Can Linux be a practical alternative OS to Windows for gamers? Paul Sullivan investigates the issue by talking about the C-based Simple DirectMedia Layer (SDL), the Linux equivalent to DirectX. The article also got a lengthy interview with Sam Lantinga, the man responsible for the SDL initiative.
SDL – The Little Engine That Could
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
SDL is a cross-plattform library – not a “Linux-DirectX”.
SDL wraps the calls – it’s not an architecture of its own. For example, SDL for Windows uses DirectX.
Even the article says it:
“1. FiringSquad: Did you initially envision SDL as a direct competitor to Direct X?
Sam Lantinga: No, SDL was originally designed as an API to provide the services that multimedia applications need across many different platforms. This is still the intent, it just happens that it works very well for developing games on Linux where there isn’t a DirectX equivalent.”
Someone who responded to that article brought up an interesting point …
If more commercial games were available for Linux, would Linux users (many of whom demand that every f**king app they use be free and open source) actually buy/pay for them?
The text was copied/pasted by FiringSquad’s introduction text, which is also can be used for third parties who link to their articles. It is not my text.
Although Loki failed for many reasons, one of the big ones was the fact that nobody bought the games. (If I remember correctly) The Unreal Tournament special release (in the tin boxes etc) sold around 5,000 units. Loki had ordered 500,000. You can blame managment for some of it, but the community voiced very strong support for the game(s). Unfortunately, people failed to put their wallet where their mouth is. This seems to be a trend [in the Linux community]. Alot of people will spout off about one thing or another, but they consistently fail to back up those “claims” (couldn’t think of a better word) with cold hard cash.
Then again, very few commercial packages out there are actualy worth paying for….
Again, a Linux fanatic on a holy war …
“For some, the whole point of choosing Linux is because it is NOT Windows. They want apps that don’t have any ties to Windows whatsoever, and I think I am starting to see myself in that category.”
‘nough said about this guy credibility … Many say they are not fanatics, but their REAL priorities are :
#1 – Being NOT Windows
#2 – Being of quality
as in …
#1 – Only *MY* God is right
#2 – My God is about love and peace
It goes like this :
“Mandrake ?” – “Kewl, dude !”
“Games ?” – “Kewl, yeah, all of them!”
“Buying? with money ?” – “Not cool, nah, later”
The best example was BeOS and Gobe Productive – million downloads of FreeBe and probably $0 revenue from Gobe Productive sales.
I know that the “Civilization:Call to Power” game from Wildcard Design, while it was well marketed in the BeOS community, and we at BeNews helped as much as we could with ads and news stories, it only sold less than 500 copies.
I think the problem is that these markets are very small, not *just* because the people do not want to buy. In the alternative OS markets, there are two problems:
1. People are expecting all alternative products to be free, just because Linux is free and in these days, it flags the “alternativeness”.
2. The markets are very small. Just yesterday we were discussing how Linux holds less than 1% of the desktop market (with WebSideStory actually saying that they have only about 0.25%, but I believe it’s got more than that, having just a bit less than 1%). Naturally, there are not many people that can create a market with this 1%, especially when these people expect everything to be free in that “alternative universe”.
Yours is a valid point – but – there is allways a but.
Lets use some over simplified numbers…
Lets say that the Linux community consists of 100,000 users.
Lets say that 10% of those users buy games. (for sake of simplicity, one game)
..Running this thourgh my 64 bit abacus….
that’s 10,000 copies.
So you’re right – the market is *tiny* compared to the other gaming markets, but I think that that would be enough for a niche market company to survive. If all the linux kiddies actualy spend money on the cause that they so much believe in.
Could it be?
Could it be that %90 of the linux community consists of loud-mouthed 13 year old running windows? Me thinks that’s quite possible….
BTW
Does any one know how much it costs to license a game for porting? I think the key to this market is *tiny* companies doing one port at a time. If it’s something reasonable, say $25-$30 grand, I see no reason why someone along with a couple of their closes hacker friends couldn’t pull something like that off.
Latest stats show that the Linux community actually consists from about 10 million users.
Problem is that not many users buy games. Games do not sell as much office & related software sells. And the people who do buy games, they mostly buy it for their Windows boxes (everyone keep saying: “I keep Windows only to play games”), and mostly, for their consoles. In fact, 80% of the people who actually have Linux install, they do not have GLX/DRI loaded in their stock XF86Config-4 file, their 3D drivers are not setuped at all. And they don’t bother either, because most of the time, depends on the gfx card (and not all gfx cards under X11 have free 3D support), it is a tedious job to make 3D work. Even I had problems under Gentoo making the Voodoo5 work in hardware mode (the changes did not stick, I was losing hardware acceleration every time I was rebooting – tdfx.o module was correctly loaded btw).
