[Bf-committers] msvc6 build broken due to fluid sim

[bjornmose]

Nils Thuerey wrote:
> Hello,
> 
> bjornmose wrote:
> 
>>1. STL <vector> and <map> variables should have namespace .. e.g:
>>std::vector .. (yadada) since in C++ allows to redifine 'em in any
>>namespace. So a R3::vector.. might be a class repesenting something we
>>know from linear algebra ..
> 
> The STL vector and map are only used within the elbeem code, and there I
> took care that its not confused with the algebra vectors - those are
> called _Vec e.g. ParamVec.
> 
well, as i said msvc6 refuses to compile ( quits with error ) unless STL 
vars are prefixed std::... --sigh--

> 
>>2. MSVC6 does not allow -->
>>{
>>for (int i .. )
>>{}
>>for (int i .. )
>>{}
>>}
> 
> That's inconvenient - I think it's valid c++ code, and used extensively
> in the solver.
> 
> 
>>3. there are macros for shortcutting 'for' statements
>>like *FORDF1* ( finding this i gave up compiling with msvc6 )
>>which finnaly make repairing msvc6 builds impossible (see 2)
> 
> That's a common operation for an LBM solver (do {} for all distribution
> functions), so it increases readability of the source code quite a bit.
> 
> Btw. - is the MSVC6 compiler still used a lot? I tried the MSVC++
> Toolkit2003 (thus MSVC7  compiler), and that works :) Maybe the solver
> code should be disabled for MSVC6 for now?

Yes and no, i have a legal copy of msvc6 and i love the debugging 
options i have. Plus i know quite some -silent- users / 'build watchers' 
  who would be very dissapointed if we'd drop 'em.

OTH i could live with a 'external ready compiled lib' blender links to.
But in this case i'll refuse any coding&scientific support.

Well i didn't do in the past ( though hos poked me to mentor your 
project .. but i knew i wouldn't have the time to do so )
Beeing a pro on liquid statics  ( interfacial dynamics is what i'm 
working on right now ) i'm digging my way into navier-stookes with semi 
open boundary conditions to estimate the effect of inner dynamics 
compared to the interfacial forces. It looks like the fluid dynamics are 
getting in the time scale we'd like to observe surface energy .. humm .. 
i'd better shut up now.

Cutting it short: i know quite a bit about ODE/PDE solvers 'n stuff.
So it's on you having me 'in' or 'out'.

BM

> 
> Regards,
> -> Nils
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at projects.blender.org
> http://projects.blender.org/mailman/listinfo/bf-committers