Main Area and Open Discussion > Living Room
Are You Ready to Switch to GNU/Linux?
40hz:
@TMan & Ren - You guys are starting to generate a bit more heat than light right now. Could we maybe go back to where it was still friendly and interesting? :)
Tuxman:
Well, it's cold outside!
f0dder:
C++ is not "slow". Compare the startup time of a C++ and a Java application and shut up.-Tuxman (January 01, 2011, 12:19 PM)
--- End quote ---
Startup time is not necessarily the best indication of execution speed anyway - there's scenarios where both Java and .NET will likely be a better fit than normal-style C++ code.
Edvard:
RE: Linux games.
I am going to stand on this point until a hole wears in my foot, so please bear with me. ;)
1- I am not and never have suggesting somebody develop solely for Linux with commercial ventures in mind.
All the cases I've quoted and claims I've made are coming from a cross-platform mentality.
So, OF COURSE you're going to get your bread and butter from Windows customers, sheer numbers make that argument for you, it would be more than stupid to think otherwise.
My point is that if you can get bread and butter with a little jam on the side for minimal extra investment, why would you NOT do it?
I know, I know, time is money, and there's this great joke about Linux being "free" if your time is worth nothing...
Yes, it'll take time to get familiar with another operating system, but many people have already done that developing cross-platform for Mac, and once you learn a programming language, the battle is half won.
I know it doesn't universally apply to every application, but in many cases it has been proven to work, and the more this happens, the quicker the myth dies and everyone benefits.
2- Small shops vs. big devs.
Every case I've linked to are to game production houses that are arguably not all that big.
Can anyone argue that 2DBoy is bigger than Adobe, or Frictional Games better staffed than Intuit?
No! Yet they are the ones ACTUALLY MAKING MONEY from cross-platform development.
This is not an argument of $1 from 1,000,000 people vs. $100 from 100 people, this is being able to charge $1.50 if you invite 10,000 more people to the party (1% of 1,000,000 is 10,000... just sayin').
You say the data is bad because there aren't that many Linux games, but you're forgetting one point; they're not selling to Linux users per se, they're selling to GAMERS, and gaming is a fairly saturated market yet it consistently sees more commerce than even Hollywood.
The fact that people purchasing games to run on Linux even makes a BLIP in the radar is beyond amazing.
BOTTOM LINE -
If somebody doesn't WANT to develop cross-platform or even just not for Linux (I see a LOT more PC-Mac versions than PC-Mac-Linux), that's perfectly fine.
It would be nice to just be told the truth, rather than rapidly-disappearing myths of revenue lost on development to a non-paying miniscule market share.
What software company (yes, even small developers) HASN'T lost money on development costs before catching stride?
Pull up your pants, look Linux users in the eye and say "I don't like you people and you are not worth my time".
We will go elsewhere, and Elsewhere stands to make a tidy profit off of us, I can assure you...
Renegade:
Looking at the guys at Hemisphere games, wow. They're serious heavyweights.
Anyways, I think I didn't phrase things very well there.
If you can pick up an extra 10%, like some game developers mentioned above, well, 10% is 10%. The only real consideration is whether you can get it there for less than that 10%. Which is pretty much determined by the size of the company/revenue.
e.g. For a company with 10 people and $1 million in revenue, it's a much more important decision than for a company with 10 people and $5 million in revenue. Resource allocation in the first case is much tighter.
For a lot of developers, they have 1 or 2 people, so the consideration is much more important and needs much more consideration.
If you have revenue of $100,000 from a product, an extra $10,000 may or may not be worth it. If you can get another platform version done in 2 months, then it's certainly worth it. If it takes you 6 months... Things are getting scary because you're messing with the food on your table at that point.
A sort of twisted inverse is also true -- not the size of the developer/revenue, but the size of the product. Very large products (code base that is) are more difficult. e.g. Could Adobe port it's Creative Suite to Linux? The complexity there is pretty much a show stopper.
I came off too quick saying, "...this is a bad argument... At the moment." The qualifications for that don't come until later.
Most developers are very small shops. Once you start going through and looking behind a lot of them, it's "a guy in a garage". It's really them that I'm talking about.
Incidentally, the last game I bought is running on Linux. Well, Android, but still Linux. :)
I'm currently facing the cross-platform issue and will need to decide in a couple months or so. Go truly cross platform? Cross-platform then fork? Go with independent versions? Different languages?
I've been thinking about this for a very... very... very... very... very... LLLLLLLOOOOOONNNNNNNNGGGGGGGG time...
Here's an example that I've pulled out from the source code of GDT. This particular snippet was written about 4 years ago:
--- Code: C# ---// F**k... Have to use the Windows API here -- no managed C# way to do this. // For Windows Mobile, replace user32.dll with coredll.dll [DllImport("user32.dll")] [return: MarshalAs(UnmanagedType.Bool)] static extern bool SetForegroundWindow(IntPtr hWnd);
The first comment line is echoed everywhere I use it.
I'm hoping that I can go back and find something new to replace that though.
Still, even years ago I saw Mono and knew that it would be an answer in the future.
I like the idea of having a little conditional compilation as possible, and having the same source base. I really don't want to have a C#, C++, and Objective-C version.
I also like managed code, but will certainly use unmanaged if the case warrants it. Here's a line from the Duplicate Photo Finder I wrote for NANY 2011:
--- Code: C# ---public unsafe UInt32 Hash(Byte[] dataToHash)
And it not working :( in Ubuntu:
Are You Ready to Switch to GNU/Linux?
The file browsers don't work there because it's coded for the Windows file system. I've not investigated how to fix that.
But the application is very simple and already there are problems. File systems. Yuck. What a lame reason for something not to work properly... Sigh...
The number of problems that need to be solved for different platforms makes things tough.
It boils down to risk management and resource allocation (human resources and developer resources, not memory :) ).
I'm quite certain that Linux offers significant profit. The real question for developers is if they believe that their time investment will pay off.
You're quite right though about the market being for gamers there. It's saturated, and difficult. I know guys that do games, and they're constantly plugging away to make things work. It's hard.
With technology advances, the bar comes down though. The size of the developer/revenue is a good indication of when they'll be able to jump over the bar. e.g. For a free and open source program like Duplicate Photo Finder, the motivations for me to get a Linux version done are pretty small, and not really enough to get me to jump over that bar and do it. However, it would be an excellent exercise for me to do that I could take information away and use elsewhere. Got me thinking now... Sigh... I need to stop and get the house cleaned. It's a disaster.
In any event, it looks to me like Mono is still one of the best options. This is the year that I'll be diving in and (hopefully) swimming the icy cross-platform channel. At the moment I think I'll be starting at the end of January.
And with any luck, I may be able to get you to give me some money~! :D :P
I think I'll start on the laundry first...
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version