Rabbi David Botton <[email protected]> writes: > The next step for Gnoga is a set of controls/widgets that will make > layouts more natural for applications, but that aspect is only for us > nerds.
Let’s see ’em. ;) > For pro apps with designers available HTML 5 works out _better_ for > desktop apps and many new apps are using HTML 5 since the layouts will be > usable on other platforms such as mobile. I must admit I still did not find such widget set to be good-enough for the desktop except for very simple apps which are actually more suited for tablets/smrtphones etc. > What is the nature of the app you are developing. For real time high > frequency monitoring on a UI even I would agree :) Something like this one: http://saravali.de/screenshots.html but it should have more extensive feature set, be friendly for research, iow. capable of analyzing big database of chart data, performing complex queries on it and, most imoportantly, to have capability to quickly perform not-real-time simulation of planetary transits as well as calculations which goes along. There are lot of computations involved and base library to do it is 3rd party Swiss Ephemeris library (http://www.astro.com/swisseph/swephinfo_e.htm) which we need to provide bindings in desired language. > And perhaps it will take years, but better than no one trying :) I really appreciate your pioneering work in this field. > However the effort is much less when you take out the need to do > optimizations and code generation which is handled by LLVM or other > backends. Well, GNAT is quite solid and performant compiler producing quite good code, afaict. If there would be no performance issues in this project, I’d simply write everything in Python with wx/Qt bindings, but Python is not quick enough and its dynamic nature is probably not a good fit for such a project. So, I’d prefer having native code and/or C/LLVM back-endy only. :-) > Sadly there was once bigger efforts but post gnat 3.14p the licensing > issues put a big damper on community efforts. Hopefully that trend will > reverse. I must admit I am totally ignorant about that part of Ada’s history, but based on several rankings tables I can see Ada used to be much popular than today. > Having spent a considerable amount of time (more than I hoped) exploring > options, none of the "up and coming" are as well designed as Ada. I also spent (too) much time exploring the options not wanting to prematurely jumped onto some option and than being forced to rewrite later…I also did explore FP-options like Haskell/OCaml, but none were satisfactorily-enouugh, mostly due to lack of stable GUI-part. > Rust has some promise but not general purpose enough, Rust has many interesting options (like FP stuff, safety) besides active community, open-source compiler etc. but its syntax is simply something I (still) cannot adjust to…Who knows, maybe I’ll be forced to if none of the alternatives will pass 'final test’. :-) > and nim is buried in trying to provide too much syntactic sugar to every > problem. I believe one is not forced to use everything provided by the language, it’s actively developed with growing community, but, similarly, there is no (yet) good-enough GUI solution (besides GTK) - there are some attempts to provide Qt (QML) and wx bindings. It’s interesting that if HTML5-based GUI could fit the purpose of the project and/or, at least, my perception, then we’d be in much better shape with more viable options to choose from. :-) If you have some good examples of ’Pro’ desktop apps done with HTMl5, I’m more than enthusiastic to explore that option more seriously since it might be that I simply have (wrong) prejudices in regard. Sincerely, Gour -- Whenever and wherever there is a decline in religious practice, O descendant of Bharata, and a predominant rise of irreligion — at that time I descend Myself. ------------------------------------------------------------------------------ _______________________________________________ Gnoga-list mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/gnoga-list
