Preliminary Editor GUI poll
Posted: 21 Aug 2011, 21:50
Zini wrote: Its a bit early for thinking about a custom editor, but I would like to gauge general interest anyway.
The choice of GUI framework is largely a question of personal taste, but we definitely want to use a mainstream multi-platform framework.
Of these I only consider Qt and Gtkmm viable. Theoretically WxWidgets would fall into this category too, but from my limited experience with it I can safely say, that it is a headache we don't need.
Please vote only, if you are seriously interested in continuing to work on the OpenMW after the 1.0 release (i.e. you plan to contribute more than a single commit or some minor fixes). Since this is a question of personal taste, our choice should be determined by the majority of those people who will do most of the actual work. If you don't fall into this category, feel free to post a reply anyway, but please don't vote.
gus wrote: I have no opinion over the library, but couldn't we use Ogitor?
Zini wrote: I don't think so. I haven't looked into Ogitor yet, but as far as I can tell, it is an extensible editor for OGRE based 3D scenes.
With MW (and OpenMW) the actual 3D scene is only a fraction of what needs to be edited. We have scripts, we have ID lists, we have class lists, we have faction lists, we have race lists and tons of other stuff. Building a new editor around a 3D scene editor, with the other editor functions attached to it as an afterthought would be a very bad idea IMHO.
Chris wrote: I'd say Qt. It's designed to take advantage of C++ from the ground up, while Gtkmm is mainly just a C++ wrapper over GTK's C API. Plus, I've seen GTK apps have quite a bit of trouble in WIndows, and don't tend to feel very "native".
Star-Demon wrote: I would say Qt. Although it would be helpful if the editor was very good with ogre so users can more directly work with the game.
I'm not very familiar with the technical side of it. I can't make an well-informed decision.
Greendogo wrote: As I'm writing this, there are 10 votes. Do we actually have ten developers or did most people not bother to read the part that says not to vote if you aren't writing code?
Anyways, if I was a developer and a contributer, I probably would have picked Qt from some usability comments I've heard from some friends
Anyway, I think the editor is something that should be included with 1.0. It is a very essential part of what makes Morrowind what it is. Providing the editor on Day 1 of the launch shows that it is a viable platform to move to from the original early, so that mod development that incorporates new features of the new engine doesn't stagnate right from the beginning. It's development may also help the development of the main engine in ways you can't think of right now.
Anyway, Qt is fine. The way Gtkmm is used for Gimp looks pretty nice too. So, I don't think it matters between the two, so far as aesthetics goes. You may want to take into account that the editor is a creative one, and so you might want it to 'look' creative. I dunno.
Zini wrote: Seems we have a firm majority for Qt. And I am the only Gtkmm proponent here. Kinda sad
Well, okay. The number is indeed suspiciously high. I guess some people have voted without reading.
Well, the plan was to do OpenMW 1.0 first and then the Editor (before any post OpenMW 1.0 work).Anyway, I think the editor is something that should be included with 1.0. It is a very essential part of what makes Morrowind what it is.
So once OpenMW 1.0 is done, we will wait for the editor (OpenCS? We sill need a name for it) to reach 1.0 too. And then we will develop them in parallel. From 1.0 on we probably will be doing mostly feature enhancements, which will usually result in a modification of the esm format. That means the changes in OpenMW must be reflected in the editor.
I got the impression, that we have some people who do not feel comfortable with working on something as complex as OpenMW, but who have decent Qt skills. So we may give the editor a head start, while OpenMW 1.0 is still in development (no point in letting developers sitting by idly). But only, and that is the important point, if this does not draw away resources from OpenMW 1.0.
btw. If anyone thinks about starting work on the editor right now, please don't. We need to do some planning or we end up with the same mistakes as the original TES-CS. From my work on the Redemption project I have tons of experience with the workflow, that should go into the design document (I guess I should write one at some point).
ap0 wrote: I don't think an open vote is a good solution.
Maybe just counting the posts here will be enough ?
Zini wrote: I would love to see a different outcome, but that is not realistic.
While the total number of votes might be questionable, Qt's victory isn't. No one else but me has voted against it and no one has voted against doing the editor either. Assuming we have at least 2 real developers among the 9 votes for Qt, the outcome of the vote won't change.
Greendogo wrote: You are the lead Zini, I suggest taking the preference for Qt into account, but that you make the decision of which one to use yourself. If you have experience with both, you might want to list the pros and cons of both, from you own usage. And then after an at least two or three day thread discussion, re-poll it.
There is no sense going with Qt if everyone that voted for it isn't informed about the opposition and without a big read warning at the top of the main post that says:
"DON'T VOTE IF YOU ARE NOT A DEVELOPER!!!"
Zini wrote: To quote myself:
Both Qt and Gtkmm would be fine for the task. Really, this is less a matter of what is better or worse than of what you feel more comfortable with and have more experience with. Therefore a simply majority vote works well enough, even if it is slightly by people voting, who should not have voted.The choice of GUI framework is largely a question of personal taste
But okay, maybe we should get a census on how many people are out there. Maybe using an anonymous voting system wasn't the best choice.
If you are a developer, who is seriously interested in participating in the editor development, please post a reply with one of these lines.
I want to help developing the editor and I prefer Qt.
I want to help developing the editor and I prefer Gtkmm.
I want to help developing the editor and I have no specific preference.
pvdk wrote: I want to help developing the editor and I prefer Qt.
As I'm the developer of the launcher it should come as no surprise that I prefer Qt. I don't have any experience with GTKmm but GTK+ apps are never really cross-platform, they just don't fit in. Also, using Qt for the launcher and GTKmm for the in-house editor is a bit illogical.
I played around with Qt+Ogre integration and it's not that hard to do. There are even projects around which do exactly that, QtOgreFramework for example.
But before we start writing a CS we should have some hefty discussion on the interface (yes I'm looking at you Zini!)
Einarin wrote: It is my understanding that gtk has incomplete support for quartz, the OSX graphics framework. Qt does support it however.
Lordrea wrote: I want to help developing the editor and I prefer Qt.
Zini wrote: Okay. Now we have two people who want to work on the editor and are not (substantially) involved with the development of the main OpenMW application, so getting them started will not draw any manpower from OpenMW 1.0. That should be good enough.
But we definitely should finish the launcher, before getting serious about the editor.
I will start making some design postings in the following days. As usual my time is a bit limited and I won't be able to go too fast. Please don't step ahead. We should try to handle this orderly. With OpenMW we managed somewhat okay with limited organisation, because with Morrowind we kinda have a live-design specification. With the editor, that won't work.
btw. should we get a separate forum section for the editor? Development (Editor) just below the Development section?
Ace (SWE) wrote: I would love to help with the editor, haven't worked much with qt but after looking through some of their samples it doesn't seem too hard to learn.
Greendogo wrote: @Zini: Yeah, I think a separate section would help focus stay on the editor and it would help keep things specific to it from being lost. It would also make it easier to see when something editor specific has been posted.
Hircine wrote: I would like to Learn C++ and help a bit out with the Editor. Not anything to do with things like Terrain editing, but maybe working on the menus/layouts etc.
I have a few years experience in what you would call amateur programming.
I am indifferent to Qt or otherwise. This way i can use my skills to help openmw.
So thats 4? (if you include ace)
to work on the Editor.
a separate forum section would be cool.
EDIT: dw about subforum. forum isn't very full atm