Also, there is one more problem here. Games are selling based on personal taste. With Office suites or graphics apps, it is not quite the same. An application either does or does not do some requested features and based on its (in)ability, it will sell or not. While with games, everyone has his/her personal likings (some people like arcade while others like FPS games), so they purchase according to these likings, which makes the overall alternative gaming market even more limited.
I buy the games I like, I bought many psx games recently and no linux games, I’ll buy NeverwinterNights because I do like it and I can run it on my system (ok GNU/linux).
> Lets say that the Linux community consists of 100,000 users. Lets say that 10% of those users buy games. (for sake of simplicity, one game) That’s 10,000 copies.
The 10,000 users who are gamers and who are willing to spend money on Linux games, do not buy the same game. As I said above, the same game does not appeal to all gamers. This really cripples sellings, a lot. Plus, all the above reasons we discussed here, really makes this small market, already problematic to try selling (especially) games.
It is free in almost all basic distributions, is not reliant on the Microsoft juggernaut and is heralded as being faster and more stable than any version of Windows except for Windows 2000
Linux is slower and less stable compared to Window 2k???
don’t just say, back it up with a reason
anyone know, please enlighten me
I completely agree with everything that you’re saying. But the cross eyed optimist in me thinks that it’s possible to make money selling to niche markets. I’m not saying one will get rich, but I think that only matters to VCs.
In you personal opinion, it’s not possible to make a linux game and have it turn a profit even if it’s tiny? Even if you hit only 1% of 10 mil, at $60 a pop, that’s quite a bit of money.
I dunno… I give up
1. Sharing with friends. “May I borrow your CD” doesn’t work well in linux world – your system may look the same Red Hat as your neighbour’s but the game doesn’t go well – missing patches, incorrect libraries etc.
2. License key protection – you can get serial code for almost every Windows product from the web. The dark side of popularity.
3. Compare number of sold copies of “Linux for dummies” vs. “Windows for dummies” – it would be a good estimation for gaming industry too.
It would be interesting to see how well enterprise level soft for linux is selling – I mean Oracle and other database vendors.
> Linux is slower and less stable compared to Window 2k???
don’t just say, back it up with a reason. anyone know, please enlighten me
I find my WindowsXP to be faster even more than the fastest Linux distribution, Gentoo Linux, when it comes to UIs.
HOWEVER.
“Faster” is a very vast and general word. DEPENDS what do you mean by saying “faster”. When it comes to serving web pages or using it as a server, and having Linux on text mode all the time, quite possibly, Gentoo Linux might be faster than WinXP in these jobs. But, when using it as a desktop, XP’s UI is much more responsive than XFree/KDE/Gnome. And don’t tell me that Blackbox is faster than KDE/Gnome. IT IS faster than KDE/Gnome, and this is why I use it. But you can’t and you should not compare a full UI like XP’s that does so many things in the background, with Blackbox which is merely a window manager plus a small dock taskbar.
So, depends what you measuring as “fast”. Some things are not measurable because they are different in the way they work. But overall, as my desktop, I would not change my XP for Linux. Only BeOS had managed to have a more responsive UI than XP’s. However, BeOS lacked speed in almost all other areas. Operating system is not just the UI you know.
from my experience gnome is the best for speed…
I can even compile some heavy program like a Xfree DRI or a mozilla ans still listen to my mp3 on xmms and webbrowse on galeon or mozilla, some less heavy compiles let me play movies on mplayer at decent framerates/smoothness.
do that on XP and you’ll just see the UI crawl…
BTW how is possible that you like that playmobil dummy interface?
What playmobil dummy interface? Please use the Header/Synopsis to “Re:”, so at least we know to whom you are replying to.
>from my experience gnome is the best for speed…
Gnome is faster than KDE 3 indeed (in fact I wrote that earlier today: http://www.osnews.com/comment.php?news_id=1016&offset=15&rows=20#15… ). But that does not make it good, UI-wise. Because it is not.
