I am giving a short talk on the subject from a programmer's perspective. I was thinking about talking about UI design along 3 axes:
Proximity: How close the code you write is to the UI
Interactivity: How flexible and powerful the UI can be
Complexity: How difficult it is to build a user interface (technical/implementation)
I was then going to discuss the following UIs:
Command Line:
High on proximity, somewhat high on complexity, low on interactivity
VB / MFC / Windows Forms
Low on proximity (deal with a big API/controls/UI Model), medium interactivity - you can build it in but with a high penalty in proximity (3rd party libraries/controls), bigger range on complexity; simple applications are easy to build.
Web
Low on proximity (browser does the legwork, different browsers behave differently), high on interactivity (DOM/CSS/Javascript/Browser), high on complexity (lots of moving parts)
Silverlight (perhaps other RIAs)
Higher proximity than RAD/MFC/VB style interfaces as well as Web, high on interactivity, lower on complexity (e.g. animation in this environment versus OpenGL etc)
Questions
Do these 3 axes make sense? I know there are some generalizations one could make such as a tradeoff of interactivity and complexity that apply in every environment but I see strengths and weaknesses inherent to each.
Is there another axis or trait in UIs that makes sense to discuss?
Future implications on how we develop (RIAs become dominant, etc...)
Any other feedback (+/-)
I think the key to your approach is to know your audience, are they other developers, or people who would like to know more about the development process?
I think the absolute key to UI is usability, everything else (including ease of development) is subordinate to that. Progammability is normally a result of tools, and if the technology is pervasive enough then the tools will come, windows dev used to be awful, and then a few years later everyone was a VB coder.
I think portability also is worth mentioning. Quite often, the UI components are what prevents the application from being ported to other platforms.
For the interactivity axis, I think usability needs to be discussed because flexibility and power are only helpful as they make a UI more usable. A user needs to rapidly be able to determine how to use the software, it ought to be intuitive.
Related
I'm looking for systematic review of the algorithms used for GUI layout. I'm particularly interested in the algorithms that favor speed over complexity, but it's hard to find anything useful other than "don't use constraint sovers".
So far the best one I've found is Constraint Solving for Beautiful User Interfaces: How Solving Strategies Support Layout Aesthetics by Clemens Zeidler,Christof Lutteroth and Gerald Weber. Unfortunately it is very short and is not very specific.
The layout algorithm landscape looks to me like this so far:
There's everybody's darling Cassowary that made its way just about everywhere.
There's the Auckland Layout Model which no one seems to use but thanks to its authors being prolific writers is well known.
There's the universally hated HTML/CSS combo that completely destroys the former two when it comes to the flow layout (especially the text flow).
Next come the algorithms used in desktop GUI frameworks, with WPF, Qt and MigLayout for Swing lauded the most. (I haven't figured out their performance characteristics yet, but I remember being disappointed by WPF layout speed compared to WinForms - the gains from faster drawing were offset by slower layout)
Than there's a handful of game engines UIs that no one seems to know anything about.
The rest are the ancient GUI toolkits that typically have an unsophisticated grid container as their most potent layout tool.
Have I missed anything significant?
this may seem like more of a theoretical question, but I see the term bare metal thrown around a lot when talking about performance.
I sometimes see people say how x has bare metal access (not necessarily on this site). I looked up a little bit and found some places describing it as just a machine without an OS. I also would assume it has something to do with the other hardware (CPU/GPU etc.)
Is this the same bare metal that people refer to with programming and languages?
When people talk about bare metal access for performance, is there a disadvantage to this?
Bare metal is a metaphorical expression of something which happens on very low level. You can imagine that modern computers are architected with a nested hierarchy of abstraction layers. On the lowest level in this hierarchy, we have electronic components, transistors, electrons if you like. Medium levels provide communication and presentation services. The top level goes beyond single applications and covers distributed IT systems.
To access the low levels directly can be a way to do something very efficiently in terms of performance. However, this is in practice only true for very limited contexts. The price to pay is high development cost, lack of flexibility/portability, long development time, greater risk of failure and outages.
It takes quite a lot of knowledge and experience to implement something with higher performance compared to what established libraries and frameworks are achieving.
