none
Aurelia.io - LightSwitch Replacement RRS feed

  • General discussion

  • Hello All,

    The following discussion to Jay Schmelzer has morphed to the topic of Aurelia / Breeze / KendoUI. So, I thought I would just start a new thread.

    Nik Ivancic from Telerik stated the following:

    "As a member of the Aurelia Tools team that created the Aurelia-KendoUI bridge (term bridge is used as an alternative to the word "interface") I want tp provide some additional information.

    This project is open source, maintained in Github as https://github.com/aurelia-ui-toolkits/aurelia-kendoui-bridge. Please see the related README file for all usual instructions about this project. Alternatively, please check Rob Eisenberg's blog http://blog.durandal.io/2016/01/28/aurelia-and-kendo-ui/

    I also want to provide a comment on "The demo/docs site is so involved and wordy that it's hard to absorb" made by joshbooker, which refers to our Aurelia application hosted on Github, written by using the Aurelia KendoUI bridge as one of its plugins.

    In this application, we tried very hard to present KendoUI "programming model" that became possible because of Aurelia (see http://aurelia-ui-toolkits.github.io/demo-kendo/#/samples/grid/basic-use as an example of what I mean by that). This application, named KendoUI components catalog is verbose by design trying to attract Aurelia application developers to use KendoUI controls presented as Aurelia Components - as well as to  attract KendoUI customers to Aurelia - best JavaScript application development framework.

    Seeing several references to Lightswitch, I could not resist the temptation to say that there is no better framework to revive Lightswitch as the best RAD environment available anywhere (yes, I am biased :-)"

    Nik also talked about the possibility of creating a designer for Aurelia / KendoUI if it would be a viable business decision.

    So, if there are any comments or suggestions on this, please post them here.

    Thanks

    Sunday, January 31, 2016 5:02 PM

All replies

  • "Nik Ivancic from Telerik"

    Not from Telerik - my company is Congral, but in this context I appear as a member of the team that created the Aurelia KendoUI Bridge project


    Nik

    Sunday, January 31, 2016 6:33 PM
  • Hi Dave,

    good post, but the controls there has long been of Syncfusion and Telerik. But where is the editor in Aurelia.io for the screen and the fields as well as in LightSwitch(LSML)?
    So we begin again to organize the whole! We search a "RAD"-System like LightSwitch.

    Where are the benefits from Aurelia???? We work with LightSwitch and Syncfusion Controls  and generate the reports with Devexpress xtraReports in the backend and can fast develop.

    Regards

    j.nord


    • Edited by J.Nord Monday, February 1, 2016 8:14 AM
    Monday, February 1, 2016 8:11 AM
  • Nik, given the sad state of affairs regarding LightSwitch, I believe you will get a lot of support from the members in this community, should you take up the challenge of creating an open source LightSwitch alternative, say with commercial support to support your effort. I think the community will be hesitant to get involved in something that is closed source, given the LightSwitch saga - once bitten, twice shy. The way to take this forward is probably to take leadership and propose what you want to do and ask the community for feedback on such a proposal. Timing is of course critical as people have projects to implement and you'd want to attract them before they move on. Best of luck, looking forward to what you I'll come up with and to support the effort where possible.

    Regards, Xander. My Blog

    • Edited by novascape Monday, February 1, 2016 8:59 AM
    Monday, February 1, 2016 8:56 AM
  • Just to clarify my statement above regarding commercial support and open source: I'm happy to pay for an alternative tool, either in license form or support form, I just don't want to end up in a situation where the tool is discontinued and we have no way forward, hence my insistance on open source.

    Regards, Xander. My Blog

    Monday, February 1, 2016 9:05 AM
  • Agreed,

    The complete absence of any form of communication, despite statements of intent to the contrary, let alone any comment from MS regarding open-sourcing LS (obviously they're afraid of something!) puts many, if not all, of us in the position outlined by you Xander.

    MS now has no credible RAD LOB dev tools, at least none that it is prepared to even acknowledge, let alone support; so it is vital that any suggested alternative does not ultimately leave those who may invest their trust in it, in such a terrible position again!

    Open source it must be


    Ian Mac

    Monday, February 1, 2016 12:11 PM
  • Where are the benefits from Aurelia???? 

    Not present at the moment, meaning that I might have made a mistake by making a reference to Aurelia in the content of the discussion "We want LightSwitch and we want it now". However considering Microsoft's lack of response to such requests, I wanted to see whether it would make sense to build a LightSwitch-like RAD system on Aurelia platform. To further substantiate the power of Aurelia, I referenced the Aurelia KendoUI bridge (open source) project (see Dave's initial post above.

    So, the purpose of this discussion is to find out whether there is enough interest to build a new Lightswitch and what are the requirements for such task 


    Nik

    Monday, February 1, 2016 2:30 PM
  • Xander LightSwitch second generation as open source is the only way to do this, we are in full agreement. I would be happy to bring Rob Eisenberg, creator of Aurelia and CEO of Durandal into this discussion, once we find out what is the real demand for such tool, what UI toolkits need to be supported (KendoUI, Syncfusion, ...) how many developers would join the effort to build LS, etc.

    Nik

    Monday, February 1, 2016 2:36 PM
  • Hi Nik,

    I don't think we need to build a RAD tool.  I believe there will be enough interest outside of the LS community.  Matt Duffield has already started on one.  Check out his video here.

