You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Erik Huelsmann <eh...@gmail.com> on 2006/10/08 20:53:10 UTC
[PATCH] Make configure report the components detected
Our current configure script warns at the end if BDB hasn't been
found. Given that it's now just one of the components that can be used
to build Subversion on (and that FSFS isn't so immature anymore), it
would be more logical to report which components *are* used. From
that, the user can decide if the list is correct or not.
Here's a patch to achieve that. Comments?
Log:
[[[
Make configure report the components detected.
* configure.in:
Write out a series of NOTICEs stating what's been detected, instead
of warning about not finding BDB.
]]]
Index: configure.in
===================================================================
--- configure.in (revision 21833)
+++ configure.in (working copy)
@@ -785,19 +785,50 @@
dnl Hence, print a warnings about what we did and didn't configure at the
dnl end, where people will actually see them.
+AC_MSG_NOTICE([Subversion has been configured using the following
components:])+AC_MSG_NOTICE([ APR $apr_version])
+AC_MSG_NOTICE([ APR-util $apu_version])
+case "$svn_lib_neon" in
+ "no" )
+ AC_MSG_NOTICE([ neon no])
+ ;;
+ "yes" )
+ AC_MSG_NOTICE([ neon $NEON_VERSION])
+ ;;
+esac
+AC_MSG_NOTICE([ serf $serf_found])
case "$svn_lib_berkeley_db" in
"no" )
- db_version="$SVN_FS_WANT_DB_MAJOR.$SVN_FS_WANT_DB_MINOR.$SVN_FS_WANT_DB_PATCH"
- AC_MSG_WARN([we have configured without BDB filesystem support
-
-
-You don't seem to have Berkeley DB version $db_version or newer
-installed and linked to APR-UTIL. We have created Makefiles which
-will build without the Berkeley DB back-end; your repositories will
-use FSFS as the default back-end. You can find the latest version of
-Berkeley DB here:
- http://www.sleepycat.com/download/index.shtml
-])
-;;
+ AC_MSG_NOTICE([ BerkeleyDB no])
+ ;;
+ "yes" )
+ AC_MSG_NOTICE([ BerkeleyDB $db_version])
+ ;;
esac
-
+case "!$APXS!" in
+ "!no!"|"!!" )
+ AC_MSG_NOTICE([ Apache (module) no])
+ ;;
+ * )
+ AC_MSG_NOTICE([ Apache (module) yes])
+ ;;
+esac
+dnl How to output 'Apache (static)'?
+case "$SWIG" in
+ "none" )
+ AC_MSG_NOTICE([ SWIG no])
+ ;;
+ * )
+ AC_MSG_NOTICE([ SWIG $SWIG_VERSION_RAW])
+ ;;
+esac
+AC_MSG_NOTICE([ zlib $zlib_found])
+AC_MSG_NOTICE([ sasl $svn_lib_sasl])
+case "$JAVA" in
+ "" )
+ AC_MSG_NOTICE([ java no])
+ ;;
+ * )
+ AC_MSG_NOTICE([ java $JAVA])
+ ;;
+esac
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Daniel Rall <dl...@collab.net>.
On Thu, 16 Nov 2006, Erik Huelsmann wrote:
> On 11/16/06, Daniel Rall <dl...@collab.net> wrote:
> >On Thu, 16 Nov 2006, David Glasser wrote:
> >
> >> On 11/16/06, Erik Huelsmann <eh...@gmail.com> wrote:
> >> >On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >> >> What happened to this patch?
> >> >
> >> >Nothing: it's still sitting in my working copy. But, I think I should
> >> >apply it, given Peter's reaction.
> >>
> >> Yes, please do -- I haven't tested the patch itself, but I keep
> >> wishing I had that data there (when testing releases, etc).
> >
> >I recall us all being in favor of the concept, but there were a few
> >comments about the implementation, no?
>
> There was only the question about using double quotes to check for
> empty strings. I think that was only out of interest, not out of
> concern.
Looking through my mail, I also mentioned:
"If we're going to report on Java (which I'm in favor of), we should
report on Python (via get-py-info.py?) and Ruby and Perl as well."
Re: [PATCH] Make configure report the components detected
Posted by Erik Huelsmann <eh...@gmail.com>.
On 11/16/06, Daniel Rall <dl...@collab.net> wrote:
> On Thu, 16 Nov 2006, David Glasser wrote:
>
> > On 11/16/06, Erik Huelsmann <eh...@gmail.com> wrote:
> > >On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> > >> What happened to this patch?
