[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Forward - sigyn@lokigames.com: Re: E-Mail interview for Linux Game Developers Center]

I received Loki's answers. There are a few issues I'd like to respond to
before writing an article though. Perhaps there are a few _you_ would like
to respond to as well.

Discuss it on the list and I'll get back to them asap.

Rob
-- 
Rob Kaper | mail: cap@capsi.com + cap@capsi.cx
          | web: http://capsi.com/ + http://capsi.cx/
          | "With a bushel of apples, you can have a hell of a time with the
          | doctor's wife."

-- BEGIN included message

>Hi, first of all let me thank you asa gamer for your efforts porting games
>to Linux.
>
>I am a contributor to the Linux Game Development Center
>(http://www.sunsite.auc.dk/linuxgames/) and during a "we need articles"
>session on our mailinglist I came up with the idea of having one called "10
>Questions to Loki".
>
>We came up with 13 questions in the end, even when trying to avoid the
>questions that have been asked too often already. We would really appreciate
>it if Loki could give answers to them. Of course you get to review the
>article before it is published.

Apologies for the delay in responding.
Please see below our answers, as provided by president/founder Scott 
Draeker.

>1) Why do you think Linux game development has a future? Did the install
>   base increase enough for you to make profit or is it simply because the
>   install base consists mostly of game playing geeks?

The installed base of Linux users is already large enough to support our 
efforts. And the number of Linux users grows every day. There is also a 
lot of overlap between traditional Linux users and game players.

What's interesting is that more and more new Linux users are not what you 
would call traditional hard core gamers. They are people who want a 
faster, more stable system for such everyday tasks as reading email and 
writing letters. These are not people who are necessarily going to pick 
up this week's must have 3D shooter, but they are excited about 
Solitaire. 

As Linux continues to grow, we'll continue to expand the user base and 
with that the breadth of titles that are available. Don't be too 
surprised when you see a Linux version of Barbie or Sticky Bear...

>2) Do you expect more companies such as Loki to appear? Apparently your port
>   of Civ:CTP was very easy and succesful, so you must expect some
>   competition.

What do you mean, easy?

Particularly after Red Hat's successful IPO, there are going to be more 
people clamoring to do something on Linux. That doesn't change anything 
for us. Loki's success depends on our continuing to release Linux 
products which are as good as or better than the original and on 
supporting the Linux community by releasing as much source as we can. 

>3) Which SDK's did you study to be used in Civ:CTP? Why did some of them
>   make it and others not? Do you think this area of Linux game development
>   needs much improvement?

We looked at all of the likely candidates: GGI, PenguinPlay, etc. We 
chose SDL because it was the most mature. Where more work needs to be 
done, we are doing it and releasing the results back to the community.

Did I mention that our lead programmer, Sam Lantinga, wrote SDL?

>4) You probably also have noticed the increase of semi-high quality games
>   for Linux made by the 'regular' coders. Are there any plans to develop a
>   game together with them in addition to the ports Loki makes, or at least
>   cooperate on certain ussues?

More high quality, AAA games for Linux is a good thing. And we applaud 
game developers who are producing Linux versions of their products. The 
Linux game market is still evolving, so we're continually looking at ways 
to cooperate with traditional Windows publishers as well as developers 
who are interested in doing something on Linux out of the gate.

>5) Is Loki planning to improve on, and then send back, even more open-source
>   libraries to be used for gaming and multimedia in general?

Absolutely. We've already released SMPEG, SMJPEG and extensive work on 
SDL. We finally have an installer we're really proud of, and that will be 
released soon as well. 

We're also actively working on a project which is our most ambitious to 
date. More on this in the next couple of months as the project matures 
and we begin to get usable code.

Open source at Loki is one of the things I'm most excited about. It's 
typically difficult for a company to justify spending development dollars 
on code such as an MPEG player, only to give it away. We on the other 
hand, can't justify charging people for an MPEG player. That's not our 
business at all. So when we do develop these tools for our own use 
there's never any questions whether we'll release the source.

>6) You were able to convince Activision to let you make a Linux port of
>   their game. How hard was that to do? What should Linux users do to get
>   more games ported other than buying the ones that have been? Do you think
>   petitions and requests will work?

In the summer of 1998 when I first started putting Loki together there 
were plenty of petitions flying around, but no games. The reason 
petitions don't work is twofold: 1) a voluntary, online petition is not 
the kind of marketing information that large, public corporations will 
use in their decision making, and 2) a petition doesn't address the 
fundamental questions that a Linux port raises. Supporting Linux is not 
as simple as 
changing the source code. You need to test the port, package it, figure 
out how to sell it and offer technical support to end users. We're set up 
to do all of these things for Linux, whereas other publishers are set up 
to do it for Windows.

>7) Do you think there will be a time that companies like Loki will no longer
>   be necessary because the software houses have their own Linux developers
>   working on releases?

If (when?) Linux supplants Windows as the dominant OS, then game 
developers will focus on creating Linux products. Until then, they'll 
continue to focus on Windows products. Some will come out with Linux 
versions just as some companies do Mac versions. Those who don't are 
welcome to work with us.

>8) If you had to praise and critisize one area of Linux (game) development,
>   what would they be?

Sound support is lacking. And there are issues supporting all the 
different Window Managers.

On the positive side, I would praise the flexibility of the system. Linux 
was not designed to 
play games, but it handles games just fine. And I believe that it won't 
be long before the same game on the same hardware will be faster, more 
responsible and more stable on Linux than on any other OS.

Of course there aren't enough games for Linux today, but we're rapidly 
solving that problem.

>9) Which games are on your wishlist to be ported, if any?

I have a few in mind. The good news is that I think we'll get to do many 
if not all of them. One game I'd really like to do is Age of Empires II...
 
>10) Do you have any advice for Linux game developers creating their own game
>    or making a port/clone?

Just because your game runs on Linux doesn't mean that it will be fun to 
play. Pretty graphics and sound are great, but gameplay is still king.

Also, Setuid is evil...

>11) Why is there no libc5 support in Civ:CtP? Was it to hard to get
>    it working or was it just that the user base with libc5 is
>    extremly small, not worth the effort?

CivCTP requires the threading support in glibc 2.0 and later. The good 
news is that CivCTP works fine on libc5 based systems. The binary which 
ships on the CivCTP disk is statically linked and works on both libc5 and 
glibc 2.0 systems. The 1.1 patch includes an update to the statically 
linked binary, but also includes a dynamically linked binary which works 
on glibc 2.1 systems.

>12) As far as I understood, Civ:CTP does not follow the FHS
>    completly. Why? Data would normaly belong to share/games/CivCTP
>    and only the binary should go to games/.

None of the CivCTP data is shared with other applications, so we opted 
for a single directory install. This makes uninstalls really easy to do, 
as you don't have to keep track of where all the individual files went. 
RPM can also handle this, but not every distribution supports RPM.

>13) What things could Linux OS developers do to make life easier
>     for games writers/porters?

Improved sound support and sound card configuration would be helpful. 
Standardizing on a single C library would also be nice. Finally, there's 
still quite a bit of work needed at the driver level. It's one thing to 
be able to use a brand new video card for simple tasks, and quite another 
to support full 2D/3D acceleration and advanced features in the hardware. 
This work is being done, but of course we're impatient :-)

Sincerely,

Kathryn Sorhaindo
Loki Entertainment Software

-- END included message