headers
subversion@HamsterRepublic.com | 1 Apr 2006 11:15
Favicon

subversion commit 542 (james)

james
2006-04-01 01:15:06 -0800 (Sat, 01 Apr 2006)
130
moved various code out of game.bas to avoid abnoxious QB out-of-memory error. Subified titlescreen (a
step towards fixing bug 49)
---
U   wip/game.bas
U   wip/yetmore.bas
U   wip/yetmore2.bas
Permalink | Reply | 0 comments |
headers
subversion@HamsterRepublic.com | 1 Apr 2006 10:13
Favicon

subversion commit 541 (james)

james
2006-04-01 00:13:47 -0800 (Sat, 01 Apr 2006)
306
Fixed an ancient bug in the arrow-key handling in both the Sprite editor and the Tile editor. The
keyboard-controlled cursor no longer lags one pixel behind when the mouse is over the drawing area. Note
that because of bug 174 you probably won't be able to notice this fix yet anywhere except customqb.exe
---
U   wip/drawing.bas
U   wip/whatsnew.txt
Permalink | Reply | 0 comments |
headers
Simon Bradley | 1 Apr 2006 19:32
Picon

user defined types (QB vs FB)

On 4/1/06, Mike Caron <caron.mike at gmail.com> wrote:
> On 3/31/06, Bob the Hamster <Bob at hamsterrepublic.com> wrote:
> > On Fri, Mar 31, 2006 at 06:33:21PM -0500, Mike Caron wrote:
> > > On 3/31/06, subversion at hamsterrepublic.com
> > > <subversion at hamsterrepublic.com> wrote:
> > > > james
> > > > 2006-03-31 12:03:37 -0800 (Fri, 31 Mar 2006)
> > > > 440
> > > > subified more of the tile editing code (still more to come).
> > > > I created a TileEditState user-defined-type that carries all the
> > > > various editor state variables. The subs that make up the tile editor
> > > > only need to pass around one instance of this, rather than a myriad
> > > > of different variables. Unfortunately UDTs forbid dynamic arrays and
> > > > variable-length strings (which for our purpose are ALL arrays and
> > > > strings) but UDTs are still useful here.
> > > > ---
> > > > U   wip/drawing.bas
> > > >
> > >
> > > I wonder why that is. I mean, wouldn't an array simply be a pointer?
> > > Sorry, typo. *Shouldn't* it be a pointer. Stupid QB.
> > >
> > > When we drop QB, we can do all that stuff...
> >
> > I'm not really sure. The FB docs on TYPE are a little ambigious about
> > this. http://www.freebasic.net/wiki/wikka.php?wakka=KeyPgType
> >
> > We will have to do some experimentation to find out what the limits are
> > on FB types (and then update the FB wiki)
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
subversion@HamsterRepublic.com | 2 Apr 2006 00:42
Favicon

subversion commit 543 (simonb)

simonb
2006-04-01 14:42:28 -0800 (Sat, 01 Apr 2006)
166
Fixed system day and system month plotscript commands, and
changed system hour, minute and second to use time$ instead
of timer, which works differently in freebasic.
---
U   wip/yetmore.bas
Permalink | Reply | 2 comments |
headers
Raekuul | 2 Apr 2006 05:05
Picon
Favicon

Hidden Bug

<Snip> and fixed (at least) one hidden bug in
> the process

What bug?

 
http://www.hamsterrepublic.com/ohrrpgce
The main OHR site.

http://www.castleparadox.com
The OHR community site.

http://elitrix.net/raekuul/portal.php
MyPhpBb forum! 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com
Permalink | Reply | 1 comment |
headers
Bob the Hamster | 2 Apr 2006 05:17
Favicon

Hidden Bug

On Sat, Apr 01, 2006 at 07:05:54PM -0800, Raekuul wrote:
> <Snip> and fixed (at least) one hidden bug in
> > the process
> 
> What bug?

The FreeBasic versions of storeset and loadset require a buffer that is 
just slightly larger than the buffer required by the assembly-language 
versions. The tile code uses a small buffer with those commands to make 
the "undo" copies, and we forgot to increase it.

This meant that a couple bytes of map data pixel data was getting 
written over some other variables every time the undo buffer was updated 
-- I have no idea which variable(s) were being overwritten. Probably 
something unimportant. When I refactored the code, I changed the order 
in which things get DIMed, and that buffer started overlowing into 
something important; the TileEditState.x variable, which causes some 
really funky behaviour. The cursor would jump to the left every time you 
tried to draw, and all your lines and boxes would end way off in the 
wrong place. Had me really confused for awhile.