And if the “playmobil dummy interface” was for XP, I do not agree with you at all. Microsoft has paid millions in UI research and R&D, personally I find the WindowsXP UI to work very, very well. The only changes I would do to it, is to clean up a bit the new Start Menu and that’s about it.
Anyways, our subject over here is SDL.
XP UI just sucks, a great way to waste money so=P
the broad border is just a wast of screen estate as well the huge icons crammed sidebars…
Just to tell few of the things I noticed
1st: if you don’t like the xp ui, make it look like win2k
2nd: xp has the same problem as all windows versions have ever had, the registry. yes, when you first install it, it’s quick. then, after you use it for a while, install/uninstall apps, the registry start to grow, A LOT, and the overall performance the the OS declines. this always has happenned to me, and, even though I don’t use xp, my co-workers do, and have been for a while, and they complain about this issue
3rd: isn’t this an SDL discussion
4th: see number 3
I have bought several linux games, but honestly I’m not a huge huge gamer. The games I’ve gotten lots of play on have been Unreal Tournament (the windows version and dl’d the Linux installers), Q3 Arena (got the Loki tin). Currently I have TacOps, Urban Terror (love this one), Myth2, RTCW all on my system.
The big problem is that it doesn’t seem like there’s really a lot of killer killer games out there, and the few ones that ARE killer haven’t been released for linux (halflife, system shock 2, deus ex, baldur’s gate 2, etc). For me I refuse to dual boot or anything so I decided not to play any of those games.
I recall reading a report that Richard M Stallman of the FSF had gone on the record saying that Computer Gaming was one area where on any platform the development would probably end up being commercial, as the Free Software movement wasn’t really up to producing the kind of software Gamers want – nor did it have the momentum to develop final versions as gamers on the whole werent interested in being long-term beta testers
I think part of the problem is anyone who makes something rather big and complicated for linux like a game or office suite is likly to want to get money for, IE their a company IE loki, and soon GOBE. To make any profit especialy with such a small market their price will be high, and since linux people in general (not all, but really in general) don’t care for buying stuff and 50-100 bucks for something is just to much to them and those companies go bust. Soon as you get to smaller apps you have an explosion of options and tons of free ones which leaves you with no market there. So you can’t get any cheap software that a company might make money with. If the $1-25 arena could exist many companies could get a foothold on something. Linux users will pay this much for something. But anything that would cost this much has a free option out there. This probly was an obivious statement but hey no one else had said it
Unless Linux games are released kinda in the same time frame as the Windows version – why would I buy it? As with Q3A, RTCW, UT, etc… you only need a binary + a Windows CD of the game…
Apart from that, I don’t believe in this packaged game thinggy any more. I bought Linux Q3A twice and I understand a lot of money was spent creating it. But as someone already mentioned for UT, this is risky to do on Linux. Why not consequently stay with D/L binaries only? A lot more people would buy a download binary as an add on for their Windows version for about 15 USD instead of the whole boxed game _again_ half a year later…
As well, when Q3A and UT hit the shelves, Linux was waaay back compared to today in terms of installation and graphics drivers, if you take a current Suse as a measure…
I remember myself as a newbie trying hard back then but I always failed – and many people don’t even see a reason to try if they already have Windows…
I see the note above telling me that my comments will not show up instantly… but now I dropped my GF at work, came back home and half an hour later, it is still not in the thread – are you guy’s pre-reading them for a possible censor ship or wot?
I like the XP look, I run the standard look (except I make the titlebar on windows smaller, and I rearrange my explorer to be tighter, and move the taskbar to the top and use the old start menu (since I know how to ^_^)) and it’s great. I work with computers so I want it to be easy to use, not being able to add a million themes that all look like crap (the only good ones I’ve seen is a few for KDE, and the new vector icons are just great, go for it!).
SDL, well, wrapping something is never fast, which is why DirectX came about, and why windows needs to be scaled down (as in removing the old, nothing else), partially rewritten and emerge better. It gives you yet another layer, and layers are only good when speed isn’t your main concern. Games need speed (belive me, I want the FPS).
SDL is a C library. This unfortunatly makes me not want to touch it, as I am OO all the way. I am not trying to start a C/C++ war (please don’t), mearly pointing out that because of this some will not want it. It’s worth a thought.
Most linux based OS that runs on machines today run as servers, or on students computers, etc. These are not markets that are strong when it comes to software sales. This is why Microsoft is way less concerned about linux than you think. The only way you are going to get your games on linux is emulation (of DirectX for instance), SDL (or similar), or the grace of the developers that will use up their companies money on a linux version that won’t pay.
