Thanks for the nice and detailed SRU template, and all the testing.
While looking at this, particularly the changes in the packaging,
there are a few points that seem to warrant clarification, since
are apparently not related/supporting the upstream bug fixes, or
are not justified as to why included in a SRU (ie, which issues
or bugs are being addressed).
I checked most of these points with Andreas yesterday, and he too
agreed more details would be required before proceeding with this.
Could please address the points below? That's very appreciated.
There are some changes in debian/ that may warrant some clarification
as they do not fit into the SRU policy for minor release updates
(ie, bug-fix only) or to be related to the bugs fixed by upstream
(ie, to support a bug-fix).
I've looked for similarities in the changelog of previous releases,
but apparently haven't found them as precedents. This is the first
time I'm looking at this package, so I might be missing something.
I'll note specific changes in changelog/rules below, for clarity.
1)
d/changelog
18 + - don't try to install dbreport config modules into
19 + -uiconfig-report-builder if the report builder is disabled
Is this possibly a separate bug, as apparently it's not related to
an upstream bug fix; it would be nice to have some clarification.
Maybe this is covered in the SRU policy under "it is also acceptable to upload new microreleases with many bug fixes without individual Launchpad bugs for each of them", but I considered such 'many bug fixes' to be related to the upstream bugs, and this change is on packaging.
2)
d/changelog
20 + - switch OOO_BASE_ARCHS/OOO_REPORTBUILDER_ARCHS variables
21 + (build base on all architectures instead of Java architectures
22 + and report-builder-* only on archs where Java is enabled, not on
23 + "all base archs"); there's -sdbc-{mysql,postgresql,firebird} anyway
Similarly; it would be nice to have some clarification.
And also the reason to enable it in an SRU? Maybe some relation to
an upstream bug or ubuntu bug? Any changes at behavior/runtime?
(apparently just an internal build dep; why is this needed, and
isn't the archive version (older) sufficient?)
Hi Rico,
Thanks for the nice and detailed SRU template, and all the testing.
While looking at this, particularly the changes in the packaging,
there are a few points that seem to warrant clarification, since
are apparently not related/supporting the upstream bug fixes, or
are not justified as to why included in a SRU (ie, which issues
or bugs are being addressed).
I checked most of these points with Andreas yesterday, and he too
agreed more details would be required before proceeding with this.
Could please address the points below? That's very appreciated.
Thanks again for all the effort on this!
cheers,
Mauricio
...
Changes:
$ debdiff libreoffice_ 7.6.2-0ubuntu1. dsc libreoffice_ 7.6.3-0ubuntu0. 23.10.1. dsc | filterdiff -i '*/debian/*'
There are some changes in debian/ that may warrant some clarification
as they do not fit into the SRU policy for minor release updates
(ie, bug-fix only) or to be related to the bugs fixed by upstream
(ie, to support a bug-fix).
I've looked for similarities in the changelog of previous releases,
but apparently haven't found them as precedents. This is the first
time I'm looking at this package, so I might be missing something.
I'll note specific changes in changelog/rules below, for clarity.
1)
d/changelog
18 + - don't try to install dbreport config modules into report- builder if the report builder is disabled
19 + -uiconfig-
d/rules
413 +ifeq "$(ENABLE_ REPORTBUILDER) " "y" -common/ $(OODIR) /share/ config/ soffice. cfg/modules/ dbreport \ -uiconfig- report- builder/ $(OODIR) /share/ config/ soffice. cfg/modules -common/ $(OODIR) /share/ config/ soffice. cfg/modules/ dbreport
414 mv $(PKGDIR)
415 $(PKGDIR)
416 +else
417 + rm -rf $(PKGDIR)
418 +endif
Is this possibly a separate bug, as apparently it's not related to
an upstream bug fix; it would be nice to have some clarification.
Maybe this is covered in the SRU policy under "it is also acceptable to upload new microreleases with many bug fixes without individual Launchpad bugs for each of them", but I considered such 'many bug fixes' to be related to the upstream bugs, and this change is on packaging.
2)
d/changelog
20 + - switch OOO_BASE_ ARCHS/OOO_ REPORTBUILDER_ ARCHS variables mysql,postgresq l,firebird} anyway
21 + (build base on all architectures instead of Java architectures
22 + and report-builder-* only on archs where Java is enabled, not on
23 + "all base archs"); there's -sdbc-{
d/rules
356 -OOO_BASE_ARCHS := $(OOO_JAVA_ARCHS) archs,OOO_ BASE_ARCHS) ) DER_ARCHS := $(OOO_BASE_ARCHS) DER_ARCHS := $(OOO_JAVA_ARCHS)
357 +OOO_BASE_ARCHS := $(OOO_ARCHS)
358 $(eval $(call gen_no_
359 -OOO_REPORTBUIL
360 +OOO_REPORTBUIL
Similarly; it would be nice to have some clarification,
and perhaps explain the trailing note about '-sdbc-* anyway'.
3)
d/changelog
24 + - don't try to build Jar_{OOoRunner, test,Connectivi tyTools} if Java is
25 + disabled as an extra safety net (instead of checking for junit only)
Similary, it would be nice to have some clarification.
d/rules
399 -ifeq "$(BUILD_ TEST_PACKAGES) " "y" TEST_PACKAGES) " "y" PATH=$( BUILD_LD_ LIBRARY_ PATH) ARCH_FLAGS= $(ARCH_ FLAGS) TMP=`mktemp -q -d` $(MAKE) Jar_{OOoRunner,tes t,ConnectivityT ools}
400 - ifeq "$(ENABLE_JUNIT4)" "y"
401 +ifeq "$(ENABLE_JAVA)" "y"
402 + ifeq "$(BUILD_
403 + ifeq "$(ENABLE_JUNIT4)" "y"
404 PATH=$(BUILD_PATH) LD_LIBRARY_
405 + endif
4)
d/changelog
26 + - re-enable cmis; bump libcmis build-dep to >= 0.6.1
Similarly; it would be nice to have some clarification.
And also the reason to enable it in an SRU? Maybe some relation to
an upstream bug or ubuntu bug? Any changes at behavior/runtime?
(apparently just an internal build dep; why is this needed, and
isn't the archive version (older) sufficient?)
d/rules
343 @@ -356,7 +356,7 @@ /usr/share/ hunspell /usr/share/ hyphen /usr/share/ mythes
344 DICT_DIR=
345 HYPH_DIR=
346 THES_DIR=
347 -ENABLE_LIBCMIS=n
348 +ENABLE_LIBCMIS=y
386 @@ -1423,7 +1429,7 @@
387 endif
388 ifeq "$(ENABLE_LIBCMIS)" "y"
389 ifneq (,$(filter libcmis, $(SYSTEM_STUFF)))
390 - BUILD_DEPS += , libcmis-dev (>= 0.5.2~), libcmis-dev (<< 0.6~)
391 + BUILD_DEPS += , libcmis-dev (>= 0.6.1~), libcmis-dev (<< 0.7~)
392 endif
393 else
394 CONFIGURE_FLAGS += --disable-libcmis
5)
27 + - build against libcurl4- openssl- dev instead of gnutls
334 @@ -271,7 +271,7 @@ gnutls openssl
335 USE_GSTREAMER=y
336 endif
337 ENABLE_CURL=y
338 -CURL_SECTYPE=
339 +CURL_SECTYPE=
Similarly; it would be nice to have some clarification on its reason,
and potential changes to runtime behavior or functionality.
This is a significant change.
6)
30 + * debian/ control. ure.in:
31 + - make ure-java actutally Architecture: %OOO_JAVA_ARCHS%
Similarly, it would be nice to have some clarification.