> > >
> > >Nothing: it's still sitting in my working copy. But, I think I should
> > >apply it, given Peter's reaction.
> >
> > Yes, please do -- I haven't tested the patch itself, but I keep
> > wishing I had that data there (when testing releases, etc).
>
> I recall us all being in favor of the concept, but there were a few
> comments about the implementation, no?
There was only the question about using double quotes to check for
empty strings. I think that was only out of interest, not out of
concern.
bye,
Erik.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Daniel Rall <dl...@collab.net>.
On Thu, 16 Nov 2006, David Glasser wrote:
> On 11/16/06, Erik Huelsmann <eh...@gmail.com> wrote:
> >On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >> What happened to this patch?
> >
> >Nothing: it's still sitting in my working copy. But, I think I should
> >apply it, given Peter's reaction.
>
> Yes, please do -- I haven't tested the patch itself, but I keep
> wishing I had that data there (when testing releases, etc).
I recall us all being in favor of the concept, but there were a few
comments about the implementation, no?
Re: [PATCH] Make configure report the components detected
Posted by David Glasser <gl...@mit.edu>.
On 11/16/06, Erik Huelsmann <eh...@gmail.com> wrote:
> On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> > What happened to this patch?
>
> Nothing: it's still sitting in my working copy. But, I think I should
> apply it, given Peter's reaction.
Yes, please do -- I haven't tested the patch itself, but I keep
wishing I had that data there (when testing releases, etc).
--dave
--
David Glasser | glasser@mit.edu | http://www.davidglasser.net/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Erik Huelsmann <eh...@gmail.com>.
Nope. I'm sorry, but I don't have the time to...
bye,
Erik.
On 2/28/07, Hyrum K. Wright <hy...@mail.utexas.edu> wrote:
> Did this every get committed?
>
> -Hyrum
>
> Erik Huelsmann wrote:
> > I can. I'll do that later today.
> >
> > On 1/18/07, Daniel Rall <dl...@collab.net> wrote:
> >> Hopeully Erik can just commit this patch, saving us the overhead of an
> >> issue.
> >>
> >> On Tue, 16 Jan 2007, Hyrum K. Wright wrote:
> >>
> >> > Erik Huelsmann wrote:
> >> > > On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >> > >> What happened to this patch?
> >> > >
> >> > > Nothing: it's still sitting in my working copy. But, I think I should
> >> > > apply it, given Peter's reaction.
> >> >
> >> > I don't think this ever got applied; shall I put it in the issue
> >> tracker?
> >> >
> >> > -Hyrum
> >> >
> >> > >> With regards
> >> > >> Kamesh Jayachandran
> >> > >> Kamesh Jayachandran wrote:
> >> > >> > Erik Huelsmann wrote:
> >> > >> >> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >> > >> >>> I like this change.
> >> > >> >>> One small suggestion, Can you attach the patch instead of inline
> >> > >> text?
> >> > >> >>>
> >> > >> >>> Erik Huelsmann wrote:
> >> > >> >>> > Our current configure script warns at the end if BDB hasn't
> >> been
> >> > >> >>> > found. Given that it's now just one of the components that
> >> can be
> >> > >> >>> used
> >> > >> >>> > Index: configure.in
> >> > >> >>> >
> >> ===================================================================
> >> > >> >>> > --- configure.in (revision 21833)
> >> > >> >>> > +++ configure.in (working copy)
> >> > >> >>> >
> >> > >> >>> > -You don't seem to have Berkeley DB version $db_version or
> >> newer
> >> > >> >>> > -installed and linked to APR-UTIL. We have created
> >> Makefiles which
> >> > >> >>> > -will build without the Berkeley DB back-end; your
> >> repositories
> >> > >> will
> >> > >> >>> > -use FSFS as the default back-end. You can find the latest
> >> > >> >>> version of
> >> > >> >>> > -Berkeley DB here:
> >> > >> >>> > - http://www.sleepycat.com/download/index.shtml
> >> > >> >>> > -])
> >> > >> >>> > -;;
> >> > >> >>> > + AC_MSG_NOTICE([ BerkeleyDB no])
> >> > >> >>> > + ;;
> >> > >> >>> > + "yes" )
> >> > >> >>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
> >> > >> >>> > + ;;
> >> > >> >>> > esac
> >> > >> >>> > -
> >> > >> >>> > +case "!$APXS!" in
> >> > >> >>>
> >> > >> >>> > + "!no!"|"!!" )
> >> > >> >>> Why we have cruft of "!" here?
