From: S. C. <ste...@gm...> - 2008-12-05 17:39:15
|
We had talked about releasing 3.0.4 on the weekend of Dec. 6-7. Benny reminded me of this on IRC today: (01:29:08 AM) bmcage: no, I just wanted to ask if you still plan release this weekend or will postpone (01:31:42 AM) EspenBe: release this weekend? then I should revise the Norwegian translations straight away (02:03:48 AM) bmcage: EspenBe: it did pass the mailing list that 6 december would be a good aim (02:04:37 AM) EspenBe: only had 2 strings to modify, so it is not a whole lot of work (02:04:46 AM) bmcage: it is the submission under the subject when 3.0.4? (02:04:54 AM) bmcage: from 25 november (02:05:35 AM) bmcage: ubuntu 9.04 freeze is 24/12 Are we good for this? I know we have some database fixes, but if someone else knows of a problem where we should wait, or has additional fixes pending, please speak up now. Otherwise, I'll put 3.0.4 together this weekend. Stéphane |
From: Luigi T. <lui...@ti...> - 2008-12-06 00:45:18
|
Stéphane Charette ha scritto: > We had talked about releasing 3.0.4 on the weekend of Dec. 6-7. > [..] > Are we good for this? I know we have some database fixes, but if > someone else knows of a problem where we should wait, or has > additional fixes pending, please speak up now. Otherwise, I'll put > 3.0.4 together this weekend. > Hi, I have some fixes and (small) suggestions. Even if some of them are not particularly urgent, I prefer to report them now, otherwise I'll forget them :) == Bugs: - GrampsDisplay.help is sometimes called with the wrong number of parameters (see ./plugins/FindDupes.py:149, ./ScratchPad.py:1255). I suppose that this issue is due to the behavior changes of GrampsDisplay.help (as it uses no more the gnome help but it refers to the website). Probably the first parameter (target) can be dropped and all the invocations update accordingly. I haven't checked if other calls to "help" are affected by this bug. - The translation of the following message: msgid " has 1 of 1 individual (100.00% complete)\n" gives a gettext validation error (test msgfmt with -c) with some languages (including for example Italian, French and Norwegian Bokmaal, but not German and Catalan): po/fr.po:8117: 'msgstr' is not a valid Python format string, unlike 'msgid'. Reason: In the directive number 1, the character ')' is not a valid conversion specifier. I think that it tries to interpret it as a special python converter (in fact, I cheated with the current Italian translation removing the "#,python-format" comment, and it complains no more, even if something else must be broken, I suppose...) == Minor bug: - the .desktop file doesn't conform to the specification (test it with desktop-file-validate); the patch is really trivial, if you want I can commit it. == Translation issues - If I understand correctly, the central person is sometimes referred to as "home person" and sometimes as "default person". I suggest to use always "default person" instead of "home person". - It'll be nice to differentiate (with a context?) "none" from src/gen/lib/date/ and the one in src/const.py; in italian they have different genders. The same for "Generation 1" and "Generation %n" from *Report.py and DefaultGramplet.py. - Are "Repo Reference Editor" and "Repository Reference Editor" the same string? The same for "Last Changed" and "Last modified" (and maybe "Last Change"). - CSV import/export: I don't think that the column headers should be translated; it makes more difficult data exchange between users with different locales, imho. == Other - I tried to write the class for displaying and parsing italian dates, "stealing" code from the Catalan one (probably the simplest one). It seems to works, and I'd like to commit it. But I'd also like to know if the extra code I see in other date handler is really important or not (i.e. the French one contains code to better handle the revolution calendar). Ok, that'd be enough :) Regards, -- Luigi |
From: Douglas S. B. <db...@cs...> - 2008-12-06 02:14:32
|
Luigi, Great bug reports and suggestions! Can you post each one to the bug tracker, and possibly even attach diffs/patches to each? These will get forgotten if we don't track them. I'd be the one to take a couple of these, but I'm not sure of the solution, so if you have fixes for any, that would be great. -Doug > Stéphane Charette ha scritto: >> We had talked about releasing 3.0.4 on the weekend of Dec. 6-7. >> [..] >> Are we good for this? I know we have some database fixes, but if >> someone else knows of a problem where we should wait, or has >> additional fixes pending, please speak up now. Otherwise, I'll put >> 3.0.4 together this weekend. >> > Hi, I have some fixes and (small) suggestions. Even if some of them are > not particularly urgent, I prefer to report them now, otherwise