If there would be a large demand for linux games, rest asure that they will be made and thus sold and then more be made. Think about that the next time you warez a game and complain (if you do, if you don’t, you are an unusual but good person. Keep at it).
OK… under XP, I got the movie running already, but how do I start compiling..?!
Everyone seems to mis the media players (spec. DVD) which use SDL, for me these are more important than games.
If you want games, at the worst just use a console.
The reason so few games sell on Linux is rarely due to “zealots” demanding everything be zero-cost or open source. The problem I personally found is trying to get the damn things in the first place – I don’t have a credit card (never needed one), and none of the shops in my home city sell a single Linux game because the market is so small – of course, there are thousands of Windows games, many of which don’t sell very well at all. I haven’t come across a single Linux user who demanded that Quake 3 be free (rather they are happy to pay for it).
btw – Loki had problems with mismanagement of funds – they did sell games quite successfully to Linux users, but the market isn’t big as anyone who wants to play games in the first place is either going to dual-boot Windows or have a console.
btw – could we stop all this nonsense about anyone using linux being “zealots” – it seems that anyone who uses linux is automatically called a “zealot”. Trading in childish insults implies that there is no genuine substance to the argument. Yes, there are Linux users who want everything to be free/libre, but most are pragmatic and just use it as an OS that works the best for them, same as some of you like MacOS, Windows or BeOS (or whatever!). I sometimes wonder who the real idealogogues (sp?) are.
Just do not put their money where their mouth is. Let me qualify that by saying that I am the founder of a Linux Users Group and also started a UNIX Study Group at a major international corporation.
I personally buy all of the software that I can and actually want, when it comes out on Linux. I bought Quake III, RailRoad Tycoon II and will be buying Star Office 6.0, when it comes out.
I will do all of this, even though I am currently employed as a Network Administrator running a mostly Window environment.
My feelings are that Linux does stand a good chance of succeeding, it just needs the actual backing of its users in order to do so. That means that the OS must be purchased, even if you can and have downloaded it. Think of that as a thank-you to the company that worked hard at providing the easy to install distribution of Linux.
I am a Mandrake-Club member. I have also bought SuSE Linux, Red Hat and Mandrake-Linux in the past 6 months. Sure, I bought the lower cost releases of those Operating Systems. The thing that matters is that I did buy those Operating System.
If there were more Linux Users like myself, I am certain that there would be more robust, commercial software, games and support in corporations for the Linux Operating System.
Pesonally, I like ID Software’s approach to linux games. They dont just sit there and wait for someone to port their game for, they do it themeselves. Which causes somewhat of a problem, well…in my case atleast. Most often they only release the binaries and we must wait for ages before sources get out. If you dont have an x86, you cant run those binaries, so the sources do get handy.
“Only BeOS had managed to have a more responsive UI than XP’s.”
You forgetting AmigaOS? It was more responsive on a 30mhz machine with 2 meg ram than OS’s using 20 x faster cpus today.
SDL rocks; I’m using it under Windows 2000 to keep my brain from imploding and it’s just great.
Even greater, PyGame ( http://www.pygame.org/ ), implemented with SDL, and PyOpenGL ( http://pyopengl.sourceforge.net/ ) mean I don’t have to battle Visual Studio.
My nascent CRPG engine cranks over 200 frames/second (dual P3 666, Radeon 64 DDR) rendering with OpenGL from Python. For an interpreted language and a totally unoptimised engine, that kicks butt.
– chrish
Eugenia:
What does a Desktop buy me that I don’t get with a Window Manager? Maybe I am clueless, but I used IRIX on an SGI Indy all through college with a CDE-like WM. It had multiple desktops, a pretty background, configurable Right-Click menus, and icons. What else do you need?
What does Gnome/KDE provide besides a taskbar and something to click on in the lower left that guides me through all of the programs I have installed?
Keep in mind that I’m trying to be serious, not sarcastic.
>What does a Desktop buy me that I don’t get with a Window Manager?
Did I ever said the opposite? I even use BlackBox. The thing is, a desktop environment has more things in it, it has shortcuts, it has an *integrated* filemanager/desktop and maybe even a browser, support for another layer over the X root window (in order to support icons and background images), it has “tsr” or other smaller programs in memory etc.
What I was trying to say is that it is not fair to compare the speed of a simple windowmanager to the speed of a whole desktop environment. In a desktop environment, a lot more things going on underneath, that can require more power or memory. Most of the time, users do not realize that these things run underneath.
