I think I owe this man some kind of expensive liquor for MVC--especially MVC3. I truly loathed ASP.NET projects until MVC3. Some will argue that people use it wrong and/or the traditional Web Forms have their advantages, but I can't really think of many situations where I'd rather do it the old way.
I owe him more than that, MVC is epic, and has pretty much fuelled my love for .NET. If it were scientifically possible (Arnie didn't really father a child in Junior, right?) I would have his baby. I'm nice like that.
But alas he sought pastures greener and I am happy for him. My dabbling in git from my windows machine hasn't been smooth to say the least. Hopefully Haack's new role will remedy that, and for that I will fund further research into male pregnancy.
Rob Conery is teasing. And Rob Connery, not Phil Haack does the "This Developer's Life" Podcast. Hence the joke. If Anders Hejlsberg replied, the thread would be complete.
On a side note, Anders is really interesting person to listen to when he unveils new features and gives talks. He seems really modest and knows the limitations/advantages of a managed language. But, most of all, I always get the impression he is in tune with the needs/desires of the community.
I'm unable to edit my blunderous original post, but as others have pointed out, Rob Conery is the host of This Developer's Life (an excellent podcast) and not Phil Haack. Honest mistake.
Yep, I had nothing to do with TDL. Wish I could claim it, but I can't. I mean, I guess I could and lie about it, but no. Won't do that. Not yet anyways. :)
Honest mistake guys, I've listened to almost every episode and had no idea. All the jokes about the Hawaiian shirts and I've been thinking of the wrong person.
Github is my social network. I wake up into it, i follow it, push to it during the day, and I survey it before sleep. I love Github (in fact i'm inlove with Github). I depend on Github. I don't do facebook, g+ or twitter for socializing.
I used to admire Haack aswell. He's a great guy. However I fell out of love with .net.
I am a polyglot programmer, with strong background in .net.
To understand these Haack's "accomplishments", please look at history (google).
There is nothing genious about MVC3, towards other MVC web frameworks, and its 'advanced' 3 version, it sucks really bad in testing. (I'm also an author of a .net web framework that pits against it pretty well in this regard).
I'm pretty sure that .NET developers won't take away from its awesomeness. Quite the contrary, I think that if more .NET developers are opening up to how things work outside of their environment, we'll see a lot more openness and collaboration coming from .NET devs.
And then maybe this whole us versus them thing will stop.
There is nothing genious about MVC3
As a Python developer, I don't think there's anything genius about Ruby on Rails either. It's just a bunch of borrowed features that were added with good taste + marketing + community enhancements. Overall the whole package is great, but it does stand on the shoulders of giants nonetheless. Borrowing stuff is not a sin - if anything I blame Microsoft for the often displayed NIH syndrome.
Your claims aren't to the point at all. Even if he didn't come up with those things originally, he introduced them to a massive community of developers, that's already one great step!.
Being a great guy as he is, I'm sure he'll push to positive in anything he'd do!.
When I went looking for a git win32 binaries a while ago the attitude was basically "lolz get a mac/linux box". Which is why I use hg instead of git.
It's great to see an open source community, that isn't engulfed in the sanfrancisco linux/mac mono culture bubble, reaching out to Win developers for the benfit of everyone.
I totally see where you're coming from, and I personally think hg is great. I just don't want your comment to scare windows developers off. For the record: running git on windows is great.
Disagree. It's gotten better, but the tooling is awkward, the commonly used UI (TortoiseGit) is wonky, and not being a "native citizen" on the OS isn't acceptable--MSYS is a mess to have around on a Windows machine and even the "minimal subset" that you can install with Git is more than I consider acceptable. (I realize that entirely too much of Git is UNIX-based plumbing, but that's their problem.)
I gave Git another shot when Bitbucket started to support it (I won't use GitHub, as it happens), but it didn't shake out for the above reasons. I like being able to open stuff up on any of my machines and the lack of care I perceive in Git on Windows drives me to stay with hg a lot more than it would otherwise.
