[abakoboParticipant]

I’ve created a github organization: https://github.com/mx2DocsCommunity

I invited the people who said they want to be involved in a community doc work (and I knew their GitHub Name (invited as owners)i.e.: nerobot, AdamStrange, peterigz (I didn’t search for long though))

If you think it’s a good idea to work on a fork to add our #monkeydoc modifications and later let Mark do what he wants merging, accept invitation or ask for an owner/member invitation to start workin together. If Mark doesn’t like it we’ll still be able to build those docs and share it.
Of course Mark if you want to be invited you’ll be welcome but the aim of this fork is not to take your time and work on our side so I didn’t invite you.

I created a ‘docsWork’ branch based on the develop branch. If we work only on that one (or other new ones based on docsWork) and not on master,develop,scratch branches we’ll be able to continuously update from the original dev branch without too much hastle.
Added a DocsTODO.txt file for sharing things we think should be done. It’s more or less the list I put on the previous post for now.

If you think it’s a good initiative please tell and I’ll create a new topic on the forum to announce it and invite people to participate..

[abakoboParticipant]

Best thing would be if many people work together. Mark could check the
spelling and mistakes, but at least people could help.
Or, Mark gives direct access to the sources to 5 to 10 people, to enhance
the docs. Let’s call it ‘documentation team’.

We can fork the dev branch and modify only the monkeydocs then Mark can make a pull requet himself if he likes the things we’ve written.. so we manage ourselves the direct acces.. by not modifying code we can update our fork easily.

I had started to list some undocumented things a while ago:

language:
———

extention keyword

where keyword

array tips and syntax for initialisation (1,2,3) multidim ((1,2,3),(7,8,9))?

protected private public, shared things when in the same .monkey2 file
(& virtual?)

infos on methods onrender and onwindowevent are not correctly created during doc building (seek that kind of ‘gaps’ and correct doc code)

? and ‘ternary’ conditions “if blah ? blah else blah”

[0] -> dereferencings syntax, are there other ways to dereference?

mojo:
—–

OnBlah stuff mx1 style is now implemented as ‘signals’ in AppInstance

mojox:
——
mojox in general?

c2mx2:
——
is not copying the api’s ///docs to #rem monkeydocs, can clang return ///docs so c2mx2 copy it?

sharing such a list somwhere would help us to make a todo list. and modify/pimp the docs

[abakoboParticipant]

IMHO this can be a kind of bug as Floor is made to transform a real variable to an integral one. So it should return an integral type not a double(wich is a real).
Ceil has the same problem..

oops no

http://stackoverflow.com/questions/15348180/why-doesnt-floor-return-an-integer

so you have to explicitly convert to an integer by yourself… and don’t play with numbers larger than MAX_LONG/MAX_INT or NaN or Infs! Or check for it before trying the conversion.

[abakoboParticipant]

writing

Monkey

[crayon-5cba9e203caf8004559710 inline="true" ]local X:int = Floor(v) & 255 or local X:int = Int(Floor(v) & 255)

1 2 3

[crayon-5cba9e203caf8004559710 inline="true" ]local X:int = Floor(v) & 255 or local X:int = Int(Floor(v) & 255)

[/crayon]

is I think exactly the same..
It will first compute “Floor(v) & 255” then convert it to an int if it is not an int.

The problem here is that there is not implicit conversion with the & operator and that Floor is returning a double but should return an int or long (integral).

if you:

Monkey

[crayon-5cba9e203cafe607361468 inline="true" ] print 3.0 & 255

1 2 3

[crayon-5cba9e203cafe607361468 inline="true" ] print 3.0 & 255

[/crayon]

you’ll have the same problem, one of the types is not integral

if you:

Monkey

[crayon-5cba9e203cb03086644598 inline="true" ] print int(3.0) & 255

1 2 3

[crayon-5cba9e203cb03086644598 inline="true" ] print int(3.0) & 255

[/crayon]

it’ll first make an explicit conversion from 3.0 to 3 then “&” is in a “int & int” situation therefore have integral types hence is happy.

