Re: C++ debugger mailing list

[Andrew Cagney <cagney@redhat.com>]

Elena Zannoni wrote:
> Andrew Cagney wrote:
>> I'd like to propose that discussions and planning for this new 
>> project (or rebooted frysk as some have called it) be on a dedicated 
>> mailing list (it can be bootstrapped using frysk's membership), and 
>> then we extend a hand to gdb@ developers inviting them to participate.
>>
>> I think an invitation to interested gdb developers to something 
>> clearly not frysk and not gdb, would be a strong symbolic gesture to 
>> all concerned that the intent here is inclusion.
>>
>> Andrew
>>
>>
>
> Can somebody elaborate?
> We'd like to participate in these ongoing discussions for the new 
> project, whatever that is.

The goal of the frysk project is to implement an allways on monitoring 
and debugging tool, with key focus points being visual or graphical 
interface, non-stop and saleability.  And in particular, to not just 
re-implement GDB.  This drove decisions such as language choice and 
architecture, directions and goals, and also the decision to start with 
a relatively clean slate.

Red Hat have decided to withdraw their support for the frysk project and 
its goals, and are now instead focusing their efforts on a more clearly 
and narrowly defined goal of implement the Best C++ debugger, and in 
C++.  Tom is leading this new effort.

This of course leads us to a number of questions that I'm sure you're 
too familiar with (having struggled with them at the start of the frysk 
project):
-> do you start a new community around a new code base
-> do you fork an existing community and code base; GDB or frysk
-> do you work with an existing community, in particular GDB

A lot has changed over the last few years; for instance GOLD; GCC moving 
to C++; and GDB's increased level of activity and functionality.   An 
evaluation of the above in light of this new environment, is important.

Finally, a mistake I must recognize with frysk was not trying to engage 
the GDB community, and I look forward to seeing that mistake redressed 
with this new effort; and that I guess is my motivation for proposing a 
neutral discussion ground.   Shall we set something up, gdb++ frysk++ ...?