Cool, great. Except that the Windows command line sucks and bash is uncomfortable to the point of unusability on Windows. I will use a shell (not bash) on Unix platforms where it is properly integrated and meshes with the rest of my workflow; this is not Windows, and attempting to attach it to Windows with baling wire does not work for me.
My experience with source control systems--you know, including ones that aren't git--is that you can generally find a good UI for them. SourceTree on Mac is great for hg and git. Tower is great for git, too. Even TortoiseHg is a country mile past TortoiseGit in usefulness.
Powershell + msysgit work just fine. It puzzles me why people still keep referring to ye olde command prompt as "the windows command line" when Powershell has actually been around for years now and (while significantly different to bash) is extremely capable. My only complaint really is the lack of dynamic resizing without first consulting the options box.
After using Tower on Mac you may change your mind. If there was a client like that on Windows it would win. Currently hg integrates cleanly and more of a svn replacement since it is more open. Note: I code mainly on osx but some .net only on windows.
I always thought it was cool that Mercurial is written in Python, part of that cross platform magic.
I think poor IDE support is pretty unilateral because git is supposed to used from a command line. It is great to work with a tool where the UI is an after thought because the UI will always be cumbersome with something as complex as source control.
msysgit works great, has for years. The install is simple enough a non-technical person can do it. I've used git on projects with non-technical people checking in from the command line. It is pretty surprising to see a PM have an easier time than a dev only because they don't have years of UI driven source control working against them. Using source control from the command line... well that is the trick. Though I haven't met many great developers who cannot.
Also, git extensions is pretty good (better than tortoise). I usually install it because viewing the history from the IDE is a nice way to keep your flow going. You can do some creative things searching through the logs as well.
I initially struggled with Git on Windows (TortoiseGit was buggy and I'm not a fan of command line) and passed it up for Hg. I took a look back in April (since I liked where GitHub & Bitbucket were going whereas the equivalents for Hg were not as impressive) and since then I've been using GitExtensions + the Git Source Control Provider for Visual Studio with zero issues and have thoroughly enjoyed the experience.
Sounds like we had similar experiences. I went to HG but then gave Git another attempt because I didn't like feeling defeated. I've since been using TortoiseGit and it's OK.
I've heard of that VS Git extension before. Maybe it's time I give it a go.
From the cmd.exe prompt, some of them are still working, as long as the cygwin bin folder is in the PATH variable, and the executables end with .exe, and are not cygwin symlinks (the file would be very small and contain something like !<symlink ...>) - then they won't execute, but you can still execute them from cmd.exe, like this:
sh -c "gitk --all"
Hopefully git is still normal executable (gcc is not, for that: sh -c "gcc --help"
There are WIN32 binaries linked to from the official GIT site. I mainly use mac but I have had to use the windows binaries a few times in the last few months and everything works fine.
Phil did a great job both before and after his MS career. He will do well at GitHub as well. He's one of the good guys in the C# world, and both GitHub and Microsoft will benefit from this.
Phil is good in the sense of being very community-oriented, among other attributes. It's not that others are "bad", just that Phil is a great ambassador for C# developers in the world-at-large.
Awesome to see a hugely popular and powerful development platform getting noticed by Github. It's great technology that deserves more love and positive mindshare in the open source world.
The last startup where I was CTO was largely based on the Microsoft .NET stack and we loved almost every ounce of it as a technology. Until, of course, we tried to hire startup and open-source minded engineers in the Microsoft-hate land of Silicon Valley. That's practically impossible.
StackOverflow and (hopefully) now Github will ideally do a lot of good in the next few years to break the stigma against C#.
It seems like the C# community at large contains Enterprise-oriented folks, who only like to consume open source and not give back. They just want to get their projects done and make a profit. I wish we could see more altruistic entities and startups embrace the technology. @haacked - fix it! Github seems like an ideal platform for change in this regard.
