On Monday, March 14, 2016 at 8:01:01 PM UTC+1, Uwe Fechner wrote:
>
> Hello,
>
> perhaps I could become mentor of such a project?
> At least I know a little bit of Julia and a little bit of QML.
>
> I am currently implementing optimizations and simulations
> in Julia, and a QML GUI would be nice.
>
> What would be needed to offer such a project?
> How much workload would be expected for a mentor?
>
> Regards:
>
> Uwe Fechner, TU Delft
>
> On Monday, March 14, 2016 at 2:23:32 PM UTC+1, Maurice Diamantini wrote:
>>
>>
>>
>> Le lundi 14 mars 2016 13:24:13 UTC+1, Uwe Fechner a écrit :
>>
>>
>>> What Julia is missing is mainly an easy to use GUI toolkit, and QML
>>> could play that role.
>>>
>>
>> That's exactly why I think a QML binding would be easier to do than a Qt5
>> one, and sufficient for most use case from Julia : no need to be able to
>> use yet another string type et socket or http or..., just application GUI
>> with callback to Julia methods.
>>
>>
>>> Implementing a Julia - QML binding sound not so difficult as there is
>>> already a C binding.
>>>
>>> I do not understand the following sentence:
>>> "Being able to build Qt interface directly from Julia without the need
>>> of the anaconda
>>> would be a great advantage for Julia users."
>>>
>>
>> Because I know there already exists a Python binding Qt4 (or there will
>> be some day for Qt5) but it require about 1.2 G bytes in the
>> `pkgs/v0.4/Conda`
>> of the Julia directory.
>>
>>
>> If you would write a QML wrapper for the C binding, the C/ C++ code from
>>> the C wrapper
>>> would still need to be compiled included in the package in some way
>>> (binutils). So there
>>> would still be a dependency on C/C++ code (pre-compiled or not), that is
>>> always a little
>>> bit difficult to install and to maintain.
>>>
>>
>> Yes, but this extension should be more robust because less sensible to
>> the Qt core changes...
>>
>> -- Maurice
>>
>>
