Translating...

Cancel

Demis Bellot
10 months agoPublic
Hey All,

Just want to give the community an early update on ServiceStack's future:

To ensure its long-term sustainability, ServiceStack will transition to an annually-supported paid product similar to Xamarin's products. Starting from the next major v4.0 release of ServiceStack.* on NuGet, which will be the first commercial-only binary distribution (with an exception for OSS projects and a "free-tier" for small projects).

The next major v4.0 will also be a "breaking-release", with new features protected under an "AGPL/FOSS License Exception" and include a number of breaking changes that have built up over time, including upgrading all projects (to .NET 4+), major code and NuGet package re-factoring, clean-up and removal of deprecated code, better consistency / simplification of some concepts and implementation of some roadmap features I've been wanting to add but never had time for. Fixes and releases will continue to v3.x and be maintained in a separate v3 branch until v4.0 comes out of beta.

After v4.0 is released, to ensure continued free/unrestricted usage, if you have a NuGet dependency on ServiceStack, it should be constrained to the unrestricted v3.x BSD releases. Whilst the core team wont be supporting v3 directly, we'll look for someway to host builds of the v3 branch so pull requests with critical fixes can still be merged and resulting binaries accessible. Any direct support for legacy v3 branches will be available under a support contract.

I'd like to iterate that the ServiceStack community and its wider ecosystem largely benefits from the efforts of a few experienced key members here who've actively donated their spare time helping others, who I'd like to thank by extending a license to the future v4.x product, once released - more details about this shortly.

In summary, whilst the current v3 of ServiceStack is a solid release that continues to power thousands of back-end systems, future development of ServiceStack will continue under an AGPL/commercially-supported framework. Given this, it may be a good time to assess your companies future technology directions and evaluate the value offered by the other great (and free) alternative .NET OSS web frameworks (http://www.asp.net/mvc/open-source) which may provide better value.

Full-time development will begin on ServiceStack v4.0, starting next month.

I've also just deployed the latest v3.9.59 of ServiceStack to NuGet. 

Thanks,

- Demis

Open Source : The Official Microsoft Site
asp.net
8
Justin Nel10 months ago
This sounds great, but how will you manage to juggle this and your current day job? I can't wait to see what you come up with though!
Demis Bellot10 months ago
+Justin Nel I've just resigned from Stack Exchange to be able to work on ServiceStack full-time - my last day is at the end of the month.+2
Justin Nel10 months ago
Wow... I never would have thought you would have left there. Well none the less, I hope all goes well with this. From the level of popularity (and how often I hear about people using SS in their businesses), I am pretty sure it will thrive.
Dave Nay10 months ago
I eagerly await pricing information! I have no issues whatsoever paying for a valuable package.
Demis Bellot10 months ago
Thanks Justin, SS is at the sink or swim stage - basically had to give up one or the other and decided to try and make SS work first. Dave: I'll work out and announce pricing near v4's release.
Dan Barua10 months ago
OWIN for v4?
Demis Bellot10 months ago
+Dan Barua Not sure yet - don't like chasing tech fads with little end-user benefits. Only wanted to add OWIN if the new production self-hosted HTTP Server was found to be better/faster than HttpListener. What other features of OWIN do you wish you had in SS?
Michael West10 months ago
I've had a great time working with the framework and look forward to seeing the continued success. 
Dan Barua10 months ago
Self hosting SS and SignalR in an .exe without the overhead of IIS would be really useful.
Justin Nel10 months ago
+Dan Barua why not host that way on Linux and not have the overhead of Windows?
Demis Bellot10 months ago
+Dan Barua yeah maybe, tho surprised self-hosting SignalR is a popular option - a lot of the surrounding tools / templates still seems to be around IIS.
Demis Bellot10 months ago
+Justin Nel from everything I've seen, Mono/Linux is still a fair bit slower than IIS/Windows
Justin Nel10 months ago
+Demis Bellot I'm surprised at that and was always under the impression of the inverse, however have never really run any performance stats against it as it's just worked fine for me on Linux (however, debugging with Visual Studio isn't exactly easy when hosting on Linux).
Dan Barua10 months ago
+Demis Bellot Demisfor sure, I did a freelance project that picked SS due to self hosting (shrink wrapped product) and cutting IIS cuts a lot of the support workload. They're using simple polling on that one. Also self-hosting means you can run on a desktop pc and sell to customers who don't want to shell out for a server! SignalR works nicely in Asp.Net with SS auth, I need to get around to adding that to the examples.+1
Demis Bellot10 months ago
+Justin Nel For other langs/platforms, most things run faster on Linux. But in general I've found Mono is a fair bit slower for most things vs .NET/IIS.
Demis Bellot10 months ago
+Dan Barua Yep I expect self-hosting, win-service / embedded / productized apps is a popular option (and shrink-wrapping is also one of the areas I want to focus on with SS), just didn't think self-hosting SignalR was a popular option for this given max utilization/scalability isn't the primary focus for intranet/embedded servers. IMO long polling would be the more appealing option given you can get away with the extra complexity and overhead of a introducing and maintaining a separate tech stack.

