Description of problem: Mageia 5 provides Gramps under version 4.1.1 New users coming from Ubuntu used Gramps under version 4.1.3. New data structures of 4.1.3 are not reverible to 4.1.1. http://www.mageialinux-online.org/forum/topic-20674+gramps-4-1-3.php Version-Release number of selected component (if applicable): Today, Gramps is delivered under version 4.2.0. Should this version be delivered by Mageia ? Thanks for the fama and accuracy of Mageia. How reproducible: always Steps to Reproduce: 1. 2. 3. Reproducible: Steps to Reproduce:
Gramps has no maintainer, so CC'ing most committers.
CC: (none) => jani.valimaa, lmenut, luigiwalser, mageia, marja11, olav, pterjan, shlomif, thierry.vignaud, zen25000Summary: Gramps provided by Mageia 5 is not up-to-date => Mga5 Gramps causes problems for users coming from Ubuntu, because the data structures of their vs 4.1.3 cannot be reverted to our vs 4.1.1
(In reply to Marja van Waes from comment #1) > Gramps has no maintainer, so CC'ing most committers. 4.1.3 builds ok in mga5, but is the above enough reason for update in stable release?
(In reply to Barry Jackson from comment #2) > 4.1.3 builds ok in mga5, but is the above enough reason for update in stable > release? IMO yes. Though upstream should have increased the minor number when breaking database compatibility IMO...
(In reply to Rémi Verschelde from comment #3) > (In reply to Barry Jackson from comment #2) > > 4.1.3 builds ok in mga5, but is the above enough reason for update in stable > > release? > > IMO yes. Though upstream should have increased the minor number when > breaking database compatibility IMO... OK, 4.1.3 pushed to Cauldron and 5/core/updates_testing. ######################################## Update Advisory ######################################## gramps-4.1.3-1.mga5.noarch.rpm has been submitted to 5/core/updates_testing ######################################## Affected rpms in 5/core/updates_testing: gramps-4.1.3-1.mga5.noarch.rpm ######################################## Source rpm: gramps-4.1.3-1.mga5.src.rpm ######################################## Testing It may be best to ask one of the reporters to test this as old data is required. I have tested briefly using an old data file which I had from an earlier version and it was converted successfully (AFAICT) by this version. There are warnings to make backups before attempting the data update.
QA Contact: (none) => qa-bugs
You can add an OK for i586/x86_64 if you like Barry. We're being pragmatic about updating things like this since mga5 took so much longer to release than normal. eplip, as reporter are you able to confirm this corrects the issue for you please. Thanks
And it's a leaf package.
(In reply to Barry Jackson from comment #4) > There are warnings to make backups before attempting the data update. There were no such warnings on my mga-5-64 system. The data base loaded normally (as if the format had not changed). Gramps seems to be functioning OK. On any previous update to the data base format I did see a warning before the data base was updated. After downgrading to gramps-4.1.1-2.mga5, the "updated" data base loaded normally and Gramps seems to be functioning normally. Are we sure that this update is really necessary or is it perhaps not achieving what is intended?
The "rpm package" reference in this bug report is to gramps-4.2 which is not what seems to be asked for in comment 0. Has anyone actually tried to load an Ubuntu 4.1.3 data base in gramps-4.1.1-2.mga5. If so what error message is reported? I don't think that the data base format has changed between 4.1.1 and 4.1.3. (Version 4.2 is an entirely different matter.)
The RPM reference to gramps 4.2 was just an indication from the bug reporter, I've fixed it. There is no plan to upgrade to 4.2 for the time being. I asked the original reporter on the French forum to test the update candidate.
Source RPM: http://sourceforge.net/projects/gramps/files/Stable/4.2.0/gramps-4.2.0.tar.gz/download => gramps-4.1.1-2.mga5
The original poster in the forum thread does not give the basis for his claim that the data base format has changed between 4.1.1 and 4.1.3. He does not say that he tried and failed to load his Ubuntu data base in Mageia, only that the Ubuntu version of gramps is different from Mageia's. If I'm correct in thinking that the data base format has not changed between 4.1.1 and 4.1.3, there would appear to be no justification for making an exception to our updates policy.
