[Clam-devel] [PATCH] segmentation in chord extractor as a seperate class

[Roman Goj]

hmm, I got your message a minute before I sent my last patch, didn't
notice it, sorry ;)

David García Garzón wrote:
> Doxygenating back
> A attribute1; ///< the documentation for attribute
> Doxygenating front
> /// the documentation for attribute
> A attribute1; 

Thanks! Nice to know, should've noted that when you explained that to
bennett I think some time ago...

I actually removed the doxygen comments, because I realised they were
for parameters and I though maybe that's not such a good idea, but I'll
bring them back then :)

> std::endl implies std::flush

Oh, very nice to know too, thanks! :)

> And yes, commented out code is a hell. It is better to use svn and patches as 
> backups. Ill commit it so you can remove it on the next patches.
> 
> Your next step, doing the pool filling in a loop after having filled it all 
> instead filling it incrementally taking the content of those new vectors. 
> Which stone is safer for you? Implementing that on the destructor (you should 
> make available the pool and the extractor to the destructor) or implementing 
> it on the dumper (you should provide const accessors to the new vectors).
> 
> As you removed pool and extractor members, the first option may seem a back 
> step, it is a safer longer way. Depending on how you perform the second 
> option, it may be a dangerous jump but it is more close to the end and you 
> can do it carefully. You are the one on the rope.

The last patch takes the second route... (though I tried a third one on
my way up... and it was definitely the wrong one, had to back up and
start again omitting the faulty footholds ;) )

Of course accessors are coming soon.

romcio