would harm them in that regard, but your right, Healthcare might not be the best example.
I am sure though, that many enterprises in the future will run most of their stuff from the cloud because of the many benefits, and when they do its probably going to be Azure because it ties in nicely into their stack.
But it is a silent majority. Because the majority of .NET software is not written to be public facing; but rather written for some internal IT need. Everyone who needs some little custom piece of software that will plug their special domain-specific software into their accounting system has it written in .NET.
Yes, Microsoft only has a small fraction of total public webservers. But theres so much that's not public that you don't see.
Github isn't a very good metric; Git is a lot more popular on Mac OS X and Linux than it is on Windows. Furthermore, most of the code written in C# isn't open source, so an open source hosting site isn't a good way to measure it; and the stuff that is, is more often hosted on Codeplex, Microsoft's code hosting site.
"Because the majority of .NET software is not written to be public facing"
I'd be very surprised if the majority of C software is either. Most C software devs I know have no interest in publishing FOSS, and write software for business use inside enterprises.
I don't think we can really claim to know anything from some stats about web servers or github.
You are sampling outside of Microsoft's market. For eg.
* Instead of GitHub they use Codeplex [0]
* Instead of web servers they sell network servers, a much larger market. Windows still holds 73% of this market, and growing. It is worth $20B a year and is much larger than the web server market [1]
* Outlook is still the most popular mail client in the world (27%)
* Office is Office, one of the most profitable business units in the world
* 550M PC's are sold each year and 85% of them run Windows
* Still 39% of the browser market [3]
* There are Microsoft clients in the Fortune 100-500 that store more data than the entire public internet (taken as being 5 petabytes - most investment banks store data in petabyte scale)
* Visual Studio is not only the most popular IDE, but also the highest selling and grossing [4]
Your data points are based on open source and web servers, which are only a tiny fraction of the total computing and networking market. It is the remainder of that market where Microsoft dominates. I don't think they are even going to pretend to attempt to infiltrate that market, they have a much larger market to defend.
I don't work within the Microsoft ecosystem any longer, but I did for over a decade. I agree with OP that Azure will win, all of my Microsoft development and admin friends are raving about it and moving their clients/organizations onto it.
Is there anyone using Azure with no Microsoft Technology on their stack? I'm really curious about this. I know there maybe some former .Net devs that are doing node.js on Azure. But is there any pure Linux/LAMP/php/python/ruby devs choosing Azure over AWS and others.
60 comments
[ 4.5 ms ] story [ 99.7 ms ] threadWait, what? Either this is off the mark or I'm missing a major initiative that could change the internet and save Microsoft. Is there a great article lingering about explaining why Azure is going to win?
MS is positioned very well. In the Enterprise world. Not consumer world.
Outside of the xbox (nd that has it's issues as well) MS has already lost the consumer market.
Consumers don't buy PC cause of Windows. The get Widows becuase it's the default OS on the PC they are buying. Most care-less what OS is running as long as they can check their email, play a few game and look at look/listen/create media.
Less than ten percent of consumers are not moronic enough to buy something better the first time.
The businesses buy windows because its good enough and already works for what they need and they don't want to learn it all again.
I live in an expensive bit of London where people can afford i-this and that and everyone still had cheap pink laptops running windows. Phones are different but there is a definite shift towards blackberry and windows phone around me from iPhones. Most people whinge about their iPhone these days.
We have some kit in it (Linux and windows which might sound odd), but it seems to be pretty reliable. Our AD forest root servers are in it and two OpenLDAP roots for clients as is our web site (which is 3rd party PHP).
It basically does everything. It's sort of GAE's app model and AWS's server + network model in one.
We did have a network/vlan issue but they sorted that on the phone in literally 10 minutes.
Keep up the good work, you guys in the ASP.NET team are doing great :)
We all read hacker news.
If you divide coders into two distinct categories - ".Net devs" and "hacker news readers", then you are mixing up at least 2 different things that aren't that strongly related.
I get that feeling that alot of the hip webdevs these days think they are something special by using ruby/node/vim etc as if a experienced .NET programmer isnt worth their salt. Tell you what, John Carmack and Notch, two of the most respected programmers in the world program on Windows with (yikes!) IDEs like Visual Studio and Eclipse.
