SRL Archive - General

V 2.0 MSI & Several Developments


Sir R. M8gic1an

2010-08-06 07:07:19

Ok, so for those who didn't see it, here is V 1.0 MSI & Several Developments, which is pointless to read now, unless you want a quick overview of what is to come in this post.

MSI is now quite big, and on the way to become much bigger, very quickly.
MSI is a script, of a scale never seen before. (atm 2 AIO in 1)

Therefore, big scale projects fit in MSI.

What might these be?
- - Color world walker - several of these have been written, but never put to proper use nor testing, for lack of a use for it.
- - Reflection anywhere walker

Those are somewhat 'obvious' ideas.

But the big thought of this post is… development.
And using MSI as a testing set. MSI is the perfect place to test 'new technologies'

AutoResponder? (yes, with the possibility to turn it on/off)

Testing new random solvers?
- - This is one of the main things. Randoms are hard to test, due to not always getting the random we want. But in MSI, we could just simply try the solver person X wrote. It would mean faster development for the randoms, as now they have a 'base of people' testing their solver.

MSI can (soon) also be the house to other community projects, eg BareBones by Boreas. Give the people a chance to run their projects.

MSI was a project I wanted to run, got the chance to do it, and it's going rather well. Why not give other people the chance? One of you might even want to run some sort of project.




Now, enough about the what, time to get to the how.

We'd need GIT. That's the first and most essential point. Without git this would never work as it would be impossible to coordinate all the sub-projects.

Then, the way to "switch on" / "switch off" people's projects would be through DEFINEs. I've made a post with something like this a while back. Not crucial to have that sorted out right now, but we'd have a Defines.Simba where you can undefine any module you don't wish to use.

And the other crucial point is: The people would make it MSI compatible, not the SLs.
The thinking behind this is very simple. Whoever is developing will want testers, and therefore will be willing to make it MSI compatible. Yes, we will be there to help, but we should give as much information as possible for them to be able to make it work by themselves. And considering we'll be using git, they will most likely be compiling it with MSI themselves.


And so what is the end result of this? Lots and Lots of community development, with minimal involvement by the SLs.

And I'd encourage everyone to also run their own projects, if they wish to :)


So, thoughts?

~RM

Coh3n

2010-08-06 11:31:40

From: Rasta Magician

- - Color world walker - several of these have been written, but never put to proper use nor testing, for lack of a use for it.

That would be HARD…

From: Rasta Magician

Testing new random solvers?
- - This is one of the main things. Randoms are hard to test, due to not always getting the random we want. But in MSI, we could just simply try the solver person X wrote. It would mean faster development for the randoms, as now they have a 'base of people' testing their solver.

Does this mean we would have to have our own FindNormalRandoms?

There are some things about the "other projects" development where I'm not sure how they would work. I may be getting the wrong idea here, but are you saying we'd have a list of different projects wrapped in defines? Something like:

{$IFDEF MSI}
Mainloop := @MSI_Mainloop;
{$ENDIF}

{$IFDEF BAREBONES}
Mainloop := @BareBones_Mainloop;
{$ENDIF}

{$IFDEF AUTORESPONDER}
AutoResponder := @MSI_AutoRespond;
{$ENDIF}

Sir R. M8gic1an

2010-08-06 12:25:41

very quick post:

with TeleRandoms[4] do //
begin
Name := 'evil bob''s island';
Music := ['evil bob''s island'];
TypeID := [1,0];
IsInRandom := @InEvilBobsIsland;
Texts := ['ScapeRune','ask to come here'];
TextColors := [clMessage,0];
TextFonts := [SmallChars,NPCChars];
TextBoxes := [IntToBox(MCX1,MCY1,MCX2,MCY2),IntToBox(MCX1,MCY1,MCX2,MCY2)];
SolveRandom := @SolveEvilBob;
RandIndex := rand_ScapeRune;
end;

We could point the functions to the ones people want tested. This is already in AntiRandoms.scar

- Mixster did a color bitmap world walker

- And I was thinking differently, but I prefer what you have for the third point;


//AutoResponders
{$DEFINE POMS_AutoResponder}
//{$DEFINE MSI_AutoResponder}
//{$DEFINE SRLComp_AutoResponder}

//Experimental Randoms
{$DEFINE NCDS_EvilBob}



////////////////////////////////

{$IFDEF POMS_AutoResponder} AutoResponder := @POMS_AR; {$ENDIF}



~RM

NCDS

2010-08-06 12:45:36

Love what I'm seeing after a quick read through. I'm short on time at the moment but I will come back to this thread later and have a 'meaningful' post.


P.S. @RM EvilBob != ScapeRune; ;)

Sir R. M8gic1an

2010-08-06 13:18:13

From: NCDS
Love what I'm seeing after a quick read through. I'm short on time at the moment but I will come back to this thread later and have a 'meaningful' post.


P.S. @RM EvilBob != ScapeRune; ;)


my bad :p been short on time to read all the new things through :redface:

~RM

Sir R. M8gic1an

2010-08-07 07:14:45

no more input on this..?

~RM

Coh3n

2010-08-07 09:39:44

From: Rasta Magician
no more input on this..?

~RM

It looks fine to me. Was this something you were looking at getting started right away? Because honestly, I would rather "finish" the core before we start adding things like this. There's not that much more to do but we do still have that pointer thread I made, script switching, Magic, Fishing and Runecrafting that are being worked on.

