From: Karl E. <ke...@gn...> - 2003-10-25 03:08:22
|
I will forward your mail to our technical discussion list. -ke- -=-=-=-=-=-=-=-=-=-=-=-=-=- cut here -=-=-=-=-=-=-=-=-=-=-=-=-=- Message-ID: <3F9...@tu...> Date: Thu, 23 Oct 2003 13:43:26 +0200 From: Christian Stimming <sti...@tu...> To: gnu...@pr... Subject: Solution for ambiguous msgids planned? Dear gettext developers, recently in some bigger software project (www.gnucash.org) we again stumbled over the well-known problem of ambiguous msgids as frequently seen in GUIs. In the gettext manual this problem is described in section 10.2.6 http://www.gnu.org/manual/gettext/html_chapter/gettext_10.html#SEC151 "How to use gettext in GUI programs", and a possible workaround is outlined as well by suggesting to define a de-mangling (prefix-removing) wrapper sgettext() inside the application. However, this workaround only works if the actual gettext calls happen inside the application developer's code. But in recent GUI toolkits and User Interface builders this is not the case. Instead, in GUIs developed by e.g. the User Interface builder Glade, the user interface is stored in an XML file, and at runtime the library LibGlade http://www.daa.com.au/~james/software/libglade/ will load that XML file and create the GUI. This means the gettext() translations of the msgids found in the Glade XML file will be done inside libglade's code. In this case it is impossible for an application developer to apply the demangling workaround as described in the gettext manual. There was some discussion on gnome-i18n http://mail.gnome.org/archives/gnome-i18n/2001-April/msg00058.html but the results were never implemented http://mail.gnome.org/archives/gnome-i18n/2001-September/msg00108.html But the problem persists. The application developer in theory could go talk to the libglade developers, trying to convince them to include the described demangling workaround in the libglade code. Nevertheless this will keep on showing up, not to mention that this would mean implementing the same function over and over again. Therefore I would like to ask you, the gettext developers: Are there plans to include such a prefix_gettext() function into the gettext library? Are there archives of any discussion around this issue (I didn't find any)? If there was a decision not to do this, what were the reasons? This issue *has* to come up every now and then, and I couldn't believe there still isn't a durable solution around. Best regards, Chistian Stimming -=-=-=-=-=-=-=-=-=-=-=-=-=- cut here -=-=-=-=-=-=-=-=-=-=-=-=-=- -- | ,__o | _-\_<, http://www.gnu.franken.de/ke/ | (*)/'(*) |