On 3/4/06, Michael Thaler <[EMAIL PROTECTED]> wrote: > On Saturday 04 March 2006 09:33, Sašo Kiselkov wrote: > > > And not only that, I think the separators are a style element which set us > > appart from the plain Apple style, and we need to be different so our menu > > system won't be an obvious rip-off. > > I really like the Apple style menu on top of the screen. I think for most > people used to Windows or MacOSX (or KDE and GNOME) the Apple style menu is > more useable then the the vertical Nextstep menus.
Even if I like the NeXT menus, I must admit that it's probably a better idea to use horizontal menus: people are used to them (I used them on my Atari ST !) , we need anyway a place to put a systray, top screen is easy to reach (fitt's law), etc. NeXT menus are usually better with large screens, because they are consistent and submenus can be easily teared-off and pinned, because they don't take the whole width, also because you can get them with a right-click. But those few things can be offsetted easily; we just need an easy way of having tear-off menus with the horizontal menu bar, and we can anyway keep vertical menus with the right-click (unless a contextual menu is defined for the widget you clicked on, obviously), and the space is anyway used easily with the systray, so the saving is not that huge... And horizontal menus works better with small resolutions too. Anyway, I'm restating the already known here, as we already said that étoilé will use horizontal menu by default (even if you could switch back to NeXT menu) > But I have to admit that I really don't like these vertical seperators. They > just make the menu look more busy and less elegant. I don't know anybody who > does not understand the concept of a menu. Yes, I think I agree. While I was curious about the introduction of vertical separators, I think they indeed make the menu look busier. And indeed a menu is a toolkit item since the end of the seventies, it's not like people aren't familiar with the idea :-) > I think it really doesn't make any sense to introduce style elements which > sets the Etoile menus apart from the plain Apple style if this does only make > the menu look more busy and less elegant. The Amiga did have an Apple style > menu on top of the screen (see http://toastytech.com/guis/amigaaweb.gif) and > also KDE can use the Apple style menu on top of the screen (see > http://people.kde.nl/images/boud/desktop_big.png) and they both did not > introduce new style elements because it is not necessary and (at least in my > opinion) just makes the menu look worse. > > Basically, just doing things different from Windows/MacOSX/KDE/Gnome for the > sake of being different is one of the main problems I have with > gnustep/Etoile. Many things in Windows/MacOSX/KDE/Gnome might not be perfect, > but they work reasonably well for many people and people can get their work > done with it. Hold on, we don't want to make thing different for the sake of making them different ! But, we're certainly not afraid of examining other ideas/concepts and to implement them. If we do so, it's hopefully for the better, not because we want to be different. Here saso's remark was about the "look" (and yes we'd like to have our "own" look), but also about consistency in the user interface -- buttons should look like buttons, that was his point. To that, as you said, we can counteract that menu items are specific UI elements that are well-known since the dawn of computer's UI, and thus can be (and probably must be) different than buttons. > With gnustep/Etoile this is not possible at the moment. For me, the first step > for gnustep/Etoile should be to create a framework and applications so that > peope could actually use gnustepEtoile. For this it is perfectly fine (at > least for me) to just use some ideas from Windows/MacOSX/KDE/Gnome that work > reasonably well. Then people should start to implement new ideas and improve > gnustep/Etoile. I agree with you, and that's one of the goal of étoilé, and that's what we're trying to focus on now (we had lenghtly discussions at FOSDEM with saso and quentin, and we started to highlight a proper path to do that I think). I must say that if you want to use a "gnustep" desktop, you can simply take gworkspace, gnustep apps, windowmaker, and you're here. It works ok. We don't want that though, we want something better :-) > I am one of the contributors of Krita (a KDE painting/image editing > application). I thought a number of times about creating a simple pixel > painting application like Kolourpaint/MS paint for gnustep. I suppose coding > such an application shouldn't even be that hard. > > I have gnustep installed on my computer and I guess I also have the coding > skills to write an application like this. But my problem is, that I cannot > use gnustep. Basically all the esential applications like a webbrowser are > missing, KDE/Gnome applications look totally foreign in wmaker + gnustep and > gnustep applications are hardly useable in KDE. And I also don't like to > write an application if noone at all will use it and I don't know anybody at > all that uses gnustep right now. yes, it's the GNUstep's little vicious circle... GNUstep is a development environment for developing GNUstep apps -- hold on, what's the point ? Nobody uses a GNUstep-based environment, and current GNUstep-gui can't make GNUstep apps "blend" into other environments such