www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Ada Vs C (with some D mixed in)

reply bearophile <bearophileHUGS lycos.com> writes:
This is an oldish article (1995) that compares development costs of C and Ada:
http://www.adaic.org/whyada/ada-vs-c/cada_art.html

D is generally a safer language than C, and more productive too. On the other
hand Ada has some safeties that D lacks still (and I think D may enjoy having
some of them), like for example regarding safety of type conversion, integral
overflows, and other things (regarding performance, Ada is similar to
statically compiled languages). D requires less lines of code than C and Ada
(especially if you use higher order functors like std.algorithm of D2 or my
ones), so I think overall the development costs of D may be intermediate
between C and Ada ones :-)

Regarding the SafeD I want to say something: I like the idea of SafeD, but
safety comes from many things, and not just a single source. So I think SafeD
has to improve other things too. Restricting the "safe" of SafeD to just one
things (that is just eliminating pointers, I think), goes against the concept
of safety. If you want a safer D, then you have to look at doing other things
too in a safer way. Because the end purpose of a programmer that wants to use
SafeD may be to write safer code, and in real programs bugs doesn't come from a
single source! (that is, pointers).

Just for example, Safe D may enforce a safer indenting of code, to avoid the
"dangling else" bug (it seems I was quite right, and GCC designers have had the
same idea of mine, take a look at recently added warnings of GCC, -Wparentheses
and -Wsequence-point here,
http://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html  I think such things
have to be built-in in SafeD), take better care of integral overflows, etc.

Bye,
bearophile
Jan 19 2009
parent reply Walter Bright <newshound1 digitalmars.com> writes:
bearophile wrote:

 Just for example, Safe D may enforce a safer indenting of code, to
 avoid the "dangling else" bug (it seems I was quite right, and GCC
 designers have had the same idea of mine, take a look at recently
 added warnings of GCC, -Wparentheses and -Wsequence-point here,
 http://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html  I think such
 things have to be built-in in SafeD), take better care of integral
 overflows, etc.


SafeD is about guaranteeing memory safety, not other issues like integer 
overflows. Memory safety is a fairly specifically defined thing.
Jan 19 2009
parent reply bearophile writes:
Walter Bright:

 SafeD is about guaranteeing memory safety, not other issues like integer 
 overflows. Memory safety is a fairly specifically defined thing.


Then maybe the name of SafeD isn't too much good, because when I hear that name
I think about a safe(r) language, and not about memory safety only. So maybe an
alternative name can be invented... (but I have no good name to suggest you so
far).

Bye,
bearophile
Jan 19 2009
parent reply Lutger <lutger.blijdestijn gmail.com> writes:
bearophile wrote:


 Walter Bright:

 SafeD is about guaranteeing memory safety, not other issues like integer 
 overflows. Memory safety is a fairly specifically defined thing.


 
 Then maybe the name of SafeD isn't too much good, because when I hear that
name I think about a safe(r) language, and not about memory safety only. So
maybe an alternative name can be invented... (but I have no good name to
suggest you so far).
 
 Bye,
 bearophile


I've always understood the 'safety' here as safe from buffer overflow exloits
and such, instead of safety against bugs in general. (security).
Jan 19 2009
parent reply Walter Bright <newshound1 digitalmars.com> writes:
Lutger wrote:

 bearophile wrote:
 

 Walter Bright:

 SafeD is about guaranteeing memory safety, not other issues like
 integer overflows. Memory safety is a fairly specifically defined
 thing.


 Then maybe the name of SafeD isn't too much good, because when I
 hear that name I think about a safe(r) language, and not about
 memory safety only. So maybe an alternative name can be invented...
 (but I have no good name to suggest you so far).
 
 Bye, bearophile


 
 I've always understood the 'safety' here as safe from buffer overflow
 exloits and such, instead of safety against bugs in general.
 (security).


Exactly. http://en.wikipedia.org/wiki/Memory_safety
Jan 19 2009
parent bearophile <bearophileHUGS lycos.com> writes:
Walter Bright:

 Lutger wrote:

 I've always understood the 'safety' here as safe from buffer overflow
 exloits and such, instead of safety against bugs in general.
 (security).


 
 Exactly. http://en.wikipedia.org/wiki/Memory_safety


OK, I'll wait for the coming of SecureD then :-)

Bye,
bearophile
Jan 19 2009