I'll > forget them :) > > == Bugs: > - GrampsDisplay.help is sometimes called with the wrong number of > parameters (see ./plugins/FindDupes.py:149, ./ScratchPad.py:1255). I > suppose that this issue is due to the behavior changes of > GrampsDisplay.help (as it uses no more the gnome help but it refers to > the website). Probably the first parameter (target) can be dropped and > all the invocations update accordingly. > I haven't checked if other calls to "help" are affected by this bug. > > - The translation of the following message: > msgid " has 1 of 1 individual (100.00% complete)\n" > gives a gettext validation error (test msgfmt with -c) with some > languages (including for example Italian, French and Norwegian Bokmaal, > but not German and Catalan): > po/fr.po:8117: 'msgstr' is not a valid Python format string, unlike > 'msgid'. Reason: In the directive number 1, the character ')' is not a > valid conversion specifier. > I think that it tries to interpret it as a special python converter (in > fact, I cheated with the current Italian translation removing the > "#,python-format" comment, and it complains no more, even if something > else must be broken, I suppose...) > > > == Minor bug: > - the .desktop file doesn't conform to the specification (test it with > desktop-file-validate); the patch is really trivial, if you want I can > commit it. > > == Translation issues > - If I understand correctly, the central person is sometimes referred to > as "home person" and sometimes as "default person". I suggest to use > always "default person" instead of "home person". > > - It'll be nice to differentiate (with a context?) "none" from > src/gen/lib/date/ and the one in src/const.py; in italian they have > different genders. > The same for "Generation 1" and "Generation %n" from *Report.py and > DefaultGramplet.py. > > - Are "Repo Reference Editor" and "Repository Reference Editor" the same > string? The same for "Last Changed" and "Last modified" (and maybe "Last > Change"). > > > - CSV import/export: I don't think that the column headers should be > translated; it makes more difficult data exchange between users with > different locales, imho. > > == Other > - I tried to write the class for displaying and parsing italian dates, > "stealing" code from the Catalan one (probably the simplest one). It > seems to works, and I'd like to commit it. But I'd > also like to know if the extra code I see in other date handler is > really important or not (i.e. the French one contains code to better > handle the revolution calendar). > > Ok, that'd be enough :) > > Regards, > -- > Luigi > > ------------------------------------------------------------------------------ > SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, > Nevada. > The future of the web can't happen without you. Join us at MIX09 to help > pave the way to the Next Web now. Learn more and register at > http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ > _______________________________________________ > Gramps-devel mailing list > Gra...@li... > https://lists.sourceforge.net/lists/listinfo/gramps-devel > |
From: Jérôme <rom...@ya...> - 2008-12-06 07:23:32
|
Luigi, > msgid " has 1 of 1 individual (100.00% complete)\n" > gives a gettext validation error (test msgfmt with -c) with some > languages (including for example Italian, French and Norwegian Bokmaal, > but not German and Catalan): > po/fr.po:8117: 'msgstr' is not a valid Python format string, unlike > 'msgid'. Reason: In the directive number 1, the character ')' is not a > valid conversion specifier. > I think that it tries to interpret it as a special python converter (in > fact, I cheated with the current Italian translation removing the > "#,python-format" comment, and it complains no more, even if something > else must be broken, I suppose...) French translation uses something which could be translated by : " 1 individual of 1 (complete to 100.00%) \ n" There is a comment on translation file, when gramps uses a python variable (#, python-format => %s, %d, ...) Does it mean we need to use a space after '%' for distinguishing percent and python variable ? > - If I understand correctly, the central person is sometimes referred to > as "home person" and sometimes as "default person". I suggest to use > always "default person" instead of "home person". 