paper on OO performance with gfortran

[Arjen Markus]

Hi Damian,

interesting article. I noticed a few things though:
- He does not seem to be aware that Fortran does do a form of garbage
collection (allocatables being cleaned when they get out of scope)
- He does not mention how much more time they spend on writing the
Fortran code. Such time would highly depend on the programmer's
experience of course and it would be hard to generalise. Still, an
indication might have been worthwhile.
- The source code seems no longer available.

Regards,

Arjen

2015-08-12 6:54 GMT+02:00 Damian Rouson <damian@sourceryinstitute.org>:
>
>
>> On Aug 11, 2015, at 4:10 PM, Jerry DeLisle <jvdelisle@charter.net> wrote:
>>
>> Many moons ago I started digging into DTIO for gfortran and at the time I
>> thought it would not be that difficult to do, but tedious.  Its been so long
>> since I looked at it I will have to start over from scratch and try to
>> understand it again.  I want to restart soon, now that I am home and no longer
>> commuting 1200 miles to work.  PaulT, I may be contacting you to bounce some ideas.
>>
>
> Wow.  This is a very encouraging development.  If you work on DTIO and Paul worlds on PDTs,
> gfortran will be headed toward full Fortran 2008 compliance with a reasonable fraction of Fortran 2015
> already implemented too.
>
> Damian
>