Deadlock presentation outcome

Tom S h3r3tic at remove.mat.uni.torun.pl
Wed May 16 15:45:56 PDT 2007


Bill Baxter wrote:
 > The heaviest thing about about Python is the libraries, and if you don't
 > need 'em you don't have to ship 'em.  The core python24.dll is 1.78MB.
 > That's not tiny but still it's a trivial size for a PC game.  The people
 > who favor Lua are the console developers, where memory and space is
 > actually a little tight.  But for PC development there's nothing wrong
 > with Python's size.

That's a valid point, but by heavy I also meant that it's slower than Lua.
Lua itself isn't very pretty, but I've recently stumbled across this: 
http://metalua.luaforge.net/ and it seems really interesting.



 > If you don't like the syntax, well that's another issue.  I go back and
 > forth between D and Python (and C++) a lot and don't find it to be too
 > much of a problem.  I do tend to forget to type semi-colons for a while
 > when I switch back to D, though.

Nah, I do like the syntax, and I used to code in Python, but I'm more 
concerned about all the rest of team0xf and the other folks that might 
write a script or two and don't find Python very friendly

Anyway, If my evil binding plan turns out to work, adding Python to the 
list should be near trivial :)



Walter Bright wrote:
 > There's dmdscript, the javascript engine written in D.

That has two problems:
- I absolutely dislike ECMA 262
- someone would have to benchmark it against other scripting langs :>



Unknown W. Brackets wrote:
 > Don't feel bad.  I had the same experience - I wrote a game with a full
 > fledged scripting language built in, years ago when I was in high
 > school.  It was a lot of lines, and represented a lot of work.
 >
 > I lost to a GPA calculator.  This was a program where you typed in your
 > grades and it told you your average.  It was like 200 lines or something
 > silly like that.
 >
 > Even the guy who won told me I clearly should have won.  Everyone
 > competing against me did.  Oh, well.  It happens.  Silly
 > only-knew-Fortran judges.

Ok, so our case wasn't that bad... Thanks for remdinding me that silly 
judges are there all over the globe ;)
Gee you must've been mad... losing to a GPA calculator... that's like a 
slap in the face.



Georg Wrede wrote:
 > Of course, the ability to code, refactor, manage a software project,
 > join technologies, and a passion for the art are important for an
 > employer who's looking for a new programmer.
 >
 > (...)
 >
 > Already with the game as your merit, you can be sure to get hired
 > wherever you seek a job. But convincing them of all the other stuff
 > gives you a much better job at the same company, and an outlook on a
 > career with increasing responsibilities and salary, and of course
 > influence among your colleagues.
 >
 > So, finish your education with good grades, and you'll thank yourself
 > many times in the years to come. Oh, and don't let any greedy employer
 > steal you from the school. Such employers will tell you that your fame
 > from this game will fade, and Right Now is the time to get a good job
 > while the fame lasts. (I've been in this business for more than 20 
years.)

Thanks for the clues :)
While I'm not going to run out towards greedy employers, the university 
and its program itself is giving my collegues and me hard time... Not 
sure if this is specific just to this department, but the level is not 
very high. Basically, for a long time there wasn't a Computer Science 
department at my univ. After some discussions, they decided to create it 
at the department of mathematics. I guess they didn't like it that much, 
as now mathematicians are blocking all computer science, and think 
they're the most important there :/ For example, the department gets 
lots of cash every month or so, and they won't even sponsor some ACM or 
IEEE memberships for even single students. There have been many 
initiatives that would need marginal founding, and they aren't willing 
to give just a few bucks. The mathematicians get everything :S
Most of the computer-science teachers/lecturers/tutors possibly knew 
their stuff 20 years ago, but now they're just bored, 
reluctant-to-do-anything dinosaurs. Only the young folks are more 
lively, but they have to conform to the program that the dinosaurs set 
out, or even worse, they can't invent new cool classes because the 
matematicians don't like it.
And how shall we find motivation to study at this univ... *sigh*



Walter Bright wrote:
 > How about posting to gamedev.net?

Sure! But we must get the long-discussed webpage praising D online ;)



Mikola Lysenko wrote:
 > Great job!

Thanks :)

 > I really admire your work on the game, and more importantly the 
remarkable story that got you there.  Overall it is a great victory for 
D at large, not to mention impressive as hell for a student project. 
You should post an image of the day on GameDev.net.

An IOTD is the plan. I think it's what gets the most attention on 
GameDev.net :)



Chad J wrote:
 > Good to hear things went mostly excellent.  Congrats!
 > I like the backgroud story too.  A politically incorrect project...
 > something that rubs the wrong people the wrong way, but undeniably kicks
 > ass.  Overcoming adversity by placing second with hostile judges.  I
 > like it.
 >
 > Rock on.

Hahaha, your comment made my day ;) Thanks!



Lutger wrote:
 > Congratulations with your work, very impressive! Very nice artwork on
 > the website with the whole cyberpunk theme, pretty cool.

Thanks :] We'll try to get the game up to the theme.


 > Whilst obviously only a small part of your project, I'm curious how
 > immediate mode gui worked out. Did you find it much better than retained
 > mode gui, and do your think this is mainly good for games or also useful
 > in other types of applications?

I'd dare to say that it would be useful in other application types. And 
to be fair, our GUI isn't exactly an Immediate GUI, but rather a hybrid 
of IM and RM GUIs. It works pretty well, but I'm not sure that I like 
the fact that some of the state info has to be stored in global vars, or 
that special checking must be made to determine when a value has 
changed. Maybe I'll find a way to make this aspect of the GUI better :)

Anyway, the lack of init<->callback separation in IM GUIs is really 
cool. The code looks pretty clean and self-explanatory and no special 
signal-slot stuff is needed. It gets especially nice when combined with 
sinking-bubbling event processing :) Definitely makes for the sexiest 
radio button code 'evar' ;D

Here's one of my test apps of the GUI (Windows only ATM):
http://h3r3tic.googlecode.com/files/guiDemo-2007.05.17.7z

And its source:
http://paste.dprogramming.com/dpyg6dkc.php

The code may look a bit alien at first... But it's pretty easy to write 
and understand, at least IMHO.

I'm working on the GUI as a mini side-project, planning to extend it to 
the most-frequently used widgets and separate it from Deadlock to make 
it usable in other projects.




Again, thanks for the feedback :)


-- 
Tomasz Stachowiak
http://h3.team0xf.com/
h3/h3r3tic on #D freenode



More information about the Digitalmars-d-announce mailing list