Basic API usage

[Jan Bruns]

Pekka Paalanen wrote:

 >> What I can tell about wl now is that I haven't yet identified lines of
 >> text or code about it, that might probably be of not so much relevance,
 >> given specific sets of workflows. That's bad news to me, a
 >> disappointment, because making things easier for the distant future was
 >> my only motivation for focussing on wl a bit.
 >
 > Sorry, but I do not understand what you are talking about. If your goal
 > is not to write a program that uses Wayland in your preferred language,
 > then what is it?
 >
 > What workflow are you referring to?

Well, why was it almost necessary to ask this mailing list about basic 
API usage?

Thanks to your great tips I was able to get to the point where I can say 
I have the option to use the API (this was the motivation: the goal was 
to evaluate if there is such an option):

mkdir sometestdir
cd sometestdir
svn co svn://abnuto.de/wayland

Without that human interaction, this wouldn't have been possible at all, 
even though we're really talking about basic API usage, the rough things 
an API-doc should really focus on. They're dealing with it as if it were 
some piece of code that could easily replaced by something else (like, 
for example, some libbigint), but it's pretty much a system-API.

This lack of information (about what the community does) also translates 
to lack of specification, that contributes to making things (other 
people's work) worse.

For example, the vulkan API claims to support a "wayland" interface. 
Great words, relative to the situation the protocol probably doesn't yet 
deliver the functionality required to fulfill that claim. What that 
mentioned interface really is about is called libwayland. But what 
happens if the wayland protocol gets to the point of supporting the 
claim directly, without libwayland interaction? Would they introduce 
structures named like

VK_STRUCTURE_TYPE_WAYLAND_NOLIBWAYLAND_SURFACE_CREATE_INFO_KHR or 
VK_STRUCTURE_TYPE_DIRECTWAYLAND_SURFACE_CREATE_INFO_KHR, in addition to 
the existing VK_STRUCTURE_TYPE_WAYLAND_SURFACE_CREATE_INFO_KHR?

 > If someone had the time, our email discussion would be a good
 > indication for what to improve in the Wayland documentation.

For my part, I'll probably work through more advance API usage patterns 
next, probably not the best point in time to recommend specific 
documentation-patches.

Gruss

Jan Bruns