Bug 12779 - [efl] update efl from 1.7.9 to 1.7.10
Summary: [efl] update efl from 1.7.9 to 1.7.10
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: 4
Hardware: All Linux
Priority: Normal normal
Target Milestone: ---
Assignee: QA Team
QA Contact:
URL:
Whiteboard: MGA4-64-OK MGA4-32-OK advisory
Keywords: validated_update
Depends on:
Blocks:
 
Reported: 2014-02-15 11:55 CET by trem
Modified: 2014-03-03 21:45 CET (History)
5 users (show)

See Also:
Source RPM:
CVE:
Status comment:


Attachments

Description trem 2014-02-15 11:55:29 CET
Description of problem:
We use efl (eina, ecore, ...) 1.7.9, but there is new version of
efl 1.7.10. This new version is only a bugfix version, so I think
that it should be a good idea to use it.

Version-Release number of selected component (if applicable):
list of efl packages (1.7.9) :
- eina
- eet
- evas
- ecore
- embryo
- edje
- efreet
- e_dbus
- eeze
- expedite
- evas_generic_loaders
- eio
- emotion
- ethumb
- elementary

How reproducible:


Steps to Reproduce:
1.
2.
3.


Reproducible: 

Steps to Reproduce:
trem 2014-02-15 11:56:00 CET

Assignee: bugsquad => tremyfr

Comment 1 trem 2014-02-17 17:23:44 CET
I've submited this new version on core/updates_testing,
and tested it on mga4. 

May you can move them to core/updates please ?
trem 2014-02-17 17:24:19 CET

Assignee: tremyfr => qa-bugs

Comment 2 Manuel Hiebel 2014-02-17 17:27:56 CET
Hello please read https://wiki.mageia.org/en/Updates_policy#QA_Team_.28How_to_check_and_validate_an_update.29
QA needs and advisory, list of srpm, and list of packages for both archs. Thanks for them.
Comment 3 trem 2014-02-17 17:44:55 CET
List of SRPM :
- eina-1.7.10-1.mga5.src.rpm
- eet-1.7.10-1.mga5.src.rpm
- evas-1.7.10-1.mga5.src.rpm
- ecore-1.7.10-1.mga5.src.rpm
- embryo-1.7.10-1.mga5.src.rpm
- edje-1.7.10-1.mga5.src.rpm
- efreet-1.7.10-1.mga5.src.rpm
- e_dbus-1.7.10-1.mga5.src.rpm
- eeze-1.7.10-1.mga5.src.rpm
- expedite-1.7.10-1.mga5.src.rpm
- evas_generic_loaders-1.8.1-1.mga5.src.rpm
- eio-1.7.10-1.mga5.src.rpm
- emotion-1.7.10-1.mga5.src.rpm
- ethumb-1.7.10-1.mga5.src.rpm
- elementary-1.7.10-1.mga5.src.rpm
Comment 4 claire robinson 2014-02-17 17:54:22 CET
Hi trem

I think Manuel meant this one..
https://wiki.mageia.org/en/Updates_policy#Maintainer_.28or_any_interested_packager.29

We need a list of srpm's and rpm's and an advisory please, there is a link to an example on the wiki page. 

The advisory is sent to updates announce ML and added to http://advisories.mageia.org so should be in natural language rather than changelog style as far as possible. Just basically giving the reasons for the update and what has changed. 

The srpm's are used to know what to push and the rpm's are used to know what to test, so we do need both please.