He does say "the databases of those two versions [4.1.1 and 4.1.3] are incompatible". But indeed he does not give an error message, and maybe that's just an assumption of his. So this bug may simply be INVALID, though if it is confirmed that 4.1.3 is a bugfix release in the 4.1.x branch, it might still be validated as such. Here are the changelogs, which do mention database-related fixes, but not necessarily breaking backwards-compatibility: https://gramps-project.org/2015/03/gramps-4-1-2-released/ https://gramps-project.org/2015/05/gramps-4-1-3-released/
well, if it works, since it's a bugfix release & a leaf package, it's good to go
(In reply to Thierry Vignaud from comment #12) > well, if it works, since it's a bugfix release & a leaf package, it's good > to go I agree. We should just mention in the advisory that it is a general bugfix update, and not specifically an update meant to fix comment 0, as the latter is unconfirmed. Speaking of advisory, the one you gave is not enough Barry. The description part of the advisory is intended at end users, and should give the motivations for the update in plain English. See the advisory example: https://wiki.mageia.org/en/Update_Advisory_Announcement_Example
(In reply to Rémi Verschelde from comment #13) > > I agree. We should just mention in the advisory that it is a general bugfix > update, and not specifically an update meant to fix comment 0, as the latter > is unconfirmed. > > Speaking of advisory, the one you gave is not enough Barry. The description > part of the advisory is intended at end users, and should give the > motivations for the update in plain English. > > See the advisory example: > https://wiki.mageia.org/en/Update_Advisory_Announcement_Example Oops sorry - yes, I will do another below later. I have an Ubuntu VM since yesterday which I was using for other tests so I will attempt to verify the OPs claim with a new data file.
CC: lmenut => (none)
The latest Ubuntu 15.04 LTS offers gramps-4.1.1 in the repo so this was AFAICT invalid, https://en.wikipedia.org/wiki/List_of_Ubuntu_releases ######################################## Update Advisory ######################################## gramps-4.1.3-1.mga5.noarch.rpm has been submitted to 5/core/updates_testing ######################################## Description: This is a bug fix update. ######################################## References: https://gramps-project.org/2015/03/gramps-4-1-2-released/ https://gramps-project.org/2015/05/gramps-4-1-3-released/ ######################################## Affected rpms in 5/core/updates_testing: gramps-4.1.3-1.mga5.noarch.rpm ######################################## Source rpm: gramps-4.1.3-1.mga5.src.rpm ########################################
As suggested in comment#5, I have validated this update, even though I have only tested on one arch. The advisory needs to be uploaded by a qa-committer and the package can then pushed to updates
Keywords: (none) => validated_updateCC: (none) => sysadmin-bugsAssignee: bugsquad => qa-bugsQA Contact: qa-bugs => (none)Whiteboard: (none) => MGA5-64-OK MGA5-32-OK
We need a better advisory explaining what's been fixed.
CC: (none) => davidwhodgins
(In reply to Dave Hodgins from comment #17) > We need a better advisory explaining what's been fixed. Really? Fixes and changes since 4.1.1: Error converting python2 utf-8 strings to python3 str when loading data from database Removing a parent place from a place leaves a dangling reference Error during checking the database Stubborn blank space in database wonât be removed, fix removing rows in flat list views Database upgrade fails if default media path is not set Error converting database after upgrade to Gramps 4.1.1 Error in a single place within the places section Entries from the add-or-choose selector of Place/Source/Media/Note cannot be dragged Enclosing places tab should work like other similar tabs, new place reference editor Association editor refuses dropped persons Error on opening twice an object from clipboard Incorrect spacing in export assistant file chooser New Event types are saved as a disordered list Always display main participants Place titles can now be generated on-the-fly by a place displayer, default is still to use the place title field GEDCOM import of embedded notes attached to media does not work Crash on Ancestry.com .ged import; consistent. Errors handling owner/submitter information in GEDCOM files. Only import researcher from GEDCOM or XML if the family tree was originally empty. GEDCOM export does not export media attached to citations. The fanchart view crashes if max generation is set to 1 away. Sidebar Filters do not match placetypes in new placeview, two new filter rules (HasTitle, HasData) Fix bug when family has no parents Fix bad handle in explanation note for unknown event Some labels now fit better on citations sidebar filter Views in Geography should not always use the last option set by the user Request for keyboard-controlled zoom on Geography view Error geography view â Displaying main menu Configure screen needs a file selector to select directory for âoffline modeâ files Filter panel on geography view displays improperly Detailed descendant report crashes, bibliography (citations) Canât disable box shadow in SVG descendant tree Descendant tree graphical