IMHO this can be a kind of bug as Floor is made to transform a real variable to an integral one. So it should return an integral type not a double(wich is a real).
Ceil has the same problem..

[abakoboParticipant]

Floor is returning a double! So for me it’s kind of buggy.. Should return int or long.
The second example works because there’s an implicit convertion before calling the & operator?

[abakoboParticipant]

Can Ted21 run latest mx2cc?

I was sad when I saw all your beautiful work on a broken ship. And I must say I really hoped that effort will converge on one IDE.

Can we test Ted21? Is there a github repo?

The fontSprite is a nice and light way to create basic 2d gfx,Thanks!

[abakoboParticipant]

I’ve wandered on the blitz forum and it looks like a legacy AND live forum.
What I’ve seen on all those discussions about BRL products is that some people are frustrated because they wanted a max3d and now 3d will happen with MX2. Some also don’t want/like a strongly typed language, MX2 is strongly typed and that’s what I like with it! For duck typed I recommend Ruby..
Any comparison with Unity or Godot have no sense for me. If I want to create a clone I use a game engine/constructor (construct2 is great too). If I really want to create and play with maths and programming I must use a fully featured language because the sky will be the limit.
Anyway I like the fact that Mark is posting only constructive and forward-walking posts.

I really hope that kind of discussion won’t take too much place on this MX2 forum. I’m not from the blitz sphere and if I had seen too much posts like that (and like those on the blitz forum) when I started Monkey-x, I don’t know what would have happened but I’m not sure I would be a patreon now..

And.. Mark has to get some fun too!
GOOO Mark! You got the style!

[abakoboParticipant]

I never used blitzmax nor went on the forum but it would be sad that someone looking for help for blitzmax would lost all the help database a dedicated forum is. It looks like a fine language too..

[abakoboParticipant]

I personaly think I’ll keep my patreon unless I become homeless or mark stop to post for several months.. or ask us to stop donating! It’s the first language I really love! Hope it’ll become a really mature language with full crossplatform features.

[abakoboParticipant]

One thing I can probably do quite easily is wrap cpSpaceBBQuery so it accepts an mx2 function, allowing you to use it with methods, lambas etc.

I’m looking forward to see the mx2/C code for it. I really wonder how those things can be connected..

That is some bad news, my next step on playing with monkey was learning callbacks and continue some chipmunk demos..!
But this is something, chipmunk is using a lot of C type callbacks.

If you show the way for wrapping these queries, I’m ready to manually do the others if you don’t have the intention to automate the process. But I can work on Sundays only for some months.

here is a (I think exhaustive) list of the callback function types, I didn’t count the queries using them..

Monkey

[crayon-5cba9e2050004765373176 inline="true" ]'chipmunk_structs.h cpSpaceArbiterApplyImpulseFunc 'cpBody.h cpBodyVelocityFunc cpBodyPositionFunc cpBodyShapeIteratorFunc cpBodyConstraintIteratorFunc cpBodyArbiterIteratorFunc 'cpDampedSpring.h cpDampedSpringForceFunc 'cpDampedRotarySpring.h cpDampedRotarySpringTorqueFunc 'cpConstraint.h cpConstraintPreSolveFunc cpConstraintPostSolveFunc 'cpMarch.h cpMarchSampleFunc cpMarchSegmentFunc 'cpSpace.h cpCollisionBeginFunc cpCollisionPreSolveFunc cpCollisionPostSolveFunc cpCollisionSeparateFunc cpPostStepFunc cpSpacePointQueryFunc cpSpaceSegmentQueryFunc cpSpaceBBQueryFunc cpSpaceShapeQueryFunc cpSpaceBodyIteratorFunc cpSpaceShapeIteratorFunc cpSpaceConstraintIteratorFunc 'cpSpatialIndex.h cpSpatialIndexBBFunc cpSpatialIndexIteratorFunc cpSpatialIndexQueryFunc cpSpatialIndexSegmentQueryFunc cpBBTreeVelocityFunc