Thanks
Comment 5 trem 2014-02-17 18:01:11 CET
List of RPM for i586 :
eina-debuginfo-1.7.10-1.mga5.i586.rpm
libeina1-1.7.10-1.mga5.i586.rpm
libeina-devel-1.7.10-1.mga5.i586.rpm
eet-1.7.10-1.mga5.i586.rpm
eet-debuginfo-1.7.10-1.mga5.i586.rpm
libeet1-1.7.10-1.mga5.i586.rpm
libeet-devel-1.7.10-1.mga5.i586.rpm
evas-1.7.10-1.mga5.i586.rpm
evas-debuginfo-1.7.10-1.mga5.i586.rpm
libevas1-1.7.10-1.mga5.i586.rpm
libevas-devel-1.7.10-1.mga5.i586.rpm
ecore-1.7.10-1.mga5.i586.rpm
ecore-debuginfo-1.7.10-1.mga5.i586.rpm
libecore1-1.7.10-1.mga5.i586.rpm
libecore-devel-1.7.10-1.mga5.i586.rpm
embryo-1.7.10-1.mga5.i586.rpm
embryo-debuginfo-1.7.10-1.mga5.i586.rpm
libembryo1-1.7.10-1.mga5.i586.rpm
libembryo-devel-1.7.10-1.mga5.i586.rpm
edje-1.7.10-1.mga5.i586.rpm
edje-debuginfo-1.7.10-1.mga5.i586.rpm
libedje1-1.7.10-1.mga5.i586.rpm
libedje-devel-1.7.10-1.mga5.i586.rpm
efreet-1.7.10-1.mga5.i586.rpm
efreet-debuginfo-1.7.10-1.mga5.i586.rpm
libefreet1-1.7.10-1.mga5.i586.rpm
libefreet-devel-1.7.10-1.mga5.i586.rpm
e_dbus-1.7.10-1.mga5.i586.rpm
libe_dbus1-1.7.10-1.mga5.i586.rpm
libe_dbus-devel-1.7.10-1.mga5.i586.rpm
eeze-1.7.10-1.mga5.i586.rpm
eeze-debuginfo-1.7.10-1.mga5.i586.rpm
libeeze1-1.7.10-1.mga5.i586.rpm
libeeze-devel-1.7.10-1.mga5.i586.rpm
expedite-1.7.10-1.mga5.i586.rpm
expedite-debuginfo-1.7.10-1.mga5.i586.rpm
evas_generic_loaders-1.7.10-1.mga5.i586.rpm
evas_generic_loaders-debuginfo-1.7.10-1.mga5.i586.rpm
eio-debuginfo-1.7.10-1.mga5.i586.rpm
libeio1-1.7.10-1.mga5.i586.rpm
libeio-devel-1.7.10-1.mga5.i586.rpm
emotion-1.7.10-1.mga5.i586.rpm
emotion-debuginfo-1.7.10-1.mga5.i586.rpm
libemotion1-1.7.10-1.mga5.i586.rpm
libemotion-devel-1.7.10-1.mga5.i586.rpm
ethumb-1.7.10-1.mga5.i586.rpm
ethumb-debuginfo-1.7.10-1.mga5.i586.rpm
libethumb1-1.7.10-1.mga5.i586.rpm
libethumb-devel-1.7.10-1.mga5.i586.rpm
elementary-1.7.10-1.mga5.i586.rpm
elementary-debuginfo-1.7.10-1.mga5.i586.rpm
libelementary1-1.7.10-1.mga5.i586.rpm
libelementary-devel-1.7.10-1.mga5.i586.rpm
Comment 6 trem 2014-02-17 18:23:16 CET
I have uploaded a patched/updated package for Mageia 4.


Suggested advisory:
========================

Use lastest stable efl version (1.7.10).

This is a bugfix version.

Updated packages in core/updates_testing:
========================
lib(64)eina1
lib(64)eina-devel
eet
lib(64)eet1
lib(64)eet-devel
evas
lib(64)evas1
lib(64)evas-devel
ecore
lib(64)ecore1
lib(64)ecore-devel
embryo
lib(64)embryo1
lib(64)embryo-devel
edje
lib(64)edje1
lib(64)edje-devel
efreet
lib(64)efreet1
lib(64)efreet-devel
e_dbus
lib(64)e_dbus1
lib(64)e_dbus-devel
eeze
lib(64)eeze1
lib(64)eeze-devel
expedite
evas_generic_loaders
lib(64)eio1
lib(64)eio-devel
emotion
lib(64)emotion1
lib(64)emotion-devel
ethumb
lib(64)ethumb1
lib(64)ethumb-devel
elementary
lib(64)elementary1
lib(64)elementary-devel

Source RPMs:
eina-1.7.10-1.mga4.src.rpm
eet-1.7.10-1.mga4.src.rpm
evas-1.7.10-1.mga4.src.rpm
ecore-1.7.10-1.mga4.src.rpm
embryo-1.7.10-1.mga4.src.rpm
edje-1.7.10-1.mga4.src.rpm
efreet-1.7.10-1.mga4.src.rpm
e_dbus-1.7.10-1.mga4.src.rpm
eeze-1.7.10-1.mga4.src.rpm
expedite-1.7.10-1.mga4.src.rpm
evas_generic_loaders-1.8.1-1.mga4.src.rpm
eio-1.7.10-1.mga4.src.rpm
emotion-1.7.10-1.mga4.src.rpm
ethumb-1.7.10-1.mga4.src.rpm
elementary-1.7.10-1.mga4.src.rpm
Comment 7 Lewis Smith 2014-02-18 20:53:26 CET
Tested MGA4 on real 64-bit hardware

Installed all pkgs cited in comment 6 from Updates Testing, version 1.7.10-1 just by selecting the base E17 meta pkg. (I had not previously installed E17 desktop).

I have used it a few hours & run through all applications in the menus & Dock. Fast. Oh, there are oddities (the most serious of which are:
- Both Install/Remove software and MageiaControlCentre do *not* work from the menu, but do from command line, asking for root authorisation. This is the same authorisation problem that affected all other desktops pre Mag4 release. Cannot find the fixed bug, which related to libc I think.
- If [?] there is a SysTray, applications which lodge themselves there are not visible. So they cannot be activated nor ended. This principally affects a few KDE applications which do that.
- the Accessories sub-menu disappeared after a time, but it was all there initially. This would really matter in real life).

I imagine it works as expected in its current state. Most importantly, most applications & E17 things work, exceptions noted above. There are other minor things like a few apps yielding an error msg on ending: ClawsMail, Expedite, AcquireImages, Digikam, KDEenlive, MateTerminal. SysTray invisible: Amarok, KWalletManager, Nepomuk, NetApplet. The "!" Dock icon Connection Manager Settings wants 'EConnMan'.

