Winter is coming [upcoming release 0.10]

Winter is coming [upcoming release 0.10]

[David Bremner]

[-- Attachment #1: Type: text/plain, Size: 717 bytes --]


Hi All.

According to our sortof-schedule, we should release 0.10 soonish.  I'd
like to freeze for a few days first, so that means we have time for
maybe one or two more non-trivial patches before the freeze.

- id:"1320846275-28520-1-git-send-email-amdragon@mit.edu" is a
  candidate, but needs review

- id:"8771bb7ee67586515ff5114747fd680005ccfc5a.1320615322.git.jani@nikula.org" 
  Needs a review of the cli part

- id:"1320599856-24078-1-git-send-email-amdragon@mit.edu" is probably
  not a candidate. I'm not ready to push database format changing
  patches without some feedback from cworth.  It also needs some more
  review in general.

Other suggestions/comments?

David





  

[-- Attachment #2: Type: application/pgp-signature, Size: 315 bytes --]

[PATCH] Don't link libnotmuch if libutil isn't linked in properly.

[Tom Prince]

For some reason, on my machine, the link is picking up
/usr/lib/libutil.so instead of util/libutil.a. This causes there to be
undefined symbols in libnotmuch, making it unuseable. This patch causes
the link to fail instead.
---
 lib/Makefile.local |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
 
On Sun, 13 Nov 2011 10:19:14 -0400, David Bremner <david@tethera.net> wrote:
> According to our sortof-schedule, we should release 0.10 soonish.  I'd
> like to freeze for a few days first, so that means we have time for
> maybe one or two more non-trivial patches before the freeze.

libnotmuch isn't linking properly to libutil for me. This seems to be
due to $(xapian-config --libs) including -L/usr/lib64, which then causes
-lutil to pick up /usr/lib64/libutil.so (from glibc) instead.
I noticed this failing because symbol-test fails to compile, causing
both symbol-hinding test to fail.

The following patch doesn't fix this, but does cuase the build to fail
earlier.

diff --git a/lib/Makefile.local b/lib/Makefile.local
index d58552c..cb53781 100644
--- a/lib/Makefile.local
+++ b/lib/Makefile.local
@@ -30,7 +30,7 @@ LIBRARY_SUFFIX = so
 LINKER_NAME = libnotmuch.$(LIBRARY_SUFFIX)
 SONAME = $(LINKER_NAME).$(LIBNOTMUCH_VERSION_MAJOR)
 LIBNAME = $(SONAME).$(LIBNOTMUCH_VERSION_MINOR).$(LIBNOTMUCH_VERSION_RELEASE)
-LIBRARY_LINK_FLAG = -shared -Wl,--version-script=notmuch.sym,-soname=$(SONAME)
+LIBRARY_LINK_FLAG = -shared -Wl,--version-script=notmuch.sym,-soname=$(SONAME) -Wl,--no-undefined
 ifeq ($(LIBDIR_IN_LDCONFIG),1)
 ifeq ($(DESTDIR),)
 LIBRARY_INSTALL_POST_COMMAND=ldconfig
--
1.7.6.1

[PATCH] Link libutil using filenmae, rather than using -l.

[Tom Prince]

glibc includes a libutil, so if the wrong -L options get passed, we
will pick up glibc's version, rather than our own.
---
 lib/Makefile.local |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/lib/Makefile.local b/lib/Makefile.local
index cb53781..54c4dea 100644
--- a/lib/Makefile.local
+++ b/lib/Makefile.local
@@ -70,7 +70,7 @@ $(dir)/libnotmuch.a: $(libnotmuch_modules)
 	$(call quiet,AR) rcs $@ $^
 
 $(dir)/$(LIBNAME): $(libnotmuch_modules) notmuch.sym
-	$(call quiet,CXX $(CXXFLAGS)) $(libnotmuch_modules) $(FINAL_LIBNOTMUCH_LDFLAGS) $(LIBRARY_LINK_FLAG) -o $@ -L$(srcdir)/util -lutil
+	$(call quiet,CXX $(CXXFLAGS)) $(libnotmuch_modules) $(FINAL_LIBNOTMUCH_LDFLAGS) $(LIBRARY_LINK_FLAG) -o $@ util/libutil.a
 
 notmuch.sym: $(srcdir)/$(dir)/notmuch.h $(libnotmuch_modules)
 	sh $(srcdir)/$(lib)/gen-version-script.sh $< $(libnotmuch_modules) > $@
-- 
1.7.6.1

Re: [PATCH] Don't link libnotmuch if libutil isn't linked in properly.