> >> > >> >>
> >> > >> >> Because I didn't know how to test for the empty string, but I
> >> think
> >> > >> >> that could just be "", right?
> >> > >> >>
> >> > >> >> bye,
> >> > >> >>
> >> > >> >> Erik.
> >> > >> >>
> >> > >> > case "$APXS" in
> >> > >> > "no"|"")
> >> > >> > echo "no"
> >> > >> > ;;
> >> > >> > * )
> >> > >> > echo "default"
> >> > >> > ;;
> >> > >> > esac
> >> > >> >
> >> > >> > The above snippet works fine in my bash. If testing the empty
> >> string
> >> > >> > is really a problem, should we not do the same for others too like
> >> > >> > "$SWIG"?
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by "Hyrum K. Wright" <hy...@mail.utexas.edu>.
Did this every get committed?
-Hyrum
Erik Huelsmann wrote:
> I can. I'll do that later today.
>
> On 1/18/07, Daniel Rall <dl...@collab.net> wrote:
>> Hopeully Erik can just commit this patch, saving us the overhead of an
>> issue.
>>
>> On Tue, 16 Jan 2007, Hyrum K. Wright wrote:
>>
>> > Erik Huelsmann wrote:
>> > > On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
>> > >> What happened to this patch?
>> > >
>> > > Nothing: it's still sitting in my working copy. But, I think I should
>> > > apply it, given Peter's reaction.
>> >
>> > I don't think this ever got applied; shall I put it in the issue
>> tracker?
>> >
>> > -Hyrum
>> >
>> > >> With regards
>> > >> Kamesh Jayachandran
>> > >> Kamesh Jayachandran wrote:
>> > >> > Erik Huelsmann wrote:
>> > >> >> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
>> > >> >>> I like this change.
>> > >> >>> One small suggestion, Can you attach the patch instead of inline
>> > >> text?
>> > >> >>>
>> > >> >>> Erik Huelsmann wrote:
>> > >> >>> > Our current configure script warns at the end if BDB hasn't
>> been
>> > >> >>> > found. Given that it's now just one of the components that
>> can be
>> > >> >>> used
>> > >> >>> > Index: configure.in
>> > >> >>> >
>> ===================================================================
>> > >> >>> > --- configure.in (revision 21833)
>> > >> >>> > +++ configure.in (working copy)
>> > >> >>> >
>> > >> >>> > -You don't seem to have Berkeley DB version $db_version or
>> newer
>> > >> >>> > -installed and linked to APR-UTIL. We have created
>> Makefiles which
>> > >> >>> > -will build without the Berkeley DB back-end; your
>> repositories
>> > >> will
>> > >> >>> > -use FSFS as the default back-end. You can find the latest
>> > >> >>> version of
>> > >> >>> > -Berkeley DB here:
>> > >> >>> > - http://www.sleepycat.com/download/index.shtml
>> > >> >>> > -])
>> > >> >>> > -;;
>> > >> >>> > + AC_MSG_NOTICE([ BerkeleyDB no])
>> > >> >>> > + ;;
>> > >> >>> > + "yes" )
>> > >> >>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
>> > >> >>> > + ;;
>> > >> >>> > esac
>> > >> >>> > -
>> > >> >>> > +case "!$APXS!" in
>> > >> >>>
>> > >> >>> > + "!no!"|"!!" )
>> > >> >>> Why we have cruft of "!" here?
>> > >> >>
>> > >> >> Because I didn't know how to test for the empty string, but I
>> think
>> > >> >> that could just be "", right?
>> > >> >>
>> > >> >> bye,
>> > >> >>
>> > >> >> Erik.
>> > >> >>
>> > >> > case "$APXS" in
>> > >> > "no"|"")
>> > >> > echo "no"
>> > >> > ;;
>> > >> > * )
>> > >> > echo "default"
>> > >> > ;;
>> > >> > esac
>> > >> >
>> > >> > The above snippet works fine in my bash. If testing the empty
>> string
>> > >> > is really a problem, should we not do the same for others too like
>> > >> > "$SWIG"?
Re: [PATCH] Make configure report the components detected
Posted by Erik Huelsmann <eh...@gmail.com>.