    Monday, February 1, 2016 3:12 PM
  • Nik,

    I'm glad to here you're not from Telrik.  Don't get me wrong there a great company.  Whatever solution we find we need to make sure it's vendor agnostic.  I personally like Syncfusion.  There community edition is free and they give back to the community in other ways.

    Monday, February 1, 2016 3:17 PM
  • Hi All,

    I cross posted this thread in the 'town hall' thread, so be prepared - the flood gates of disenfranchised LS users (if they are still listening to these forms) are now open! ;-) 

    I would definitely support the effort.  Something tells me Mr. Eisenberg has plans for some tooling after release, but doesn't hurt to bring some ideas of what was successful in LS to the table.

    PS.. Nik, regarding my 'hard to absorb' comment... please know that was not meant to be critical.  Your demo site is very impressive.  It's prolly just my thick skull.  ;-)

    Cheers,

    Josh

    Monday, February 1, 2016 3:23 PM
  • I did check Matt's video so, I can say that our plans are more ambitious and fully synchronized with Aurelia's CEO. We are also discussing collaboration with Syncfusion as one of the premier supporter's of LightSwitch.

    Nik

    Monday, February 1, 2016 3:23 PM
  • Yes, I agree there are more pieces that need to be considered in this RAD ecosystem.  But it is a start.  

    I can envision Durandal developing and marketing a product that would be that tool.  And knowing Rob he would make sure that whatever code that gets generated would be configurable.  This would mitigate issues when new technology becomes available or you want to generate Angular2 instead of Aurelia.

    Monday, February 1, 2016 3:35 PM
  • Hi Josh You indicated the obvious fact that an About page written for "managers" is not the best way to introduce a technology tool - so this screenshot below is the preview the new default "About" page.


    Nik

    Monday, February 1, 2016 3:35 PM
  • Hi Nik,

    Off topic:  I did find that page lacking info specific to your project, but my real trouble began because I went directly to the 'Bridge developers | Setup' page and found the first couple paragraphs referring back to 4 other earlier sections.  So I quickly got lost with the feeling I had a lot of reading ahead in order to get the high level of what it takes to develop a 'bridge'.  Serves me right for skipping ahead I suppose.  ;-)

    Anyway, I'll quit cluttering this thread with my deficiencies...if I have suggestions as I endeavor to absorb, would github issues be a good place?

    Thanks again. 

    On topic: I look forward to seeing where this might go...

    Cheers,

    Josh

    Monday, February 1, 2016 3:57 PM
  • Hello LightSwitch developers. Rob Eisenberg here. Nice to make your acquaintance.

    I've been alerted by several individuals of the current state of LightSwitch and the need for a viable alternative.  Aurelia would be an ideal foundation upon which to build a next-generation LightSwitch alternative. In fact, it was designed from the beginning to be a sort of platform that others could build new platforms on top of...and we've seen people do this already in specific vertical markets.

    We have planned for some time to build a set of tools that could be used to create Line of Business applications. Our idea was not quite the same thing as LightSwitch though there were some similarities in a few parts. However, we would be open to consider developing such a product. That said, it would be quite time consuming and probably expensive for us to do this. So, it would need to be commercial in nature. We may be able to find a way to do it as an open source commercial project through a carefully constructed license. There are other options as well...such as potentially us running a Kickstarter to build the tool. That would fund the effort and then it could be made open source afterwards with a commercial support model. There are many options, but before we do anything we would want to know that this is really wanted and needed by enough developers to justify the effort. Make no mistake, I very much want to build this...but we have to also consider our business carefully here.

    I'm meeting with the other executives this evening to discuss business planning. I will talk this over with them. We are a diversified group so I'm not sure what everyone's opinions will be. In the mean time, I would love to hear more feedback from the LightSwitch community.

    Our ears are open and we're earnestly seeking if there's a way we can partner together with you for the mutual benefit of all those involved.

    Thank you,

    Rob Eisenberg - CEO, Durandal Inc.


    Rob Eisenberg

    Tuesday, February 2, 2016 1:55 AM
  • Rob,

    Amazing that you have responded to the call, so long unheard (or ignored) by those that should have been listening.

    Fully understand your rationales and caveats but it's just great to know you're even considering what role your business could play if things were to pan out and be commercially viable for you guys too.

    Count me in - just hoping there is a veritable flood of other responders!

    Thanks 


    Ian Mac

    Tuesday, February 2, 2016 11:09 AM
  • This is all very exiting, I hope more community members show interest in this. My knowledge is nowhere near to you guys but I am in with whatever help I can do.

    Divyang Vyas

    Tuesday, February 2, 2016 12:12 PM
  • Hi Rob,

    I would be very interested in this. I too like J.Nord and Dave Vorgang use SyncFusion a lot.

    Thanks,

    Andrew


    A Fleet

    Tuesday, February 2, 2016 2:31 PM
  • "If you want to leave something great, let him free"
    open source is the answer, you can count me too
    Tuesday, February 2, 2016 3:22 PM
  • Thanks Rob for your willingness to explore this option.

    I think your idea to start a Kickstarter to fund the initial build is very wise.  For one, it would put people’s money where their mouth is. Also, I know for myself, I’ll be more involved if I have a stake in it.  I would assume you would have certain levels of incentives based on the amount.  Maybe support hours or contract time in addition to the main product.