Oh, yeah, that startup was acquired and they threw away all our back end tech for no other reason than it was Microsoft based. It's not all bad, though, I do love my shiny Macbook Pro. ;)
Feature requests:
- CodePlex migration tool (I have a project there...)
- Github client for Windows
- Github VS integration
Have you tried GitExtensions? It has a Github plugin. Together with the Git Source Control Provider extension in VS, I've had no problems with Github integration (that being said I'm leaving Github for Bitbucket since Github's arbitrary limits on the number of private repositories pisses me off... for $7/mo at Github I'm getting way less than I do with Bitbucket's free plan).
Microsoft needs to become more outwardly focused and influential softies leaving the fold and pushing from the outside will do that. Right now the two good things Microsoft has going for it is MVC and C#. Both of which were inspired or heavily influenced by outside forces (MVC by Castle and C# 4 was influenced heavily by Python).
Phil Haack moving to Github will hopefully push Microsoft to embrace open collaboration systems a little better and because he's outside of Microsoft he won't have to deal with "you might kill off VS Team Foundation" hindrances.
All this in turn MIGHT give Microsoft a chance at becoming relevant again in the developer space (and please don't say "a lot of people use Microsoft" because I'm one of those people and I know a lot of those people and they're all looking at their future with worried eyes)
> All this in turn MIGHT give Microsoft a chance at becoming relevant again in the developer space (and please don't say "a lot of people use Microsoft" because I'm one of those people and I know a lot of those people and they're all looking at their future with worried eyes)
If you read the blog post of this very submission, it cites C# as #4 and climbing on the TIOBE language index. I know and work with a bunch of .NET developers, also, and I can't name one that's looking at the future with worried eyes.
In what way are they irrelevant in the developer space?
In my particular case, I tend to shy away from Microsoft development because the stack isn't really free, top to bottom and it seemed like MS was following trends instead of leading them (MVC was an obvious response and I see a lot of Railisms in it).
Well free stacks aren't really free either but we could argue that point until we're blue in the face.
The issue with Microsoft right now is they seem to want to throw away everything they've created over the last decade and embrace the web in a desperate attempt to bring new developers in.
For example, if you watch the Build sessions you see a lot of focus on C++ where as Microsoft used to treat C# like it's preferred language. That's got a lot of people thinking Microsoft's looking to sideline C# in the same way they did VB 10 years ago.
Visual Basic was obviously more popular that C# at that time (being C# hadn't existed before that). So the Tiobe index doesn't mean much (and honestly I'm always dubious of Tiobe in that I don't think PHP is more popular than Ruby and Python combined in modern development)
3 years ago Microsoft was telling developers to use ASP.NET Web Forms, WPF, and Silverlight. Today all that is essentially looked on as bad practice. So yes, there are people in the MS Developer community who are nervous.
I did watch many of the BUILD sessions, but never saw any indications that C# was going away or even being de-emphasized. I think you may be misunderstanding. Could you give examples of what you mean?
Microsoft isn't really "throwing everything away", they're shifting product strategies in a rather confusing and convoluted way. However, the latest stuff their pitching is C# and stuff from Silverlight (even though it's "dead") -- it's largely the same technologies. A Silverlight or WPF programmer will find a relatively painless transition to WP7 (which is essentially WPF/Silverlight) or W8 Metro apps.
I think your idea that C# is being de-emphasized or going away is a very, very incorrect statement contrary to reality.
Correct me if I'm wrong since I'm no longer .NET regular (but was in the past around 2005-2007 ish).
Lately the development in .NET revolves mostly around Sharepoint (customizing), cloud deployment, and probably Dynamics AX/GP/CRM.
The era of Microsoft shops building independent products seem to slowly/gradually being phased out by the web-app shops (the SaaS type). Lately the trend for MS shops are largely around "toolkits" companies such as Telerik, DevForce, and Infragistics or the "webparts" (SharePoint) companies like Bamboo Solutions. That's pretty much it...
