[CinCV TNG] Cinelerra code analysis

Nicola Ferralis feranick at yahoo.com
Wed Dec 14 20:40:14 CET 2016


I second that. When time allowed me to in the past, I used coverity as a 
learning opportunity to deal with Cinelerra's code. While I was never 
shy in stating that I am no more than a learning student in coding, I 
submitted a few patches based on what came out from the coverity 
analysis. I always hoped (and got) the expert say of Einar and Hannes on 
the viability/need/relevance of such patches. All I am saying is that 
the peer-review system we have for (and it's important to say *only* 
for) CV allows people to experiment in their venturing with these 
analytic tools as well as learning the inner working with cinelerra. So 
if you feel like trying, I'd say go for it, as long as the expectation 
is not to revolutionize the code, but rather learn from it. If there is 
one thing I learned from dealing with coverity is the respect I have for 
the core community of cinelerra's developers (including their never 
ending patience).

Nicola

On 12/14/16 1:58 PM, Einar R√ľnkaru wrote:
>
>
> On 12/14/2016 08:22 PM, Johannes Sixt wrote:
>> We've had Cinelerra submitted to Coverity, IIRC, and fixed some bugs.
>>
>> However, as long as the code base does not use std library, smart
>> pointers, etc, i.e., modern C++, you are fighting and losing a battle.
>>
>
> Why are you think that 'modern' tools help? There are some companies 
> with enormous free resources - they can use any tool invented, but 
> they send regularily updates - fix bugs. Tools may help, but the do 
> not create miracles.
>
> Einar
> _______________________________________________
> Cinelerra mailing list
> Cinelerra at lists.cinelerra-cv.org
> http://lists.cinelerra-cv.org/cgi-bin/mailman/listinfo/cinelerra




More information about the Cinelerra mailing list