    Obviously this would be a product that would extend beyond the LightSwitch community and there would need to be additional research to see what the interest level would be for other types of programmers.

    One of the major benefits I see is that we basically have the specifications written for us.  If you check out https://msdn.microsoft.com/en-us/vstudio/gg491708 you can explore the LightSwitch architecture.

    Thanks again for just considering this venture.

    Dave Vorgang


    • Edited by Dave Vorgang Tuesday, February 2, 2016 6:09 PM Spelling
    Tuesday, February 2, 2016 5:58 PM
  • I am very interested in LightSwitch-like Aurelia LOB applications and will follow this closely.  From my perspective the crucial need is for a new flexible and extensible client side framework that can generate screen templates by inferring from a data model.  I believe most developers could produce a web service, whether through LightSwitch OData or through EF and Web API.  Rather than another IDE to design services, the focus in my opinion should be on a RAD client that can consume those services and still be fully customizable.  The current LightSwitch client is too tightly coupled to jQuery mobile, is difficult to customize, and has very limited navigation options.
    Tuesday, February 2, 2016 6:45 PM
  • I echo Hessc (and all). 
    I was really hoping even after all this time, that something would be done/improved regarding the HTML client. 
    I've been following and playing around with Aurelia for the past year and it looks amazing.
    I am however, a little concerned with the "counting swords" - but I can understand it is necessary. From what I've experienced, there's a LOT more devs using Lightswitch than what one feels here in the forum. And a lot more using whatever RAD tools. 
    I would ask "Is the market lacking a RAD tool on the lines of LS?" 1000%Yes.
    Tuesday, February 2, 2016 8:46 PM
  • Rob, welcome to the LightSwitch community! This thread, even just as an idea for now, is quite possibly the best news to reach the LS community in the last 14 months or so.

    I'm quite willing to make a contribution to a Kickstarter campaign should that prove the best way to get this off the ground.

    Although I (personally) would ultimately like to remove my dependence on Microsoft altogether, .NET and SQL Server is still used quite a bit on the backend, so it would be good for a lot of developers if this endeavor can include a backend service as well, whether RESTier, WebAPI, OData, Etc.

    The key concept here is a "RAD tool for Aurelia" that will allow us to be as productive (or more) in Aurelia as we have been in LightSwitch.


    Regards, Xander. My Blog

    Tuesday, February 2, 2016 9:04 PM
  • I also believe the LS design experience has a lot to offer in terms of 'what works' for RAD dev.

    Agree with Hessc...let's focus on the UI design experience - rather than backend stuff.  LS server tier produces OData so those who want to continue to use it for the service will be fine if we focus on a UI app builder.

    Much of the simplicity in LS UI design experience is due to the hierarchical model of the content item tree.  The hierarchy of groups/controls is easy to understand at design time and it's container-child model makes perfect sense when rendering an adaptable & responsive UI at runtime.

    Those with greater understanding of Aurelia will know best, but it seems to me a LS 'content item' would translate well into an Aurelia Component and because components can contain other components, the tree model would be a nice fit.

    That's my 2 cents for now.  Please let us know how we might better collaborate.

    Cheers,

    Josh

    Tuesday, February 2, 2016 9:10 PM
  • Josh, I could not agree more and 2nd your 2 cents!
    Tuesday, February 2, 2016 9:20 PM
  • Agree with Hessc...let's focus on the UI design experience - rather than backend stuff.  LS server tier produces OData so those who want to continue to use it for the service will be fine if we focus on a UI app builder.

    The UI design experience is obviously where the pressing need is, but I'd also campaign for a new backend technology - unless there is something else that we can piggyback onto. The LS backend has not evolved and is really not that scalable (can't do serverside caching for example) and you need to add RIA services for anything slightly more complex. Plus I'm not convinced that the LS data context is perfect for RIA - using an EF context directly works better IMO, but then you double up which is not ideal.

    Most complex LS applications IMO need to be built on top of a RIA service rather than a LS native database data source. RIA is powerful but slow to develop and there are numerous restrictions when used from LS, like not being able to refresh related entities without jumping though numerous hoops. RIA as a technology is also on the way out AFAIK.

    OData seems to work well, but is it that popular outside of the Microsoft community? It would be interesting to get a view on this from outside the Microsoft community. If it is indeed the most suitable form of REST to use then so be it. Perhaps RESTier with support for custom/shaped entities will be a good backend solution, depending on where RESTier goes.


    Regards, Xander. My Blog


    • Edited by novascape Tuesday, February 2, 2016 9:47 PM
    Tuesday, February 2, 2016 9:46 PM
  • The UI design experience is obviously where the pressing need is, but I'd also campaign for a new backend technology - unless there is something else that we can piggyback onto.

    I totally agree new backend tech is wanting and one of the best rapid dev aspects of LS is it's ability to design, build and publish a 3 tier app in one efficient tool set.  This gain in developer efficiency came at the cost of flexibility.  Ultimately, due to the lack of advancement by MSFT, the tightly- coupled architecture of a monolithic LS project has become a dead end. 

    Thankfully they did leave us a clean separation between server and client by using OData v3 protocol between them.

    This avails us to proceed with any client tech that can speak OData while still reusing the LS backend.

    It also allows us later to create a new backend that exposes OData which'll just work with the new client.  So LS apps can be incrementally transitioned completely away from LS.