[Tom Prince]

Perhaps the g++ step in symbol-hiding should in fact be a test. Right
now, that step failing isn't captured by the test-suite.

  Tom

Re: Winter is coming [upcoming release 0.10]

[Austin Clements]

Quoth David Bremner on Nov 13 at 10:19 am:
> - id:"1320599856-24078-1-git-send-email-amdragon@mit.edu" is probably
>   not a candidate. I'm not ready to push database format changing
>   patches without some feedback from cworth.  It also needs some more
>   review in general.

This was originally Istvan Marko's patch back in
id:"m3sjsv2kw2.fsf@zsu.kismala.com", where it received a decent amount
of discussion (and then some more starting at
id:"m3pqlfhrkk.fsf@zsu.kismala.com" ).

I've poked cworth about the schema change on IRC (and CC'd him here).

Re: [PATCH] Link libutil using filenmae, rather than using -l.

[David Bremner]

On Sun, 13 Nov 2011 12:05:03 -0500, Tom Prince <tom.prince@ualberta.net> wrote:
> glibc includes a libutil, so if the wrong -L options get passed, we
> will pick up glibc's version, rather than our own.

pushed,

d

Re: Winter is coming [upcoming release 0.10]

[David Bremner]

[-- Attachment #1: Type: text/plain, Size: 2098 bytes --]

On Sun, 13 Nov 2011 10:19:14 -0400, David Bremner <david@tethera.net> wrote:

> According to our sortof-schedule, we should release 0.10 soonish.  I'd
> like to freeze for a few days first, so that means we have time for
> maybe one or two more non-trivial patches before the freeze.

We are now frozen for 0.10 release. As is my habit, I tagged 0.10_rc1,
but didn't upload tarballs. I also upload 0.10~rc1 packages to Debian
experimental.

Being frozen means that pushes to master after the tag 0.10_rc1 won't
automatically be included in the release. Mainly this effects the
bindings and contrib maintainers.  Please send me email or catch me on
irc about anything urgent to include. And no, I don't mean all
outstanding patchs.

> - id:"1320846275-28520-1-git-send-email-amdragon@mit.edu" is a
>   candidate, but needs review

didn't make it, next round.

> - id:"8771bb7ee67586515ff5114747fd680005ccfc5a.1320615322.git.jani@nikula.org" 
>   Needs a review of the cli part

made it.

> 
> - id:"1320599856-24078-1-git-send-email-amdragon@mit.edu" is probably
>   not a candidate. 

My various objections were overcome, and this made it.

As usual, we miss some NEWS items:

Ali Polatel:
    Some announcement for notmuch-deliver in contrib

Austin Clements:
      emacs: Use a single buffer invisibility spec to fix quadratic search cost.

Daniel Schoepe (2):
      emacs: Turn id:"<message-id>" elements into buttons for notmuch searches
      emacs: Tab completion for notmuch-search and notmuch-search-filter

David Bremner:
      contrib/nmbug: new script for sharing tags with a given prefix.
      - patch to follow.

Dmitry Kurochkin: ?
      emacs: add invisible space after the search widget field in notmuch-hello

Jameson Graef Rollins:
      emacs: add notmuch-show-worker function for specifying crypto processing directly

Michal Sojka:
      Do not query on notmuch-search exit

Pieter Praet (2):
      emacs: add keybind and function to stash Message-ID without prefix
      test: stashing in notmuch-{show,search}



[-- Attachment #2: Type: application/pgp-signature, Size: 315 bytes --]

[PATCH] NEWS: discuss contrib and nmbug

[David Bremner]

