cppcon 2015 submission?

[Nicolas Guillemot]

Hi folks. As you may already know, submission for cppcon 2015 talks was recently opened. (http://cppcon.org/submissions/) It might be worth submitting a talk related to the discussions in this group.

I've been thinking about potential topics....
An overview of classic C++ gamedev design patterns that shows how they can be improved with C++11/14, and how they could be improved further with C++17?
An analysis of the constraints in game development, in order to teach the context and motivation to people who don't usually do game development with C++?
Some explanations and demonstrations of proof of concepts related to the proposals here?

What do you think?

[Scott Wardle]

I think a talk would be good.  Maybe more useful then the paper really.  You have to submit the topic by when May 22 was it? 

Scott Wardle スコット・ワードル

--
You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.
To post to this group, send email to unofficial-r...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/7f73f5af-4b72-4adb-bb26-abb38fdd3bef%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Michael Wong]

A talk is almost a forgone conclusion at this stage, given that we had a number of game talks as well as one of the keynote at CPPCon 2014. But it will be a much better talk as we develop the paper, through the next few months and with feedback from the committee for the next few months, especially now that we have a starting point. Many of the other talks at CPPCon all started life as a paper developed through the committee.  I always find it better if we have both as I sometimes develop the talk from the written paper, but I also sometimes start with a presentation and write the words so there is no defined order. But keep in mind that talks happen once a year (assuming its only CPPCon, but papers and mailing, happen 6  times a year minimum. A paper will keep this in people's mind more then a talk. 

Also another reason for both. The talk evangelizes and attracts attention. The paper is how things get changed in ISO C++ (unless that is not the point of this group). As it turns out, this paper has so far already attracted almost as much attention as Transactional Memory as I have gotten 3-4 emails asking to join. This often surprises me as to how many people actually read papers.

On Friday, April 17, 2015 at 6:27:29 PM UTC-4, Scott Wardle wrote:

I think a talk would be good.  Maybe more useful then the paper really.  You have to submit the topic by when May 22 was it? 

Scott Wardle スコット・ワードル

On Apr 17, 2015, at 14:23, Nicolas Guillemot <nlgui...@gmail.com> wrote:

Hi folks. As you may already know, submission for cppcon 2015 talks was recently opened. (http://cppcon.org/submissions/) It might be worth submitting a talk related to the discussions in this group.

I've been thinking about potential topics....
An overview of classic C++ gamedev design patterns that shows how they can be improved with C++11/14, and how they could be improved further with C++17?
An analysis of the constraints in game development, in order to teach the context and motivation to people who don't usually do game development with C++?
Some explanations and demonstrations of proof of concepts related to the proposals here?

What do you think?

--
You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.

To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-time-cxx+unsub...@googlegroups.com.

[Michael Wong]

Making sure we don't forget this. Please keep track of the final date for submission. We can build a somewhat abstract high level talk proposal based on the work of this group. Nothing so specific that it would disallow us to not add content as the design develops. But something open but provocative like:

What is wrong with C++ for Games, embedded, real-time etc ...

But not that long. And something better ....

[Michael Wong]

HI I just put in a submission for CPPCON 2015 on atomics and memory model:
C++11/14/17 Atomics the Deep dive: the gory details, before the story consumes you!
MY mentor will put in Part 2 which talks about how to implement memory order consume while I will talk about all the other ones (seq_cst, relaxed, acq, rel, acq_rel)

Given that Herb and I are plannign a dedicated Games Track day at CPPCon, may be we should start discussing a possible submission here. Let me start with a possible title here and fill in the details + make it catchy?

The Birth of ISO C++ Games Dev/Low Latency SG14
.....

[Michael Wong]

Or
Why does ISO C++ suck for games and what can you do about it?

[Michael Wong]

Does ISO C++ still suck for games or just OK?

[Klaim - Joël Lamotte]

Standard (ISO) C++ Low-latency issues: the gamedevs perspective.

--

You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.

To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.

To post to this group, send email to unofficial-r...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/5331c7bb-5add-4752-8935-1f9f912629a5%40googlegroups.com.

[Michael Wong]

When is the submission deadline? I am ok with any of these titles.
We should fill in a basic abstract content....here is a suggestion although I don't quite like it still. May be someone can improve on it.


C++ is paramount for games development, and low-latency real-time applications everywhere. But has it recall improved since C++98/03? What features in C++ 11/14 has helped, and what else do we still need from C++ 17/22? In last year's CPPCon, there was clearly demand for even better support for this community as evidenced by the large number of games submission and an impromptu BoF.

But this industry has a demanding schedule and cannot freely attend ISO C++ standard meetings. So when the prophet cannot go to the mountain, the mountain has decided to come to the prophet!

This year, we like to announce the formation of an official SG14 that will go where the community is, at CppCon 2015 will be an official SG14 meeting followed by an SG14 meeting at GDC 2015 hosted by Sony.

This talk will describe the initial findings of the unofficial cox real time google group where we considered improvements such as flat map, intrusive container, and other suggestions for better support for games development and low latency in ISO C++.

[Paul Pedriana]

At some point I want to write an EASTL retrospective. How it's used, how it differs from C++11/14, what I would have done differently, etc. I don't know if that's for a presentation or just a writeup.

--
You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.
To post to this group, send email to unofficial-r...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/01b4df2c-c6a8-4a90-b5c9-ce7d23ece77e%40googlegroups.com.