    For SQL Server + .NET backend, EF code-first migrations make a lot of sense.  Tack on RESTier and you have OData plus rich domain logic - back-end done.  Assuming RESTier gets to GA.

    So IMHO, any useful app builder should speak OData (self-describing data protocol with relationship support, concurrency, batching etc - all built-in). A dependency on something like breeze.js (client side OData, caching, offline, etc.) wouldn't bother me a bit.

    As Xander, points out, OData isn't for everyone.  It needs to do plain old REST with some mechanism for passing entity schema from server to client.  Perhaps the ability to get schema from swagger would be a nice alternative to OData(?)

    I wonder if makes sense to build a custom breeze DataServiceAdapter for swagger?

    ...2 cents more.

    Cheers,

    Josh

     
    Tuesday, February 2, 2016 11:41 PM
  • I was thinking that the binding would be done with simple Web API routes.  The Web API would be codegen as wrappers for the actual underlying model.  That underlying model would also be codegen but it could be swapped out to use whatever technology you want.

    So, the Web API interface would be standardized but underlying technology would change.

    My .2 cents.

    Wednesday, February 3, 2016 12:41 AM
  • I was thinking that the binding would be done with simple Web API routes.  

    Web API and code gen is great, Dave. Did you know that ASP.Net WebAPI paths and response schema can be described in swagger using swashbuckle which comes OOB in an ASP.Net WebAPI project?

    Swagger is a simple yet powerful representation of your RESTful API. With the largest ecosystem of API tooling on the planet, thousands of developers are supporting Swagger in almost every modern programming language and deployment environment. With a Swagger-enabled API, you get interactive documentation, client SDK generation and discoverability.

    Keyword discoverability.  So you could use Microsoft WebAPI backend and I can use a nodejs backend - we both describe our APIs with swagger - the client will know how to interact with both.

    Plus swagger ui puts a nice interface on your API. 

    Here's Northwind db WebAPI OData using RESTier and swashbuckle to expose swagger ui:

    http://bit.ly/1Qb5L51

    and here's what the swagger doc looks like:

    https://gist.github.com/joshbooker/de7e40be3d6def1731b5

    Josh 

     

    • Edited by joshbooker Wednesday, February 3, 2016 2:28 AM
    Wednesday, February 3, 2016 2:05 AM
  • Wow, what a great thread. Good to see Rob actually contribute here.

    +1 for a new RAD tool that does Aurelia.

    To add my 2p's worth, I vote for a tool that also serverside. Web API wrappers around a rich model works really nicely. I've got a little tool I've been playing with called 'Agilis Lux' (stick it into google translate from Latin) that scaffolds out an application from an entity model using T4. I currently have an ASP.NET MVC Razor views version and an Angular 1.x version. I had planned to at some point also try to develop Angular 2.0 and Aurelia versions.

    Wednesday, February 3, 2016 10:32 AM
  • Josh,

    Swagger looks really interesting.  I'll play around with it later today.

    Thanks

    Wednesday, February 3, 2016 4:39 PM
  •  I've got a little tool I've been playing with called 'Agilis Lux' (stick it into google translate from Latin) that scaffolds out an application from an entity model using T4. I currently have an ASP.NET MVC Razor views version and an Angular 1.x version. I had planned to at some point also try to develop Angular 2.0 and Aurelia versions.

    Delordson, 

    Do you have a link to this tool or is it something you developed strickly for your own development?

    Thanks

    Wednesday, February 3, 2016 5:03 PM
  • Hi Dave,

    It's something I developed for myself as a quick way of scaffolding out an application. I plan to get it to github at some point once its in a presentable state. With the kind of licence Rob was talking about to allow some commercial aspect to it but making it open source.



    • Edited by Delordson Wednesday, February 3, 2016 5:43 PM
    Wednesday, February 3, 2016 5:42 PM
  • I think the aurelia + swagger route is the most viable option.

    Swagger is a great indirection instrument, it makes the client completely independent (Rob likes that, I start to understand him:) ) from the server (obviously, both sides must 'speak' rest) and the swagger metadata can

    1. help the client tooling to generate the 'touchpoints' or the client proxies.

    2. help the client tooling to scaffold screens.

    I'm questioning if "the new lightswitch" still needs a feature rich designer as it has currently in visual studio (I guess rob want to be independant from VS as well, ... i hope so). Most trouble which LS comes in via this designer. You will understand why I believe that the tool doesn't need a fancy designer when observing following markup related to the tool named 'aurelia interface' maestro Eisenberg spoke about in following blog post:

    http://blog.durandal.io/2016/02/01/big-february-status-update/

    I love this markup, it is super lean and contains everything you need. Amazing. Compare this markup with angular 2 or worser, LS html client...

    So, if some scaffolding tooling would generate based on the swagger meta data, crud screens with the above markup, we would have a great starting point.

    Definitely, if you see that the above markup leads to:

    my 2 cents...

    cheers


    paul van bladel ==independent enterprise application architect== http://blog.pragmaswitch.com




    Thursday, February 4, 2016 12:22 PM
  • Nice Post Paul!  Glad you've joined the discussion.  Agree a full featured designer is not necessary and VS dependency is certainly not desirable. 

    I would argue that all data-centric apps have a bunch of attributes common to all screens and would benefit from a simple web-based designer to build, say a json model of screens. 

