This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Problem with shared libraries, exceptions, and global constructors


On Mon, Aug 18, 2003 at 10:43:19PM +0200, Gabriel Dos Reis wrote:
> | Anyway, what would you suggest?  If the dependencies are so 
> | complex you can't express them as a DAG, then your project
> | is Just Broken, IMNSHO.
> 
> From compiler implementors point of view, it might be fine to try to
> phrase things that way.  However, from application developers point of
> view, the solution you outilined earlier with its requirements are
> just plain too complex to be effective.  And that is not speculation.

This is nonsense.

I am not speaking from a "compiler implementors point of view".

If, from ANYONE's point of view, there is a cycle in the dependency
graph, then there exists NO fully correct initialization sequence.

Period.

And, if I may be so bold, from a software engineer's point of view,
if there exists such a cycle, then the project is badly designed.  It
is a sign of layering violations or poor problem decomposition.  If
the application works at all, it will be by accident.  

If a developer cannot be bothered to create the dependency graph at
all, merely by linking each library to the libraries it uses, it is
not a matter of complexity, it is a matter of laziness.

I have no sympathy.



r~


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]