On 04/25/2013 04:52 PM, Alexis Lopez Zubieta wrote:
> Hello:
>
> John, we do have a web page for our project but right now its down due to 
> some troubles with the hosting,
 > if you want to know more about our work take a look at 
http://es.wikipedia.org/wiki/Nova_%28linux%29.

can you also share the link to the currently offline page ? i will look 
if its reachable in a few days.

>
> About C++, I have to say that it support in a native form the object oriented 
> programing
 > what enable to have the data and the functions related to it together 
and also encourage the code reusage.

1) putting data and functions together is considered harmful
http://research.scee.net/files/presentations/gcapaustralia09/Pitfalls_of_Object_Oriented_Programming_GCAP_09.pdf

but it's perfectably doable in C so i don't see your point

2) that C++ encourages code reusage and C doesn't, is a propaganda lie.
the reality looks quite different, C++ adds a lot of hidden dependencies 
that make code reusage *harder*. think about inherited classes. in order 
to know how to use them, i have to know all about the involved classes. 
when some code i'm trying to debug passes around a base class object, i 
can not be sure which descendant of the class was originally 
instatiated. this makes debugging C++ code that was written by other 
people very hard.

> And over all it enables to put more of the business logic into the code, 
> (less comments are needed).

huh ? citation needed.

> Comparing C and C++ in runtime the second is just a bit (very small bit)
> more resource hunger wich is the cost of the metadata asociciated to C++ 
> objects.

there's much more than that, as soon as you use templates you get a ton 
of specialised functions (it is sufficient to use STL).

in the end, even a dynamically linked binary that uses some form of 
templates will end much bigger than it's C counterpart.
and bigger binary also means bigger RAM consumption since the binary has 
to be mapped into RAM.

> I have done a more detailed comparison but it is in spanish, if you are 
> interested I can translate it to english.

sure, would be interesting to read about that.

>
> In order to avoid that the next LXDE become a giant air ball I propose a 
> monolithic architecture extensible
 > through modules and use lightweight threading instead of FORK.

now you're talking about fork(), i was refering to forking lxde (the 
last release that works with gtk+2).

when you talk about modules, do you think about dlopen() ?

P.S. next time you come to IRC, stay a bit longer please :)
i am not always in front of the PC...



------------------------------------------------------------------------------
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
_______________________________________________
Lxde-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lxde-list

Reply via email to