[Geany-Devel] Adding a plugin

[Roger Booth]

On 06/09/2013 04:48 AM, Frank Lanitz wrote:
> On Sun, 09 Jun 2013 03:54:05 -0700
> Roger Booth <rbooth at kabooth.com> wrote:
>
>> I assume you are surprised by this. How dare the user of our docs not
>> study to great lengths the implementations of the shit we (and others
>> that have come before us) have produced. Nobody should have the
>> effrontery to attempt to interface our code without spending a
>> sufficient amount of time appreciating the edifice we have erected.
> Please don't be a party-puper. Geany is FLOSS and everybody is welcome
> to improve ... also the documentation. You recognized that the
> documentation might not be clear in all cases. .. yepp. That's true.
> But it's hard to find such things for somebody who knows the code from
> deep inside the heard that Geany-Version -> API-Version -> ABI-Version.
> so no need to document it for them.
> Please take a look at github and improve the parts of documentation.
> Complaining on the mailing list that the whole documentation suckz just
> is making people spending a lot of spare time sad and sending them into
> aggro-mode.
>
> Would be really happy to see a patch improving that aspect from
> documentation as well as the Plugin-HowTo.
>
> cheers,
> Frank
>
>
> _______________________________________________
> Devel mailing list
> Devel at lists.geany.org
> https://lists.geany.org/cgi-bin/mailman/listinfo/devel
But if I'm going to *improve* the docs, I *still* have an outstanding 
(in a manner of speaking) question:

Why does the function plugin_init() accept the one parameter GeanyData*?

The reason that is a question is that the three global variables

GeanyPlugin         *geany_plugin;
GeanyData           *geany_data;
GeanyFunctions      *geany_functions;

are available to a plugin and you will notice that one of those global 
variables is geany_data which is suspiciously similar to the parameter 
to plugin_init().

So either there is no reason to pass the parameter to plugin_init() or 
there is a good reason to pass the parameter to plugin_init().

Either way, I'd like to know. Maybe others would to.

A previous reply from Matthew Brush made me think that I need to really 
spell out what I mean here. Say I have the following

GeanyData           *geany_data;
extern "C" void plugin_init(GeanyData *bob)
{
     if(*geany_data == *bob) {
         exit(1);
    } else {
        exit(2);
    }
}

What I'm trying to ask by this example, and remember, I'm a moron, is

     'is geany_data the same as bob?'

Please don't get all wishy-washy on me here. geany_data and bob are both 
pointers. My question is

     Do bob and geany_data point to the same location in memory.

Now you could all say "geez, moron, just do a comparison". I know how to 
do that. But that would only say that this version is blah.
What I want to know is about API guarantees. I want to be able to 
document the plugin_init() function and say

     geany_data == bob

or not.


P.S.

>> Complaining on the mailing list that the whole documentation suckz

Um, I was commenting on one macro for the plugin API. Actually, I think 
in general the docs are superb. I want to improve the areas where I see 
they can be improved, and I volunteer to do so and have already done so. 
What pissed me off was my impression that yall thought this particular 
area of the docs was just peachy and I was just blowing smoke.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geany.org/pipermail/devel/attachments/20130609/8ef99ea9/attachment-0001.html>