[armedbear-devel] Spurious errors about redefinitions
Since Erik's autoload patches of about a week ago hit the tree that started to optimize the FASL loading, I see a lot of spurious errors about redefining symbols like the following when loading SLIME: STYLE-WARNING: redefining SWANK::%MAKE-FUZZY-MATCHING in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING-P in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING.SYMBOL in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING.PACKAGE-NAME in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) I thought this was taken care of with [svn 12324][1], but this doesn't seem to be the case. I haven't yet been able to isolate the behavior yet, but if someone knows a quick fix, that would be great. [1]: http://code.google.com/p/abcl-dynamic-install/source/detail?r=ad17b2b2733ee9... -- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
On 1/5/10 10:53 PM, Mark Evenson wrote:
Since Erik's autoload patches of about a week ago hit the tree that started to optimize the FASL loading, I see a lot of spurious errors about redefining symbols like the following when loading SLIME:
Actually these are "spurious warnings", and non-fatal. Just verbiage that is confusing so shouldn't be generated. -- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
Hi Mark, On Tue, Jan 5, 2010 at 10:53 PM, Mark Evenson <evenson@panix.com> wrote:
Since Erik's autoload patches of about a week ago hit the tree that started to optimize the FASL loading, I see a lot of spurious errors about redefining symbols like the following when loading SLIME:
STYLE-WARNING: redefining SWANK::%MAKE-FUZZY-MATCHING in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING-P in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING.SYMBOL in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING.PACKAGE-NAME in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL)
I thought this was taken care of with [svn 12324][1], but this doesn't seem to be the case. I haven't yet been able to isolate the behavior yet, but if someone knows a quick fix, that would be great.
I've run tests on Slime with a version just before my changes. I *think* I saw the same slew of warnings. Can you tell me which abcl version is known not to generate them (using the current SLIME head version)? I'll have a look at them then. Bye, Erik.
I'm seeing them without slime. -Alan On Tue, Jan 5, 2010 at 4:57 PM, Erik Huelsmann <ehuels@gmail.com> wrote:
Hi Mark,
On Tue, Jan 5, 2010 at 10:53 PM, Mark Evenson <evenson@panix.com> wrote:
Since Erik's autoload patches of about a week ago hit the tree that started to optimize the FASL loading, I see a lot of spurious errors about redefining symbols like the following when loading SLIME:
STYLE-WARNING: redefining SWANK::%MAKE-FUZZY-MATCHING in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING-P in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING.SYMBOL in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL) STYLE-WARNING: redefining SWANK::FUZZY-MATCHING.PACKAGE-NAME in #P"/Users/evenson/work/slime/contrib/swank-fuzzy.lisp" (previously defined in NIL)
I thought this was taken care of with [svn 12324][1], but this doesn't seem to be the case. I haven't yet been able to isolate the behavior yet, but if someone knows a quick fix, that would be great.
I've run tests on Slime with a version just before my changes. I *think* I saw the same slew of warnings. Can you tell me which abcl version is known not to generate them (using the current SLIME head version)?
I'll have a look at them then.
Bye,
Erik.
_______________________________________________ armedbear-devel mailing list armedbear-devel@common-lisp.net http://common-lisp.net/cgi-bin/mailman/listinfo/armedbear-devel
On 1/5/10 10:57 PM, Erik Huelsmann wrote: […]
I've run tests on Slime with a version just before my changes. I *think* I saw the same slew of warnings. Can you tell me which abcl version is known not to generate them (using the current SLIME head version)?
I'll have a look at them then.
A reasonable request: it'll take a bit of time to run the regression tests (and I'm probably done for the night). -- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
On 1/5/10 10:57 PM, Erik Huelsmann wrote: […]
I've run tests on Slime with a version just before my changes. I *think* I saw the same slew of warnings. Can you tell me which abcl version is known not to generate them (using the current SLIME head version)?
It turns out to start with [svn r12306], and it never went away. I thought that you had corrected this with [svn r12324] based on your comment, but had not actually verified it in my installation. Reproducing this problem is simple. 1. Create a file to compile named "foo.lisp" containing the form (defun foo () 42) The contents of the file don't seem to matter. 2. Start ABCL, compile the file CL-USER> (compile-file "foo") 3. Quit ABCL. 4. Restart ABCL, issuing CL-USER> (load "foo") You'll see an warning along the lines of STYLE-WARNING: redefining COMMON-LISP-USER::FOO in #P"/Users/evenson/work/abcl/foo.lisp" (previously defined in NIL) emitted. -- "A screaming comes across the sky. It has happened before, but there is nothing to compare to it now."
On Wed, Jan 6, 2010 at 1:44 PM, Mark Evenson <evenson@panix.com> wrote:
On 1/5/10 10:57 PM, Erik Huelsmann wrote: […]
I've run tests on Slime with a version just before my changes. I *think* I saw the same slew of warnings. Can you tell me which abcl version is known not to generate them (using the current SLIME head version)?
It turns out to start with [svn r12306], and it never went away. I thought that you had corrected this with [svn r12324] based on your comment, but had not actually verified it in my installation.
Reproducing this problem is simple.
1. Create a file to compile named "foo.lisp" containing the form
(defun foo () 42)
The contents of the file don't seem to matter.
2. Start ABCL, compile the file
CL-USER> (compile-file "foo")
3. Quit ABCL.
4. Restart ABCL, issuing
CL-USER> (load "foo")
You'll see an warning along the lines of
STYLE-WARNING: redefining COMMON-LISP-USER::FOO in #P"/Users/evenson/work/abcl/foo.lisp" (previously defined in NIL)
emitted.
Erik and I fixed the issue; the fix is now on trunk. However, the problem appeared to only affect macros, not functions like in you example. Can you confirm? Bye, Alessio
On Wed, Jan 6, 2010 at 4:59 PM, Alessio Stalla <alessiostalla@gmail.com> wrote:
Erik and I fixed the issue; the fix is now on trunk. However, the problem appeared to only affect macros, not functions like in you example. Can you confirm?
Bye, Alessio
It was for functions too. Will try with new trunk. -Alan
On Wed, Jan 6, 2010 at 6:06 PM, Alan Ruttenberg <alanruttenberg@gmail.com> wrote:
On Wed, Jan 6, 2010 at 4:59 PM, Alessio Stalla <alessiostalla@gmail.com> wrote:
Erik and I fixed the issue; the fix is now on trunk. However, the problem appeared to only affect macros, not functions like in you example. Can you confirm?
Bye, Alessio
It was for functions too. Will try with new trunk.
Gone. Guess it fixes functions too. -Alan
participants (4)
-
Alan Ruttenberg -
Alessio Stalla -
Erik Huelsmann -
Mark Evenson