Steel Bank Common Lisp

Jonathon McKitrick <jmckitrick@...> writes:
> I tried following the steps listed in the manual for running sb-
> cover, but the end result is an empty report, as though I had not
> evaluated the DECLAIM form with the sb-cover option.

Where did you evaluate the DECLAIM form? On the REPL before doing the
compilation?

> 1.  I assume sb-cover works on Intel Macs, but the screenshot on
> Lemonodor could be a ppc machine.  Am I wrong?  I'm running sbcl
> 1.0.5.49.

It'll work on all architectures.

> 2.  I'd like to avoid recompiling all the fasl files that belong to
> sbcl, if possible, unless this is required by sb-cover.  Since I
> don't work in my root account, I have to manually ACCEPT all the read-
> only files that cannot be recompiled by the :FORCE T in the ASDF load
> command.  Is there a workaround?

You can remove your own fasl files, and then just do a normal
asdf:load-op without force.

The current behaviour of :FORCE T doesn't seem very useful to me, and
the comments in asdf.lisp actually imply that the intended behaviour
is different than the observed one. (:FORCE T -> force recompilation
of this system, :FORCE 'some-non-t-true-value) -> force recompilation
of dependencies).

> Could these un-recompiled files be
> why no report is generated?

Sounds unlikely.

-- 
Juho Snellman

On May 23, 2007, at 6:15 PM, Juho Snellman wrote:

> Jonathon McKitrick <jmckitrick@...> writes:
>> I tried following the steps listed in the manual for running sb-
>> cover, but the end result is an empty report, as though I had not
>> evaluated the DECLAIM form with the sb-cover option.
>
> Where did you evaluate the DECLAIM form? On the REPL before doing the
> compilation?

Yes.  Then I started including them in the first file loaded when I  
though that might have been the problem.

>
>> 1.  I assume sb-cover works on Intel Macs, but the screenshot on
>> Lemonodor could be a ppc machine.  Am I wrong?  I'm running sbcl
>> 1.0.5.49.
>
> It'll work on all architectures.

I was able to get it working after recompiling everything, uncluding  
the sbcl fasls.

>
>> 2.  I'd like to avoid recompiling all the fasl files that belong to
>> sbcl, if possible, unless this is required by sb-cover.  Since I
>> don't work in my root account, I have to manually ACCEPT all the  
>> read-
>> only files that cannot be recompiled by the :FORCE T in the ASDF load
>> command.  Is there a workaround?
>
> You can remove your own fasl files, and then just do a normal
> asdf:load-op without force.

I'll try that instead.


--
Jonathon McKitrick
jmckitrick@...