    For example, all screens have a title bar, all screens have command bar with buttons and associated methods, maybe you want the commands in the header or the footer.  Tabs and grouping of controls into containers is important for reactive layout.

    As earlier noted, I just think the way LS screen designer does the contentItem hierarchy is brilliant in it's simplicity with Group 'controls' containing child controls which layout as either Rows or Columns. 

    https://msdn.microsoft.com/en-us/library/gg445202.aspx

    How hard could it be to design a tree builder like that?

    Screen types: Browse (collection), AddEdit (singleton), View (singleton)

    Screen template hierarchy:

    >Screen

    >>Header 

    >>>CommandBar

    >>Tabs

    >>>Tab

    >>>>Group Control

    >>>>>ChildControl

    >>Footer

    >>>CommandBar

    A simple tool that builds the contentItem tree in json.  Then json model is used to generate or scaffold Aurelia templates would be nice way to speed up dev and it would make Aurelia approachable for much of the citizen-developer crowd who benefit from LS. 

    Heck, we could even import a LS HTML app by translating model.json from LS into the new AureliaApp.json. 

    By 'we', I mean those of you with more skills than I, of course.  ;-)

    Josh 





    • Edited by joshbooker Thursday, February 4, 2016 2:15 PM
    Thursday, February 4, 2016 2:03 PM
  • Completely agree Josh. 

    Maybe, aurelia interface has similar intentions? 


    paul van bladel ==independent enterprise application architect== http://blog.pragmaswitch.com


    Thursday, February 4, 2016 6:39 PM
  • Just because I started this thread it does not mean that I make any decisions or that I speak for this group.  Most of you know more then I.

    But, it would be good to let Rob know, as a group, any concrete decisions.

    Can we propose to him that the designer should look and function similar to LightSwitch's Designer?

    Let me know if I'm being out of line here. 

     
    Thursday, February 4, 2016 7:33 PM
  • All,

    Great conversation going on here. It's neat to see your excitement about possibilities. There are a number of different commercial products my company is considering building. We can go in a lot of different directions. I have a few questions that would really help us...

    * How long was LS in development before it's v1 was released?

    * Do you think that this type of product could produce at least $500k in revenue annually?

    * Do you think a Kickstarter could raise $500k up front to produce a product in 1yr?

    * Do you think Microsoft would be interested in helping to promote a Kickstarter and/or product and "bless" it as a migration path for LS users?

    * Would you be ok with a closed source product IF the licensing stated that it be built with open source technology and should the company ever discontinue development, it must open source everything?

    * Would you be ok with only a Web/Aurelia "output" for the app (assuming support for browser, desktop and mobile outputs through tech like PhoneGap and Electron)? Or do you need multiple other outputs? For example, do you need an Aurelia output and an Angular output? Does that matter or do you only care about being able to run cross-platform, and not so much what "framework" is used under the hood?

    I've got a lot more questions, but these are a few that came to my mind while scanning the thread. Thanks! (Forgive me if this sort of research isn't appropriate here.)


    Rob Eisenberg

    Thursday, February 4, 2016 8:13 PM
  • * How long was LS in development before it's v1 was released?

    I would guess 18 - 24 months based on early videos and discussions about the development of the product. However, they started from scratch with just about everything (except Silverlight and EF that existed) so you will have a huge head start given the state of Aurelia today. 

    * Do you think that this type of product could produce at least $500k in revenue annually?

    The question is what would the cost be and how many developers can be attracted. A $500 pa license fee (say subscription based) would be acceptable to me which means you need 1000 developers to come on board. So I think the answer is yes.

    * Do you think a Kickstarter could raise $500k up front to produce a product in 1yr?

    Difficult to say/predict - I would think that you will at least be able to raise $250K

    * Do you think Microsoft would be interested in helping to promote a Kickstarter and/or product and "bless" it as a migration path for LS users?

    We will need Microsoft to comment, but I suspect that they won't comment and won't bless it either. They simply don't seem to care about developers (my cynical view).

    * Would you be ok with a closed source product IF the licensing stated that it be built with open source technology and should the company ever discontinue development, it must open source everything?

    I would prefer opensource, but I could be convinced of this model too. I think you and your company do have enough credibility to be able to back this, but you'd have to put safe guards in place in the event of a buy-out.

    * Would you be ok with only a Web/Aurelia "output" for the app (assuming support for browser, desktop and mobile outputs through tech like PhoneGap and Electron)? Or do you need multiple other outputs? For example, do you need an Aurelia output and an Angular output? Does that matter or do you only care about being able to run cross-platform, and not so much what "framework" is used under the hood?

    Personally I would be absolutely fine with Aurelia only. As already mentioned, I think there should be something for the back end as well, to provide a full-stack offering.

    You may also want to explore options to reach out to the IronSpeed community, given that it is in the process of closing down and those developers are also looking for an alternative. You probably also want to be aware of the reasons for their closure:

    ISD Alternatives


    Regards, Xander. My Blog

    Thursday, February 4, 2016 9:04 PM
  • I know that Beta 1 was released on August 23rd 2010 and the Official release July 26th, 2011.  I was using Beta 1 soon after it was released in a production environment.

    I don't think $500k would be reasonable if you just include LS developers.  I think it would bring in revenue of at least $500k if this is embraced by the programming community at large.  I think it also has the potential to raise revenue in other areas like support and custom contract work.  Once you have an end-to-end programming environment it changes everything and this will bring more developers.  (If you build it, they will come).