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

[crayon-5cba9e2050004765373176 inline="true" ]'chipmunk_structs.h cpSpaceArbiterApplyImpulseFunc   'cpBody.h cpBodyVelocityFunc cpBodyPositionFunc cpBodyShapeIteratorFunc cpBodyConstraintIteratorFunc cpBodyArbiterIteratorFunc   'cpDampedSpring.h cpDampedSpringForceFunc   'cpDampedRotarySpring.h cpDampedRotarySpringTorqueFunc   'cpConstraint.h cpConstraintPreSolveFunc cpConstraintPostSolveFunc   'cpMarch.h cpMarchSampleFunc cpMarchSegmentFunc   'cpSpace.h cpCollisionBeginFunc cpCollisionPreSolveFunc cpCollisionPostSolveFunc cpCollisionSeparateFunc cpPostStepFunc cpSpacePointQueryFunc cpSpaceSegmentQueryFunc cpSpaceBBQueryFunc cpSpaceShapeQueryFunc cpSpaceBodyIteratorFunc cpSpaceShapeIteratorFunc cpSpaceConstraintIteratorFunc   'cpSpatialIndex.h cpSpatialIndexBBFunc cpSpatialIndexIteratorFunc cpSpatialIndexQueryFunc cpSpatialIndexSegmentQueryFunc cpBBTreeVelocityFunc

[/crayon]

[abakoboParticipant]

I haven’t read all the topic… but I made some little things with chipmunk: https://github.com/abakobo/chipmunk_monkey2

It includes a modified hellochipmunk.monkey2 with MouseJoint/DistanceQuery and a debugdraw with rounded end fat segment and rounded corner polys that were missing on Mark’s debugdraw.
The debugdraw also includes a “complete”/”fast draw” and a camera using affineMat3 matrix that you can test in the stress_test example. I plan to work on it a bit more but have no time for now..

For the mouse joint I use filter this way:

Monkey

[crayon-5cba9e2054225229910872 inline="true" ]' This Filter is for collision and queries filtering, it's the filter received by default by shapes and shall collide with any other. ' The two last parameter are shown as bits because they are bitmasks Local FullGrabFilter:=cpShapeFilterNew(ULong(0),UInt(StringToULong("11111111111111111111111111111111",2)),UInt(StringToULong("11111111111111111111111111111111",2)))

1 2 3 4

[crayon-5cba9e2054225229910872 inline="true" ]' This Filter is for collision and queries filtering, it's the filter received by default by shapes and shall collide with any other. ' The two last parameter are shown as bits because they are bitmasks Local FullGrabFilter:=cpShapeFilterNew(ULong(0),UInt(StringToULong("11111111111111111111111111111111",2)),UInt(StringToULong("11111111111111111111111111111111",2)))

[/crayon]

Used stringToULong because there’s no stringToUInt…

No data however..

As c2mx2 is not copying the api docs I use the original official C api doc: https://chipmunk-physics.net/release/ChipmunkLatest-API-Reference/

[abakoboParticipant]

I’d be interested to know what everyone else has got planned for this year too, even if it’s not monkey2 related.

Working on an new apartment, from empty space so a bit of work still… After that I plan to work as much as I can on a game scenario I started in 2015.

Also, what would people like to see added most to monkey2?

-Mobile multitouch
-IOS emulator ability
-Accelerometer.
-End user friendly setup of targets.
-3D.

[abakoboParticipant]

http://stackoverflow.com/questions/5532595/how-do-opengl-texture-coordinates-work

I should go into GL but I’m afraid it’s going to eat me a lot of time…

can we already put a texture on a triangle or quad with canvas’ drawQuad and drawTriangle?

[abakoboParticipant]

I’d like to add primitive Draws to mojo.canvas. I don’t feel like I want to subclass the big mojo thing..

But rebuildin using makemods for mojo only is actually pretty fast

What would this do?

Pextension would let me create extension that can call some private methods (addvertex in this case) without having to git the mojo module with my own changes each time I update mx2..

so I can have a simple canvas.DrawBezier(blah) or canvas.DrawRoundedCornerPoly(blah) or … in my onrender

could be uesfull in other situations and I thought it can maybe very easy to add..

[abakoboParticipant]

I had a try but some assets are missing from the “performance test” “component” and “gameobject” samples

nice effects!

[Author]

Posts