As I said, I am using Blackbox myself on my Unices, which is a simple windowmanager and nothing more. But you can’t and you should not compare the speed of simple window managers with XP/MacOSX/KDE/Gnome desktop environments. It is not fair. Apples and oranges, even if the end result may or may not be the same for a user.
My 2 1/2 cents:
I think Eugenia is right, the gaming market is small for the alternative OSes, so perhaps it isn’t worth it to port games over. Thinking broadly, if you could get a linux user to pony up $60 for a game (hell, you can’t even get them to buy a distro for $30-70, look at Mandrake!), then maybe you get $60k out of sales.
If you need to have an office, computers, programmers to port the game, etc., to commercially produce a Linux game, it’s not a viable business plan.
Keep in mind that bulk ordering packaging, CDs, etc., is less expensive the greater the units ordered (inverse relationship-more expensive to produce 10,000 units that 100,000).If you produce 10,000 copies, the the packaging of the code would probably run about 60% of that $60, around $36K per run. And that’s without purchasing the license for the game from the original developers, or even paying for continuing operations. You get behind the 8 ball quick, and can’t get out.
The transgaming idea is better-emulate directx via wine, and play the games emulated. The down side is speed, rendering, and in my case, lesser game play (my Dell just ain’t quick enough!). But it is a start.
As for Desktops, yes, Windows XP’s UI is a memory hog, and not all that useful to boot. Most of the time, I don’t even use it. I have a shell switcher installed(SHELL ON) that allows you to kick explorer out of the way and load Talisman, Litestep, Winstep, etc., instead.
I really like the Litestep interface, which is alot like NEXT/Windowmaker in functionality. Small foor print too.
There is even a port of blackbox for windows that works well, although is a little buggy under XP. Not enough to stop me using it, mind you, just not as stable as Litestep.
All I can say is-get rid of explorer, and be happy!
If more commercial games were available for Linux, would Linux users (many of whom demand that every f**king app they use be free and open source) actually buy/pay for them?
I believe they would and I believe they do when available.
I think the percentage of Windows users who buy computer games is probably equivilant to the percent of Linux users who buy games. However, the percentage of Linux users is a lot less than the percentage of Windows users. Therefore, there will obviously be less games purchased for Linux than there are for Windows.
Although Loki failed for many reasons, one of the big ones was the fact that nobody bought the games. (If I remember correctly) The Unreal Tournament special release (in the tin boxes etc) sold around 5,000 units. Loki had ordered 500,000.
It was not Unreal, it was QuakeIII (which was a disappointment in my mind). I would have bought the Linux version of the game had the game been worth buying.
I read the Loki story, and as far as I can tell, the problem wasn’t really that nobody bought games, but rather the company and its finances were poorly managed.
“SDL is a C library. This unfortunatly makes me not want to touch it, as I am OO all the way”
From the main page at http:://www.libsdl.org :
“SDL is written in C, but works with C++ natively, and has bindings to several other languages, including Ada, Eiffel, ML, Perl, PHP, Python, and Ruby.”
‘nough said about this guy credibility … Many say they are not fanatics, but their REAL priorities are :
#1 – Being NOT Windows
#2 – Being of quality
I think its funny how people that hate MS and like Linux are fanatical, religious zealots and people who hate Linux and like Windows are level headed, open-minded folk who have transended the human existance to a more enlightened state of being.
It reminds me a little of the same mentality and open-mindedness of animal rights activists, vegetarians, environmentalists, pro-choice advocates, homosexuals, etc. when compared to the fanatical majority of the Earth (only in the Windows vs. Linux case, the Linux crowd is a minority).
First, I’ll plug XFce again as a Linux/BSD desktop. It’s very fast, and while it’s not as feature-rich as Gnome/KDE it shares a lot of their strong points and it’s considerably more responsive.
Second–sure you can make money selling games in a niche market, but you have to be realistic about it. The boxed retail market is probably not realistic no matter what your title is—you’re making, what, 50% of the box’s cover price after distribution and store shares, and you’ve spent at least 50% of that on various packaging costs. And spending tens of thousands of dollars on licensing a port from another company, waiting on their development team, getting your own team working on it, and shipping 6-18 months after the original version shipped is not realistic. At that point you’re competing with the same game in the $12.99 bin.