    Kickstarter: I don't know, $500k seems like a lot. I have no idea.

    Microsoft: I don't know.  Microsoft doesn't even want to talk about LS.  In fact, there was some litigation brought against Microsoft.  I think it was all settled.

    I would be okay with a closed source product.  I see this more as an IDE or an extension to Visual Studio.  Also, if something does go wrong you open source it.

    I think it needs to be able to generate various frameworks.

    Just my opinion.

    Thursday, February 4, 2016 9:05 PM
  • * Do you think that this type of product could produce at least $500k in revenue annually?

    * Do you think a Kickstarter could raise $500k up front to produce a product in 1yr?

    I have no Clue.  I can tell you though it would vastly improve adoption of Aurelia by the masses.  And if you had a paypal donate button, I'd tap it repeatedly. ;-)

    * Do you think Microsoft would be interested in helping to promote a Kickstarter and/or product and "bless" it as a migration path for LS users?

    Doubtful.  I did hear someone elude to MS possibly doing a migration path to something, but that was a year ago and I've no reason to believe that is true at this point.

    * Would you be ok with a closed source product IF the licensing stated that it be built with open source technology and should the company ever discontinue development, it must open source everything?

    Dev tool yes - framework and resulting app must be open source HTML5+js and host-able anywhere without ties.  For example, form.io is neat but I'm not a fan of the fact my 'app/api' must be hosted with them.  It appears to be a consequence of their paymodel, but I could be wrong.  I'd be happy to pay for the dev tooling, but want resulting app to be purely HTML5+JS, self hostable, moveable etc.  When tooling breaks or goes out of business or I just get tired of paying monthly fees, I want my app to work even if I must code by hand thereafter.

    * Would you be ok with only a Web/Aurelia "output" for the app (assuming support for browser, desktop and mobile outputs through tech like PhoneGap and Electron)? Or do you need multiple other outputs?

    Sure but ideally the tool has configurable code gen - not required v1 of course.  Yes it would be important to access native device features via phonegap, cordova, etc.

    *For example, do you need an Aurelia output and an Angular output?

    No need for angular here, but you might get some converts by letting them gen both Aurelia and angular - then compare....Aurelia wins!

    Thanks Rob for your interest.

    Have a great day!

    Thursday, February 4, 2016 9:48 PM
  • LightSwitch - both the Silverlight version and the HTML version - typically allowed us to create 80% of the application  in 20% of the time. But then the remaining 20% took 80% of the time.

    Aurelia RAD should build on the 80% in 20% of the time, but then allow us to do the remaining 20% in 20% of the time, resulting in the overall development to only take 40% of the time it usually takes to hand craft an application.

    If that can be achieved, it will be a big winner.

    On the client side, a one-way flexible wizard-type code generation or scaffolding tool - along the lines of what Delordson is building - is perhaps not a bad idea. If it provides a powerful eco system of templates then it could work well.

    A client side designer like the one LightSwitch uses is nice, but not essential IMO - especially given the straight forward syntax that Paul gave an example of above.

    The ability to build your own reusable client side components through composition will be great too.

    Form layout is one of the areas that our developers are sometimes struggling with. Bootstrap grids only get you so far. DevExpress seems to have realised this in their latest HTML offering by providing nice form layout components. This would be nice too.

    On the server side it is perhaps better to use a "designer" approach with appropriate hooks where the the model can easily be updated for changes. This process should also update the javascript client side model - this is very important.


    Regards, Xander. My Blog

    Friday, February 5, 2016 1:07 AM
  • Oh, and one of the most important server side features for me is the ability to shape your entities and create additional custom entities to make UI development easier. Database entities do not always match nicely with UI view models.

    In LightSwitch the only way we can do this is via 1) database views and 2) RIA domain services entities - this works but is not ideal.

    Ideally the server side will expose your EF (or DAL) entities by default, but then allow you to shape them if required and add additional custom entities and be able to easily declare relationships. Hooks can be provided server side to implement the CRUD methods for those custom and/or shaped entities.


    Regards, Xander. My Blog

    Friday, February 5, 2016 1:13 AM
  • * How long was LS in development before it's v1 was released?

    I think I remember there being about 1 year or so between the initial beta and v1 of the SilverLight version. The html client seems to have been done faster but I don't have any inside info.


    * Do you think that this type of product could produce at least $500k in revenue annually?

    * Do you think a Kickstarter could raise $500k up front to produce a product in 1yr?

    Hard to say for sure. I had a lot of fun building and selling LightSwitch extensions back in the day but we were always talking a few thousand dollars a year though it did drive more revenue through consulting. My best guess would be that a fully featured LightSwitch replacement at say $99 a year could make that amount but it might need to come with a LightSwitch converter/upgrader

    * Do you think Microsoft would be interested in helping to promote a Kickstarter and/or product and "bless" it as a migration path for LS users?

    I would say no. Having said that, it might not be such a bad thing. I think LightSwitch developers are more likely to listen to take note of an endorsements from the community leaders here rather than Microsoft at this stage.

    * Would you be ok with a closed source product IF the licensing stated that it be built with open source technology and should the company ever discontinue development, it must open source everything?

    Personally? Yes. I can't speak for the rest of the community but my guess would be once bitten twice shy. My instinct is that an open source community version along side a licensed Pro version with more configurability would be the way to go. 