report, syntax error in svg output Regression: Complete Individual report has partially-untranslated output Events Page in Narrative Report not working Gramps freeze after defining a report style with German cm values Various problems with docgen.TextDoc.add_media_object Report event attribute name is not translated Records Gramplet uses wrong text Closing detached gramplet causes python to crash Cannot reduce size of gramplets detached from a gramplet bar Gramplets donât fill window when detached from dashboard ImageMetadata doesnât show metadata typo on GLib call, used by an addon only Fix error setting gramplet tab label Check for active person in session log gramplet Spurious spaces in CLI List Family Trees, tab delimited output. Print statements changed to assemble the whole line before output. Gtk3 warning and custom undoableentry widget, see bugzilla_id 644927 Warnings: deprecated Gtk properties and errors loading theme icon. Fix: database manager dialog is inconsistent for older gtk+3 versions. Warnings: deprecated Gtk properties and errors loading theme icon. Fix: Error loading theme icon âgtk-applyâ gramps fails to start with gtk+-3.13.3 Places in data.gramps are not in the new Place hierarchy Upgrade the version of some dependencies for Mac OS and Windows OS. Keywords entry in gramps.desktop does not work âAvailable Gramps Updates for Addonsâ window not on top Some text not translatable in context menu fancharts Fix for Unit test date inflections in _datehandler.py, update for Ukrainian, Russian, Croatian Better support for Serbian and Turkish locales New translation: Icelandic Re-enable Turkish support after a major review. Updated translations: cs, de, eo, fi, fr, hr, hu, it, nb, nn, ru, sk, sr, sv, uk, zh_CN Fixes and changes since 4.1.2: Fix db upgrade failure GtkDialog mapped without a transient parent [Gedcom} SUBN and SUBM record handling [Gedcom] Import/export round trip causes lost information [Gedcom] Entering a witness to an event such as marriage might be ignored [Gedcom] Gramps canât import estim. date period exported by itself [Gedcom] 1/4 and 1/2 ANSEL characters not supported on importing ANSEL [Gedcom] Importing file containing multibyte UTF-8 characters fails [Gedcom] Import fails for ANSI file under python 3 [Gedcom] Failure importing ANSEL encoded gedcom file [Gedcom] Characters ignored on a Gedcom encoded ANSI (cp1252 West Europe, USA) [Gedcom] NameError in importer [Gedcom] Event address is lost on import, i.e. disconnected from event Crash on geneweb export with python3 GuiColorOption missing avail-changed event handler Bad generation of [timeline report] ODT files since 4.0.0 Fix bad handle in explanation note for unknown event Fix spurious generation of empty âAlternative Nameâ in place.merge() Support creating directories in various scenarios Attempting to add a bookmark causes an error Long series of âunhandled exceptionâ popup boxes while doing a check & repair Crash when trying to link existing place as an enclosing place using P0001 number HTML view fails to load Relationship Graph crashes Python3 needs new_subpixbuf not subpixbuf Regression: running gramps from crontab fails tag_map is not initialized Some labels now fit better on citations sidebar filter Event columns in web narrative are too narrow Problem by start program (launcher) Translation string missing in Not Related tool for help and close button Date format month/year is not well reported at editing time [in Italian] Fix unknown gender relationships handler for the french locale Fix a handle type bug on sidebar filter Tidy up About dialog Cleanup on some man files Convert some remaining unicode literals Fix mac menubar setting Enable python3 to run po/update_po.py Updated translations: cs, de, fr, is, nl Do you really want all that in an advisory?
It doesn't need to say any more than "The gramps package has been updated to version 4.1.3, which fixes several bugs. See the upstream release announcements for details." with the two URLs in Comment 11 as references.
(In reply to Barry Jackson from comment #18) > (In reply to Dave Hodgins from comment #17) > > We need a better advisory explaining what's been fixed. > > Really? > > Fixes and changes since 4.1.1: [...] > > Do you really want all that in an advisory? Please don't play dumb, that doesn't lead us anywhere. "This is a bug fix update" is clearly a bit short. Sometimes upstream gives a short list of the most important bug fixes in the release, and then it can be used in the advisory. When they don't and only list many bug fixes, then something like David's suggestion is already more user-friendly that "hey dude, this is an update".
Advisory committed to svn based on comment 19 and comment 11.
Whiteboard: MGA5-64-OK MGA5-32-OK => MGA5-64-OK MGA5-32-OK advisory
(In reply to Rémi Verschelde from comment #20) I don't see why 20 words are needed when 5 say exactly the same thing. The references were given in the Advisory and the version number is in the package name. I was keeping it concise following your lead in comment #13. However if that is what is required then I will try to remember for next time.
An update for this issue has been pushed to Mageia Updates repository. http://advisories.mageia.org/MGAA-2015-0085.html
Status: NEW => RESOLVEDResolution: (none) => FIXED