'Home person' is the central person on the database 'Central person' is the active person, often used on reports 'Default person' is ... Yes, seems there is a mix on translation strings. Sometimes 'default person' is 'home', sometimes 'active person' ! > i.e. the French one contains code to better handle the revolution calendar). "better handle", I don't know ... but there is a conversion mistake (Gregorian-> French revolution -> Gregorain) and negative years (before 1792 - first year on French revolution calendar) http://www.gramps-project.org/bugs/view.php?id=445 http://gramps.svn.sourceforge.net/viewvc/gramps/trunk/src/DateHandler/_Date_fr.py?r1=8939&r2=8944 I did not add changes on main DateParser because I was not very happy with this code... Also, there is some other issues with French calendar : http://www.gramps-project.org/bugs/view.php?id=2303 Note, looking at results on debug tool (DateParserDisplayTest), there is an problem using Gregorian/Julian Calendar and '|' on year with _Date_fr.py... Seems to be an error on DateDisplayer (present on branch but fixed on trunk !!!) maybe related to http://www.gramps-project.org/bugs/view.php?id=2260 Data are safe, parsing is OK, that's why I let code and will make changes on trunk only. Otherwise, there is changes to avoid a dot issue, because fr_CH uses German date format ! I do not know if there is a it_CH locale ? Jérôme Luigi Toscano a écrit : > Stéphane Charette ha scritto: >> We had talked about releasing 3.0.4 on the weekend of Dec. 6-7. >> [..] >> Are we good for this? I know we have some database fixes, but if >> someone else knows of a problem where we should wait, or has >> additional fixes pending, please speak up now. Otherwise, I'll put >> 3.0.4 together this weekend. >> > Hi, I have some fixes and (small) suggestions. Even if some of them are > not particularly urgent, I prefer to report them now, otherwise I'll > forget them :) > > == Bugs: > - GrampsDisplay.help is sometimes called with the wrong number of > parameters (see ./plugins/FindDupes.py:149, ./ScratchPad.py:1255). I > suppose that this issue is due to the behavior changes of > GrampsDisplay.help (as it uses no more the gnome help but it refers to > the website). Probably the first parameter (target) can be dropped and > all the invocations update accordingly. > I haven't checked if other calls to "help" are affected by this bug. > > - The translation of the following message: > msgid " has 1 of 1 individual (100.00% complete)\n" > gives a gettext validation error (test msgfmt with -c) with some > languages (including for example Italian, French and Norwegian Bokmaal, > but not German and Catalan): > po/fr.po:8117: 'msgstr' is not a valid Python format string, unlike > 'msgid'. Reason: In the directive number 1, the character ')' is not a > valid conversion specifier. > I think that it tries to interpret it as a special python converter (in > fact, I cheated with the current Italian translation removing the > "#,python-format" comment, and it complains no more, even if something > else must be broken, I suppose...) > > > == Minor bug: > - the .desktop file doesn't conform to the specification (test it with > desktop-file-validate); the patch is really trivial, if you want I can > commit it. > > == Translation issues > - If I understand correctly, the central person is sometimes referred to > as "home person" and sometimes as "default person". I suggest to use > always "default person" instead of "home person". > > - It'll be nice to differentiate (with a context?) "none" from > src/gen/lib/date/ and the one in src/const.py; in italian they have > different genders. > The same for "Generation 1" and "Generation %n" from *Report.py and > DefaultGramplet.py. > > - Are "Repo Reference Editor" and "Repository Reference Editor" the same > string? The same for "Last Changed" and "Last modified" (and maybe "Last > Change"). > > > - CSV import/export: I don't think that the column headers should be > translated; it makes more difficult data exchange between users with > different locales, imho. > > == Other > - I tried to write the class for displaying and parsing italian dates, > "stealing" code from the Catalan one (probably the simplest one). It > seems to works, and I'd like to commit it. But I'd > also like to know if the extra code I see in other date handler is > really important or not (i.e. the French one contains code to better > handle the revolution calendar). > > Ok, that'd be enough :) > > Regards, |
From: Benny M. <ben...@gm...> - 2008-12-06 07:57:44
|