I’d like to see some data to back up his claims that Azure “has the Big Mo” and that it’s displacing Amazon + Salesforce.
This is likely the result of spending too much time inside a Microsoft cocoon and has no idea that most of the outside world isn't built on Azure, or even windows servers. The only thing Azure is going to win at is being the default/preferred cloud provider for Microsoft MVPs and other VS.NET devs primarily based on a Microsoft tech stack.
P.S. I haven't used Azure platform yet so I can't comment on product quality of Azure
Not so sure this is true.
Unless you mean to include the whole software industry in 'startups world', but the likes of IBM are not really startups or bay based.
You might also like to think about whether 'gold standard' or 'default' is more accurate.
There is a ton of business IT that is entirely based around the Microsoft platform, and they are making good inroads there with their cloud services. Why run your own Sharepoint? Why buy everyone a desktop license for Office, when you can run it in the cloud and easily sync?
Which is not the same as "outside the bay area and startups world", which is all I was saying.
We have a couple of Linux instances on azure to give you an idea how odd it is.
http://www.slideshare.net/fullscreen/newmovie/what-istheserv...
The current Hetzner server is looking good over 176 days uptime having processed more than 9.7M+ commands: http://www.servicestack.net/RedisAdminUI/AjaxClient/
There have multiple Azure outages within the same time-frame my Linux servers have been happily chugging along.
It's trivial to setup PostgreSQL/MySql on Linux. I personally don't deploy my own stuff on expensive SqlServer instances.
In our case, a trial with Hetzner found their support to be somewhat lacking if something went wrong.
The azure outages didn't affect us at all.
Out load characteristics are somewhat higher than that. We can shift 10 million http hits an hour quite happily.
This.
This reminds me of the article about Heroku's "intelligent routing" and a discussion about scaling/scalability and performance.
We pay more for mitigating and reducing risks. Cloud services, particularly Iaas and Paas, such as Azure, AWS, GAE reduce (or are at least suppose to) the downtime when crap hits the fan and also enable more flexibility, speed, and agility in building out solutions.
The interesting part about the Heroku discussion pertained to the following:
Assume we have a straight forward, minimum layered architecture and that the average response time is 1.45 seconds. Using average responses times is a bad idea for determining true performance but let's continue for purposes of illustration. To get this average lets say 19 out of 20 requests return in 1 second, 1 out of 20 takes 10 seconds (Please correct me if my math is off/wrong). If we could add a layer that would shorten our long tailed curve that would be an improvement. However, adding a layer will incur some overheard. Lets assume for illustration purposes this layer will add 500ms to every request but will reduce the worst case from 10 seconds to 5 secords. So, 1 out of 19 take 1.5 sec, 1 our of 20 takes 5 for an average of 1.675 seconds. The average time is worse! However, the worst case scenario is much better. We've mitigated the costs of the worst case.
Using cloud services and vendors that have expertise and the ability to quickly address hardware issues is suppose to reduce the risk and cost when things go bad. Yeah you might have hiccups here and there that put you down for a few hours. But that is better than being down for days! It's really all about risk mitigation.
Why do you believe this? Just because it's from Microsoft?
NB Not being snarky - really curious as to why people would think this.
I've been burned by AWS and GAE support which is sometimes non existent. If you're running your business off it, you can't afford that risk.
I'm not concerned who it comes from.
Windows pays the bills for a lot of people.
One of our clients has 17,000 windows desktops, 3,500 windows laptops, 500 windows servers and a team of 80 people looking after it all. None of this is connected to the internet or is tongue-wagged about profusely. There are thousands of companies that size hiding that no one hears about here be use they are not clever, vc funded or bleeding edge.
The world is built on lots of things, but a chunk of it certainly is built on windows, possibly more than you can see with the naked eye.
Some rough numbers to provide some meaningful context of Microsoft's overall server market share:
Microsoft only has 11.62% of the total web server market share of active sites, 12.86% activity from the top 100M busiest sites: http://news.netcraft.com/archives/2013/03/01/march-2013-web-...
