colingilbert / projectzombie Goto Github PK
View Code? Open in Web Editor NEWAutomatically exported from code.google.com/p/projectzombie
Automatically exported from code.google.com/p/projectzombie
Currently, we have the concept of ZEntity, which is a class which represent
ZEntities in the system. Various system then use this as a base to create their
own view of the ZEntity. The current system has only the bare minimums in terms
of corresponding between the various systems. We need to make this better. For
Render Entities, this is fine, because it is done through the Command and
Control systems, and it operates directly on the ZEntities themselves (i.e.
they have strong cohesion, object oriented cohesion.) However, for the
correspondence between OpenCL buffers and ZEntities is not so clear. We have a
function which converts ZEntities managed by EntitiesManager which converts the
ZEntities into data buffers. It works right now. We may want to make this
better by implementing a system which does this conversion better.
Original issue reported on code.google.com by [email protected]
on 19 Aug 2010 at 12:06
VolumeMap works. But it's time for a refactor. Here is a quick 5 min run-down
of what needs to be change.
1.)
VolumeMap internally should use page x,y. Try to eliminate the page ID to long
pageX, Y conversion as much as possible.
2.)
PageRegion should be refactored so it stores the computed value such as page x,
y. Currently, there is a lot of conversion going, conversions that can be saved
and reused across different methods.
3.)
We need to move all the addressable to be vector3s instead of all vector2 as we
are doing now. Even if we are completely 2d at the moment.
Original issue reported on code.google.com by [email protected]
on 11 Feb 2011 at 11:54
This is an overall code review request:
-EngineController:
Review the entire EngineController system.
-Game States:
Review the game state system.
-Command and Control:
Review the command and control system.
Note:
For each system, please input any issues, objections, and thoughts.
Original issue reported on code.google.com by [email protected]
on 21 Aug 2010 at 12:40
What steps will reproduce the problem?
1. Do nothing.
2. Happens on shutdown.
3.
What is the expected output? What do you see instead?
We expect no leaking elements from libRocket. Instead during shutdown libRocket
debugger reports:
47 leaked elements detected.
(0) #text -> #text < tab:selected
(1) #text -> #text < tab
(2) #text -> #text < tab
(3) tab -> tab
(4) div -> div
(5) form -> form#hdrhighform < div
(6) select -> select#tone_mapper_select_id < form#hdrhighform < div
(7) selectarrow -> selectarrow < select#tone_mapper_select_id < form#hdr
highform < div
(8) selectvalue -> selectvalue < select#tone_mapper_select_id < form#hdr
highform < div
(9) selectbox -> selectbox < select#tone_mapper_select_id < form#hdrhigh
form < div
(10) select -> select#glare_type_select_id < form#hdrhighform < div
(11) selectarrow -> selectarrow < select#glare_type_select_id < form#hdr
highform < div
(12) selectvalue -> selectvalue < select#glare_type_select_id < form#hdr
highform < div
(13) selectbox -> selectbox < select#glare_type_select_id < form#hdrhigh
form < div
(14) tab -> tab
(15) tab -> tab
(16) div -> div
(17) form -> form#cineform < div:focus
(18) select -> select#camera_select_id < form#cineform:focus < div:focus
(19) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(20) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(21) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(22) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(23) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(24) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(25) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(26) #text -> #text < option < selectbox < select#tone_mapper_select_id
< form#hdrhighform < div
(27) #text -> #text < option < selectbox < select#glare_type_select_id <
form#hdrhighform < div
(28) #text -> #text < option < selectbox < select#glare_type_select_id <
form#hdrhighform < div
(29) #text -> #text < option < selectbox < select#camera_select_id:focus
< form#cineform:focus < div:focus
(30) #text -> #text < option < selectbox < select#camera_select_id:focus
< form#cineform:focus < div:focus
(31) #text -> #text < selectvalue < select#camera_select_id:focus < form
#cineform:focus < div:focus
(32) selectarrow -> selectarrow < select#camera_select_id:focus < form#c
ineform:focus < div:focus
(33) selectvalue -> selectvalue < select#camera_select_id:focus < form#c
ineform:focus < div:focus
(34) selectbox -> selectbox < select#camera_select_id:focus < form#cinef
orm:focus < div:focus
(35) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(36) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(37) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(38) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(39) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(40) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(41) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(42) option -> option < selectbox < select#tone_mapper_select_id < form#
hdrhighform < div
(43) option -> option < selectbox < select#glare_type_select_id < form#h
drhighform < div
(44) option -> option < selectbox < select#glare_type_select_id < form#h
drhighform < div
(45) option -> option < selectbox < select#camera_select_id:focus < form
#cineform:focus < div:focus
(46) option -> option < selectbox < select#camera_select_id:focus < form
#cineform:focus < div:focus
GuiController::shutDown()
What I think is going on:
It has to do with creation elements (Selection in this case) with c++ code.
LibRocket has this reference counting thing but you have to manually call
ReleaseReference on objects. Since I'm new, I'm not sure when to call this
ReleaseReference. I think you are supposed to call it whenever you add an
object to something, since the place where you added this object to now has a
reference of this object, you need to call ReleaseReference after adding,
manually. This is confusing.