There might be a bunch of "worried" .NET developers that don't want to work on Sharepoint customizations. But I'm pretty sure there are plenty consultants that won't mind doing that kind of works.
But again, this is just my personal point of view.
> Right now the two good things Microsoft has going for it is MVC and C#.
Right now the two good things Microsoft developer tools has going for it are those.
Let's not forget products that make Microsoft's big money - like windows and office. Not to mention other investments like azure, xbox, and ms research.
This is great news for .NET devs- the Windows GitHub client (and VS integration) is not great.
This is awful news for .NET devs- ASP.NET MVC is the best thing that's happened to ASP.NET since it began. Losing the lead guy on that project is worrying.
It's a bit distressing that a lot of the most interesting .NET things (such as mono[touch,droid] and asp.net MVC) are either black sheep or 3rd party projects. .NET is a solid technology, but I fear it may end up having a limited lifespan due to being so heavily tied to windows, especially as the next generation of systems (tablets, non-traditional PCs, etc.) starts taking over the market.
I've been a staunch .Net defender for a long time, but have almost entirely thrown in the towel at this point. MVC is awesome, C# is awesome, F# is awesome...
but, unfortunately Mono just doesn't cut it (Boehm GC? Really? s-gen is a nightmare too, before you counter with that) and Windows isn't a realistic option for a lot of people.
Can you explain what your concerns are with Mono? I haven't used it but as .NET developer hoping to deploy some apps on Mono to cut costs I'd like to understand your criticism better. Are your concerns regarding performance, and within which area? Is it poor memory management and what consequences do you have in mind?
We went guns-blazing from .Net on physical Windows servers to Linux/Mono/EC2. Mono performance is okay -- we still run a lot of C# utility stuff on Linux with it, no problem (build on windows, push to s3, servers self-deploy).
Forget about web stuff though. Mono fastcgi is not production ready (lots of little things, the interplay of nginx and Mono's fastcgi server 'lost' some routes, particularly scriptmethods).
Mod_mono is better... until you start getting OutOfMemoryExceptions. My understanding of the problem is that the Boehm GC is non-compacting, and thus, unless you (somehow) craft your application memory use patterns perfectly, will run out of memory if you churn through enough allocations.
From the mono ASP.NET FAQ:
Why does the memory consumed by the Mono process keep growing?
Mono currently uses a conservative, non-moving, non-compacting garbage collector. This means that the heap is not compacted when memory is released. This means that applications can produce memory allocation patterns that will effectively make the process grow, just like C, C++, Perl, Python applications would.
It is hence important to not get into patterns that would create these holes, for example such a hole could be created if you create a block of size SIZE, release it, and then create two blocks of size SIZE/2+1.
ASP.NET in Mono is particularly vulnerable to this kind of memory problems because it is easy for developers to define APIs that transfer large blobs of data like entire image files, these would allocate a lot of memory that can easily be fragmented.
A simple solution is to try to write your software in a way that large data blocks are not allocated, but instead your application handles them in blocks (like writing a "copy" command).
So the new GC, S-Gen is available, but for us at least, ~6 months ago it was really buggy. Some of our code would inexplicably cause mono crashes in the allocator. We couldn't run our apps with it, at all.
So ultimately we were left with restarting our webserver process every 30 minutes or so, or we could switch back to Windows web servers. Mono was decently performant, but going back to Windows we had a slight but noticable performance increase.
Don't get me wrong, Mono is awesome for a lot of things, and it kicks fuckin' ass to run my C# on Linux. But the ASP.NET side of things is just not mature enough. It's possible I'm just a bad programmer and if I'd taken the time to re-architect our allocation patterns to play nice with Boehm this story would have a happier ending, but alas :)
P.S. None of this stuff is fresh in my mind right now, so shoot me an email you have more questions or run into any specific situations or whatnot.