Only a fraction of the top internet sites are deployed on a Microsoft platform: http://www.seomoz.org/top500
C# (the most popular .NET language) is the 12th most popular language on GitHub: https://github.com/languages/C%23
I don't say that .NET is a majority, or minority, or whatever. I sincerely don't have the data. But those numbers you gave don't prove anything.
GitHub is not a good measure of popularity of a proprietary language mainly aimed at internal and closed source software. In fact, if you added our 4.2mloc of total c# code to githib I reckon the stats would be different.
Yes, many enterprises also makes use of MQ's and NoSQL, which due to their weak offerings and poor positioning is relatively non-existent on Microsoft platforms.
So hadoop with SQL server (even ESENT if you want a key-value store) and MSMQ/BizTalk are not used? Only the entire health service in the UK runs on their 'poor offerings'...
Deep sigh. I think there are two views of reality.
I said "majority" of services going on the cloud will be HTTP-based services - feel free to provide any evidence that contradicts this (and no, "But we host our AD on Azure!" doesn't count). That's also where the most of their Azure marketing efforts are going (which has historically had a strong impact in how .NET devs behave)
I said "relatively non-existent", not that they're not used at all. i.e. MQ's are sparingly used in .NET than say compared to the JVM platform which has a metric ton more quality MQ solutions than what's offered from Microsoft. Also NoSQL is effectively thriving on most other platforms which benefit from active communities and solid language bindings, in contrast, most of .NET still uses SQL Server for most things.
Github and the web development world is a very bad metric to measure this.
It's strange to see you've got this far and still haven't worked out this thread about hosted services on Amazon/Azure. Thanks for dropping a useless anecdote about the German healthcare windows infrastructure, please enlighten us with what % of the sensitive data and infrastructure will the German healthcare system be moving to an external cloud hosting provider? When did it become commonplace for health institutions to host their sensitive/confidential information on an external hosting provider?
Yes, Microsoft only has a small fraction of total public webservers. But theres so much that's not public that you don't see.
Github isn't a very good metric; Git is a lot more popular on Mac OS X and Linux than it is on Windows. Furthermore, most of the code written in C# isn't open source, so an open source hosting site isn't a good way to measure it; and the stuff that is, is more often hosted on Codeplex, Microsoft's code hosting site.
I'd be very surprised if the majority of C software is either. Most C software devs I know have no interest in publishing FOSS, and write software for business use inside enterprises.
I don't think we can really claim to know anything from some stats about web servers or github.
Or java. Or C++.
* Instead of GitHub they use Codeplex [0]
* Instead of web servers they sell network servers, a much larger market. Windows still holds 73% of this market, and growing. It is worth $20B a year and is much larger than the web server market [1]
* Outlook is still the most popular mail client in the world (27%)
* Office is Office, one of the most profitable business units in the world
* 550M PC's are sold each year and 85% of them run Windows
* Still 39% of the browser market [3]
* There are Microsoft clients in the Fortune 100-500 that store more data than the entire public internet (taken as being 5 petabytes - most investment banks store data in petabyte scale)
* Visual Studio is not only the most popular IDE, but also the highest selling and grossing [4]
Your data points are based on open source and web servers, which are only a tiny fraction of the total computing and networking market. It is the remainder of that market where Microsoft dominates. I don't think they are even going to pretend to attempt to infiltrate that market, they have a much larger market to defend.
I don't work within the Microsoft ecosystem any longer, but I did for over a decade. I agree with OP that Azure will win, all of my Microsoft development and admin friends are raving about it and moving their clients/organizations onto it.
[0] http://codeplex.org
[1] http://blogs.computerworld.com/16263/windows_widens_lead_ove...
[3] http://royal.pingdom.com/wp-content/uploads/2012/01/web-brow...
[4] http://www.internetnews.com/stats/article.php/3616626/Whats+...
The .Net devs that I know all use github, not codeplex. Is there any data behind this? Both are in use (and others), but in what proportion?
But lets be honest for a minute, HN isn't exactly the right place to speak up about cool stuff you're doing with Microsoft tech.