Sir R. M8gic1an

2010-08-07 11:38:25

From: Coh3n
It looks fine to me. Was this something you were looking at getting started right away? Because honestly, I would rather "finish" the core before we start adding things like this. There's not that much more to do but we do still have that pointer thread I made, script switching, Magic, Fishing and Runecrafting that are being worked on.


Well pointed out. Any proper project will require planning.
There are two things that come to my mind:
- Some of these projects can match whatever we are working on right now
- - eg WorldWalkers & Script Switching

- Some of these projects do not need to affect the main flow at all
- - eg AutoReponder: People can work on a separate branch, and then we can just add it to MSI_AntiRandoms
- - (keep git in mind)
- - Same goes for antirandom development

On a different note, a thought I had was to make a sticky in the Test Corner, letting people know that MSI would be accepting development projects, under rules that would require minimal SL involvement.

It's the people's choice whether to use the "developments" or not, and any that we do want to use & find reliable enough we can just simply include it by default, with ways of customizing it.

And of course that with anything we include by default, we'll check it through for any possible flaws. :)

I think the main point here to your concerns is minimal SL involvement.. ?

~RM

Coh3n

2010-08-07 12:22:55

Well yes, the idea is to have minimal SL involvement, but I was more or less saying that I don't want to have too many projects on the go when there's not that many people to develop them.

Sure, other people can have the option to start their own, but us SL's I think should concentrate on the current projects before moving on to the next (yes, I've been guilty of it).

The thing is, when something new is brought up, a lot of people are interested at first, but then after a couple days it dies off. It takes someone to step up and do some work on it before people become more involved again. I would just like to get our current projects up and running before we start new ones.

Know what I mean?

Sir R. M8gic1an

2010-08-07 13:13:25

From: Coh3n
Well yes, the idea is to have minimal SL involvement, but I was more or less saying that I don't want to have too many projects on the go when there's not that many people to develop them.

Sure, other people can have the option to start their own, but us SL's I think should concentrate on the current projects before moving on to the next (yes, I've been guilty of it).

The thing is, when something new is brought up, a lot of people are interested at first, but then after a couple days it dies off. It takes someone to step up and do some work on it before people become more involved again. I would just like to get our current projects up and running before we start new ones.

Know what I mean?


:) Let's get the current ones kicked off the ground first then.

~RM

NCDS

2010-08-08 21:02:50

From: Rasta Magician

On a different note, a thought I had was to make a sticky in the Test Corner, letting people know that MSI would be accepting development projects, under rules that would require minimal SL involvement.

It's the people's choice whether to use the "developments" or not, and any that we do want to use & find reliable enough we can just simply include it by default, with ways of customizing it.

And of course that with anything we include by default, we'll check it through for any possible flaws. :)
~RM


I really like that idea :) and feel that it may increase some activity for MSI. The cup seems like it already has :p

Gu3st...

2010-08-09 05:47:25

My world walker could be implemented in MSI as it's all raw code (no plugins / files needed).

I just have to improve it a bit and make it in MSI-Standards

E: It's color btw

Coh3n

2010-08-09 09:37:11

From: Zyt3x
My world walker could be implemented in MSI as it's all raw code (no plugins / files needed).

I just have to improve it a bit and make it in MSI-Standards

E: It's color btw

Link? I'm interested to see a color world walking. :p

Gu3st...

2010-08-09 16:16:30

From: Coh3n
Link? I'm interested to see a color world walking. :p
http://www.youtube.com/watch?v=-lyzYBmrwF8

Blumblebee

2010-08-09 16:33:51

From: Zyt3x
http://www.youtube.com/watch?v=-lyzYBmrwF8


your's requires the opening of the World Map every time though does it not? I would prefer to have a world map bitmap saved, look at the minimap and try to pinpoint it on the bitmap and then figure out the walking scheme from there.

Coh3n

2010-08-09 17:07:38

From: Zyt3x
http://www.youtube.com/watch?v=-lyzYBmrwF8

That's pretty sweet actually. Looks like it needs some tweaking, but it has potential. I'm surprised I didn't see this before.

But yeah, it would be nice to do it without having to open the map every time.

Gu3st...

2010-08-09 19:02:43

From: Coh3n
That's pretty sweet actually. Looks like it needs some tweaking, but it has potential. I'm surprised I didn't see this before.

But yeah, it would be nice to do it without having to open the map every time.
A tiny tweak on MSI to make it store information on where on the world map it is would make it only have to open it once, and that's the first version I made :) If I remember correctly the code was at about 70 lines or so…

The concept is really simple actually.. Open the world map and the tiny mini-map and find the red square telling you where you're at, then go to the location wanted, and compare the two points, returning angle and how many times it has to click, then click that amount of times, and it should be at it's position.

Coh3n

2010-08-09 19:31:35

From: Zyt3x
A tiny tweak on MSI to make it store information on where on the world map it is would make it only have to open it once, and that's the first version I made :) If I remember correctly the code was at about 70 lines or so…

The concept is really simple actually.. Open the world map and the tiny mini-map and find the red square telling you where you're at, then go to the location wanted, and compare the two points, returning angle and how many times it has to click, then click that amount of times, and it should be at it's position.

The only issue I see with that is that it could get stuck easily if a door is shut on the player or something. I can't see that happening that often though. Honestly, I think it's worth further development.