So what is realistic? Following a model like Ambrosia Software on the Mac, or some of the earlier PC shareware companies. Do completely original stuff rather than ports, sell it only online, and sell it at reasonable prices.
The biggest problem with that on Linux is that, no matter how much Linux fans like to protest otherwise, they are pretty enchanted with “free as in beer” as a concept. The advantage games might have over other kinds of applications is simply that there aren’t many commercial-class free games out there. AbiWord 1.0 is approaching competitive with other “non-Word” word processors, Gnumeric is a better spreadsheet than gobeProductive’s, and NEdit is pretty comparable to BBEdit and Pepper. But there ain’t no good freeware answer to Phobos Rising, Myst or Tempest 2000, much less any recent Final Fantasy, Counterstrike or Half-Life.
I disagree with you Eugenia. I have two identical machines, one with XP and one with Debian. The Debian box boots faster, is more stable and loads programs that run both on Linux and Windows faster than my Windows XP machine.
It is not fair to compare the launch speed of IE on Windows to Mozilla on Linux (or Word vs. any of the myriad Linux office suites) because Microsoft loads parts of these programs into memory when you start the machine (making boot time longer, which sucks on an OS that constantly needs rebooting). This offers the user the optical illusion that Word starts fast, when in reality it is only loading a small part of the program when you click the Word icon.
Therefore, I use programs that run both on Linux and Windows in order to compare the speed and stability of both of these systems. In most cases, Linux is superior in speed and stability (even while running KDE 3). 3D is still better under Windows (as far as I can tell) but that is about it.
In the realm of stability, Windows doesn’t even compare!!! Yes, XP is better than previous versions of Windows (except for Windows 2000), but I really USE my machines and expect a lot out of them. The longest I have had XP up before it either crashed or became so incredibly slow that I had to reboot, is four months (which is good compared to versions prior to 2000). However, I have had a debian machine running without a reboot for over two years without any slowdown. My current home installation of Debian has been up since February without a reboot (that’s when I installed it).
Four months up time may be great for you, but knowing that the system stability is slowly degrading during that time should be of concern. It is to me.
Also, you mention XP vs. Blackbox wasn’t a fair comparison because Blackbox doesn’t do a lot of crap in the background. I again disagree with your statement. Linux can do just as much in the background as Windows can. Not everything Windows is churning away on is GUI related. Also, isn’t it nice to be able to decide if you want to waste CPU cycles on shading and crap or not? I think it is.
I use Windows all the time for work. I like that it makes money for me and will use it as long as such is the case. However, my personal time is spent using Linux. Since my time is also of value to me, I choose not to waste it rebooting Windows.
Finally, you have made the statement that XP is faster numerous times in these posts, but not once have you said what exactly it is that runs faster for you. I would be interested to hear.
> I have two identical machines, one with XP and one with Debian. The Debian box boots faster
Are you telling me that Debian boots in 15 seconds?
> It is not fair to compare the launch speed of IE on Windows to Mozilla on Linux (or Word vs. any of the myriad Linux office suites)
Did I ever compared them? But overall, it is something that strikes the user indeed.
<sarcasm> It’s not fair when MS pre-loads the most commonly used app in memory, because then it loads too fast.</sarcasm>
As if it was a “feature” that a full-featured browser should take long to load.
> Finally, you have made the statement that XP is faster numerous times in these posts, but not once have you said what exactly it is that runs faster for you. I would be interested to hear.
I do the same as I do under Linux or BeOS these days. Browse the web, use a graphics app, use ICQ/MSN/Y!/AIM, use a text editor to write PHP or HTML, use the filemanager, play mp3s and I am also using an email client. That’s about it, believe it or not. Simple things an internet-oriented user would do. And as this “simple internet user”, XP does it better for me. Working with these simple apps, it is easier/simpler/faster/better under XP than any other OS. I like how the UI behaves (*not* how it looks – I believe its new window manager is too “fat”, but overall it looks ok) and how it does what I expect it to do. Under other OSes or window managers, I don’t get the same amount of what XP delivers. KDE is close, but it is overly bloated in its menus that does the whole thing more frustrating chasing all these little options somewhere in some conf panel. Gnome is ugly as hell and its UI doesn’t make sense. I can’t stand Gnome for more than 5 minutes. The rest are minimalistic and does not offer a complete solution for a desktop (however, I DO user Blackbox or XFCE most of the time – faster). Different shortcuts for Gnome apps under Blackbox than the KDE ones. A mess.
