Do we need our own library?

136 views
Skip to first unread message

Guy Davidson

unread,
May 1, 2015, 10:27:49 AM5/1/15
to unofficial-r...@googlegroups.com
A large chunk of the standard library started life in Boost.  Here are some questions which I think merit discussion:

1. Would we benefit from our own public library which offers solutions for real-time programmers?

2. What might we put in there besides containers and allocators?

3. Would IP issues limit contributions to personal development rather than industry hardened classes?

Cheers,
G

Sean Middleditch

unread,
May 11, 2015, 8:06:31 PM5/11/15
to Guy Davidson, unofficial-r...@googlegroups.com
Hello Guy,

On Fri, May 1, 2015 at 7:27 AM, Guy Davidson <elgu...@gmail.com> wrote:
> A large chunk of the standard library started life in Boost. Here are some
> questions which I think merit discussion:
>
> 1. Would we benefit from our own public library which offers solutions for
> real-time programmers?

Why could this not be part of Boost? Not saying it should be - and I'd
be for something separate just to avoid the negative connotations of
Boost - but I do feel a little like the XKCD comic about standards
(https://xkcd.com/927/) every time I start a new library... :)

>
> 2. What might we put in there besides containers and allocators?

Math? When working on a shared low-level library for multiple
engines/games, the big items tend to be:
- STL replacement (containers, allocators, strings)
- math library
- low-level debug hooks (assertions, crash handlers, stack traces, etc.)
- reflection

> 3. Would IP issues limit contributions to personal development rather than
> industry hardened classes?

Probably. I know I wouldn't be able to supply anything without at
least an uphill battle (WG has released code to GitHub before under
permissive licenses, but not frequently).

>
> Cheers,
> G
>
> --
> 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/6e8491fc-9041-4b89-88f2-8b33f7016939%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.



--
Sean Middleditch
http://seanmiddleditch.com

Michael Wong

unread,
May 13, 2015, 10:45:12 AM5/13/15
to Sean Middleditch, Guy Davidson, unofficial-r...@googlegroups.com
That was my thought too. This sounds like a great Boost library proposal to get started immediately, which usually evolves into a Standard technical report. But now that we have our own SG14, the library could be devloped as a library for the SG14 Games Dev/Low Latency TR.

But either route is great. The boost route gets more eyes on it if we did not have an SG, but now, I am sure a lot of people will be looking at it.

As for IP, anything going to standard/ Boost must have IP clearly stated. One can have a patent/copyright on the implementation, but clearly the interface must be Open.

Guy Davidson

unread,
May 22, 2015, 3:27:41 AM5/22/15
to unofficial-r...@googlegroups.com, se...@middleditch.us, elgu...@gmail.com
Sorry about the delay: Total War is getting interesting right now...

Then I guess my question is: how would I start the SG14 library?  Is there a "standard form" for names and identifiers, organisation, location &c. or would I simply create a GitHub repo and shout loudly across the web?

I don't have much to add in yet but offering somewhere for submissions is surely a good thing.

Cheers,
G
--
Sean Middleditch
http://seanmiddleditch.com

--
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.

Sean Middleditch

unread,
May 25, 2015, 5:46:39 AM5/25/15
to Guy Davidson, unofficial-r...@googlegroups.com
On Fri, May 22, 2015 at 12:27 AM, Guy Davidson <elgu...@gmail.com> wrote:
> Sorry about the delay: Total War is getting interesting right now...

Haha, yeah, that's probably why game developers have so little
interaction with the committee... who of us has the time and energy
with these work schedules? :)

> Then I guess my question is: how would I start the SG14 library? Is there a

My first question is: a library for what? Is this to flesh out a new
implementation of a proven design to build a proposal around it? To
test out new ideas and experiment with solutions to previously
unsolved problems? These are very different things.

> "standard form" for names and identifiers, organisation, location &c. or

I'd lean towards the conventions of the C++ standard document where
possible. Makes it easier to write those proposals, I'd think. :)

This is what Boost did. At least for the Boost libraries I've
interacted with (not sure if there's a Boost-wide standard form or
not).

> would I simply create a GitHub repo and shout loudly across the web?

Wouldn't be a bad start. It can move under a single GitHub
Organization, perhaps. The committee, Michael, or an officially
designated trustee should create/own said Organization and any other
resources (official mailing list, etc.) I think, otherwise I'd go
ahead and create them.

Guy Davidson

unread,
May 25, 2015, 9:01:47 AM5/25/15
to unofficial-r...@googlegroups.com, elgu...@gmail.com
I imagined the library would be for depositing proven ideas for things like contiguous containers, "fast" rather than "correct" maths implementations and so on.  Having said that, I guess it's a little bit of both of what you suggest: proposals can be built around finished submissions, but also submissions can be developed and finished prior to proposal generation.  Perhaps the namespace can be divided between developed and developing.

At the very least, there are several ideas swirling around that need somewhere to exist, even if it's just a header with a class name and an overview comment.

Regarding the naming, I was thinking more of the hosting than the implementation: is there a standard form for the name of the library or the hierarchy of the contents &c.  Following the standard is fine, but not sufficient.

When I get a moment I'll create an SG14 repo and see what happens.

Cheers,
G
Reply all
Reply to author
Forward
0 new messages