I can. I'll do that later today.
On 1/18/07, Daniel Rall <dl...@collab.net> wrote:
> Hopeully Erik can just commit this patch, saving us the overhead of an
> issue.
>
> On Tue, 16 Jan 2007, Hyrum K. Wright wrote:
>
> > Erik Huelsmann wrote:
> > > On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> > >> What happened to this patch?
> > >
> > > Nothing: it's still sitting in my working copy. But, I think I should
> > > apply it, given Peter's reaction.
> >
> > I don't think this ever got applied; shall I put it in the issue tracker?
> >
> > -Hyrum
> >
> > >> With regards
> > >> Kamesh Jayachandran
> > >> Kamesh Jayachandran wrote:
> > >> > Erik Huelsmann wrote:
> > >> >> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> > >> >>> I like this change.
> > >> >>> One small suggestion, Can you attach the patch instead of inline
> > >> text?
> > >> >>>
> > >> >>> Erik Huelsmann wrote:
> > >> >>> > Our current configure script warns at the end if BDB hasn't been
> > >> >>> > found. Given that it's now just one of the components that can be
> > >> >>> used
> > >> >>> > Index: configure.in
> > >> >>> > ===================================================================
> > >> >>> > --- configure.in (revision 21833)
> > >> >>> > +++ configure.in (working copy)
> > >> >>> >
> > >> >>> > -You don't seem to have Berkeley DB version $db_version or newer
> > >> >>> > -installed and linked to APR-UTIL. We have created Makefiles which
> > >> >>> > -will build without the Berkeley DB back-end; your repositories
> > >> will
> > >> >>> > -use FSFS as the default back-end. You can find the latest
> > >> >>> version of
> > >> >>> > -Berkeley DB here:
> > >> >>> > - http://www.sleepycat.com/download/index.shtml
> > >> >>> > -])
> > >> >>> > -;;
> > >> >>> > + AC_MSG_NOTICE([ BerkeleyDB no])
> > >> >>> > + ;;
> > >> >>> > + "yes" )
> > >> >>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
> > >> >>> > + ;;
> > >> >>> > esac
> > >> >>> > -
> > >> >>> > +case "!$APXS!" in
> > >> >>>
> > >> >>> > + "!no!"|"!!" )
> > >> >>> Why we have cruft of "!" here?
> > >> >>
> > >> >> Because I didn't know how to test for the empty string, but I think
> > >> >> that could just be "", right?
> > >> >>
> > >> >> bye,
> > >> >>
> > >> >> Erik.
> > >> >>
> > >> > case "$APXS" in
> > >> > "no"|"")
> > >> > echo "no"
> > >> > ;;
> > >> > * )
> > >> > echo "default"
> > >> > ;;
> > >> > esac
> > >> >
> > >> > The above snippet works fine in my bash. If testing the empty string
> > >> > is really a problem, should we not do the same for others too like
> > >> > "$SWIG"?
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Daniel Rall <dl...@collab.net>.
Hopeully Erik can just commit this patch, saving us the overhead of an
issue.
On Tue, 16 Jan 2007, Hyrum K. Wright wrote:
> Erik Huelsmann wrote:
> > On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >> What happened to this patch?
> >
> > Nothing: it's still sitting in my working copy. But, I think I should
> > apply it, given Peter's reaction.
>
> I don't think this ever got applied; shall I put it in the issue tracker?
>
> -Hyrum
>
> >> With regards
> >> Kamesh Jayachandran
> >> Kamesh Jayachandran wrote:
> >> > Erik Huelsmann wrote:
> >> >> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >> >>> I like this change.
> >> >>> One small suggestion, Can you attach the patch instead of inline
> >> text?
> >> >>>
> >> >>> Erik Huelsmann wrote:
> >> >>> > Our current configure script warns at the end if BDB hasn't been
> >> >>> > found. Given that it's now just one of the components that can be
> >> >>> used
> >> >>> > Index: configure.in
> >> >>> > ===================================================================
> >> >>> > --- configure.in (revision 21833)
> >> >>> > +++ configure.in (working copy)
> >> >>> >
> >> >>> > -You don't seem to have Berkeley DB version $db_version or newer
> >> >>> > -installed and linked to APR-UTIL. We have created Makefiles which
> >> >>> > -will build without the Berkeley DB back-end; your repositories
> >> will
> >> >>> > -use FSFS as the default back-end. You can find the latest
> >> >>> version of
> >> >>> > -Berkeley DB here:
> >> >>> > - http://www.sleepycat.com/download/index.shtml
> >> >>> > -])
> >> >>> > -;;
> >> >>> > + AC_MSG_NOTICE([ BerkeleyDB no])
> >> >>> > + ;;
> >> >>> > + "yes" )
> >> >>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
> >> >>> > + ;;
> >> >>> > esac
> >> >>> > -
> >> >>> > +case "!$APXS!" in
> >> >>>
> >> >>> > + "!no!"|"!!" )
> >> >>> Why we have cruft of "!" here?