BeOS was closer but it does not have a good or fast browser (Bezilla is still slow) or a good gfx app. MacOSX is closer too (but IE is not as fast or stable as under XP). X11 and its all wms are not close at all as I described above. X11 has the applications I need, but the overall feeling of the UI *behaviour* is not there. Too much fragmentation and different looks between all these toolkits and also X feels clunky.
If you are living in the Bay Area, you are welcome to come to our home and I will show you on my X11 what I mean by “clunky”.
XP wins *as my desktop*, hands downs. That’s the reason it most people like it. There is nothing superficial or a connspiracy behind it. IT IS GOOD.
*As for a server*, our home server is FreeBSD. Not XP. Reason: FreeBSD runs faster in text mode on our K6-450 Mhz/64 MB RAM, and most of all, it is free.
What does a Desktop buy me that I don’t get with a Window Manager? Maybe I am clueless, but I used IRIX on an SGI Indy all through college with a CDE-like WM. It had multiple desktops, a pretty background, configurable Right-Click menus, and icons. What else do you need?
What does Gnome/KDE provide besides a taskbar and something to click on in the lower left that guides me through all of the programs I have installed?
Keep in mind that I’m trying to be serious, not sarcastic
>>>>>>>>>>
1) An integrated environment. Apps look the same, behave the same, use common services, etc. This includeces a common communications, clipboard, scripting, preferences, filetypes, help, component, and copy-and-paste model. This is stuff that should be common to *all* graphical applications (not just those using the same toolkits) but often isn’t.
2) Better APIs: KDE and Qt offer tons of developer APIs that encompass everything from basic functionality (threads, memory allocation, etc) to XML processors, etc. The advantage of one toolkit offering all these services is that apps don’t reimplement what they don’t have to, and improvements in one part of the library percolates to all applications using that toolkit.
3) Powerful abstractions: KDE abstracts local and remote filesystems, compressed file hierarchies, etc.
Now, given all that, desktop environments are “good” for end-users. It is expected that they will be slower. However, Windows 2k does just as much as KDE and GNOME, yet is faster. *That* is the root of the problem, not whether or not Blackbox is faster than Win2K.
1. Microsoft spends millions on UI research, but that’s futuristic stuff like speech processing, AI, etc. Anyone with any HCI experience will tell you that if microsoft has done any UI research, that research is certainly not being put into the design of their products(with XP, one or two things might have changed, but why did it take them sooooo long?) Any sysadmin, system programmer, command-line geek (i.e. the technical people who will argue that they’ve heard that microsoft spends zillions on UI, so they must know what they are doing), please furnish me with an explanation for Microsoft shennanigans like window-in-window MDI, multi-row tabs, and Office 2000’s adaptive menu system, all of which have received intense criticism from computer usability experts.
but it is overly bloated in its menus that does the whole thing more frustrating chasing all these little options somewhere in some conf panel. Gnome is ugly as hell and its UI doesn’t make sense. I can’t stand Gnome for more than 5 minutes
2.KDE people can’t seem to understand the concept of sane defaults. They’re like “well, if you dig here, dig there, go to this hierarchical menu, go to that one, talk to the one-armed man down the street and click there, then you can change this stupid feature”. You are also very right on KDE being very cluttered, too. The more clutter there is in an interface (espcially with millions of ridiculously tiny toolbar buttons), the longer time on average a user spends doing a visual search, even if they know where to look for things. Unfortunately, the folks doing KDE’s UI’s don’t seem to understand this.
But what you can you expect from people who pride themselves on blindly copying Microsoft? You don’t expect intelligence from those who cheat off the stupidest kid in class.
3.As for GNOME ‘s ugliness, I think that one of the big things for many people are the insane widgets that don’t really scale very well to different sizes (i.e. the pencil-thin buttons that stretech half way across the screen”) One of the problems with gnome is that designing a gnome interface is kind of a like designing an interface with html tables. Everything wants to either expand to the other side of the planet or wants to take up as little room as possible. Just as with HTML tables, with enough fiddling and alignment voodoo you can get things to roughly where you want them. But most GNOME programmers (stupidly) don’t have enough patience to practice this voodoo, yet they can’t seem to understand why people don’t have the patience to use their software. If GNOME revised the mechanics of the widget layout, probably some of the ugliness due to bad scaling would go away. If you can’t get programmers to work to design things well, you should try making designing things well less work.
