Description of problem: scim-rawcode package, that provides the RAW Unicode input method, is not automatically installed when non-english locales are selected during installation. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. Install Fedora rawhide version using Anaconda 2. At the Package Selection=>Languages=>Select any non-english locale that uses a different script (e.g Bengali Language Support) Actual results: scim-rawcode package is not installed along with the other scim-* packages used for language support Expected results: scim-rawcode package should be installed, if any language using a script other than English is being selected for installation. Additional info: This package is required to be installed because, often rarely used characters are not put on the standard keymaps for a language due to resource constraints and are required to be input using the RAWCODE.
Would it not be better to put those rarely used characters in the input maps? Otherwise I feel some GUI would be better for this. Normal users should not have to use rawcode, and it is not that hard to install scim-rawcode if it is needed. (gtk also supports input of unicode with Ctrl-Shift-u.)
Jens, it is not really feasible to put all these characters on the keymap. e.g. old currency symbols that were used about a century back. These are not used in the regular writing but is used when documenting about history. Same goes for the character (U+09FA ishwar). It is used to write the name of a deceased person.
Runa, have you tried Ctrl-Shift-u for gtk apps?
Yes Jens. The key combination works on gedit. Is there a similar input combination for QT applications?
I don't know of any way to do it under Qt but then I don't use KDE much. Can you provide a list of languages that might require direct unicode input or fixes? I still wonder if ordinary users would use that though? it seems to me some kind of input table would be better for that: either it should be possible to extend the current maps with some longer input strings (say using pre-edit) to allow those seldom used characters to be input, or some kind of GUI tool like scim-input-pad could do it. Asking normal users to remember and enter Unicode points directly does not seem right to me.
Jens, I do not know of any other languages that might require special characters through raw unicode input. However, we would like this to be activated for bn-IN locale in Fedora as we often need to test things out and having the option of raw unicode input is extremely helpful. Adding Jamil as a cc for this bug, in case bn locale also needs it. If required we can change the bug summary to indicate only our language. Thanks
But how many (In reply to comment #6) > I do not know of any other languages that might require special characters > through raw unicode input. However, we would like this to be activated for bn-IN > locale in Fedora as we often need to test things out and having the option of > raw unicode input is extremely helpful. It could be done, but how many people would actually be using it? I don't feel it is right to do it if it is just for a handful. :) > Adding Jamil as a cc for this bug, in case bn locale also needs it. bn?
> > I do not know of any other languages that might require special characters > > through raw unicode input. However, we would like this to be activated for bn-IN > > locale in Fedora as we often need to test things out and having the option of > > raw unicode input is extremely helpful. > > It could be done, but how many people would actually be using it? > I don't feel it is right to do it if it is just for a handful. :) Irrespective of numbers, which IMHO is never clear, the larger objective would be to focus on the fact if there are any hindrances to the usability in the locale. The option for raw unicode input was always available and its sudden exclusion would lead to confusion. > > > Adding Jamil as a cc for this bug, in case bn locale also needs it. > > bn? bn =Bengali.
(In reply to comment #8) > Irrespective of numbers, which IMHO is never clear, the larger objective would > be to focus on the fact if there are any hindrances to the usability in the > locale. Hmm, it is a little hard for me to judge. I imagine many users don't need it. Maybe you could describe some common use cases, and we can consider them. IMHO it would be better to fix the UI than just provide rawcode as a bandaid for that. > The option for raw unicode input was always available and its sudden > exclusion would lead to confusion. Well rawcode was already disabled by default since FC6, so personally I don't think asking users to install it is so big a problem. Though I agree it would have been good to install it for people upgrading from F7 or earlier. Perhaps we should do an update for F7 to make scim provide "scim-rawcode". > > > Adding Jamil as a cc for this bug, in case bn locale also needs it. bn_BD?
Actually thinking more, it seems to me this points to a larger issue that it is harder to get IMEngines up and running for us than with some other OSes. In that sense I agree with Runa, that the IMEngines should just be there, available and ready to be loaded. I think we need to address this usability issue: ie how easy it is to add and have a new IMEs up and running for users. They shouldn't have to restart their desktop, etc. But it may take some time before we're ready to install many IMEs by default (well we already do it in Fedora Live in fact, so this would be a good chance to fix this). (It is also mentioned vaguely on http://fedoraproject.org/wiki/I18N/SCIM/Wishlist.) Actually the main original motivation for not installing rawcode by default anymore was to stop scim from starting when no other IMEs were installed.