[Michael Wong]

It could be both. I like to use a paper and evolve it into a talk. Then I have a solid foundation for the talk already. A paper can also get early feedback if there was a meeting to discuss it. A slide is harder unless you give the talk first. But sometimes I also do the opposite order, and expand a talk into a full paper. In this case, its depends on timing of when each of those pieces are done. For C++ now that we have SG14, we will have to soon transition this google group to an SG14 reflector, and submit papers against SG14. There will be 3 mailing times for papers. One next Friday, One in about 2 months, and one in September 14 just before the next C++ Std meeting in Hawaii. But since SG14 will meet at CPPCon, a paper would get reviewed there , if it is sent in in the Middle mailing date( I have to find out when that is). Then it could also be converted into a presentation for CPPcon.

On Monday, May 18, 2015 at 6:42:34 PM UTC-4, Paul Pedriana wrote:

At some point I want to write an EASTL retrospective. How it's used, how it differs from C++11/14, what I would have done differently, etc. I don't know if that's for a presentation or just a writeup.

On Mon, May 18, 2015 at 10:17 AM, Michael Wong <fragga...@gmail.com> wrote:

When is the submission deadline? I am ok with any of these titles.
We should fill in a basic abstract content....here is a suggestion although I don't quite like it still. May be someone can improve on it.


C++ is paramount for games development, and low-latency real-time applications everywhere. But has it recall improved since C++98/03? What features in C++ 11/14 has helped, and what else do we still need from C++ 17/22? In last year's CPPCon, there was clearly demand for even better support for this community as evidenced by the large number of games submission and an impromptu BoF.

But this industry has a demanding schedule and cannot freely attend ISO C++ standard meetings. So when the prophet cannot go to the mountain, the mountain has decided to come to the prophet!

This year, we like to announce the formation of an official SG14 that will go where the community is, at CppCon 2015 will be an official SG14 meeting followed by an SG14 meeting at GDC 2015 hosted by Sony.

This talk will describe the initial findings of the unofficial cox real time google group where we considered improvements such as flat map, intrusive container, and other suggestions for better support for games development and low latency in ISO C++.

--
You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.

To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-time-cxx+unsub...@googlegroups.com.

To post to this group, send email to unofficial-r...@googlegroups.com.

[Michael Wong]

John Kalb just emailed and I think submission deadline is this weekend. We should start submitting. Anyone else interested in submitting to CPPCon should also submit. Thanks.

[Scott Wardle]

I hope you can do a talk too Paul.  I would love to hear your thoughts on EASTL now.   It is still critical part of all EA games.   

Here is my abstract.  I guess I have to submit this tomorrow so if you have time for feed back that would be great.  I will submit it anyways and I will see how it goes.  

My talk will be about the problems games have and how we currently solve things.  I hope people can give me suggestions on how we can do things better.  

I might talk about EASTL and STL however really most of my coding on large project uses EASTL so I don’t think I will go that deep.  The ideas that I will talk about are universal problems. 

I don’t know how much I can show our memory tools and things but so far I have good feed back from my managers on this. 

Abstract

A veteran game software engineers perspective on the current memory and C++ debugging tools used at Electronic Arts.  

  

PS4 and Xbox One have virtual memory and 64 bit address spaces, GPU and CPU are getting closer in the ability to work virtual memory.   So our tools are getting better and better and closer to PCs.  

 

Most of games memory use is art and level data like bitmap textures and polygon meshes.  So artist and designer need to understand how much their art takes up.  Giving them call stacks of memory allocations does not help.  They want to know how big is a group of building is.  Why is this group of building bigger than this one?  Maybe this one has some animation data or one of the textures is too big.  But there are 10,000s of objects.

  

Outline of content (that I will not be held to)

A veteran game software engineers perspective on the current memory and C++ debugging tools used at Electronic Arts.  

 

PS4 and Xbox One have virtual memory and 64 bit address spaces, GPU and CPU are getting closer in the ability to work virtual memory.   So our tools are getting better and better and closer to PCs.  We have tools to find memory corruption that are fast enough for games.

 

Games have millions of allocations taking up GBs of ram.  Games are written 100s of software engineers spread around the world.  We have even more artist and designer then engineers on most game teams.  Getting everyone to understand our budgets has been hard.  With 100s of people all pointing fingers at one another you have to make it clear who owns what problem. 

 

Most of games memory use is art and level data like bitmap textures and polygon meshes.  So artist and designer need to understand how much their art takes up.  Giving them call stacks of memory allocations does not help.  They want to know how big is a group of building is.  Why is this group of building bigger than this one?  Maybe this one has some animation data or one of the textures is too big.  But there are 10,000s of objects.

 

Games are on embed real time systems where if your code takes too long the code can run quite differently.  Also getting to the point in your level that you want to test takes longer then build times.  So our best debugger will often debug optimized code mostly in ASM.

 

C++ is the heart of our engine and is what makes this all possible.

 

Outline of content:

• A History game memory systems
• EASTL vs STL use at EA
• Debugging in games
• Hierarchical assets memory tracking 
• Enforcing SEs allocators use in STL
• Working around limits in new and delete
•  Memory corruption detection
• Memory leak detection 

To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.

To post to this group, send email to unofficial-r...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/16362f85-781d-4a04-8287-ae94c45c3ba3%40googlegroups.com.

[Nicolas Guillemot]

Scott, I'm sold on this talk!

To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.
To post to this group, send email to unofficial-r...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/01b4df2c-c6a8-4a90-b5c9-ce7d23ece77e%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.
To post to this group, send email to unofficial-r...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/16362f85-781d-4a04-8287-ae94c45c3ba3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "unofficial-real-time-cxx" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unofficial-real-ti...@googlegroups.com.
To post to this group, send email to unofficial-r...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/unofficial-real-time-cxx/65F999E8-07D1-4D98-BE99-3E512145687E%40gmail.com.

[Michael Wong]

+1

[Matthew Bentley]

I submitted my own talk on plf::colony but it got rejected unfortunately. Good luck for the accepted talks!