Giuseppe Verde wrote: > Phillip M. Jones, C.E.T. wrote: > >> Jonas J�rgensen wrote <<snip>>
>> complaints. (Just wondering.) > > > Erm, if it's an important enough feature, why don't *you* do the work of > coding it? I am going to open a can of flaming worms here, *but* that is easier said than done. I selected a file at random for do a code review. the file was: lxr.mozilla.org/seamonkey/source/caps/src/nsAggregatePrincipal.cpp My background is 25 years of development on distributed real-time systems. The last being a C++ distributed object database running on 7 different operating systems. -- Big License block - useless for maint. -- No file description block - What does this object do? Where is the reference to the Object Behavior Description? -- No Change log (CVS or RSC or whatever, the change log should be in the source as well) -- No descriptive comments for the method implementations (scanning first 9 functions produced none!) -- 7 descriptive comments in 492 lines. (maybe 400 lines of code) -- Not nearly ready for commit to tier 1 community libraries! Now, there are probably processes in place that are understood by the full-time developers; tools; perl scripts; some sort of design docs squirled away. But these are no help to someone you feel could be constructive working on a piece of code. A person can't just pick a file and start coding. Where are the Object Behavior Descriptions? Event diagrams? Which Callback methodology is be used? What are the naming conventions? How are the error catchers handling destructors in the multi-threaded environment? The list is endless. The architecture overwhelms the code. It may be open source, but it is also effectively obscured by the interdependance and lack of publicly structured documentation..... My 5 cents.......... > This isn't Proprietary Software, it's Software Libre. You're more than > welcome to fix or enhance someting and send a patch to the developers > or maintain the patch yourself. >
