What Do You Think The Pandora Should Feature First?

[Firefox]

I've been tinkering with a couple of fairly demanding emulators this weekend, and it's pretty obvious to me that any emulator that runs at even a semi-playable rate on the GP2X will run at fullspeed on the Pandora after only a very quick port. It's only the new emulators that weren't possible before like Playstation and N64 that are going to need more than a minimal amount of work doing on them.

I think the important thing about the GUI is that it's fast, intuitive, and reliable. Sod themeability and fancy 3D effects, they'll all come in time.

I looked at Samsung's "iPhone slayer" in a shop last week. The hardware was gorgeous, and I was very interested indeed in buying one - until the sales assistant foolishly let me play with a real one. The GUI was very slow and unresponsive, and it took a lot of working out how to do even very basic things - and it was usually a pain in the arse when I had worked out how to do whatever.

Then I went upstairs to the Apple shop, and within seconds of picking up an iPhone 3G I was paging through Slashdot, taking photos, flicking through the album of photos other people had taken on the demo model, etc etc. The GUI wasn't flashy, it just let you do things very quickly and easily.

So don't worry about the emulators, and a nice helpful GUI please is what I'm saying.

 

[ingrin]

I think showing someone a Pandora with a really slick interface running a media player, browser, PSX and CPS2 wide screen MAME, and saying soon well have N64 etc., would have a far better impact than showing them a Pandora with a wider range of emulators and an average interface. Apple seems to do this quite well; the iPhone's innovation is all in its GUI and user interface.

Show people a nicely built house with a few nicely finished rooms instead of a big pile furniture!

This sums it up my thoughts exactly. Of course, while we have a good idea of what is/might be available, emu/homebrew wise, we don't have haven't seen anything really GUI/app wise. That makes this whole voting process a bit of a farce since most everyone voting don't have enough detail to make a truly informed decision. (I'm not trying to be critical, I appreciate being asked for input)

IF we are looking at PSX, SNES. CPS2. MAME already covered, then I think thats a good starting library for the gamer centric side of the community. I haven't really seen anything for the UMPC side, so I would be keen on seeing more than just GP2X kinda stuff, and see some things that really take advantage of the hardware . Of course, some more homebrew or impressive ports (Quake3/Homeworld/Freespace/Wargus/etc...) would round that out a bit...

I really don't know how many Dev Boards there are to hand out now, but it seems like a 70% GUI/App/firmware focus would be nice.

 

[javaJake]

I voted "Fullspeed emulators" but now I wish I had voted "A good GUI and menu system".

I'd rather have a good base system with the minimal stuff installed, all set up so that it's ready to receive whatever extras you want. This is an internet device, after all. Let the user do the customizing.

This is actually how Ubuntu, and many other Linux distros, work. They provide the system, the users install the applications. I think Pandora should work the same way.

 

[icurafu]

GUI is best as long as the community is involved.

Since we are yet to know anything about the gui, it's best to focus on emulators so that you don't go to far down the path of developing an unpopular GUI.

 

[nim]

if it's about time I vote for a functional and maybe stylistic but minimalist gui (no gradients, bitmap graphics etc, just antialiased line art and text gui on a coloured background?) if not to just use what can be ported from existing compatible linux themes... and then on to whichever app projects seem urgently exciting and then maybe a more eye-catching gui later? (allthough I'd find a minimalist-stylish.. wireframish looking gui most eye-catching if it comes to taste )

 

[cb88]

we have no idea what said UI looks like right now... so it is REALLY hard to decide is a screencast possible criagix?

if the gui is not multitask friendly it might need a little more work to get that going but do you really think that it will require detracting from other efforts to get the basic functionality well all need? personally i am pro 100% functionality (launcher and task switcher) + whatever fluff you have time for other wise don't worry about it and get MK0s to the emu devs I don't see what the point of holding up the guys that are chomping at the bit to write them...

one more thing is this launcher just reading .desktop files and creating a list or do we have to navegate the file tree (rather a pain)

it would be rather ideal for this situation to adopt appdirs baically the program folder contains an AppRun script that runs the program (it doesn't even to be in the folder but it can) and an .Diricon which is the application icon and possibly other options in Appinfo.xml (rather Rox-filer specific)

you could even search the system for all appdirs to generate the menu and there would be no need to sacrifice a MK0 for this since it is mostly just bash scripting and geting the launcher to support recognising the appdir itself to run the AppRun script

there is a more complete discription here : http://roscidus.com/desktop/AppDirs

the basics of the format should be pretty simple to implement

 

[aznJun]

I voted for "a good GUI and menu system" due to that if it has a good operation menu system
we would be able to navigate through the system easier.... even if someone who just began using
the pandora. They would easily understand how to use it if it had a good UI. For example; If one of your friends had a pandora and u ask if u could play with it (you've never used one before) then you could possibly turn it on and total get lost in how to function the system....
However this is only my opinion.

-aznJun

 

[Exophase]

I voted GUI but why is this a relevant choice at all? I didn't think there was a team of skilled emu coders who were being delegated to GUI coding instead...

 

[ingrin]

I voted GUI but why is this a relevant choice at all? I didn't think there was a team of skilled emu coders who were being delegated to GUI coding instead...

craig clarified somewhere within the thread that it is to help determine whom to send the next set of MK0s to...

 

[Alpha2]

my god. I had no idea Pandora had advanced this far... I've been so far away from the GP scene that I'd not seen any of this coming.