> >> >>
> >> >> Because I didn't know how to test for the empty string, but I think
> >> >> that could just be "", right?
> >> >>
> >> >> bye,
> >> >>
> >> >> Erik.
> >> >>
> >> > case "$APXS" in
> >> > "no"|"")
> >> > echo "no"
> >> > ;;
> >> > * )
> >> > echo "default"
> >> > ;;
> >> > esac
> >> >
> >> > The above snippet works fine in my bash. If testing the empty string
> >> > is really a problem, should we not do the same for others too like
> >> > "$SWIG"?
Re: [PATCH] Make configure report the components detected
Posted by "Hyrum K. Wright" <hy...@mail.utexas.edu>.
Erik Huelsmann wrote:
> On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
>> What happened to this patch?
>
> Nothing: it's still sitting in my working copy. But, I think I should
> apply it, given Peter's reaction.
I don't think this ever got applied; shall I put it in the issue tracker?
-Hyrum
>> With regards
>> Kamesh Jayachandran
>> Kamesh Jayachandran wrote:
>> > Erik Huelsmann wrote:
>> >> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
>> >>> I like this change.
>> >>> One small suggestion, Can you attach the patch instead of inline
>> text?
>> >>>
>> >>> Erik Huelsmann wrote:
>> >>> > Our current configure script warns at the end if BDB hasn't been
>> >>> > found. Given that it's now just one of the components that can be
>> >>> used
>> >>> > Index: configure.in
>> >>> > ===================================================================
>> >>> > --- configure.in (revision 21833)
>> >>> > +++ configure.in (working copy)
>> >>> >
>> >>> > -You don't seem to have Berkeley DB version $db_version or newer
>> >>> > -installed and linked to APR-UTIL. We have created Makefiles which
>> >>> > -will build without the Berkeley DB back-end; your repositories
>> will
>> >>> > -use FSFS as the default back-end. You can find the latest
>> >>> version of
>> >>> > -Berkeley DB here:
>> >>> > - http://www.sleepycat.com/download/index.shtml
>> >>> > -])
>> >>> > -;;
>> >>> > + AC_MSG_NOTICE([ BerkeleyDB no])
>> >>> > + ;;
>> >>> > + "yes" )
>> >>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
>> >>> > + ;;
>> >>> > esac
>> >>> > -
>> >>> > +case "!$APXS!" in
>> >>>
>> >>> > + "!no!"|"!!" )
>> >>> Why we have cruft of "!" here?
>> >>
>> >> Because I didn't know how to test for the empty string, but I think
>> >> that could just be "", right?
>> >>
>> >> bye,
>> >>
>> >> Erik.
>> >>
>> > case "$APXS" in
>> > "no"|"")
>> > echo "no"
>> > ;;
>> > * )
>> > echo "default"
>> > ;;
>> > esac
>> >
>> > The above snippet works fine in my bash. If testing the empty string
>> > is really a problem, should we not do the same for others too like
>> > "$SWIG"?
>> >
>> > With regards
>> > Kamesh Jayachandran
Re: [PATCH] Make configure report the components detected
Posted by Erik Huelsmann <eh...@gmail.com>.
On 11/16/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> What happened to this patch?
Nothing: it's still sitting in my working copy. But, I think I should
apply it, given Peter's reaction.
Thanks for the heads up!
bye,
Erik.
> With regards
> Kamesh Jayachandran
> Kamesh Jayachandran wrote:
> > Erik Huelsmann wrote:
> >> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> >>> I like this change.
> >>> One small suggestion, Can you attach the patch instead of inline text?