To work on higher abstraction levels results in better development efficiency, sometimes at the cost of performance. But development on higher and higher abstraction levels is what most developers tend to do to gain economic benefits.
Years ago I discovered radial menus and realized that they could potentially be much easier to use than linear menus. At the time, I didn't have enough understanding of user experience principles (other than Fitt's law) to give them an honest evaluation. I thought that if they were indeed much easier to use in practice they should eventually show up in mainstream products, even against the inertia of the convention of linear menus.
Fast forward to today, I download ZGRViewer to visualize some graphs. It uses a radial context menu and I realize that these things really are easier to use, especially when you learn the simple mouse gestures to select nested menus.
Why hasn't the radial menu gained more traction? Is it user inertia? Framework inertia (i.e. not enough good implementations on various platforms)? Are they just not that much better in practice, overall?
In UX there are things we can do and some we would like to do but don't. Understanding how the human machine ingests information is critical in designing a great experience. A radial menu does make sense but it's not what a user has been accustom to. It's not impossible to implement but it will take the user some time to understand the context of where they were to where they are now. There are inherent behavioral patterns that can't be swept aside. You learn to push the boundaries and still make it usable just beyond the visual appeal.
I have been asked to investigate porting Wii games and some (Sony) PSOne games to OpenGL ES (can you guess what platform?).
I have never undertaken a game port like this before (and will be hiring someone to do it) but I'd like to understand the process.
Does the Wii use OpenGL? If not what does it use and how easy is it to port to OpenGL / OpenGL ES?
Are there any resources/books/blogs that will help me in understanding the process?
Will my company have to become an official Wii developer? If so where do I start that process?
Porting from the Wii or the PSOne is a complex and involved task that can be broken down into multiple separate engineering efforts working in parallel to produce a working end product. The best possible thing you can do before moving to the target hardware is to compartmentalize all of the non-portable code while ensuring that the game continues to run as expected. When you commit to moving to the new platform, your effort switches to reimplementing the non-portable compartmentalized parts.
So, to answer your question, yes, you will need to become or work with a Sony and Nintendo licensed developer in order to take this approach. In the case of Sony, I don't even know if they offer a PSOne development program anymore which presents issues. Your Sony account rep can help clarify.
The major subsystems that are likely to be the focus of your porting effort are:
Rendering Graphics code contains fundamental assumptions about the hardware it is being run on in order to perform optimally. API-level compatibility is superficial compatibility and does not get you as much as you may hope it does. Plan on finding the entry point to the renderer and determining what data you need to render a scene and rewriting all the render code from there for your target hardware.
Game Saving Game state serialization and archival will need to be separated out. Older games often fwrite() structs with #pragma packed fields. Is that still going to work for you?
Networking Wii games write to high level services that are unavailable on your target hardware. At the low level, sockets are still sockets. What network services do your Wii games rely on?
Controls From where you are coming from to where you are going, anything short of a full redesign or reimagining of input will result in poor reviews of the software.
Memory Management Console games often make fundamental assumptions about the rate the system software returns memory from the heap, how much fragmentation it will cause and the duration the game needs to operate under these conditions. These memory management assumptions are obsolete on the new platform. It is wise to write your own memory manager that provides a cushion from the operating system. Also, console games compiled for release are stripped of most error handling and don't gracefully handle running out of memory-- just a heads up.
Content Your bottleneck will be system memory. Can you fit the necessary assets into memory? With textures, you can reduce mip where necessary and with graphics hardware timing, you can pull in the far clipping plane. With assets resident in memory, you may need a technical artist to go through and reduce the face density of your models or an animation programmer to implement a more size-friendly animation codec. This is very game specific.
You also run into the standard set of problems with things like bit compatibility (though the Wii and PSOne are both 32-bit), compiler idiosyncrasies, build script incompatibilities and proprietary compiler extensions.
Games are relatively challenging to test. A good rule of thumb is you want to have enough testers on your team to run through the game in a maximum of two days, covering all major aspects of play. In games that take a long time to beat (RPGs with 30+ hours of gameplay), your testing team needs to be quite large to offer full coverage. Because you are just doing a port, you can come up with a testing plan that maximizes coverage of your new code without having a testing team punch every wall in your game to make sure it (still) has clipping. The game shipped once.