Speaking from a sensible standpoint:
Develop a good GUI first make the device look like it's got a clear function (even if it really does!) besides just being a Emubox otherwise companies like Capcom and Nintendo start giving you the eye and you dont need that.

 

[God Ginrai]

my god. I had no idea Pandora had advanced this far... I've been so far away from the GP scene that I'd not seen any of this coming.

Speaking from a sensible standpoint:
Develop a good GUI first make the device look like it's got a clear function (even if it really does!) besides just being a Emubox otherwise companies like Capcom and Nintendo start giving you the eye and you dont need that.

There shouldn't be any first and second. There should be a good GUI, and a few killer apps (SNES, MD, PSX, N64 [If Zod gets that working in time])

-God Ginrai

 

[quadomatic]

You know, I think we really need a preview of what you have in the GUI so far before we can decide if it's good enough or needs more work. We really are making a rather uninformed decision here.

I was about to post the same thing. Its difficult to know what we want when we don't know exactly what's been done already, I suppose.

 

[God Ginrai]

You know, I think we really need a preview of what you have in the GUI so far before we can decide if it's good enough or needs more work. We really are making a rather uninformed decision here.

I was about to post the same thing. Its difficult to know what we want when we don't know exactly what's been done already, I suppose.


Yea, I have been getting itchy to see what the GUI looks like.

-God Ginrai

 

[kloplop321]

All I can say about Applications Vs. GUI is: Get the base of the GUI done, because This should be theme-able, so get the framework in and have a fall back in the case that there is not a theme selected and it would be like Win95/98 style(barebones if you know what I mean)
However the framework must be finished before release because it would be very helpful towards future theming.
I would not use it much for gaming reasons, but for the general population that are possible customers, it should have considerable function in all aspects.

EDIT: Going on about the GUI framework, make it possible to do custom coded affects such as some that are available in the open source code for Compiz. Since Compiz also uses OpenGL, some specific methods should be able to be ported.

 

[quadomatic]

All I can say about Applications Vs. GUI is: Get the base of the GUI done, because This should be theme-able, so get the framework in and have a fall back in the case that there is not a theme selected and it would be like Win95/98 style(barebones if you know what I mean)
However the framework must be finished before release because it would be very helpful towards future theming.
I would not use it much for gaming reasons, but for the general population that are possible customers, it should have considerable function in all aspects.

EDIT: Going on about the GUI framework, make it possible to do custom coded affects such as some that are available in the open source code for Compiz. Since Compiz also uses OpenGL, some specific methods should be able to be ported.

I don't think that compiz is realistic...its still not totally stable and I don't think there's a version for ARM...but you could compile it for ARM from source...right?

 

[Esn]

If the general consensus really is to focus on style instead of substance, then by all means that is where pre-release development should concentrate its efforts.

I consider a well-designed, easy-to-use GUI to be "substance" rather than "style". I don't care about theming & fancy graphics; just make the basic design work well, make it easy to use and make it handle many different functions with elegance, and the rest will follow.

"If you build it, they will come"

However, I haven't voted yet because I don't think we've been given enough information about where things stand (as others have said).

 

[God Ginrai]

All I can say about Applications Vs. GUI is: Get the base of the GUI done, because This should be theme-able, so get the framework in and have a fall back in the case that there is not a theme selected and it would be like Win95/98 style(barebones if you know what I mean)
However the framework must be finished before release because it would be very helpful towards future theming.
I would not use it much for gaming reasons, but for the general population that are possible customers, it should have considerable function in all aspects.

EDIT: Going on about the GUI framework, make it possible to do custom coded affects such as some that are available in the open source code for Compiz. Since Compiz also uses OpenGL, some specific methods should be able to be ported.

I don't think that compiz is realistic...its still not totally stable and I don't think there's a version for ARM. It would certainly be cool though.


I thought I saw someone in a topic earlier on the forum mention that Compiz was too high in SysReqs. Although I could be imagining it.

-God Ginrai

 

[quadomatic]

All I can say about Applications Vs. GUI is: Get the base of the GUI done, because This should be theme-able, so get the framework in and have a fall back in the case that there is not a theme selected and it would be like Win95/98 style(barebones if you know what I mean)
However the framework must be finished before release because it would be very helpful towards future theming.
I would not use it much for gaming reasons, but for the general population that are possible customers, it should have considerable function in all aspects.

EDIT: Going on about the GUI framework, make it possible to do custom coded affects such as some that are available in the open source code for Compiz. Since Compiz also uses OpenGL, some specific methods should be able to be ported.

I don't think that compiz is realistic...its still not totally stable and I don't think there's a version for ARM. It would certainly be cool though.


I thought I saw someone in a topic earlier on the forum mention that Compiz was too high in SysReqs. Although I could be imagining it.

-God Ginrai


I'm not sure that I'd rule it out based on system requirements. Intel integrated gfx chipsets are powerful enough to run compiz just fine. Pretty much any 3d accelerated gfx hardware will run compiz. Not sure about processor reqs tho.

 

[Esn]

One thing I don't quite understand... so will things like audio & video playback, ebooks, and so on be integrated into the firmware? What if somebody later writes a better audio player than the "official" one, will it be easy to replace the one with the other?

 

[quadomatic]

One thing I don't quite understand... so will things like audio & video playback, ebooks, and so on be integrated into the firmware? What if somebody later writes a better audio player than the "official" one, will it be easy to replace the one with the other?

I imagine that you would just open up the audio player app like u do on gp2x