Yeah any tutorials around SS + SignalR would definitely be v. welcomed :)
Matt Cowan10 months ago
Great news
Steve Flitcroft10 months ago
Good luck Demis. Bold move, congrats+1
Humberto Campolina10 months ago
Great news! Sucess Demis!
Rob Ashton10 months ago
Good luck Demis - are you staying in NYC or moving somewhere else for this?
Demis Bellot10 months ago
+Rob Ashton thx Rob, yep going to stay and work from home in Brooklyn for the foreseeable future, tho as I can work on SS remotely will probably spend 2-3 months in Australia when I visit the family this Xmas.
Chris McKee10 months ago
Good luck; its a brilliant framework to work with. I won't pretend it's not going to be hard to sell a replacement to core-framework functionality for cash (to none devs why pay for something that's free, they don't see the difference), but we all know the way to solve that; SERVICESTACK ENTERPRISE EDITION ;)
Robert Greyling10 months ago
Nice one mate - I really look forward to seeing what you can do full time if you've been working part time this far :)

I've been meaning to ask - I got OAuth2 login (google, live, FB, github LinkedIn etc) working with SS, and been meaning to send a pull req for a while now. What's going to happen to those kinds of contribs from v4 onwards? Should I send a pull req to v3 or has somebody else already done it and I just didn't see?
+2
Demis Bellot10 months ago
+Chris McKee Yep, I can concede that a non-free web fx will never be the most popular option, (tho realistically its unlikely for any non-MS .NET web fx to stand a chance of relative popularity - even when free). Unfortunately this year has seen very little progress and supporting the current user-base has become unmanageable. So whilst a move to a commercial product will have fewer users, it will at least give the project a chance to continue improving and compete on features - which isn't happening atm. Basically more time for dev = a better product.
Demis Bellot10 months ago
+Robert Greyling Thx, Only contributions from the core team will be protected commercially, any external contributions will remain BSD licensed. You can also place it in ServiceStack.Contrib if you want, which will remain unchanged. Also if you can add it in by the end of next week you can make before the v3 branch which will (and its forks) remain bsd/free/un-restricted.
Chris McKee10 months ago
+Demis Bellot Yep, I totally understand your motivations and wish you the best of luck. The pricing will be the clincher at which markets decide to use the product.
(Epic Generalisation warning)
You have a double whammy of introducing core-replacement functionality on a platform that's considered widely as being expensive/overpriced in a weird mix of 'Enterprise' orgs that will only push MSFT's nasty-ass wagon of crap around (usually still in .net 2 and IIS6 with devs who should of retired to street sweeping years ago) or the 'start=ups' who will be few and far between on the MSFT stack ($$$) as most devs don't want to get their lips chapped sucking Balmers lizzard bits to get a Bizspark licence; or Agencies that are looking for new and inventive ways to avoid spending any more than they've already spent hiring staff and paying for VS.
Tough crowd in many respects!

You've built a fantastic platform that's been a pleasure to use, and I hope it stays big enough to support you and to allow you to expand supplying a framework that's a shiny light on how MSFT should of done it rather than trying to create the overly flappy thing they've pushed out.


Anyhoo; will this affect ServiceStack.Text? 
+1
Eric Williams10 months ago
Congratulations +Demis Bellot — I can't wait to see what the future brings for you and for Servicestack!
Demis Bellot10 months ago
+Chris McKee Yep, this will cover all my future .NET contributions that I plan on working on whilst I'm full-time. Tho there will also be a "free-tier" for smaller projects.+1
Demis Bellot10 months ago
+Eric Williams thx Eric, yep should be interesting - will be good to get back into SS dev again and start realizing some of the future roadmap.
Chris McKee10 months ago
+Demis Bellot You'll need a fair few licences considering how many projects ss.text is in; its probably somewhere near on par with JSON.Net 
Demis Bellot10 months ago
+Chris McKee Not likely, SS.Text isn't anywhere near as popular as JSON.NET: https://www.nuget.org/packages?q=json
Chris McKee10 months ago
+Demis Bellot JSON.net's cheating, its not popular, its bundled. Microsoft apparently unable to find a developer who could write a fucking decent serializer ><
Rudy Alvarez10 months ago
Nice!, I can't wait to see SS take the world!, Great job +Demis Bellot 
Kevin Howard10 months ago
Excellent move!  I believe the #1 reason why organizations don't choose well-designed OSS technologies such as ServiceStack is due to the lack of support and accountability. With a competitive pricing model and the excellent support that you already provide, I have no doubt it will surpass the use of the so-called "standard" options, especially once it is proven that it is flexible, scalable, and out-performs them.

This is one reason why I recently devoted some time to clean up the ServiceStack submission to the TechEmpower/FrameworkBenchmarks project on github. Others have made significant contributions too because they believe in your product.  Best of luck and I will continue to support!
Kunjan Dalal10 months ago
WOW... I feel awesome... You know I love service stack. And trust we miss you over the year and specially on twitter.. I know maintaining work and OSS side by side is tough task but now you will be full time. It will be fun to see how things progress.