    * Would you be ok with only a Web/Aurelia "output" for the app (assuming support for browser, desktop and mobile outputs through tech like PhoneGap and Electron)? Or do you need multiple other outputs? 

    I think the answer this this probably yes but...

    Most users wont mind what web client is delivered as long as it supports the modern web. Pro/Semi Pro developers would likely want extension points so they can plug in the client framework of their choice though this may not have to be in v1.



    • Edited by Delordson Friday, February 5, 2016 6:08 AM
    Friday, February 5, 2016 6:05 AM
  • Rob,

    Love the feedback that is popping up here.

    Maybe a good idea to open an separate gitter channel for this discussion, no? There are really a lot of former lightswitch devs in this group who have  in depth experience when it comes to line of business app development. If you agree and open the channel, i'll point these guys (with the help of xander and josh?) to the channel? 

    take care

    paul.


    paul van bladel ==independent enterprise application architect== http://blog.pragmaswitch.com

    Friday, February 5, 2016 8:07 AM
  • I support all of you. See Rob Eisenberg here is very promising
    Friday, February 5, 2016 9:25 AM
  • I use Lightswitch exclusively for building HTML Apps (CBA) for SharePoint online. Would this be possible using a new too that is independent from Visual Studio?

    A Fleet

    Friday, February 5, 2016 9:34 AM
  • Maybe a good idea to open an separate gitter channel for this discussion, no? There are really a lot of former lightswitch devs in this group who have  in depth experience when it comes to line of business app development. If you agree and open the channel, i'll point these guys (with the help of xander and josh?) to the channel? 

    Agree let's move out of this forum. I find gitter to be rather one track - hard to separate topics and search history.  It would be nice to have a multi-thread forum format.
    Friday, February 5, 2016 1:09 PM
  • We are looking at this from a LS perspective - our beloved tool does everything -

    1. database,
    2. ORM,
    3. EF Backend,
    4. Middle tier business logic, and
    5. HTML\JS front end. 

    It would be nice to hear from Rob and or Nik where their interest lies regarding Aurelia RAD.  Being a UI framework, I suspect you are envisioning tooling for only #5, no?

    Don't get me wrong, I agree with Xander and others that the ideal situation would be for LS backend code gen to move forward beyond WCF-DS toward WebAPI\OData, but that's not happening.

    At the same time, forced by the lack of advancement, I've found I rather like decoupling the backend and frontend tooling as long as there is a easy/reliable way to update data model\schema to the client tooling.

    So if the initial idea for Aurelia RAD is to do #5 only, let's use swagger to pass data schema and allow tooling to - 'update schema' from changes made on the server.

    When it comes to client side javascript data management, rather than recreate the wheel let's depend on a popular OSS library.  Looking at Breeze and JayData there are bits I like about each.  The former does queries more Linq-like, yet the latter seems to take-a-JSON-model-and-go-with-it a bit easier, I think.

    I'd love to see an Aurelia-JayData plugin similar to the Aurelia-Breeze.  Add the ability to init entities from swagger schema and the client side data bits are done - data from any backend (OData, WebAPI, REST, node, whatever) is available to Aurelia - assuming swagger is provided at design time.

    Friday, February 5, 2016 2:26 PM
  • I also vote for a traditional forum over a Gitter channel.

    Regards, Xander. My Blog

    Friday, February 5, 2016 9:06 PM
  • Even Issues list in an empty github repo would work - albeit not private.

    PS... a few of us are not getting notifications from this thread.  It happens from time to time on this forum.  More reason to move on I suppose.

    Friday, February 5, 2016 9:15 PM
  • I created a forum here.  Let me know if this would be a good alternative.

    If not, I'll just delete it.

    Thanks

    Friday, February 5, 2016 11:36 PM
  • The reason why I proposed gitter is that it is the communication medium for the aurelia community. 

    https://gitter.im/Aurelia/Discuss

    So, it is not only a technical matter, but also a matter of getting people on board, I.e.  the aurelia people.


    paul van bladel ==independent enterprise application architect== http://blog.pragmaswitch.com

    Saturday, February 6, 2016 8:38 AM
  • The reason why I proposed gitter is that it is the communication medium for the aurelia community. 

    https://gitter.im/Aurelia/Discuss

    So, it is not only a technical matter, but also a matter of getting people on board, I.e.  the aurelia people.

    I agree that gitter is a great communication tool and it would allow aurelia developers to get involved.  I just think it might be a little premature at this point.  Also, a forum better lends itself to organize specific topics.  With gitter I feel I need to be watching it closely so I don't miss anything. 
    Saturday, February 6, 2016 4:06 PM
  • Yeah, you are right, gitter is not optimal.

    paul van bladel ==independent enterprise application architect== http://blog.pragmaswitch.com

    Saturday, February 6, 2016 7:11 PM
  • What an awesome thread!

    First off, Mr. Eisenberg, thank you so much for doing the right -- and smart as you always are :) -- thing here and reaching out to yet another failed community of Microsoft developers who have been ignored left aside by the very company they trusted for their livelihoods and passions. FWIW, I am not even sure if you know, but I have actually listed you in my site's thank you/credits and wanted to make sure you know about that).  Anyways, you know that I am a longtime fan!

