emd error

[Hans Fugal]

I've been getting this today:

immsd: version 3.1.0-rc5 ready...
emd: Unexpected error in findBasicVariables!
This typically happens when the EPSILON defined in
emd.h is not right for the scale of the problem.

I don't know if it's new to 3.1.0-rc5 vs rc4, but I haven't seen it
before upgrading.

--
Hans Fugal ; http://hans.fugal.net

There's nothing remarkable about it. All one has to do is hit the
right keys at the right time and the instrument plays itself.
-- Johann Sebastian Bach

[Michael Grigoriev]

Hm, that's not supposed to happen. Did the schema upgrade work?
The version is supposed to be 14 (sqlite3 imms2.db 'select * from
Schema;') and the Acoustic db was supposed to have been cleared
('sqlite3 imms.acoustic.db 'select count(*) from Acoustic;').

Have fun,
Michael.

[Hans Fugal]

On Oct 5, 6:30 pm, "Michael Grigoriev" <m...@luminal.org> wrote:
> Hm, that's not supposed to happen. Did the schema upgrade work?
> The version is supposed to be 14 (sqlite3 imms2.db 'select * from
> Schema;') and the Acoustic db was supposed to have been cleared
> ('sqlite3 imms.acoustic.db 'select count(*) from Acoustic;').

Schema is empty, and Acoustic has about 1600 entries. So I moved
~/.imms and started rc5 fresh. Schema is still empty, but it doesn't
crash. I compared the schemas (sqlite3 foo.db .schema) of imms2.db and
imms.acoustic.db from my previous ~/.imms and the fresh one, and they
are exactly the same.

[Michael Grigoriev]

Upon closer examination, it appears that the logic for updating the
database would sometimes neglect to write out the new schema version,
causing further upgrades to not happen.
I think I fixed it rc6: http://www.luminal.org/files/imms/imms-3.1.0-rc6.tar.bz2

Thanks for finding this important bug.

Have fun,
Michael.

[Hans Fugal]

Michael Grigoriev wrote:
> Upon closer examination, it appears that the logic for updating the
> database would sometimes neglect to write out the new schema version,
> causing further upgrades to not happen.
> I think I fixed it rc6: http://www.luminal.org/files/imms/imms-3.1.0-rc6.tar.bz2

Looks like it worked. I see a message about updating the schema and no
crashes. Acoustic is cleared, which I suppose means it would be
profitable to run my find and analyze script again.

Speaking of which, any hints on working around this objcopy problem? Is
there maybe another way of getting the binary data (storing it in
~/.imms and loading it at runtime for example)?

Because of my environment I missed an important part of one of the
earlier patches. You can apply this to rc6:

---
configure.ac | 1 +
1 file changed, 1 insertion(+)

Index: imms-3.1.0-rc6/configure.ac
===================================================================
--- imms-3.1.0-rc6.orig/configure.ac 2008-10-05 22:55:06.000000000 -0600
+++ imms-3.1.0-rc6/configure.ac 2008-10-06 06:59:29.000000000 -0600
@@ -101,6 +101,7 @@ if test "$with_pcre" = "no"; then
AC_MSG_ERROR([PCRE required and missing.])
else
LIBS="$LIBS $pcre_LIBS"
+ CPPFLAGS="$CPPFLAGS $pcre_CFLAGS"
fi

if test "$enable_analyzer" != "no"; then

Thanks

[Michael Grigoriev]

On Mon, Oct 6, 2008 at 09:20, Hans Fugal <fug...@gmail.com> wrote:
> Because of my environment I missed an important part of one of the
> earlier patches. You can apply this to rc6:
>
> ---
> configure.ac | 1 +
> 1 file changed, 1 insertion(+)
>
> Index: imms-3.1.0-rc6/configure.ac
> ===================================================================
> --- imms-3.1.0-rc6.orig/configure.ac 2008-10-05 22:55:06.000000000 -0600
> +++ imms-3.1.0-rc6/configure.ac 2008-10-06 06:59:29.000000000 -0600
> @@ -101,6 +101,7 @@ if test "$with_pcre" = "no"; then
> AC_MSG_ERROR([PCRE required and missing.])
> else
> LIBS="$LIBS $pcre_LIBS"
> + CPPFLAGS="$CPPFLAGS $pcre_CFLAGS"
> fi
>
> if test "$enable_analyzer" != "no"; then

Thanks, applied.

Have fun,
Michael.