[Proj] Welcome to the "Proj" mailing list

jamesmikedupont at googlemail.com jamesmikedupont at googlemail.com
Mon Jan 4 12:51:24 PST 2010


my plan is to integrate the code via some templates.
I can imagine that inline templates can generate much more efficient code.
we would select the exact operations on the native data. I am going to
look into this more.

mike

On Mon, Jan 4, 2010 at 9:48 PM, Jan Hartmann <j.l.h.hartmann at uva.nl> wrote:
> Why don't you write a small stub C++ library on top of C-PROJ, just to
> enable this sort of calls?
>
> Jan
>
> On 4-1-2010 12:44, jamesmikedupont at googlemail.com wrote:
>>
>> here is my patch so far.
>>
>> I can understand the desire to keep things simple. I would have never
>> even bothered to compile this at all if I did not have problems. My
>> problem was the usage of radians that is dependant on the combination
>> of projections. This is something that could be hidden by the
>> templates and done automatically. It is done in cs2cs but not in the
>> api.
>>
>> My idea about the templates is very simple, to be able to create a
>> single function
>>
>> newcood= f<fromproj,toproj>(coord)
>>
>> it would not need any initialization of the projections, but the
>> projections would be selected via templates. That would allow for a
>> very fast usage.
>>
>> mike
>>
>>
>> On Mon, Jan 4, 2010 at 12:38 PM, Jan Hartmann<j.l.h.hartmann at uva.nl>
>>  wrote:
>>
>>>
>>> I would prefer to keep the code of the library itself as standard-C like
>>> as
>>> possible. It is scientific programming, could be done just as well in
>>> Fortran (preferably not, of course :-)). On the other hand, what someone
>>> does on top of this library (e.g. building user interfaces in Fortran),
>>> should be doable in any other language.
>>>
>>> Jan
>>>
>>> On 4-1-2010 12:27, jamesmikedupont at googlemail.com wrote:
>>>
>>>>
>>>> Yes,
>>>> Of course. My plan was to test this and submit it to the main branch.
>>>> In fact, I would not have published this at all, but was having
>>>> problems and wanted help. Now I have resolved the issues and this
>>>> branch is not needed any more.
>>>>
>>>> This is just a working snapshot.
>>>> Ideally there more work to do with c++ and proj.
>>>> Like to move the configuration of the projections to templates..
>>>>
>>>>
>>>> mike
>>>>
>>>> On Mon, Jan 4, 2010 at 12:20 PM, Jan Hartmann<j.l.h.hartmann at uva.nl>
>>>>  wrote:
>>>>
>>>>
>>>>>
>>>>> If it is just a patch to make the source code c++ compliant, IMHO it
>>>>> belongs
>>>>> in the main branch. Useful for indepentdent C++ projects that would
>>>>> like
>>>>> to
>>>>> link in the PROJ4 library.
>>>>>
>>>>> Jan
>>>>>
>>>>> On 4-1-2010 12:10, jamesmikedupont at googlemail.com wrote:
>>>>>
>>>>>
>>>>>>
>>>>>> It does not do anything more at the moment.
>>>>>>
>>>>>> This is just compiling with c++, you can apply the patch if you like.
>>>>>>
>>>>>> mike
>>>>>>
>>>>>>
>>>>>> On Mon, Jan 4, 2010 at 12:09 PM, Jan Hartmann<j.l.h.hartmann at uva.nl>
>>>>>>  wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>> On 4-1-2010 8:47, jamesmikedupont at googlemail.com wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> In addition, I have started a c++ branch of proj4, and have resolved
>>>>>>>> some compile problems.
>>>>>>>> Feel free to check it out
>>>>>>>> http://github.com/h4ck3rm1k3/ProjC--/
>>>>>>>>
>>>>>>>> mike
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> Just curious, what does this branch do that the main tree doesn' t?
>>>>>>>
>>>>>>> Jan
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>
>>>>>
>>>
>>>
>



More information about the Proj mailing list