I just drafted blog series of servicestack with F#. And how you can put javascript library like angularjs and emberjs at work only on client side. And just for that I give a face lift to my whole blog. And this news came... and feeling was awesome. It provides me necessary push to complete it even faster.

In enterprise this is an issue, specially with .Net that questions asked by managers, who will support OSS, who will answer ? But this will give new direction to project like Servicestack, NanceFX and other open source projects.

If any, any damn help needed from my end please let me know. Will be more than happy to do so. :), And even though I know that service stack will going to rock and roll, still best of luck to you and servicestack.
Stefan Poulsen10 months ago
Awesome, if possible I would like to catch you for some specifics on the future just for reference in upcoming presentations :)
Lorenzo Dematté10 months ago
Good luck Demis! Personally, I think that it is the right move for the future of SS; it needs focus to really get momentum, and making a commercial version will give you the ability to fill in the "gaps" (docs, samples, some rough/incomplete parts) of an excellend design. Too bad to see you leave SE, though!
Herdy Handoko10 months ago
Here, take my wallet!

In all seriousness, good luck Demis! Can't wait to see what kind of innovation you'll bring to the framework now that you'll be working on it full-time :)
Jonas Eriksson10 months ago
Good news, best of luck! This makes me even more confident in keeping ServiceStack as the most important server web fw for the organisation I code for.
Pauli Price10 months ago
Awesome! I think this will go a long way towards making my management comfortable with my choice of the platform. I'm eager to see what comes next.
Providing commercial support combined with cross platform functionality is a strong value proposition. Congratulations and good luck :)
Alex Mizuki10 months ago
+Demis Bellot I definitely wish you luck and deeply respect and support your decision since after all you have to pay the bills.  But perhaps you should reach out to your compatriot Rod Johnson for advice, guidance and perhaps even more.  He took the model of consulting and paid support while keeping the Apache license for broad adoption.  I embraced Spring when I programmed Java because it made Java EE tolerable, just as I embraced ServiceStack because it makes the ASP.NET stack awesome.  Having an ASL or a BSD license was definitely a part of that.  But I can see the other side as well since .NET has a very different culture than java that has always had a vibrant open source community.+1
Demis Bellot10 months ago
+Alex Mizuki Thanks but I've considered all my options very carefully before I made the decision to resign - even dropping SS which had become a liability taxing all my free time with hardly any progress this year to show for it. SS is nowhere near complete, what it needs is less support burden and more skilled/focused dev effort. The harsh reality is that I was way more productive when SS didn't have any users. Adoption is currently slowing progress, to the point I no longer had enough spare capacity to even handle the support burden. As independent .NET OSS projects doesn't have cloud or server licenses it can leverage to justify commitment of additional full-time resources, I've instead drawn inspiration from other successful independent .NET companies, notably Xamarin and RavenDB which both offer flagship commercial products. The AGPL/FOSS Exception was inspired by RavenDB, WebSharper and even MongoDB server who all use it to offer dual commercial licenses and protect IP on their core products. I initially didn't believe in it either, but tracking RavenDB's progress has proven it works, where they've been able to generate enough revenue to support full-time development and continually improve the product. AGPL/FOSS-Ex is a red-herring, very few commercial companies will actually choose to adopt the AGPL license, its just an instrument that allows dev companies to protect their IP, ensures source code remains publicly accessible (important for being able to reason about software), allows free usage in OSS projects whilst ensuring companies who benefit from it commercially purchase a license and fund future development. Almost all .NET ISV's with successful products I'm aware of benefit from a commercial license, and their products end up much better quality because of it. Xamarin is a prime example who IMO is single-handedly improving .NET's image and making C# more appealing to outside communities - I've watched their OSX products grow from the start where it was barely usable on OSX to the quality product it is now, which has no equal. Now that I no longer have full-time employment, I've decided on the path which will maximize dev effort, reduce support and ensure future efforts are protected by an IP that may appeal to future investment. I honestly have no idea if SS can sell enough commercial licenses to fund future development, even free/un-restricted, no .NET OSS web fx stands a chance at relative popularity against the latest default MS fx. So already this means SS's potential market is limited to only companies who would consider evaluating an alternative fx. At the same time I believe in "the SS way", and given the opportunity of un-interrupted full-time development believe I can deliver a lot of USP's that can justify the licensing cost based on value-added features and time saved. If in the end it doesn't work out, I'll know sooner that it would never be self-sustaining, and I can safely exit the .NET OSS business having tried my best. As I see it the possible options going forward is to stick with the current v3 version of SS (which is both stable and feature-rich as it's the result of many man years of development). If in future you find there's a feature you need in the core framework that's missing, just take the v3 branch and add the feature in your local fork. You can later decide if the future commercially supported SS delivers enough value-added features to justify its cost and effort to upgrade. Otherwise if you want stick to a free/un-restricted .NET framework then there's still a number of quality options left in .NET: http://www.asp.net/mvc/open-source. I'm going to close this thread now as I don't want to expend any more effort re-iterating this. Barring any unforeseen circumstances, SS's future will be as described above - what's left is being able to execute it.+2