Thumbs up for the first part of your message Marian. I'm more interested in a webui but that is not because of the "beauty" of ccs3/html5/ecma6. No further comments on the backend, I've shared my thoughts already. Neither .net nor abl would be my choice. But it's very very difficult to choose, lots of considerations and how to value everything...
"...only disadvantages to using a fat client these days..."
In regards to building front-ends for our applications.... It is interesting to me that software developers spend all day long using rich/full client applications like Visual Studio or Eclipse (Excel, Outlook, etc)
As long as I've been a software developer, I've heard that web-based front-ends are going to take over the world. Its not happening. In the past five years we've seen a lot of end-user mobile devices and, guess what, they normally work a lot better running native apps too!
If you are convinced that native apps are going away, then try googling Facebook's experience with HTML5. Long story short - native apps are *not* going away and web apps are *not* taking over the world. I think a developer should be open-minded about both.
dbeavon, Would you agree that to argue either extreme is just as limiting? While your last statement suggests 'open-minded about both', I infer from the content prior to this statement that you are quite opposed to client-side web apps ('How does that kind make any logical sense' & 'Its not happening' and finally 'Long story short - ...'). With proper architecture and experience, I feel that a team can have success with developing a feature-rich client web app (even a fully-realized ERP client) built on HTML5/CSS3/JS.
Yes, I think all types of apps have their place. The lines are especially blurry now that web technologies are being used to build full client apps.
Here are some relatively new web technologies that I'm excited about : (https://angularjs.org and https://cordova.apache.org and http://www.typescriptlang.org and http://electron.atom.io )
Note that I was responding to the unthinkable idea that full client apps would have "only disadvantages" over web-based apps. You are right to infer that, as a user, I would choose a classic "full client app" over a classic "web app" every time. Of course if web technologies are used to built something that looks and acts like a full client app, then I won't complain a single bit.
(I get confused when I hear a web developer summarily dismissing the types of apps that they use themselves - every single day. Instead of saying these things, go eat your own dogfood!)
Concerning the web frontend I think small libraries are the future. No angular (maybe 2.0 will be more interesting) or servoy f.e. Even 'giants' (like angular) are on the road to a more modular approach:
The telerik webui components are beauties as far as I have seen.
@dbeavon: I have exactly the same thoughts about native clients and web clients. The world is moving towards web-based frontends, but the native client is also needed and can not be forgotten. Maybe someday web-based UI is close enough to native UI in terms of usability, performance etc. but that day is not here yet. I had high hopes for Silverlight when it came out, but then Microsoft did the usual thing and changed their strategy completely.
agent_008_nlI'm more interested in a webui but that is not because of the "beauty" of ccs3/html5/ecma6.
Stefan, was avoiding hijacking the thread here and only commented on the C# part :)
I do think web/mobile can't/shouldn't be ignored, also think native are here to stay (mobile or desktop) so beside mobile/web we should probably start looking at a Qt (http://qt.io) integration for our node.js appsrv (http://akera.io)
Ten years ago or so, I played with QT and GTK in a personal fun project to have a GUI for Linux and QTopia on the Zaurus 5500 that would connect to an OpenEdge backend.
It would be fun to see what new version of QT can do. (I wonder if they would add bindings for Swift.)
> Stefan, was avoiding hijacking the thread here and only commented on the C# part :)
> we should probably start looking at a Qt (http://qt.io) integration for our node.js appsrv (http://akera.io)
Funny Marian, so you found an excuse in my reaction to hijack yourself and promote your own business. No offense, just LOL. :-)
qt.io really looks impressive and they are addressing mobile as well it seems so worth spending some time in doing a data provider for it
sure thing, knowing you this thread will go on for quite some time so thought to throw in something about what we do... kind of the cheapest marketing option ;)
how's your 'micro' framework going, ready for the future already?
My web framework you mean (thumbs up for myself :-)? That's for learning purposes in the first place. No commercial plans, but who knows what it could bring. Got to make a living, less and less oe contractor jobs in Holland.