as GNOME or KDE. Which is why we want to provide such an environment with étoilé. But we don't want a clone of KDE/GNOME that would simply happen to be written in Objective-C and using GNUstep, we want to actually take advantage of ObjC+GNUstep to make something more interesting to use, and we also want to focus on usability problems. And while we are here, a nice looking UI. > I think the gnustep/Etoile people should create a lightweight MacOSX like > desktop environement (maybe something similar like XFCE) and some of the most > important applications like a browser, an email program (probably GNUMail is > fine), a good editor (personally I would like to see something like kate, but > maybe not as bloated). I think it is perfectly fine to look at > Windows/MacOSX/KDE/Gnome and copy the best ideas from them. MS, Apple, KDE > and Gnome all copied the best ideas from eachother and from other systems and > there is really nothing wrong with that. We all learn from other people. > > After this is done and gnustep/Etoile gets a reasonable number of > users/developers, one could start to think about improving gnustep/Etoile > over existing systems like Windows/MacOSX/KDE/Gnome. But in my opinion, one > should really take one step after another and not try to make everything > different from the very beginning. This almost never works out. That's more or less the path we're taking. We won't have something completely different at first, although it will be slightly different -- if only because our inspiration is the OPENSTEP desktop rather than MS Windows. But indeed the goal now is to have something usable as a standalone desktop before next year, even if it means it's in a raw state / not as advanced as we'd like. > I am quite sure if I would have a useable desktop based on gnustep TODAY that > I can actually use to surf the web, write emails and so on, I would have > already started to develop for gnustep/Etoile. But I will never start to > develop for gnustep/Etoile if gnustep/Etoile is just about being different > from Windows/MacOSX/KDE/Gnome. It's not about beeing different for the sake of beeing different, believe me. If we introduces differences, it's in the hope that those differences will make a better environment.. But indeed we don't want to have a clone of KDE too -- what would be the point ? KDE is already an excellent project. > And I am not the only one that thinks like that. There recently was an > interview with the main developer of Krita (http://people.kde.nl/boud.html). > He talks a bit why he is developing for KDE and not the competition and he > says: > > What competition? When I started looking for something better than fvwm, > GNUStep was still in denial mode -- we're not a desktop environment, we're a > cross platform development platform! -- and little-known GREAT was a gratis > but closed source CDE clone. CDE was out of reach, so I started using KDE as > soon as I got my first computer with 64MB of memory. > And when I wanted to code a GUI application for Linux, the only GUI toolkit > that came close to being as easy to use as Visual Basic was Qt with its > Python bindings. Motif was closed, Gtk still asked me to handle the > scrollbars around a treeview myself, that funny toolkit Lyx used -- XForms -- > was closed, binary and sucked. And Athena simply demanded C coding. No way... > So my second free software application, Kura was done in PyQt. (The first was > a UUCP mail and news client in Visual Basic -- just when I had finished it, I > moved to Linux.) And when I got my first glimmer of Gnome it was a horrible, > complex, convoluted, crash-prone sliver of a shell around Enlightenment. I > remember I called it the Dark Side of User Interface Design in an article I > wrote for O'Reilly Network. That kind of political incorrectness would be > impossible nowadays, of course, so I won't repeat that remark. > And the current state of the competition doesn't inspire a desire to switch > either. XFCE isn't competition: it may have a panel and a file manager, but > there's no framework for integrating applications. GNUStep applications are > still horribly crash prone and there's still no concerted effort making a > real GNUStep desktop environment. And Gnome still looks like no fun at all to > hack on. Or work with. Interesting comments, indeed. Although he's wrong on one point, as there's at least one concerted effort to make a real GNUstep desktop environment ;-) > Krita is becoming really nice lately. Just imagine he would have started to > work on a gnustep image editing application instead of an KDE image editing > application. indeed. > I am not writing this because I want to blame gnustep/Etoile developers. I > write this because I think it is a really pity that gnustep/Etoile keeps a > shadowy existence. I think gnustep could have been THE desktop environement > for Linux/BDS/... if it would have actually tried to. On the other hand, we started étoilé about a year ago, and we choose to focus first on frameworks than applications; in one year we did a lot of things, albeit few are "visible". But I think this year that will change a lot -- things looks really exciting :-) Cheers, -- Nicolas Roard "Any sufficiently advanced technology is indistinguishable from magic." -Arthur C. Clarke _______________________________________________ Etoile-dev mailing list [email protected] https://mail.gna.org/listinfo/etoile-dev