From: David Bremner <bremner@debian.org>

I mention the possibility of a seperate license here because currently
notmuch-deliver is licensed GPL v2 only.
---
 NEWS |   14 ++++++++++++++
 1 files changed, 14 insertions(+), 0 deletions(-)

diff --git a/NEWS b/NEWS
index 7bd9987..98e07eb 100644
--- a/NEWS
+++ b/NEWS
@@ -48,6 +48,20 @@ Search avoids opening and parsing message files
 	notmuch new
 	notmuch restore notmuch.dump
 
+New collection of add-on tools
+------------------------------
+
+The source directory "contrib" contains tools built on notmuch.  These
+tools are not part of notmuch, and you should check their individual
+licenses.  Feel free to report problems with them to the notmuch
+mailing list.
+
+nmbug - share tags with a given prefix
+
+   nmbug helps maintain a git repo containing all tags with a given
+   prefix (by default "notmuch::"). Tags can be shared by commiting
+   them to git in one location and restoring in another.
+
 Notmuch 0.9 (2011-10-01)
 ========================
 
-- 
1.7.7.1

Re: [PATCH] NEWS: discuss contrib and nmbug

[David Bremner]

On Tue, 15 Nov 2011 23:13:29 -0400, David Bremner <david@tethera.net> wrote:
> From: David Bremner <bremner@debian.org>
> 
> I mention the possibility of a seperate license here because currently
> notmuch-deliver is licensed GPL v2 only.

I decided to push this patch "eagerly" rather than waiting for comments,
in order to reduce conflicts as we all edit NEWS.

d

Re: Winter is coming [upcoming release 0.10]

[Tomi Ollila]

On Tue, 15 Nov 2011 23:11:49 -0400, David Bremner <david@tethera.net> wrote:
> On Sun, 13 Nov 2011 10:19:14 -0400, David Bremner <david@tethera.net> wrote:
> 
> > According to our sortof-schedule, we should release 0.10 soonish.  I'd
> > like to freeze for a few days first, so that means we have time for
> > maybe one or two more non-trivial patches before the freeze.
> 
> We are now frozen for 0.10 release. As is my habit, I tagged 0.10_rc1,
> but didn't upload tarballs. I also upload 0.10~rc1 packages to Debian
> experimental.
> 
> Being frozen means that pushes to master after the tag 0.10_rc1 won't
> automatically be included in the release. Mainly this effects the
> bindings and contrib maintainers.  Please send me email or catch me on
> irc about anything urgent to include. And no, I don't mean all
> outstanding patchs.
> 
> > - id:"1320846275-28520-1-git-send-email-amdragon@mit.edu" is a
> >   candidate, but needs review
> 
> didn't make it, next round.

No wonder my  'notmuch tag -inbox ..1321192972' was so slow...
(and subsequent 'notmut tag -inbox from:... and not tag:inbox' were faster)


> > 
> > - id:"1320599856-24078-1-git-send-email-amdragon@mit.edu" is probably
> >   not a candidate. 
> 
> My various objections were overcome, and this made it.

Thanks, things are faster now -- I just need to start testing/using
amdragon's rewritten json.el[c] soon.


Thanks, you've done huge job to get all this integrated.

Tomi

Re: Winter is coming [upcoming release 0.10]

[Austin Clements]

Quoth David Bremner on Nov 15 at 11:11 pm:
> Austin Clements:
>       emacs: Use a single buffer invisibility spec to fix quadratic search cost.

To avoid conflicts on NEWS, here's the text to put under
"Optimizations" (which should perhaps be "Performance" instead?):


Emacs now constructs large search buffers more efficiently

Re: Winter is coming [upcoming release 0.10]

[David Bremner]

On Wed, 16 Nov 2011 10:43:54 -0500, Austin Clements <amdragon@MIT.EDU> wrote:
> Quoth David Bremner on Nov 15 at 11:11 pm:
> > Austin Clements:
> >       emacs: Use a single buffer invisibility spec to fix quadratic search cost.
> 
> To avoid conflicts on NEWS, here's the text to put under
> "Optimizations" (which should perhaps be "Performance" instead?):
> 
> 
> Emacs now constructs large search buffers more efficiently

done and done.

d

Re: Winter is coming [upcoming release 0.10]

[Pieter Praet]

Thanks David, awesome job!


