• S <S s.com> Nov 29 2007
• "Craig Black" <cblack ara.com> Nov 29 2007
• BCS <ao pathlink.com> Nov 29 2007
• S <S s.com> Nov 29 2007
• Robert Fraser <fraserofthenight gmail.com> Nov 29 2007
• Dan <murpsoft hotmail.com> Nov 29 2007

S <S s.com> writes:

I'm curious what kind of results Coverity would produce from the D frontend? 
Anybody else curious?

"Craig Black" <cblack ara.com> writes:

"S" <S s.com> wrote in message news:fin463$1gsb$1 digitalmars.com...
 I'm curious what kind of results Coverity would produce from the D 
 frontend?  Anybody else curious?


What is Coverity? 

BCS <ao pathlink.com> writes:

Reply to Craig,

 "S" <S s.com> wrote in message news:fin463$1gsb$1 digitalmars.com...
 
 I'm curious what kind of results Coverity would produce from the D
 frontend?  Anybody else curious?
 


 


a c/c++ static analisis tool

http://www.coverity.com/

S <S s.com> writes:

Craig Black Wrote:

 "S" <S s.com> wrote in message news:fin463$1gsb$1 digitalmars.com...
 I'm curious what kind of results Coverity would produce from the D 
 frontend?  Anybody else curious?


 What is Coverity? 
 
 


http://scan.coverity.com/

It would be cool to submit the D frontend to this.  I believe it's "Open
Source" enough to get free analysis.

Robert Fraser <fraserofthenight gmail.com> writes:

S wrote:
 I'm curious what kind of results Coverity would produce from the D frontend? 
Anybody else curious?


Well, it doesn't use delete (almost at all), and the #1 problem in 
Descent's Java port of it is NPEs. So I'm guessing it'd fail pretty bad.

More importantly, would anyone be interested in a Coverity-style null 
checker (i.e. not just a lint) for D?

Dan <murpsoft hotmail.com> writes:

Robert Fraser Wrote:

 S wrote:
 I'm curious what kind of results Coverity would produce from the D frontend? 
Anybody else curious?


 Well, it doesn't use delete (almost at all), and the #1 problem in 
 Descent's Java port of it is NPEs. So I'm guessing it'd fail pretty bad.
 
 More importantly, would anyone be interested in a Coverity-style null 
 checker (i.e. not just a lint) for D?


At the moment, I do most of my programming in an extended IDA Pro 5.1.  I do
this because it allows me to do machine-level call graphing with collapsing
features and step-through debugging.  I haven't had to compile more than once
now since I started using it last year - no debugging, just analysis.

While it's still far from perfect, it's better than standard "text+compiler"
authoring a program.

Not to sound arrogant, but my interest in D is sort of legacy to the days when
I thought languages were significant.