Becoming a licensed developer requires you to apply. The turnaround time, from experience, is not good. Generally speaking, priority is given to studios with shipped titles and organized offices with reasonably good security and the ability to buy the (relatively) expensive development kits. You may be better off working with a licensed developer if you do not meet these criteria.
Console and game development is challenging for people already experienced in it. There is no book that covers it all. My recommendation is to attempt to recruit an expert who has experience shipping titles in a position of systems or engine programmer. What types of programmers and skillsets exist in games is a whole different question for Stack, though.
Games consoles don't use OpenGL but their own, custom libraries. The main reason is that they are pretty slow and have little RAM. So you need to squeeze out every drop of performance you can get. And that means: Custom code. Usually, you get a framework with the developer kit which gets you started and then, you build your code from that. Eventually, you'll start replacing parts from the developer kit with your own special code to get all the speed and special effects you need.
There is a reason why PSOne games are so ugly on the PS3 despite the fact that the developers have access to the sources: Revenue just doesn't justify to touch the code.
Which is one reason why game development is so expensive: Every game is (more or less) a completely new product. Sometimes, game companies can reuse a bit of code from the last version but more often than not, they have to develop everything again. They also don't talk much with each other.
In recent years, kits have become more complex and powerful and you can get complete game engines (with all kinds of effects and 3D support) but each engine is a completely different kind of beast, so you can't even copy code from engine A to B.
Today, media content (video, audio and render sequences) are so expensive that the actual game engine is often a minor detail, so this isn't going to change any time soon.
Net result: If you want to port a game, write an emulator for the hardware (which is usually pretty simple and allows you to run all kinds of games).
[EDIT] To develop software for the Wii, see here: http://www.warioworld.com/
For a Wii emulator, see http://wiiemulator.net/
I ported a couple of games, when I was a new game programmer, from working with one version of our engine to a newer version (where backwards compatibility was neither ignored nor pursued). Even copying (and possibly renaming) the files and placing them in a home in the new project was a bit of work. Following that, the procedure was:
recompile
fix many of the hundreds of errors [in many places, with the same error occurring over and over again]
and
"wire up" calls from the new game engine to the appropriate calls in the old code
"wire up" function calls from the old code into the new game engine
deal with other oddities (ex. in the old game engine, the 2d game would "swizzle" textures itself; in the new version, the engine did it (on specific platforms))
and, while I don't recall this clearly, it was probably mixed in with a bunch of #ifdeffing out portions of code so the thing would actually compile, and possibly creating function stubs to be filled in later.
As I recall, it was three or four days until I had something that compiled. (But, it did help when we ported other games from the old version to the new one!)
The magnitude of the task will come down to what the code you are getting is like. If it has generic 3D calls that you can intercept -- add a thunking layer to -- then you are in business. It depends on the level of abstraction in the code. If it is well-behaved and has things like "RenderModel" and "RenderWorld" calls, you can replace those functions, and even the structures that they work with. If drawing is occurring all over the place, and calls are more like "Draw Polygon" and "Draw Line" or "Draw using this highly optimised data structure", then you are likely in for a long slog.
You shouldn't need a Wii dev kit. Sometimes it is nice to verify that the code you are given does indeed compile in the original environment (and matches the shipping code!), but sometimes you can just take it on faith and make it work in its new environment.
Lastly, I don't think the Wii uses OpenGL, and I really don't know where to point you for further help.
What you may want to do is to start with designing the architecture of the game, write up a detailed specification for what the new game is like.
Once you have this, since you will be rewriting the code, you may find that some of the business logic that doesn't deal with the console can be ported over. But, anything dealing with I/O, user interaction or graphics/sounds will be rewritten, so you might as well do that from scratch.
A specification is very important, to make certain that you know how the current game is working so that the new port will give the same user experience, if that is what is desired.
You may want to keep the same bugs, if that is part of the experience, as, if I know that in the Wii I can jump down and bounce off the wall to safely land, then if I can't do that in the new version then that may be bothersome.
Well porting a PS1 game to an iPhone would be quite a task they work in very different ways. I'm sure its doable but it will be a LOT of work to replace all the fixed point maths and lack of Z-Buffer based rendering to a real graphics chip.
