Any thoughts on a new name for apitrace?

[José Fonseca]

I admit I'm terrible for coming up with good names.  apitrace's
original name was d3dtrace (too narrow), so I've may have gone from
one extreme to the other.

On the other hand, fact is that the tracing infrastructure is pretty
general, and could be used to trace any API really. Not only 3D, but
2D (not hypothetical, as I've laready started Direct2D support), or
even non graphics related stuff.

Anyway, I don't know what future will bring, but at this stage I'd
prefer avoiding any name changes, as it would take away time from more
important things.

"apitrace trace" is fine

Jose


On Thu, Oct 20, 2011 at 3:52 AM, Carl Worth <cworth at cworth.org> wrote:
> As I recently started creating a new top-level binary program for
> apitrace, it occurs to me that this program could use a better name.
>
> One problem with "apitrace" is that the name is a bit too generic. What
> we have here is a program to trace (and manipulate traces) graphics
> APIs, but "apitrace" sounds much more generic than that.
>
> In looking through the existing code, there are a few other names:
>
>        gltrace
>        glxtrace
>
> both of which have the opposite problem. These are too specific since
> the current code can trace some other graphics APIs besides OpenGL.
>
> Additionally, we might prefer to have a name without "trace" in it. This
> is a minor cosmetic thing in that a command like:
>
>        apitrace trace <program>
>
> just feels a bit redundant to type.[*]
>
> So, what do you think. Do you agree we could use a better name? Do you
> have any proposals for what we might use?
>
> The one concrete proposal I was going to make was "gator" (for Graphics
> Api Trace Or Replay) which works well for something you could feed huge
> traces too (and parallels nicely with wireshark). But gator is
> unfortunately already well as existing spyware software of some sort.
>
> Anybody have any good name ideas?
>
> -Carl
>
> [*] I was tempted to make "apitrace <program>" without any "trace"
> command implicitly mean to perform tracing. But this would lead to
> problems such as a typo of "apitrace trim" as "apitrace trip" trying to
> execute and trace a program named trip. Also, to support tracing a
> program with the same name as an existing command we would need to allow
> "apitrace -- <program>". And if user's have to type that, perhaps they
> should just type the "trace" command explicitly.
>
> --
> carl.d.worth at intel.com
>
> _______________________________________________
> apitrace mailing list
> apitrace at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/apitrace
>
>