Original issue reported on code.google.com by [email protected]
on 12 Jan 2011 at 7:47
After refactoring of Screens using factories via Controller the Screen
transition code is now broken.
Why?
Reason is because it was wrong previously. You meant to have the MainMenuScreen
as the root document, but instead DebugScreen was added first thus it was the
root. So the entire pop/push transition stuff is wrong.
Original issue reported on code.google.com by [email protected]
on 6 Jan 2011 at 12:29
Need to setup staging and deployment structure.
-How to stage?
-How to deploy?
-Continuous build systems?
Original issue reported on code.google.com by [email protected]
on 13 Jan 2011 at 1:06
PolyVox::Volume does behind the scene shared_ptr memory management WRT for
PwoerOfTwo block sizes. In Page generation, for each page we also generate an
extra value on the shared edge with neighboring blocks. Due to the memory
access pattern, most of the time this is fine. However, it is possible that due
to the interleaved processing nature of threads, we could have a memory where a
PolyVox Block is leaked.
The plan for fixing this is to use Intel Thread Block's atomic feature and
basically keep a reference count. Such that, the first person accessing an edge
will atomic write to a variable that subsequent threads can check, so the first
thread will have the sole responsibility of generating the value at that edge.
It's not a show stopper ATM because the thing with memory access patterns
(basically the access is sequentially across the X's, so most of the time when
a block is address an shared edge, the neighbor would have moved on already, so
no memory leakage occurs).
Original issue reported on code.google.com by [email protected]
on 8 Dec 2010 at 1:03
Make Life-Cycle delegate registration process more robust.
Original issue reported on code.google.com by [email protected]
on 15 Dec 2010 at 6:27
We discussed the need to remove Camera from EngineView singleton because we
need to implement camera events (remember we added Camera there for testing;
well it's time to implement camera system properly).
We need to just get rid of EngineView completely. This is NOT a big change
because we were careful to limit getting stuff from EngineView only in onInit
functions of the LifeCycle abstraction.
The change is simply change onInit function declares to pass in SceneManager.
But before this can be done you need to implement the Camera system properly as
Camera is gotten from EngineView!
Original issue reported on code.google.com by [email protected]
on 13 Dec 2010 at 6:23
What steps will reproduce the problem?
1. Make the system unable to find resources need by Console.
2.
3.
What is the expected output? What do you see instead?
Expected output is the proper destruction of system resources. Instead, we see
that an Ogre exception is caught, and the destruction sequence is initiated,
but the system crashes in CommandController destructor.
Please use labels and text to provide additional information.
Original issue reported on code.google.com by [email protected]
on 17 Aug 2010 at 3:11
For some reason, we are NOT using delegate classes with forward declaration.
This leads to the problem of having to recompile many files every time a new
event is added, etc.
Solution:
We already started on this. Implment a ZPrerequisite file and stick forward
declarations there. Why wasn't this done to begin with? Well, we thought about
doing that, but somehow decided it was better to do foward declaration in each
file. But this is a pain. So let's go back to sticking it in ZPrerequisite.h.
Thanks.
Original issue reported on code.google.com by [email protected]
on 29 Dec 2010 at 5:50
When we switched over to generation the borders first, when iterating through
the region after border, we didn't change some of the values to reflect the
fact we first generated the borders. So there is a off by one error.
It's easy to correct. It's only off by one.
I think the big ugly stripes we see for SkyX ground scattering is caused by
this.
Original issue reported on code.google.com by [email protected]
on 12 Feb 2011 at 12:26
The original plan was to have the ability to intercept delegate calls. One
should be able to filter any sort of event in a generalize way so to intercept
such delegations upstream from any downstream delegates. Use case: For example,
in our GameMainState, we have various modules which correspond to the main
state of our game. Say we want to have an SDK tray. And such that we want to
disable ControlModule when SDK tray is in view. Currently, we are doing this
the EngineController level by doing multiple checking in OIS::MouseEvent*
listeners call-backs. We should implement the system such that when we register
SdkTray (using the current system) we can specify interception delegates. We
should be filter it to say we want a single delegate to handle ALL MOUSE
events, or any particular mouse events.
This is doable with the current system and SHOULD NOT involve major
refactoring.
Original issue reported on code.google.com by [email protected]
on 22 Aug 2010 at 7:13
What steps will reproduce the problem?
1. Keep hold down movement key while paging is loading.
What is the expected output? What do you see instead?
Expected output is continuation of program as normal. What one see for certain
cases is the system will crash on loading one of the pages.
Please provide any additional information below.
Okay this has been fixed temporarily by locking the movement to 60fps or
higher. The idea here is when loading pages the game slows down, and the
movement continues, which causes more pages to load, overwhelming the system
somehow. When using threading this causes problems.
Please either fix using something besides locking movement, or put up a loading
screen when loading takes too long.
Note: another fix is to move the locking movement thing to somewhere else
beside locking in the movement module.
Original issue reported on code.google.com by [email protected]
on 7 Dec 2010 at 6:54
I know onDestroy is supposed to be part of the life-cycle. But we need to use
more RAII. This is definitely doable. We tried to do this just never got around
to finishing it.
Some parts uses RAII. We need to use RAII everywhere where possible.
Original issue reported on code.google.com by [email protected]
on 23 Dec 2010 at 4:49
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.