My main points:
1. Microsoft dumb.
2. KDE dumber.
3. GNOME equally as dumb as KDE but in different ways.
>window-in-window MDI
Actually, Microsoft now tends to do everything in SDI.
>KDE…
Amen brother.
>If GNOME revised the mechanics of the widget layout
This is actually the job of the GTK+ team, not Gnome’s directly. Gnome also has other problems, it looks ugly overall, ugly tiny scrollbars, ugly arrows on the menu/submenus and the Gnome Panel is the worst of all. And do not talk about themes. 95% of the Joe Users stay with the default theme. They do not care about themes! So, the default theme should be done right! The Crux theme is not bad btw, because it was made by a REAL COMPANY (Eazel), not from a hacker somewhere in his bedroom.
Right click to the Gnome Panel and you will see all this bloat of menus, frustrating even the people who use Gnome daily. Also, check the icons on the Gnome menu (click on the foot). Under Mandrake 8.0 some icons were on 32×32 and some on 16×16. This made the menu to look so disporpotional and stuff. It is horribly done. It is CLEARLY done by Unix programmers and it has no input from real UI people. Even Gnome 2.0 is just a port to GTK+ 2.0 and it has no new real UI elements that could fix the situation.
The thing that “microsoft spends millions on UI research” it is true. But as a user, I don’t care. I don’t care if Microsoft gets their UIs with research and KDE/Gnome gets it through a deal with the devil. The end user will not say: “oh, I can live with these problems. At the end of the day, they were created by everyday people like me, I should feel happy I even have a desktop!“. It DOESN’T MATTER how and from whom it was created. What matters is the end result. And the end result on X11 is freaking dissapointing. Especially with all this fragmentation, having a desktop with 5 applications, all looking different and behaving different just because they were written with different toolkits. This is STUPID and this way the X11 desktop will NEVER replace or be satisfactory as Mac or Windows or even the already long-dead BeOS.
Are you telling me that Debian boots in 15 seconds?
Yes, what’s so mystical about boot time? My Debian box boots in 15 seconds flat. That includes loading up a couple of daemons i.e. vmware modules, alsa, nfsd, sshd, and the Houdini key server. I get a nice NVidia logo and an XDM login prompt 2 seconds after the login prompt (which would make to total boot time from POST to GUI login prompt 17 seconds). Of course, I don’t really give a fuck about boot times since I only reboot when upgrading my kernel. I’m still amazed at how much abuse my Linux box takes without ever locking up. For my XP/2000 needs VMware simply rocks. WinMX keeps my DSL line busy most of the evening.
There nothing like running Windows iconified on your Linux desktop 🙂
I’ve used SDL to develop little toy apps and play with graphics and such in C++. SDL is sweet. I haven’t seen that nice of an API since I programmed on BeOS regularly. Yeah, it’s not C++, but the conceptual layout is very clean and maps pretty nicely onto an OO design.
I must commend you for your post; it was extremely well-articulated, and I could simply not agree more, especially with your comments in regards to KDE. As much as I want to love KDE from a philosophical standpoint (which is admittedly a terrible reason to love something!), its abounding clutter drives me away every time. And its relentless hope of cloning Windows is a perfect example of tremendous amounts of misdirected effort.
But enough on that. As for SDL, I honestly don’t have really all too much to say (which is perhaps why this discussion has managed to find itself so offtopic). It’s a very nice and clean API… that’s about it. =) As far as cross-platform gaming goes, the sorry fact is that almost no company cares. (I have considerable experience and am closely involved with the game development community, for the record.) The bottom line is churning out code as quickly as humanly possible. Period. Hell, code cleanliness is far from a priority — do you think cross-platform consideration ought to be? The bottom line is that Windows is the only market for gaming, and the only one a sane company would consider support, especially should they wish to compete at the cutting edge.
Are you telling me that Debian boots in 15 seconds?
That is precisely what I’m saying. Plus, I have the added benefit of only having to boot once (unless I upgrade my kernel). You can’t say that about Windows; XP or otherwise.
Anyway, XP is your favorite OS. Great! It isn’t mine, which is great too. However, the experience of many users state beyond a doubt that Linux is far more stable than Windows. And my personal experience is that it is faster, which is why I use it at home.
Most of the reasons you stated for liking XP were asthetics. Percieved beauty has nothing to do with speed or stability.
Besides, as you said, this is supposed to be about SDL, which is very cool and works great under Linux.