A bit late, but here is the thread for 0.19.0. I'll skip going through all the open issues this time, because most of them are taken already. Quite a few by the recent wave of new developers, of which most unfortunately have not reported back yet. I want to give them a bit more time though, before I flag them as unclaimed again.
Most of the remaining unclaimed tasks fall into two areas:
1. Animation and NPC rendering, plus movement which is tied in to animations: We have a discussion about that going on in another thread right now.
2. The whole ini importer topic: This is plaguing us for quite some time now. More complex than I initially thought. Will require some changes to the launcher (both the way it works and the UI). Also, a lot of work. I really hope we can find someone to take this group of tasks, because it becomes tiring to drag it along.
We should continue to aim for monthly releases. We already have some finished features (6 out of 45 issues are closed), so I don't see a problem for that with 0.19.0.
OpenMW 0.19.0
Re: OpenMW 0.19.0
I'll take them. I can start working on them as soon as I finish the new data model for the Data Files tab. The model is nearly complete and I am currently in the process of moving my changes from a debug tab to the actual Data Files tab, hence no commits to my repo yet. The only thing that is not working properly now is drag and drop and moving of the rows when they are checked or moved through the context menu. (Here's a screenshot)Zini wrote: 2. The whole ini importer topic: This is plaguing us for quite some time now. More complex than I initially thought. Will require some changes to the launcher (both the way it works and the UI). Also, a lot of work. I really hope we can find someone to take this group of tasks, because it becomes tiring to drag it along.
For the ini importer tab I will need to make some changes to the way settings are handled, but I was planning on rewriting this anyway. We also need to discuss on how things are supposed to work and how it should look.
Re: OpenMW 0.19.0
That looks nice.
btw. no need for a debug tab. You could have done that in the regular tab in a separate branch. That is what we have git for.
When you are ready to discuss the ini topic please make a new thread in the development forum.
btw. no need for a debug tab. You could have done that in the regular tab in a separate branch. That is what we have git for.
When you are ready to discuss the ini topic please make a new thread in the development forum.
Re: OpenMW 0.19.0
Well initially this only contained a view for my new model, just like Eli2 did in the editor. It's just easier and quicker this way.
I'll let you know when we're ready to talk INI importer.
I'll let you know when we're ready to talk INI importer.
Re: OpenMW 0.19.0
General wake up call: We are close to the point where I would normally call for RC builds. But we are not there, apparently. Not enough development activity. We have not enough material for a release yet and we still have a large number of bugs, two of which I would classify as blockers and one as an almost-blocker:
The blockers:
* launcher.cfg file stored in wrong directory
* Wrong esm order in openmw.cfg
These will make OpenMW crash under certain configurations and we can not afford to have them in another release.
The almost blocker:
* Extreme shaking can occur during exterior cell transitions
Technically we could have it listed as a known issue for the next release. But it just looks awful and I don't think we are doing OpenMW any favor, if we have another release without a fix for this issue.
Obviously it would be nice to get rid of some other bugs too. There are some relatively easy ones, that would be even suitable for a beginner.
The blockers:
* launcher.cfg file stored in wrong directory
* Wrong esm order in openmw.cfg
These will make OpenMW crash under certain configurations and we can not afford to have them in another release.
The almost blocker:
* Extreme shaking can occur during exterior cell transitions
Technically we could have it listed as a known issue for the next release. But it just looks awful and I don't think we are doing OpenMW any favor, if we have another release without a fix for this issue.
Obviously it would be nice to get rid of some other bugs too. There are some relatively easy ones, that would be even suitable for a beginner.
Re: OpenMW 0.19.0
Right, for the order of esm/esp files: would sorting by modified timestamp do for now or do we need dependency sorting too?
About the config file location: I use QSettings for the launcher configuration and it doesn't support reading/writing multiple file locations. If I read from the local directory I can't write to the user directory without creating a new QSettings object, thus losing the settings read from the local cfg. I want to rewrite the whole settings stuff anyway but that's nowhere near finished for 0.19.
About the config file location: I use QSettings for the launcher configuration and it doesn't support reading/writing multiple file locations. If I read from the local directory I can't write to the user directory without creating a new QSettings object, thus losing the settings read from the local cfg. I want to rewrite the whole settings stuff anyway but that's nowhere near finished for 0.19.
Re: OpenMW 0.19.0
Dependency. Time stamp is kinda optional at this point, though apparently we need it before 1.0. We can get away with a minimal implementation for now, since we are actually only using the very first esm file and nothing else.Right, for the order of esm/esp files: would sorting by modified timestamp do for now or do we need dependency sorting too?
Pondering ... At least on Linux we will only have a launcher.cfg file in local, if we are running without installing first, so that should kinda work (still very ugly). On WIndows it should not be a problem at all, since IIRC local is equal user. No idea about the OS X situation though.About the config file location: I use QSettings for the launcher configuration and it doesn't support reading/writing multiple file locations. If I read from the local directory I can't write to the user directory without creating a new QSettings object, thus losing the settings read from the local cfg. I want to rewrite the whole settings stuff anyway but that's nowhere near finished for 0.19.
Re: OpenMW 0.19.0
I have a question about the launcher.cfg problem. I managed to get the local launcher.cfg to overwrite profiles from the user location because local takes precedence over user, but since we always write to user there is no way in changing profiles that are also in the local cfg. How should this be handled?
Should the profiles be merged or should I revert to writing the changes to the local file if that's present, else write to user?
Should the profiles be merged or should I revert to writing the changes to the local file if that's present, else write to user?
Re: OpenMW 0.19.0
Interesting question. Maybe we can completely do away with the local file? As you just found out, it doesn't make all that much sense to have it and I don't think we actually need it, once all the launcher and ini importer changes are in place.
Edit: The other options would be to merge the local settings into user or to flag the profiles from local as read only and not write them at all.
Edit: The other options would be to merge the local settings into user or to flag the profiles from local as read only and not write them at all.
Re: OpenMW 0.19.0
I'm all for getting rid of the local file. Merging the local settings and flag them read-only won't work on Windows, as local and user both refer to the same location. Also, there is no need to provide a default cfg as openmw already has some fallbacks in place when the plugins/masters are not present in openmw.cfg.