Frederik, can you look at the first problem Luigi mentions: Grampsdisplay help? You have expertise in that regard Stephane, About the rest, It is not good to do string changes so close before a release. Let's do them after 3.0.4 and some less problematic only in trunk, like that all translators have time to catch up before 3.0.5. To fix the .desktop would be great however, Luigi can you commit and notify Stephane when finished? About the % in translation text, are there no default solutions to that, like escaping it or so (\%) ? Obviously it can cause a lot of problems with translators not knowing the tranlation may not be of the form %d, ... (I now understand why python 3.1 will deprecate % for string identifiers) Benny 2008/12/6 Luigi Toscano <lui...@ti...> > Stéphane Charette ha scritto: > > We had talked about releasing 3.0.4 on the weekend of Dec. 6-7. > > [..] > > Are we good for this? I know we have some database fixes, but if > > someone else knows of a problem where we should wait, or has > > additional fixes pending, please speak up now. Otherwise, I'll put > > 3.0.4 together this weekend. > > > Hi, I have some fixes and (small) suggestions. Even if some of them are > not particularly urgent, I prefer to report them now, otherwise I'll > forget them :) > > == Bugs: > - GrampsDisplay.help is sometimes called with the wrong number of > parameters (see ./plugins/FindDupes.py:149, ./ScratchPad.py:1255). I > suppose that this issue is due to the behavior changes of > GrampsDisplay.help (as it uses no more the gnome help but it refers to > the website). Probably the first parameter (target) can be dropped and > all the invocations update accordingly. > I haven't checked if other calls to "help" are affected by this bug. > > - The translation of the following message: > msgid " has 1 of 1 individual (100.00% complete)\n" > gives a gettext validation error (test msgfmt with -c) with some > languages (including for example Italian, French and Norwegian Bokmaal, > but not German and Catalan): > po/fr.po:8117: 'msgstr' is not a valid Python format string, unlike > 'msgid'. Reason: In the directive number 1, the character ')' is not a > valid conversion specifier. > I think that it tries to interpret it as a special python converter (in > fact, I cheated with the current Italian translation removing the > "#,python-format" comment, and it complains no more, even if something > else must be broken, I suppose...) > > > == Minor bug: > - the .desktop file doesn't conform to the specification (test it with > desktop-file-validate); the patch is really trivial, if you want I can > commit it. > > == Translation issues > - If I understand correctly, the central person is sometimes referred to > as "home person" and sometimes as "default person". I suggest to use > always "default person" instead of "home person". > > - It'll be nice to differentiate (with a context?) "none" from > src/gen/lib/date/ and the one in src/const.py; in italian they have > different genders. > The same for "Generation 1" and "Generation %n" from *Report.py and > DefaultGramplet.py. > > - Are "Repo Reference Editor" and "Repository Reference Editor" the same > string? The same for "Last Changed" and "Last modified" (and maybe "Last > Change"). > > > - CSV import/export: I don't think that the column headers should be > translated; it makes more difficult data exchange between users with > different locales, imho. > > == Other > - I tried to write the class for displaying and parsing italian dates, > "stealing" code from the Catalan one (probably the simplest one). It > seems to works, and I'd like to commit it. But I'd > also like to know if the extra code I see in other date handler is > really important or not (i.e. the French one contains code to better > handle the revolution calendar). > > Ok, that'd be enough :) > > Regards, > -- > Luigi > > > ------------------------------------------------------------------------------ > SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. > The future of the web can't happen without you. Join us at MIX09 to help > pave the way to the Next Web now. Learn more and register at > > http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ > _______________________________________________ > Gramps-devel mailing list > Gra...@li... > https://lists.sourceforge.net/lists/listinfo/gramps-devel > |
From: Douglas S. B. <db...@cs...> - 2008-12-06 15:57:02
|
> About the % in translation text, are there no default solutions to that, > like escaping it or so (\%) ? Obviously it can cause a lot of problems > with > translators not knowing the tranlation may not be of the form %d, ... > (I now understand why python 3.1 will deprecate % for string identifiers) > > Benny Isn't "%%" how you escape a percent sign in an interpolated string? Is the problem that sometimes it is interpolated, and sometimes not? -Doug |
From: Benny M. <ben...@gm...> - 2008-12-07 20:43:59
|
Luigi, can you also look into this? Benny 2008/12/6 Douglas S. Blank <db...@cs...> > > About the % in translation text, are there no default solutions to that, > > like escaping it or so (\%) ? Obviously it can cause a lot of problems > > with > > translators not knowing the tranlation may not be of the form %d, ... > > (I now understand why python 3.1 will deprecate % for string identifiers) > > > > Benny > > Isn't "%%" how you escape a percent sign in an interpolated string? Is the > problem that sometimes it is interpolated, and sometimes not? > > -Doug > > |