Peace

-- 
Pieter

[PATCH 1/2] NEWS: add entries for stashing-related keybinding and tests

[Pieter Praet]

Add news entries for commits:
  f9764bfacc97457d1154c2d2a6001a6564f13ec3
  64febdf71c4184ca369f5d11d7f196704a3ec1a6
---
 NEWS |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/NEWS b/NEWS
index 044a51e..6c064da 100644
--- a/NEWS
+++ b/NEWS
@@ -8,6 +8,8 @@ Emacs tests are now done in dtach. This means that dtach  is now
 needed to run the notmuch test suite, at least until the checking for
 prerequisites is improved.
 
+Full test coverage of the stashing feature in Emacs.
+
 New command-line features
 -------------------------
 
@@ -46,6 +48,10 @@ Turn "id:MSG-ID" links into buttons associated with notmuch searches
   Text of the form "id:MSG-ID" in mails is now a clickable button that
   opens a notmuch search for the given message id.
 
+Add keybinding ('c I') for stashing Message-ID's without an id: prefix
+
+  Reduces manual labour when stashing them for use with `git send-email'.
+
 Performance
 -----------
 
-- 
1.7.7.1

[PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[Pieter Praet]

No further explanation needed.
---
 NEWS |   14 +++++++-------
 1 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/NEWS b/NEWS
index 6c064da..daec4ad 100644
--- a/NEWS
+++ b/NEWS
@@ -130,7 +130,7 @@ Ruby bindings changes
 ---------------------
 
   - Wrap new library functions notmuch_database_{begin,end}_atomic.
-  - Add new exception Notmuch::UnbalancedAtomicError. 
+  - Add new exception Notmuch::UnbalancedAtomicError.
   - Rename destroy  to destroy! according to Ruby naming conventions.
   - Update for n_d_find_message* API changes (see above).
 
@@ -159,7 +159,7 @@ Improved handling of message/rfc822 parts
   headers, is unfortunately not yet supported (but hopefully will be
   soon).
 
-Improved Build system portability 
+Improved Build system portability
 
   Certain parts of the shell script generating notmuch.sym were
   specific to the GNU versions of sed and nm. The new version should
@@ -202,7 +202,7 @@ Sebastian Spaeth contributed two changes related to unicode and UTF8:
   * message tags are now explicitly unicode
   * query string is encoded as a UTF8 byte string
 
-Build-System improvments
+Build-System improvements
 ------------------------
 
 Generate notmuch.sym after the relevant object files
@@ -217,7 +217,7 @@ Bug-fix release.
 ----------------
 
 Re-export Xapian exception typeinfo symbols.
-  
+
   It turned out our aggressive symbol hiding caused problems for
   people running gcc 4.4.5.
 
@@ -382,7 +382,7 @@ Hiding of repeated subjects in collapsed thread view
 
   In notmuch-show mode, if a collapsed message has the same subject as
   its parent, the subject is not shown.
-  
+
 Automatic detection and hiding of original message in top-posted message
 
   When a message contains a line looking something like:
@@ -448,7 +448,7 @@ Ruby bindings are now much more complete
   Including QUERY.sort, QUERY.to_s, MESSAGE.maildir_flags_to_tags,
   MESSAGE.tags_to_maildir_flags, and MESSAGE.get_filenames
 
-* Python bindings have been upodated and extended
+* Python bindings have been updated and extended
   (docs online at http://packages.python.org/notmuch/)
 
   New bindings:
@@ -467,7 +467,7 @@ Ruby bindings are now much more complete
 	s2 -= s1
 
    Removed:
-     - len(Messages()) as it exausted the iterator.
+     - len(Messages()) as it exhausted the iterator.
        Use len(list(Messages())) or
        Query.count_messages() to get the length.
 
-- 
1.7.7.1

Re: [PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[Tomi Ollila]

On Wed, 16 Nov 2011 21:37:50 +0100, Pieter Praet <pieter@praet.org> wrote:
> No further explanation needed.
> ---
>  NEWS |   14 +++++++-------
>  1 files changed, 7 insertions(+), 7 deletions(-)

The 'No further explanation needed.' should be *BELOW* the waistline (---) ;);
now it goes to the commit message which is probably not what anybody wants.