Wii would be a lot easier. The Wii API is very similar to OpenGL. However the Wii has some very nice fixed function features that just are not available on any other GL based platform. Should be doable, though ...
I'm not really sure I can say anything more than that. Have signed far too many NDAs over the years to be 100% sure of what I can and cannot say ;)
Still if you want to hire someone to do some porting work and are prepared to supply the required hardware then I might be free ;)
All you Stackoverflowers,
I was wondering why GUI code is responsible for sucking away many, many cpu cycles. In principle, the graphical rendering is far less complex than Doom (although most corporate GUIs will introduce lots of window dressing). The event handling layer is also seemingly a heavy cost, however, it seems that a well-written implementation should switch between contexts efficiently on modern processors with a lot of memory/cache.
If anybody has run a profiler on their big GUI application, or a common API itself, I'm interested in where the bottlenecks lie.
Possible explanations (that I imagine) may be:
High levels of abstraction between hardware and application interface
Lots of levels of indirection to the correct code to execute
Low priority (compared to other processes)
Misbehaving applications flooding API with calls
Excessive object orientation?
Complete poor design choices in API (not just issues, but design philosophy)
Some GUI frameworks are much better than others, so I'd like to hear varied perspectives. For example, the Unix/X11 system is much different than Windows and even than WinForms.
Edit: Now a community wiki - go for it. I have one more thing to add -- I'm an algorithms guy in school and would be interested if there are inefficient algorithms in GUI code and which they are. Then again, it's probably just the implementation overhead.
I've no idea generally, but I'd like to add another item to your list - font rendering and calculations. Finding vector glyphs in a font and converting them to bitmap representations with anti-aliasing is no small task. And often it needs to be done twice - first to calculate the width/height of the text for positioning, and then actually drawing the text at the right coordinates.
Also, most drawing code today relies on clipping mechanisms to update just a part of the GUI. So, if just one part needs to be redrawn, the code actually redraws the whole window behind the scenes, and then takes just the needed part to actually update.
Added:
In the comments I found this:
I'm also very interested in this. It can't be that the gui is rendered using only the cpu because if you don't have proper drivers for your gfx-card, desktop graphics render incredibly slow. If you have gfx-drivers however desktop-gfx go kinda fast but never as fast as a directx/opengl app.
Here's the deal as I understand it: every graphic card out there today supports a generic interface for drawing. I'm not sure if it's called "VESA", "SVGA", or if those are just old names from the past. Anyway, this interface involves doing everything through interrupts. For every pixel there is an interrupt call. Or something like that. The proper VGA driver however is able to take advantage of DMA and other enhancements that make the whole process WAY less CPU-intensive.
Added 2: Ah, and for OpenGL/DirectX - that's another feature of today's graphics cards. They are optimized for 3D operations in exclusive mode. That's why the speed. The normal GUI just utilizes basic 2D drawing procedures. So it gets to send the contents of the whole screen every time it wants an update. 3D applications however send a bunch of textures and triangle definitions to the VRAM (video-RAM) and then just reuse them for drawing. They just say something like "take the triangle set #38 with the texture set #25 and draw them". All these things are cached in the VRAM so this is again way faster.
I'm not sure, but I would suspect that the modern 3D-accelerated GUIs (Vista Aero, compiz on Linux, etc.) also might take advantage of this. They could send common bitmaps to the VGA up front and then just reuse them directly from the VRAM. Any application-drawn surfaces however would still need to be sent directly every time for updates.
Added 3: More ideas. :) The modern GUI's for Windows, Linux, etc. are widget-oriented (that's control-oriented for Windows speakers). The problem with this is that each widget has its own drawing code and associated drawing surface (more or less). When the window needs to get redrawn, it calls the drawing code for all its child-widgets, who in turn call the drawing code for their child-widgets, etc.. Every widget redraws its whole surface, even though some of it is obscured by other widgets. With above mentioned clipping techniques some of this drawn information is immediately discarded to reduce flickering and other artifacts. But still it's lots of manual drawing code that includes bitmap blitting, stretching, skewing, drawing lines, text, flood-filling, etc.. And all this gets translated to a series of putpixel calls that get filtered through clipping filters/masks and other stuff. Ah, yes, and alpha blending has also become popular today for nice effects which means even more work. So... yes, you could say this is because of lots of abstraction and indirection. But... could you really do it any better? I don't think so. Only 3D techniques might help, because they take advantage of GPU for alpha-calculations and clipping.