> >>>
> >>> Erik Huelsmann wrote:
> >>> > Our current configure script warns at the end if BDB hasn't been
> >>> > found. Given that it's now just one of the components that can be
> >>> used
> >>> > Index: configure.in
> >>> > ===================================================================
> >>> > --- configure.in (revision 21833)
> >>> > +++ configure.in (working copy)
> >>> >
> >>> > -You don't seem to have Berkeley DB version $db_version or newer
> >>> > -installed and linked to APR-UTIL. We have created Makefiles which
> >>> > -will build without the Berkeley DB back-end; your repositories will
> >>> > -use FSFS as the default back-end. You can find the latest
> >>> version of
> >>> > -Berkeley DB here:
> >>> > - http://www.sleepycat.com/download/index.shtml
> >>> > -])
> >>> > -;;
> >>> > + AC_MSG_NOTICE([ BerkeleyDB no])
> >>> > + ;;
> >>> > + "yes" )
> >>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
> >>> > + ;;
> >>> > esac
> >>> > -
> >>> > +case "!$APXS!" in
> >>>
> >>> > + "!no!"|"!!" )
> >>> Why we have cruft of "!" here?
> >>
> >> Because I didn't know how to test for the empty string, but I think
> >> that could just be "", right?
> >>
> >> bye,
> >>
> >> Erik.
> >>
> > case "$APXS" in
> > "no"|"")
> > echo "no"
> > ;;
> > * )
> > echo "default"
> > ;;
> > esac
> >
> > The above snippet works fine in my bash. If testing the empty string
> > is really a problem, should we not do the same for others too like
> > "$SWIG"?
> >
> > With regards
> > Kamesh Jayachandran
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
> > For additional commands, e-mail: dev-help@subversion.tigris.org
> >
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Kamesh Jayachandran <ka...@collab.net>.
What happened to this patch?
With regards
Kamesh Jayachandran
Kamesh Jayachandran wrote:
> Erik Huelsmann wrote:
>> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
>>> I like this change.
>>> One small suggestion, Can you attach the patch instead of inline text?
>>>
>>> Erik Huelsmann wrote:
>>> > Our current configure script warns at the end if BDB hasn't been
>>> > found. Given that it's now just one of the components that can be
>>> used
>>> > Index: configure.in
>>> > ===================================================================
>>> > --- configure.in (revision 21833)
>>> > +++ configure.in (working copy)
>>> >
>>> > -You don't seem to have Berkeley DB version $db_version or newer
>>> > -installed and linked to APR-UTIL. We have created Makefiles which
>>> > -will build without the Berkeley DB back-end; your repositories will
>>> > -use FSFS as the default back-end. You can find the latest
>>> version of
>>> > -Berkeley DB here:
>>> > - http://www.sleepycat.com/download/index.shtml
>>> > -])
>>> > -;;
>>> > + AC_MSG_NOTICE([ BerkeleyDB no])
>>> > + ;;
>>> > + "yes" )
>>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
>>> > + ;;
>>> > esac
>>> > -
>>> > +case "!$APXS!" in
>>>
>>> > + "!no!"|"!!" )
>>> Why we have cruft of "!" here?
>>
>> Because I didn't know how to test for the empty string, but I think
>> that could just be "", right?
>>
>> bye,
>>
>> Erik.
>>
> case "$APXS" in
> "no"|"")
> echo "no"
> ;;
> * )
> echo "default"
> ;;
> esac
>
> The above snippet works fine in my bash. If testing the empty string
> is really a problem, should we not do the same for others too like
> "$SWIG"?
>
> With regards
> Kamesh Jayachandran
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
> For additional commands, e-mail: dev-help@subversion.tigris.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Peter Samuelson <pe...@p12n.org>.
[Kamesh Jayachandran]
> case "$APXS" in
> "no"|"")
> echo "no"
> ;;
> * )
> echo "default"
> ;;
> esac
>
> The above snippet works fine in my bash. If testing the empty string is
> really a problem, should we not do the same for others too like "$SWIG"?
In general, /bin/sh is not bash. When you have a shell syntax
question, you really need to test on other shells. I keep dash, posh,
ksh93, pdksh, and zsh around just for this purpose.
Those shells all agree with your finding in the present case, but I
note that I don't have a real Bourne shell handy, or the infamous
Solaris /bin/sh.
Re: [PATCH] Make configure report the components detected
Posted by Kamesh Jayachandran <ka...@collab.net>.
Erik Huelsmann wrote:
> On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
>> I like this change.
>> One small suggestion, Can you attach the patch instead of inline text?