(In reply to comment #9) > (In reply to comment #8) > > Irrespective of numbers, which IMHO is never clear, the larger objective would > > be to focus on the fact if there are any hindrances to the usability in the > > locale. > > Hmm, it is a little hard for me to judge. I imagine many users don't need it. > Maybe you could describe some common use cases, and we can consider them. > IMHO it would be better to fix the UI than just provide rawcode as a bandaid > for that. > Not all users would require it and some would miss the option sorely if its absent. Fixing a keymap or introducing a new one is always an open option, but really depends on a greater consensus. However, making available the existing resources always helps in further enhancements and usability support. > > The option for raw unicode input was always available and its sudden > > exclusion would lead to confusion. > > Well rawcode was already disabled by default since FC6, so personally > I don't think asking users to install it is so big a problem. > Though I agree it would have been good to install it for people upgrading > from F7 or earlier. Perhaps we should do an update for F7 > to make scim provide "scim-rawcode". > > > > > Adding Jamil as a cc for this bug, in case bn locale also needs it. > > bn_BD? > locale is bn_BD, however the language is bn (bn and bn_IN have similar distinction as pt and pt_BR) http://translate.fedoraproject.org/teams/bn http://translate.fedoraproject.org/teams/bn_IN
(In reply to comment #10) > Actually thinking more, it seems to me this points to a larger issue that it is > harder to get IMEngines up and running for us than with some other OSes. > > In that sense I agree with Runa, that the IMEngines should just be there, > available and ready to be loaded. I think we need to address this > usability issue: ie how easy it is to add and have a new IMEs up and running > for users. They shouldn't have to restart their desktop, etc. > But it may take some time before we're ready to install many IMEs > by default (well we already do it in Fedora Live in fact, so this would be a good > chance to fix this). (It is also mentioned vaguely on > http://fedoraproject.org/wiki/I18N/SCIM/Wishlist.) > > Actually the main original motivation for not installing rawcode by default > anymore was to stop scim from starting when no other IMEs were installed. Perhaps, there could be a conditional factor put in, like including this package only when language support for languages using non-latin locales was selected during installation. This would ensure another IMEngine to be present and avoid the scenario that was the initial motivation to do away with it in the first place.
(In reply to comment #11) > (In reply to comment #9) > > (In reply to comment #8) > > > Irrespective of numbers, which IMHO is never clear, the larger objective would > > > be to focus on the fact if there are any hindrances to the usability in the > > > locale. > > > > Hmm, it is a little hard for me to judge. I imagine many users don't need it. > > Maybe you could describe some common use cases, and we can consider them. > > IMHO it would be better to fix the UI than just provide rawcode as a bandaid > > for that. > > > > Not all users would require it and some would miss the option sorely if its > absent. Fixing a keymap or introducing a new one is always an open option, but > really depends on a greater consensus. However, making available the existing > resources always helps in further enhancements and usability support. > Yes, I agree with Runa. It will be better to use the scim-rawcode option without changing the keyboard layout, always. As some keys are not used widely. > > > > The option for raw unicode input was always available and its sudden > > > exclusion would lead to confusion. > > > > Well rawcode was already disabled by default since FC6, so personally > > I don't think asking users to install it is so big a problem. > > Though I agree it would have been good to install it for people upgrading > > from F7 or earlier. Perhaps we should do an update for F7 > > to make scim provide "scim-rawcode". > > > > > > > Adding Jamil as a cc for this bug, in case bn locale also needs it. > > > > bn_BD? > > > locale is bn_BD, however the language is bn (bn and bn_IN have similar > distinction as pt and pt_BR) > > http://translate.fedoraproject.org/teams/bn > http://translate.fedoraproject.org/teams/bn_IN > > > >
Changing version to '9' as part of upcoming Fedora 9 GA. More information and reason for this action is here: http://fedoraproject.org/wiki/BugZappers/HouseKeeping
scim-rawhide should be installed by default again in f10 - I just committed such a change to comps-f10.