I would want to deploy ASP.NET MVC3-based web applications so restarting the web server sounds like it'd be my only option (I'm not going to waste my time re-architecting things if it's eventually likely to run out of memory anyway).
Then again, I haven't tackled switching away from MSSQL (aside from learning to use a new DBMS, I'm a fan of stored procs too so I'd have to learn a new SQL dialect), either. Since my time is probably worth more than Windows licenses, it looks like sticking to Windows is the logical choice for now.
Thanks for the offer to pick your brain... I may take you up on it some time :)
I really don't think the ASP.NET MVC can be classified as a black sheep. Neither are things like Linq or Entity Framework. Microsoft is still doing a lot of the interesting stuff in-house.
MVC was developed skunk works style, as was Linq. And MS has only begun embracing MVC anywhere near fully lately. Meanwhile, Linq has been effectively abandoned in preference to the bloated entity framework.
If Microsoft had their heads screwed on straight MVC would be the default ASP.NET web-site project type in Visual Studio and Linq would have gotten the love it deserved, even if it competed with EF. I fear that without Haack's leadership MVC will be mishandled and slowly die off.
Edit: To clarify, I mean Linq to SQL. It was a decent, lightweight, highly usable ORM that has now been more or less deprecated in favor of EF. For now it's not so bad because Linq to SQL only just recently wandered off into the twilight, but eventually the lack of support will make it harder and harder to use and EF will be the only option.
LINQ works /with/ entity framework, as well as with XML and generic collections. Entity Framework is not a replacement for LINQ.
Also, I think it's a natural progression for the project to start out as a "skunkworks" system, and be gradually more mainlined as it matures. It's on version 3 now, and version 4 is right around the corner.
I'd rather see it happen that way than have a huge release and a giant push onto a platform that isn't fully realized. Most "enterprise" organizations would never move onto a version 1.0 platform anyway.
MVC has to be one of the most non-skunkworks projects in Microsoft history. Scott Guthrie (senior management) was involved from the very beginning, and there were frequent preview and beta releases through the entire development cycle. Couldn't have been more open, both inside and outside of Microsoft. Maybe there's another meaning of skunkworks I'm not aware of?
Also, FWIW, a lot of Microsoft's biggest public facing sites run on ASP.NET MVC now, so even if you assumed a lot of stupidity or evilness, there's a lot of inertia behind ASP.NET MVC.
Seriously, it's easy to work with and doesn't get in my way. The fact that it may take a few more ms to do some things that probably aren't my bottleneck doesn't matter.
EF seems very much targeted towards "enterprisey" line of business apps, where ease of development is critical and performance is a side note. Microsoft just does not care to make a high-performance web stack. Even in the next generation of Visual Studio ASP.NET MVC stuff is still an optional download. How many production sites in the world are using EF? Microsoft is still very much mired in the idea of selling tools to developers who tend to be on the lowest rung of the skill-tree and who build mediocre internal apps at big companies. They're solution to scaling and efficiency is to buy bigger boxes, higher end Windows licenses, and more of them.
That strategy has certainly fattened their wallets handsomely over the last several decades, but it's ultimately self-defeating. By failing to capture the most skilled developers, and by failing to cater to the most critical development needs (high-throughput business critical web services being a prime example) they have been alienating such developers. And those developers are ultimately the ones who set the standards (in some cases literally, when they are in positions of power at large companies) that other developers tend to follow. And that momentum can be difficult to break. By the time MS realizes it they will be obsolete faster than they can react. And it won't be because the entire world changed overnight, it'll be because the world has been changing for a long long time and it will have finally caught up to their core business.
It could be a net gain for Microsoft if the Windows developer community begins to embrace github. As long as the most talented in the community don't leave the Windows platform, it could lead to a more vibrant and productive Windows developer community.
I've been using git & git-svn for all my c# projects and for committing to Lucene.Net. The msys version of git has come a long ways. I also prefer using github when possible. Some of us do branch out and embrace outside Microsoft technologies.