Tomi

Re: [PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[Pieter Praet]

On Wed, 16 Nov 2011 22:54:35 +0200, Tomi Ollila <tomi.ollila@iki.fi> wrote:
> On Wed, 16 Nov 2011 21:37:50 +0100, Pieter Praet <pieter@praet.org> wrote:
> > No further explanation needed.
> > ---
> >  NEWS |   14 +++++++-------
> >  1 files changed, 7 insertions(+), 7 deletions(-)
> 
> The 'No further explanation needed.' should be *BELOW* the waistline (---) ;);
> now it goes to the commit message which is probably not what anybody wants.
> 

That was intended :)

Empty commit messages are frowned upon (even if the commit subject is
sufficiently descriptive), so "rather a pointless one than none?" :)

> Tomi
> 


Peace

-- 
Pieter

Re: Winter is coming [upcoming release 0.10]

[David Bremner]

On Wed, 16 Nov 2011 21:37:48 +0100, Pieter Praet <pieter@praet.org> wrote:
> Thanks David, awesome job!
> 
> 

You're very welcome. I hope you still feel that way when you see how I
amended your commits ;). If not, send updates.

d

Re: Winter is coming [upcoming release 0.10]

[Pieter Praet]

On Wed, 16 Nov 2011 18:17:46 -0400, David Bremner <david@tethera.net> wrote:
> On Wed, 16 Nov 2011 21:37:48 +0100, Pieter Praet <pieter@praet.org> wrote:
> > Thanks David, awesome job!
> > 
> > 
> 
> You're very welcome. I hope you still feel that way when you see how I
> amended your commits ;). If not, send updates.
> 

Hehe :)  For the occasion, I'll temporarily relinquish my right to refuse
being the recipient of credit for that which I have not accomplished, but
don't expect it to become a habit ;)

Thanks again!

> d


Peace

-- 
Pieter

Re: [PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[Jani Nikula]

[-- Attachment #1: Type: text/plain, Size: 901 bytes --]

On Nov 16, 2011 11:39 PM, "Pieter Praet" <pieter@praet.org> wrote:
>
> On Wed, 16 Nov 2011 22:54:35 +0200, Tomi Ollila <tomi.ollila@iki.fi>
wrote:
> > On Wed, 16 Nov 2011 21:37:50 +0100, Pieter Praet <pieter@praet.org>
wrote:
> > > No further explanation needed.
> > > ---
> > >  NEWS |   14 +++++++-------
> > >  1 files changed, 7 insertions(+), 7 deletions(-)
> >
> > The 'No further explanation needed.' should be *BELOW* the waistline
(---) ;);
> > now it goes to the commit message which is probably not what anybody
wants.
> >
>
> That was intended :)
>
> Empty commit messages are frowned upon (even if the commit subject is
> sufficiently descriptive), so "rather a pointless one than none?" :)

Generally I think the message should be self-contained and understandable
without the subject line, so IMHO a full meaningful sentence is better than
one/none, even if it's repetition.

BR,
Jani.

[-- Attachment #2: Type: text/html, Size: 1276 bytes --]

Re: [PATCH] Don't link libnotmuch if libutil isn't linked in properly.

[David Bremner]

On Sun, 13 Nov 2011 11:34:42 -0500, Tom Prince <tom.prince@ualberta.net> wrote:
> For some reason, on my machine, the link is picking up
> /usr/lib/libutil.so instead of util/libutil.a. This causes there to be
> undefined symbols in libnotmuch, making it unuseable. This patch causes
> the link to fail instead.

I pushed this to master, let's see if it breaks for anyone.

d

Re: Winter is coming [upcoming release 0.10]

[David Bremner]

[-- Attachment #1: Type: text/plain, Size: 1311 bytes --]

On Tue, 15 Nov 2011 23:11:49 -0400, David Bremner <david@tethera.net> wrote:

> We are now frozen for 0.10 release. As is my habit, I tagged 0.10_rc1,
> but didn't upload tarballs. I also upload 0.10~rc1 packages to Debian
> experimental.