I guess all these things [E17 generally] escaped the Mag4 pre-release testing. Can someone more authoritative say whether this is OK to push, or whether the various faults should be bugged & fixed first? I await feedback on the issued version to see whether the problems were already there.

CC: (none) => lewyssmith

Comment 8 Lewis Smith 2014-02-19 09:51:13 CET
From Comment 7:
> - the Accessories sub-menu disappeared after a time, but it was all there initially.
It came back after a re-boot (re-login was not enough).

All its applications work OK, sole doubt being those KDE ones that put themselves in SysTray where they are not seen: Klipper, KNotes, KAlarm at least.
Comment 9 claire robinson 2014-02-21 16:12:08 CET
Are you happy to add mga4-64-ok Lewis?
Comment 10 Lewis Smith 2014-02-22 10:06:18 CET
(In reply to claire robinson from comment #9)
> Are you happy to add mga4-64-ok Lewis?
Not really. From comment 7:
"Can someone more authoritative say whether this is OK to push, or whether the various faults should be bugged & fixed first? I await feedback on the issued version to see whether the problems were already there."
This still needs 32-bit testing anyway before pushing, & anyone can add MGA4-64-OK at will if my results are deemed par for the course. Whoever does 32-bit can help by checking the various problems I noted *before* updating it. I was hoping Trem would assess my comments.
Can we leave it that MGA4-64-OK is ready to cite once the 32-bit pkg & update has been assessed? Like this, nothing is held back.
Comment 11 trem 2014-02-23 10:39:42 CET
Hi all,

I agree that e17 is far of being "bugless", but if this new version 1.7.10
don't have any regression (from 1.7.9). I think we should use it, it's a
bugfix version.

My plan is :
- move to efl to 1.7.10
- move e to 0.17.6
- bump e_modules to fix #12610

What do you think about it ?
Comment 12 Lewis Smith 2014-02-23 11:22:38 CET
(In reply to trem from comment #11)
Thanks for replying.
> I agree that e17 is far of being "bugless"
If it is recognised as such, we can go ahead as proposed.

> but if this new version 1.7.10
> don't have any regression (from 1.7.9). I think we should use it, it's a
> bugfix version.
It was just the question of possible regression that caused my restraint. Given the lack of feedback about 1.7.9, and the lack of bugs on it, I am happy to accept that 1.7.10 is at least no worse, & let it go.

> My plan is :
> - move to efl to 1.7.10
Agree. Fixes this bug.

> - move e to 0.17.6
Know nothing about this; your opinion is the best one, so implicitly yes.

> - bump e_modules to fix #12610
By all means; but under *that* bug, not this one. (BTW I wondered why it was necessary to link 0.7.14 to 0.7.15 rather than renaming them).
If you want to do these last 2 things, should it not be done as resolution of #12610 (which seems easy to specifically test; and to which I have added your comment); and not be mixed up with *this* bug? Otherwise this one will never get closed.

We still need *this one* testing on i586, which I cannot do. I am OK-ing it as it stands for x64.

Whiteboard: (none) => MGA4-64-OK

Comment 13 Oleg Kozlov 2014-02-23 12:58:57 CET
(In reply to trem from comment #11)
> Hi all,
> 
> I agree that e17 is far of being "bugless", but if this new version 1.7.10
> don't have any regression (from 1.7.9). I think we should use it, it's a
> bugfix version.
> 
> My plan is :
> - move to efl to 1.7.10
> - move e to 0.17.6
> - bump e_modules to fix #12610
> 
> What do you think about it ?

Sound good.

CC: (none) => xxblx.duke

Comment 14 Oleg Kozlov 2014-02-23 13:06:46 CET
(In reply to Lewis Smith from comment #7)
> - Both Install/Remove software and MageiaControlCentre do *not* work from
> the menu, but do from command line, asking for root authorisation. This is
> the same authorisation problem that affected all other desktops pre Mag4
> release. Cannot find the fixed bug, which related to libc I think.
It's not a bug.
You just need add PolicyKit agent to autostart. Problem will be solved. In LXDE you will have same problem that can be solved with similar action - adding PolicyKit agent to autostart.

Updated Mageia tools (include MCC, mageia online [updater], rpmdrake, etc...) need PolicyKit agent for super user authorization request.
Comment 15 Dave Hodgins 2014-03-03 13:27:02 CET
Testing complete on Mageia 4 and advisory added to svn.

Someone from the sysadmin team please push 12779.adv to updates.

Keywords: (none) => validated_update
Whiteboard: MGA4-64-OK => MGA4-64-OK MGA4-32-OK advisory
CC: (none) => davidwhodgins, sysadmin-bugs

Comment 16 Thomas Backlund 2014-03-03 21:45:38 CET
Update pushed:
http://advisories.mageia.org/MGAA-2014-0073.html

Status: NEW => RESOLVED
CC: (none) => tmb
Resolution: (none) => FIXED


Note You need to log in before you can comment on or make changes to this bug.