Let's begin by saying that writing libraries is much harder than writing a stand-alone code. The requirement that your abstraction be reusable in as many contexts as possible, including contexts which you haven't though of yet, makes the task challenging even for experienced programmers.
Amongst libraries, writing a GUI toolkit library is a famously difficult problem. This is because the programs which use GUI libraries range over a very wide variety of domains with very different needs. Mr Why and Martin DeMollo discussed the requirements placed of GUI libraries a little while ago.
Writing GUI widgets themselves is difficult because computer users are very sensitive minute details of the behavior of the interface. Non-native widget never feel right, don't they? In order to get non-native widget right -- in order to get any widget right, in fact -- you need to spend an inordinate amount of time tweaking the details of the behavior.
So, GUI are slow because of the inefficiencies introduced by the abstraction mechanisms used to create highly-reusable components, that added to shortness of time available to optimize the code once so much time has been spent just getting the behavior right.
Uhm, that's quite a lot.
The most simple but probably obvious answer is that the programmers behind these GUI apps, are really bad programmers. You can go along way in writing code which does the most bizarre things and it will be faster but few people seem to care how to do this or they deem it to be an expensive non-profitable time wasted effort.
To set things straight off-loading computations to the GPU won't necessarily fix any problems. The GPU is just like the CPU except it's less general purpose and more a data paralleled processor. It can do graphics computations exceptionally well. Whatever graphics API/OS and driver combination you have doesn't really matter that much... well OK, with Vista as an example, they changed the desktop composition engine. This engine is far better composting only that which has changed, and since the number one bottle neck for GUI apps is redrawing is a neat optimization strategy. This idea of virtualizing your computational needs and only update the smallest change every time.
Win32 sends WM_PAINT messages to windows when they need to be redrawn, this can be a result of windows occluding each other. However it's up to the window itself to figure out whats actually changed. More than so nothing did change or the change that was made was trivial enough so that it could have been just preformed on top of what ever top most surface you had.
This kind of graphics handling doesn't necessarily exist today. I would say that people have refrained from writing really efficient and virtualizing rendering solutions because the benefit/cost ration is rather low/high (bad).
Something Windows Presentation Foundation (WPF) does, which I think is far superior to most other GUI API is that it splits layout updates and rendering updates into two separate passes. And while WPF is managed code the rendering engine is not. What happens with rendering is that the managed WPF rendering engine builds a command queue (this is what DirectX and OpenGL does) which is then handed of to the native rendering engine. What's a bit more elegant here is that WPF will then try to retain any computation which didn't change the visual state. A trick if you may, where you avoid costly rendering calls for things that doesn't have to be rendered (virtualizing).
In contrast to WM_PAINT which tells a Win32 window to repaint itself a WPF app would check what parts of that window requires repainting and only repaint the smallest change.
Now WPF is not supreme, it's a solid effort from Microsoft but it's not the holy grail yet... the code which runs the pipeline could still be improved and the memory footprint of any managed app is still more than I would want. But I hope this is the kind of answer you are looking for.
WPF is able to do some things asynchronously rather decent, which is a huge deal if you wanna make a really responsive low-latency/low-cpu UI. Asynchronous operations is more than off-loading work on a different thread.
To summarize things slow and expensive GUI means too much repainting and the kind of repainting which is very expensive i.e. the entire surface area.
I does to some degree depend on the language. You might have noticed that Java and RealBasic applications are a fair bit slower than their C-based (C++, C#, Objective-C) counterparts.
However GUI applications are much more complex than command line apps. The Terminal window needs only to draw a simple window that doesn't support buttons.
There are also multiple loops for extra inputs and features.
I think that you can find some interesting thoughts on this topic in "Window System Design: If I had it to do over again in 2002" by James Gosling (the Java guy, also known for his work on pre-X11 windowing systems). Available online here[pdf].
The article focuses on the positive side (how to make it fast), not on the negative side (what's making it slow), but it is still a good read on the topic.