I have now tagged 0.10_rc2, which mainly has doc changes since _rc1,
except for two patches trying to make the test suites use of dtach more
robust.

The emacs tests are still failing on the Debian autobuilders (not sure
why) and in places where TMPDIR is set to some absurdly long path. We
may or may not need to fix this before release.

The license of notmuch-deliver is a concern. notmuch-deliver is GPL2
(only) and notmuch is GPL3+. Now, I'm fairly confident that notmuch
upstream is not going to sue, but it seems that binaries of
notmuch-deliver are probably not distributable according to a strict
reading of the licenses (or the FSF FAQ).

We still are missing a few NEWS items; some of which which we could
perhaps ignore.

> 
> Ali Polatel:
>     Some announcement for notmuch-deliver in contrib
> 
> Dmitry Kurochkin: 
>       emacs: add invisible space after the search widget field in notmuch-hello

This is probably ignorable from a NEWS point of view.

> 
> Michal Sojka:
>       Do not query on notmuch-search exit
> 

[-- Attachment #2: Type: application/pgp-signature, Size: 315 bytes --]

Re: [PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[Pieter Praet]

On Thu, 17 Nov 2011 08:21:59 +0200, Jani Nikula <jani@nikula.org> wrote:
> On Nov 16, 2011 11:39 PM, "Pieter Praet" <pieter@praet.org> wrote:
> >
> > On Wed, 16 Nov 2011 22:54:35 +0200, Tomi Ollila <tomi.ollila@iki.fi>
> wrote:
> > > On Wed, 16 Nov 2011 21:37:50 +0100, Pieter Praet <pieter@praet.org>
> wrote:
> > > > No further explanation needed.
> > > > ---
> > > >  NEWS |   14 +++++++-------
> > > >  1 files changed, 7 insertions(+), 7 deletions(-)
> > >
> > > The 'No further explanation needed.' should be *BELOW* the waistline
> (---) ;);
> > > now it goes to the commit message which is probably not what anybody
> wants.
> > >
> >
> > That was intended :)
> >
> > Empty commit messages are frowned upon (even if the commit subject is
> > sufficiently descriptive), so "rather a pointless one than none?" :)
> 
> Generally I think the message should be self-contained and understandable
> without the subject line, so IMHO a full meaningful sentence is better than
> one/none, even if it's repetition.
> 
> BR,
> Jani.

That would indeed have been the most sensible thing to do.

Could someone (who -unlike me- has already submitted his/her public SSH
key for commit access) add something along those lines to the Patch
Formatting guidelines [1] ?  Perhaps just quote Carl's statement [2] ?


Peace

-- 
Pieter

[1] http://notmuchmail.org/patchformatting/
[2] id:"87ocmwok2w.fsf@yoom.home.cworth.org"

Re: [PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[David Bremner]

On Thu, 12 Jan 2012 18:06:14 +0100, Pieter Praet <pieter@praet.org> wrote:
> That would indeed have been the most sensible thing to do.
> 
> Could someone (who -unlike me- has already submitted his/her public SSH
> key for commit access) add something along those lines to the Patch
> Formatting guidelines [1] ?  Perhaps just quote Carl's statement [2] ?
> 

Hi Pieter;

It's a wiki, anyone can push.

d

Re: [PATCH 2/2] NEWS: fix some old typos and trailing whitespace

[Pieter Praet]

On Thu, 12 Jan 2012 23:11:35 -0400, David Bremner <david@tethera.net> wrote:
> On Thu, 12 Jan 2012 18:06:14 +0100, Pieter Praet <pieter@praet.org> wrote:
> > That would indeed have been the most sensible thing to do.
> > 
> > Could someone (who -unlike me- has already submitted his/her public SSH
> > key for commit access) add something along those lines to the Patch
> > Formatting guidelines [1] ?  Perhaps just quote Carl's statement [2] ?
> > 
> 
> Hi Pieter;
> 
> It's a wiki, anyone can push.
> 

Oh, I wasn't aware of this. (changed recently?)

http://notmuchmail.org/patchformatting/ has been amended!

> d


Peace

-- 
Pieter