    Secondly, as a MSFT developer for over 15 years, I would like to encourage everyone that if they do use Aurelia (and they should -- it's awesome), to please (PLEASE) consider also switching the server side to NodeJS as well.  This is because if you do switch to a NodeJS client on top of a .NET server application, you end up breaking DRY and encapsulation.

    Additionally, you also split your organization's resources between JavaScript and .NET (since they are at present completely incompatible languages/technologies) and end up creating generalists in two languages instead of experts in one.

    Now, you might be thinking... if I am a MICROSOFT developer for 15 years, why am I pushing you to a non-Microsoft technology stack and suggesting to ditch .NET altogether... Well, like you, I have been putting up with MSFT since Silverlight's shift in 2011 and I am really starting to get to my limits here, and seeing threads like this one only accelerate this fact.  In fact, I do have a vote on Visual Studio's UserVoice that does appeal to this very problem space, but honestly if I do not see some movement soon, Aurelia, NodeJS, and Mr. Eisenberg become the more sane (and economical) choice by the day. :)

    I hope that helps.  I am also hoping to provide a cleaner, better, and tidier ecosystem for the reasons outlined above.

    Finally, I am not sure if it has been presented here yet, but the awesome part about using NodeJS/Aurelia is that you can pair it with Cordova and confidently reach the nearly 3-billion devices in the marketplace today.  All while being able to share code between the client and server... you know, like Silverlight did 5 years ago. :P


    • Edited by Mike-E_wins Wednesday, February 10, 2016 11:56 AM
    Wednesday, February 10, 2016 11:54 AM
  • FWIW, this thread along with another thread just as telling has inspired a new post: 

    The .NET to NodeJS Exodus: Sound Off

    I have included this thread in the comments.  Please feel free to follow along and/or contribute there.  As time goes on I will be looking for threads like this one (and the latter thread) to add to the evidence.

    Thursday, February 11, 2016 10:00 AM
  • Hi Everyone - (for some reason I cannot use links - please paste into browser.....)

    I have a suggestion to get 'some' response from people 'in the know' at https://live.asp.net They have a weekly 'Google Hangout' on Wednesdays whereby people can ask questions about .Net Core.

    I already tried to get a response here from them a few weeks ago https://youtu.be/niCDYdrCOu0?t=16m22s . (The live sessions get archived)

    How about this suggestion...

    (1) In this forum thread, formulate the question to be asked..something that will get their attention - short, direct -  but without sounding standoffish or rude

    ...something like

    "Can you guys have a 2 minute discussion on possible RAD tools that may come to .NET Core. Microsoft Lightswitch HTML is great for a person like me who came from a world of Microsoft Access and small 'departmental' data driven applications. What I like about Lightswitch is that with minimal coding you can get a very functional HTML5 application that is data driven with inbuilt security ... btw, Beth has moved on to a different department now 8-( "

    (2) then, on the day, without sounding like it's an ambush, everyone 'vote-up' the question in Google Hangouts and when it is posted + also post + "Gee, that is a great question!". 

    Thoughts?

    Kon.

    • Edited by kondaments2 Thursday, February 18, 2016 12:26 AM add my name at end
    Thursday, February 18, 2016 12:25 AM
  • Hi,

    sounds like a great idea.
    One should also point out the real advantage with a "shell" like LS. They can work on their low-level changes on webapi etc. until it's stable and then inroll it into LS, either as a new feature or then as a change not even noticed by the LS-users. Then they don't have to get so many complaints about breaking development for people.

    Regards Roger.


    • Edited by RogerExp Friday, February 19, 2016 8:38 AM
    Friday, February 19, 2016 8:38 AM
  • All,

    After thorough investigation and vigorous internal conversation, Durandal Inc. has decided not to build a RAD tool *at this time*. This doesn't mean we won't build something like this in the future. However, we felt that for now, we need to stay the course with our existing business plans. As it turns out, the success of our plans would lay a solid foundation for building a RAD tool in the future. A couple of things we are working on:

    1. Preparing to Beta 2 and then Release Aurelia 1.0

    2. Preparing to Release Aurelia Interface - A cross-platform, high performance, core component library for Hybrid Mobile apps and Material Design browser/desktop apps (fully themable/skinnable and with other UI platforms planned for the future).

    3. Developing Aurelia Business Components - A series of high-level data and LOB-oriented components (such as a data grid).

    4. New tooling that makes it easier for developers to get started with, build and deploy Aurelia applications (especially for Windows developers).

    We feel strongly that we need to at least see these four things through before developing a RAD tool. Hopefully it is obvious how these would be a necessary foundation for a next generation LighSwitch-type project. We've got some other things we're working on that we haven't announced yet as well, which will also help a lot in building the ecosystem around a potential RAD tool.

    Thank you for all the discussion and feedback here. Thank you for understanding our present course of action. I look forward to discussing these ideas again in the future after we've put down a bit more foundation.


    Rob Eisenberg

    Saturday, February 20, 2016 2:14 AM
  • Rob,

    I appreciate your thoughtful response.  What you say makes sense and I'm very excited to see Aurelia progress.

    I do hope, that in the future, Aurelia will have a complete end-to-end LOB RAD tool.  Either developed by your company or somebody else.

    Take care,

    Dave 

    Tuesday, February 23, 2016 7:07 PM
  • Thanks Rob...til then!  Cheers.
    Tuesday, February 23, 2016 10:01 PM