On 8 Mar 2001, at 22:41, Rob Bowes wrote:
> At 09:58 PM 3/8/01 -0500, you wrote:
> >Well, one, its not a small number of users anymore... Linux is the
> >fastest growing piece of software on the planet. Its slated to pass
> >windows for desktop PCs in like 8 years. MS is starting to worry, but
> >wont admit it publicly.
>
> I didn't say I wouldn't support it. I said the demand needed to
> justify the investment. It's quite likely that will occur over the
> upcoming years. When that happens, I'll likely do a Linux version of
> AB.
>
Ok, fair enough, just dont wait to long there is plenty of people who
do open source projects that might just make a clone of your program
that might even use the same datafiles. And it might just be open
source and thus free to boot. Something to consider.
> >Also, is it a convenience to require that the person pay extra for a
> >OS that will be used basicly for one program? (Granted most linux
> >users still run windows for games, but thats not the point).
>
> Please consider carefully your last statement above. If most Linux
> users already have an emulator to run Windows games, then most of AB's
Well, right now most people do a dual boot, which is very differant
from an emulator. THe average linux user does not mess with Wine or
any other emulator, its only normally used by those who must run a
piece of software for work or school.
> customer base won't NEED to buy a native Linux version. That means
> that the huge effort to create a native version will only be NEEDED by
> a small fraction of the already small (but growing) base of Linux
Well, most users only run a emulator or dual boot, as most software
is still only written for windows. That is rapidly changing, I know
a good number of people that work full time and have nothing but
linux on the their machines. This will become the norm in about 5
years. As linux matures a bit more so that drivers are more availble
(and you dont have to beg or hunt forever for them). It will become
alot more mainline. Question is, do you want to join the crowd, or
lead it?
> users. This means that sales will certainly not justify the investment
> at the current time. Once the sales numbers will balance out the
> investment, it's a good business decision to create a native version.
> My time is extremely limited and I have to make the best use of it.
> It's not a slight against Linux users. It's a basic reality of
> business - same as any other company developing software for any
> platform.
>
Granted, your a lone man doing alot. It will take some time, and some
effort to learn the OS. I just hope you dont get tied up and lose
track of future plans.
> >Depending on the level of ANSI compatiblity, it could be, as simple
> >as loading it up and recompling. Assuming your not using MSVC, most
> >other compliers are about 90% ansi complient. It would not be the
> >huge chore you think. I have translated larger programs from windows
> >to Linux, and am working on a few others.
>
> I do cross-platform portability as a profession, so I'm reasonably
> familiar with the issues.
AB is written to be portable, with the
> exception of a smattering of low-level facilities (easily ported),
> some key components (e.g. internet access), and the UI. It's the UI
> that will be a bear to port, since THAT piece of AB was NOT written to
> be portable (the overhead to do so would have been prohibitive). The
> complexity of hte AB UI has increased significantly over the years, so
> this is a major piece to try and port.
>
I have one word to say, Widgets, or GTK+ as its properly called.
http://www.gtk.org Its the X windows version of the visual tool kit,
it has all the stuff you need, prepackaged and ready to go. List
boxes, combo boxes (maybe under differant names, but the function is
the same), text boxes, etc.
> >Also Borland and others have compatible libraries under linux, to
> >make porting even easier. If you dont have a linux box, I would sit
> >down install it on a spare system, (takes an hour for a basic
> >install, ignoring sound/zips/burners/high powered video) gcc/g++ is
> >installed pretty much by default. YOu might just give it go one after
> >noon when your bombed out on writing code. Ask I will snail mail you
> >a copy of RedHat, (Yes its legal) just ask me off list.
>
> I'll make a point of doing this over the Summer. First I have to get
> finished with my current project, then get a new release of AB put
> together, and then deal with Origins/GenCon. After that, I can think
> about Linux.
Ok, fair enough, its on the agenda, at least.
>
> BTW, I've been using g++ and gdb on a project for client. I must say
> that the compiler is reasonably solid (sans a few quirks that are a
> pain), but gdb is a few giant leaps backwards from what I'm used to
> with the VC debugger (or even the Borland debugger). That will make
> the port a much less pleasant process. :-(
Gdb is not the major player anymore, Codewarrior is.
>
> >I happen to be a memeber of a LUG (Canton Linux Users Group), and if
> >you have some other problems that I cannot answer myself. I am about
> >90% sure I can get the answers with hours or overnight. We have some
> >heavy hitters in our little group, profession programmers, and
> >sysadmins (And a few web celebs as well).
>
> Excellent. It's nice to have a knowledgeable resource to tap. I'll
> definitely keep this in mind when the Linux effort is pursued.
Sure, anytime, and I can also point you to some reasonable online
references.
>
> >BTW, ONce you port a version to Linux, your done, it will only
> >require a recomplie to run under the many flavors and platforms that
> >linux runs under, including: Mac, Intel, Sparc, most modern
> >mainframes, and even the amiga. Oh yea, the commadore 64/128!
> >
> >Try that with any other operating system....
>
> My code already takes all that into consideration pretty cleanly. I've
> been doing cross-platform portability work (typically Windows, Unix,
> and Mac) for a LONG time, so that's not an issue for me. The gotcha is
> the UI.
Cool. Be honest I was begining to think you were a hobbiest
programmer that made it big, and was perhaps over your head.
>
> >So its worth the time, and you might see your sales grow even faster,
> >if you took a weekend and gave it a shot.
>
> So fill me in on how difficult it would be to port about 50,000 lines
> of code that is tied to the Windows API (primarily GDI). Also describe
> what libraries are in place that emulate the functionality of things
> like the "wininet" API provided under Windows. I have no idea what
> provisions/libraries are in place under Linux for this, but this is
> critical to porting the product.
All are availble in the Gnome toolkits. GTK+ (at least just about
everything, including Inet access).
>
> >Course you welcome to ignore me, and in five years I promise not to
> >look you up and say I told you so.... I will just sit back and smile.
>
>
>
Thank you for at least considering what I have to say.
> Thanks, Rob
>
> ----------------------------------------------------------------------
> ----- Rob Bowes (rob@wolflair.com)
> (650) 726-9689 Lone Wolf Development
>
www.wolflair.com
>
>
> To unsubscribe from this group, email
>
>
armybuilder-unsubscribe@egroups.com
>
> Your use of Yahoo! Groups is subject to
>
http://docs.yahoo.com/info/terms/
>
>
Fallenhunter
"They know you have the information they want, question is when will they find you."
Overheard at a bar in the Corbia sector.
