You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am not maintaining the project. I cloned the code from here http://repo.or.cz/w/qgit4/redivivus.git. The original author is Cristian Tibirna. I contacted him some time ago but he has not much time for the project. But if you look at the repo there I see some recent activities. I currently don't know how to what is the way to contribute there.
Sorry, I made a mistake: it was necessary to create a 'pull request' to rhaschke/qgit. From Cristian long wait for an answer, so I decided to work with Robert Haschke
Do you release the code from https://github.com/rhaschke/qgit to GNU/Linux distributions? It would be nice if there was some statement in the repository description that reflects your/R. Haschke's view about who is supposed to work on which repo. E.g. I ask myself if qgit is now "officially" forked?
Patches can be sent only by mail (I offered to Cristian allow third-party developers to create branches in its repository, but the idea didn't gain promotion);
Unfortunately, Cristian did not respond quickly (This is not a complaint, just ascertaining of the fact, I myself am often busy on the working);
Recently, the development of the most active leads Robert, but he also enters inaccuracies to a code. I thought it best to solve these problems directly with Robert, not through Cristian.
I hope that over time all new the improvements will pushed to the main repository (http://repo.or.cz/w/qgit4/redivivus.git)
Unfortunately, I have no concept of development for QGit. I do modifications QGit mostly out of necessity: Linux is improving internationalization, and GIT began to publish self messages in russian language, and qgit no longer understand it.
I like QGit so I decided to try to teach him how to work correctly with utf-8 and utf-16 with no-latin characters. So far it turns out badly - QGit wins :(( .
Keep on your effort. :-) In the description of this (feinstaub clone) repo I linked to this discussion thread so that the current friction of current development model becomes more transparent.
For what is worth I created an "official" github clone for qgit: https://github.com/tibirna/qgit. I'd be more than pleased (and indebted) if you all decided to use that for development coordination.
I'll try my best to be as responsive as possible, given my stringent time constraints.
An important note: the original QGit author is Marco Costalba: https://github.com/mcostalba. I'm just the "official" maintainer of the last few years.
@hkarel Internationalisation of any codebase is not easy feat. IMO a first important step is deciding that we use utf-8 for any i18n task. I'm yet to find something that utf-16 and utf-32 can do that utf-8 can't. Using utf-8 ensures transparent evolution and acceptable cross-platform compatibility.
The simplest part of the i18n job is making sure the UI uses only utf-8 strings. The hardest will be interactivity with file-system (file names). I guess, though, we can advance gradually. The most important part will be to have people with actual need for non-latin characters and, of course, writing proper unit-tests.
@ghost I personally compile in Linux with gcc or clang, with qt5. It should compile in Windows. I don't have a mac, but it should work there too. If you try these two other platforms and find problems, let me know.
I really did work to properly display messages received from the GIT into the format UTF-8. (The work is already done. The result is here: https://github.com/hkarel/qgit/commits/ru-lang). The essence of the work is that at the moment of receiving information from GIT perform the conversion from the format UTF-8 to UTF-16, and continue to use only UTF-16 format.
The work does not affect the user interface.
File names are displayed correctly (in any case, the russian-language characters)
The main development is made on Kubuntu 14.04 LTS 64 bit, Qt 4.8.6/5.6.1
Right now I tried to build the project on Win7 32 bit, Qt 4.8.6. It was succeeded to build and even to launch! :)
@tibirna : The thing that UTF-16 can do which UTF-8 can't, is "work nicely with Qt". Given that qgit is build on Qt, that's a huge win for UTF-16. (Personally I prefer UTF-8 for green-field development, but if you are working with Java, Windows, or Qt, there's a strong argument in favour of UTF-16.)
Activity
feinstaub commentedon Oct 22, 2016
I am not maintaining the project. I cloned the code from here http://repo.or.cz/w/qgit4/redivivus.git. The original author is Cristian Tibirna. I contacted him some time ago but he has not much time for the project. But if you look at the repo there I see some recent activities. I currently don't know how to what is the way to contribute there.
hkarel commentedon Oct 22, 2016
Sorry, I made a mistake: it was necessary to create a 'pull request' to rhaschke/qgit. From Cristian long wait for an answer, so I decided to work with Robert Haschke
hkarel commentedon Oct 22, 2016
Please discard this pull request
feinstaub commentedon Oct 22, 2016
I see that Robert Haschke also contributed to the presumed original repository http://repo.or.cz/w/qgit4/redivivus.git
Do you release the code from https://github.com/rhaschke/qgit to GNU/Linux distributions? It would be nice if there was some statement in the repository description that reflects your/R. Haschke's view about who is supposed to work on which repo. E.g. I ask myself if qgit is now "officially" forked?
hkarel commentedon Oct 22, 2016
I believe that there should be one repository. Well if it was a http://repo.or.cz/w/qgit4/redivivus.git
But there are several problems:
I hope that over time all new the improvements will pushed to the main repository (http://repo.or.cz/w/qgit4/redivivus.git)
Unfortunately, I have no concept of development for QGit. I do modifications QGit mostly out of necessity: Linux is improving internationalization, and GIT began to publish self messages in russian language, and qgit no longer understand it.
I like QGit so I decided to try to teach him how to work correctly with utf-8 and utf-16 with no-latin characters. So far it turns out badly - QGit wins :(( .
feinstaub commentedon Oct 22, 2016
Keep on your effort. :-) In the description of this (feinstaub clone) repo I linked to this discussion thread so that the current friction of current development model becomes more transparent.
tibirna commentedon Oct 29, 2016
Hello
For what is worth I created an "official" github clone for qgit: https://github.com/tibirna/qgit. I'd be more than pleased (and indebted) if you all decided to use that for development coordination.
I'll try my best to be as responsive as possible, given my stringent time constraints.
An important note: the original QGit author is Marco Costalba: https://github.com/mcostalba. I'm just the "official" maintainer of the last few years.
Thanks for your understanding.
Cristian
tibirna commentedon Oct 29, 2016
@hkarel Internationalisation of any codebase is not easy feat. IMO a first important step is deciding that we use utf-8 for any i18n task. I'm yet to find something that utf-16 and utf-32 can do that utf-8 can't. Using utf-8 ensures transparent evolution and acceptable cross-platform compatibility.
The simplest part of the i18n job is making sure the UI uses only utf-8 strings. The hardest will be interactivity with file-system (file names). I guess, though, we can advance gradually. The most important part will be to have people with actual need for non-latin characters and, of course, writing proper unit-tests.
Thanks
Cristian
tibirna commentedon Oct 30, 2016
@ghost I personally compile in Linux with gcc or clang, with qt5. It should compile in Windows. I don't have a mac, but it should work there too. If you try these two other platforms and find problems, let me know.
Oh! Please, follow the https://github.com/tibirna/qgit repo. This one seems to be inactive.
Thanks
feinstaub commentedon Oct 30, 2016
Yes, please leave this repo now :-) -- and open further issues here: https://github.com/tibirna/qgit
I close this issue now.
hkarel commentedon Oct 30, 2016
Colleagues, let me make a few clarifications:
MartinBonner commentedon Apr 6, 2017
@tibirna : The thing that UTF-16 can do which UTF-8 can't, is "work nicely with Qt". Given that qgit is build on Qt, that's a huge win for UTF-16. (Personally I prefer UTF-8 for green-field development, but if you are working with Java, Windows, or Qt, there's a strong argument in favour of UTF-16.)