>>
>> Erik Huelsmann wrote:
>> > Our current configure script warns at the end if BDB hasn't been
>> > found. Given that it's now just one of the components that can be used
>> > Index: configure.in
>> > ===================================================================
>> > --- configure.in (revision 21833)
>> > +++ configure.in (working copy)
>> >
>> > -You don't seem to have Berkeley DB version $db_version or newer
>> > -installed and linked to APR-UTIL. We have created Makefiles which
>> > -will build without the Berkeley DB back-end; your repositories will
>> > -use FSFS as the default back-end. You can find the latest version of
>> > -Berkeley DB here:
>> > - http://www.sleepycat.com/download/index.shtml
>> > -])
>> > -;;
>> > + AC_MSG_NOTICE([ BerkeleyDB no])
>> > + ;;
>> > + "yes" )
>> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
>> > + ;;
>> > esac
>> > -
>> > +case "!$APXS!" in
>>
>> > + "!no!"|"!!" )
>> Why we have cruft of "!" here?
>
> Because I didn't know how to test for the empty string, but I think
> that could just be "", right?
>
> bye,
>
> Erik.
>
case "$APXS" in
"no"|"")
echo "no"
;;
* )
echo "default"
;;
esac
The above snippet works fine in my bash. If testing the empty string is
really a problem, should we not do the same for others too like "$SWIG"?
With regards
Kamesh Jayachandran
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Erik Huelsmann <eh...@gmail.com>.
On 10/9/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> I like this change.
> One small suggestion, Can you attach the patch instead of inline text?
>
> Erik Huelsmann wrote:
> > Our current configure script warns at the end if BDB hasn't been
> > found. Given that it's now just one of the components that can be used
> > Index: configure.in
> > ===================================================================
> > --- configure.in (revision 21833)
> > +++ configure.in (working copy)
> >
> > -You don't seem to have Berkeley DB version $db_version or newer
> > -installed and linked to APR-UTIL. We have created Makefiles which
> > -will build without the Berkeley DB back-end; your repositories will
> > -use FSFS as the default back-end. You can find the latest version of
> > -Berkeley DB here:
> > - http://www.sleepycat.com/download/index.shtml
> > -])
> > -;;
> > + AC_MSG_NOTICE([ BerkeleyDB no])
> > + ;;
> > + "yes" )
> > + AC_MSG_NOTICE([ BerkeleyDB $db_version])
> > + ;;
> > esac
> > -
> > +case "!$APXS!" in
>
> > + "!no!"|"!!" )
> Why we have cruft of "!" here?
Because I didn't know how to test for the empty string, but I think
that could just be "", right?
bye,
Erik.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Kamesh Jayachandran <ka...@collab.net>.
I like this change.
One small suggestion, Can you attach the patch instead of inline text?
Erik Huelsmann wrote:
> Our current configure script warns at the end if BDB hasn't been
> found. Given that it's now just one of the components that can be used
> Index: configure.in
> ===================================================================
> --- configure.in (revision 21833)
> +++ configure.in (working copy)
>
> -You don't seem to have Berkeley DB version $db_version or newer
> -installed and linked to APR-UTIL. We have created Makefiles which
> -will build without the Berkeley DB back-end; your repositories will
> -use FSFS as the default back-end. You can find the latest version of
> -Berkeley DB here:
> - http://www.sleepycat.com/download/index.shtml
> -])
> -;;
> + AC_MSG_NOTICE([ BerkeleyDB no])
> + ;;
> + "yes" )
> + AC_MSG_NOTICE([ BerkeleyDB $db_version])
> + ;;
> esac
> -
> +case "!$APXS!" in
> + "!no!"|"!!" )
Why we have cruft of "!" here?
With regards
Kamesh Jayachandran
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: [PATCH] Make configure report the components detected
Posted by Daniel Rall <dl...@collab.net>.
On Sun, 08 Oct 2006, Erik Huelsmann wrote:
> Our current configure script warns at the end if BDB hasn't been
> found. Given that it's now just one of the components that can be used
> to build Subversion on (and that FSFS isn't so immature anymore), it
> would be more logical to report which components *are* used. From
> that, the user can decide if the list is correct or not.
...
I'm +1 on the concept of 'configure' reporting the dependencies it
will be using.
If we're going to report on Java (which I'm in favor of), we should
report on Python (via get-py-info.py?) and Ruby and Perl as well.