Re: [Plplot-devel] Should we uniformly use the const attribute for our input arrays?

2011-03-15 Thread Andrew Ross
Alan, For the C++ case I will make the required changes so the C++ wrapper has const PLFLT * const * in . This way explicit casting is not required. BTW a quick test case will show that this is fine if you compile with g++, but generates warnings with gcc - as expected from the relevant standar

Re: [Plplot-devel] Should we uniformly use the const attribute for our input arrays?

2011-03-15 Thread Alan W. Irwin
On 2011-03-15 13:25- Andrew Ross wrote: > > Alan, > > For the C++ case I will make the required changes so the C++ wrapper > has const PLFLT * const * in . This way explicit casting is not > required. Hi Andrew: Now that I have cooled down from my anger and dismay at what the C standard requ

Re: [Plplot-devel] Should we uniformly use the const attribute for our input arrays?

2011-03-15 Thread Alan W. Irwin
On 2011-03-15 10:32-0700 Alan W. Irwin wrote: > Of course, > we should give advice in the release notes on how to avoid those C > warnings (for the recent const modifier changes) with explicit casts while > referring to > http://c-faq.com/ansi/constmismatch.html for the justification. To Hazen a