So now the buffer is the right size. True story ;)

---
Bob the Hamster
Permalink | Reply | 0 comments |
headers
Mike Caron | 2 Apr 2006 05:21
Picon
Gravatar

Pathfinding

On 3/30/06, Adam Perry <arperry at gmail.com> wrote:
>  Yes! Those are even better than what I had in mind.
>
>
>  Bob the Hamster wrote:
>  On Fri, Mar 31, 2006 at 12:41:42AM +1200, TeeEmCee wrote:
>
>
>  On 3/30/06, Adam Perry <arperry at gmail.com> wrote:
>
>
>  CP is down :(
>
> So I want to write a pathfinding routine for OHRRPGCE. Since it's for a
> TRPG (read: FFT clone) I don't care too much if it takes a second to
> calculate, but I don't want it to take too long and I care about finding
> a fairly optimal path. I think I know a way to meet all of my needs, but
> I want input on the following:
>
> 1. I heard plotscripting has generally sped up in recent releases! This
> pleases me. This is kind of data-processing-intensive, so what can I
> expect? I don't want to code the whole thing and then find out it takes
> ten seconds to execute.
>
>  Someone made a cool maze generation script once, so I used that as a
> benchmark. It was prehaps 5-10 times faster on FB Game. However, I had
> a fairly fast computer, and some people have had FB Game run scripts
> slower! However, this is probably a bug/external.
>
>  I have a hunch that most people who report plotscripts being "slower"
(Continue reading)

Permalink | Reply | 1 comment |
headers
TeeEmCee | 2 Apr 2006 07:10
Picon

Pathfinding

On 4/2/06, Mike Caron <caron.mike at gmail.com> wrote:
> On 3/30/06, Adam Perry <arperry at gmail.com> wrote:
> >  Yes! Those are even better than what I had in mind.
> >
> >
> >  Bob the Hamster wrote:
> >  On Fri, Mar 31, 2006 at 12:41:42AM +1200, TeeEmCee wrote:
> >
> >
> >  On 3/30/06, Adam Perry <arperry at gmail.com> wrote:
> >
> >
> >  CP is down :(
> >
> > So I want to write a pathfinding routine for OHRRPGCE. Since it's for a
> > TRPG (read: FFT clone) I don't care too much if it takes a second to
> > calculate, but I don't want it to take too long and I care about finding
> > a fairly optimal path. I think I know a way to meet all of my needs, but
> > I want input on the following:
> >
> > 1. I heard plotscripting has generally sped up in recent releases! This
> > pleases me. This is kind of data-processing-intensive, so what can I
> > expect? I don't want to code the whole thing and then find out it takes
> > ten seconds to execute.
> >
> >  Someone made a cool maze generation script once, so I used that as a
> > benchmark. It was prehaps 5-10 times faster on FB Game. However, I had
> > a fairly fast computer, and some people have had FB Game run scripts
> > slower! However, this is probably a bug/external.
> >
(Continue reading)

Permalink | Reply | 0 comments |
headers
Simon Bradley | 2 Apr 2006 16:29
Picon

subversion commit 543 (simonb)

On 4/1/06, subversion at hamsterrepublic.com
<subversion at hamsterrepublic.com> wrote:
> simonb
> 2006-04-01 14:42:28 -0800 (Sat, 01 Apr 2006)
> 166
> Fixed system day and system month plotscript commands, and
> changed system hour, minute and second to use time$ instead
> of timer, which works differently in freebasic.
> ---
> U   wip/yetmore.bas
>

Just a note on this: Mike suggested last night that I just switch the
numbers for system day and system month in plotscr.hsd (the problem
was that they were the wrong way round). That wouldn't necessarily be
a bad idea if people have used the buggy versions, it would be similar
to the "npc is walking" fix, where new scripts compile with the
correct behaviour but previously compiled scripts still use the bugged
behaviour.

I'm not sure how worthwhile it is though. Only one person has ever
noticed the bug, and the day, month and year commands are missing from
plotdict.xml so I imagine no-one knew about them.

If I want to add them to the plot dictionary, is it only plotdict.xml
that I need to change?

Simon
(Continue reading)

Permalink | Reply | 1 comment |
headers
Mike Caron | 2 Apr 2006 19:18
Picon
Gravatar

subversion commit 543 (simonb)

On 4/2/06, Simon Bradley <neworiginal at gmail.com> wrote:
> On 4/1/06, subversion at hamsterrepublic.com
> <subversion at hamsterrepublic.com> wrote:
> > simonb
> > 2006-04-01 14:42:28 -0800 (Sat, 01 Apr 2006)
> > 166
> > Fixed system day and system month plotscript commands, and
> > changed system hour, minute and second to use time$ instead
> > of timer, which works differently in freebasic.
> > ---
> > U   wip/yetmore.bas
> >
>
> Just a note on this: Mike suggested last night that I just switch the
> numbers for system day and system month in plotscr.hsd (the problem
> was that they were the wrong way round). That wouldn't necessarily be
> a bad idea if people have used the buggy versions, it would be similar
> to the "npc is walking" fix, where new scripts compile with the
> correct behaviour but previously compiled scripts still use the bugged
> behaviour.

I suggested no such thing. I jokingly said I would do that if I were
lazy enough. That would be the bad way to fix this.

> I'm not sure how worthwhile it is though. Only one person has ever
> noticed the bug, and the day, month and year commands are missing from
> plotdict.xml so I imagine no-one knew about them.
>
> If I want to add them to the plot dictionary, is it only plotdict.xml
> that I need to change?
(Continue reading)

Permalink | Reply | 0 comments |

Gmane