Its possible to sway the windows developer community to branch out and hopefully more people in the community will take initiative to sway others to do so.
I think it's a win for MS for many of the reasons that @Haacked lists on this blog post.
IMO MS is too much driven by profits and control to realize expanding the .NET OSS community will be a NET win for their .NET platform and surrounding tooling / aspirations.
Phil is smart, and MVC3 helped .NET enormously but I think people are getting carried away that the guy is some sort of paid ambassador for Microsoft and .NET at Github. I imagine that he'll do whatever it takes to make Github better with the skill-set and experience he has, but that doesn't mean he has to have some mandate from Microsoft on what to change. Maybe he's just a good engineer and Github hires those? Let's give it a few months at least.
Perhaps he'll work on a whole bunch of non-Microsoft stacks now? Getting Open Source within Microsoft and the Microsoft communities is probably more an issue with culture rather than tooling - a hire won't 'fix' that. It would take many counter-culture things (and years) to change I think. Phil working at Github rather than wanting to work at Microsoft is the interesting aspect, and I can't see why that would be 'good' for Microsoft.
I really hope this leads to better Git tools for Windows developers. I've attempted to host several C# projects on GitHub in the past, but I always end up giving up and using either CodePlex or BitBucket because it's so much easier to get up and running with HG and SVN.
Yes, but you still have to install MSysGit or whatever, which is a ridiculously huge download just to do version control requests. Plus, running the commands to generate public and private keys and all that junk is super counter-intuitive, and the default steps recommended on GitHub won't work on my work machine, as Cygwin can't write to the location it wants to create the key files, and once I got to that point, I just decided to use HG.
This is great! Hopefully some good things will come out of this in regards to integrating Github into VS.NET. Right now I use Mecurial but would be nice to give Git a try.
Sorry not an OSS project (at least not yet). I've got one work project on GH and one "nights and weekends startup" project both in private repositories. But I like the workflow of using git for my source control versus something like TFS. Although I admit I initially picked it up just because I wanted to learn something new, I used GH instead of a local git repository because my side project has another collaborator.
EDIT: my beef with Codeplex is that I just don't find it very usable when I'm researching something. Purely as a repository for things like NuGet packages it's probably fine.
Well Phil, you'll be missed. I've been following MVC since the beginning years ago and love the direction things are going with it. I only just recently had the opportunity to jump into the world of Git.
I'm sure you'll do some much needed good out there as the "Windows Badass".
Formerly developed at all MS shops, now at Amazon mostly doing Java with a Linux workstation. Oh words can't describe how I miss Visual Studio and responsiveness of Windows 7.
I used VS.NET 2005 back in 2007 and been doing Java + Eclipse + Maven in both Windows 7, OSX, and Linux (Ubuntu) since 2008 until today.
I also play with Ruby (and Rails as well) on my free time.
Somehow I like my current setup. Maven 2/3 is the biggest killer tool for me (aside from CLI and Eclipse). Ruby tools like gems, capistrano, rake are all awesome for me.
Eclipse has been awesome so far for me. Tons of plugins (I use checkstyle, findbugs, m2clipse, subclipse). Eclipse also comes with JUnit runner (and Git out of the box in the recent version) while VS.NET used to be lacking unless you get the top of the line (even then, you'd have to learn a lot to know how to use the VS.NET Test aspect I believe).
Everything I can do with VS.NET I can do with Eclipse. But then again I may not be very well versed in VS.NET in the past.
I've always installed Console2, Vim, and Cygwin in all of my Windows machines.
I don't know how you use Java at Amazon and to be honest, sometime it depends on the project and the culture as well. I tend to force my team to always clean up our scripts/builds to make the overall development experience enjoyable.
- MS Lead on ASP.net MVC framework vers. 1 through 3 (current)
- Lead on the NuGet package manager framework for Visual Studio
- This Developer's Life Podcast with Scott Hanselman