I guess it's all about resources/time - if folks can make better
documentation for an implementation (or friend) stuff, or add feature X and
Y, but not both, they choose to do something, based on some criteria. Of
course, someone may agree with the choice and other may not - but the best
way to ensure things you consider important are done is to help to do them.
So, I wonder, if you could document html.editor.lib, first because you may
be the most knowledgeable of it here anyway, and also to set an example.
Thanks,
Jan
On Mon, Sep 24, 2018 at 7:13 AM Peter Nabbefeld <peter.nabbef...@gmx.de>
wrote:
> Too many nested comments inline already, sorry, so I try to summarize.
> Please feel free to remove the discussion at the bottom, if You agree it
> is not needed anymore.
>
> 1. I don't like documentation handled differently based on friend-state.
> There should always be sufficient information on every module.
>
> 2. Most people will not be able to use a module with insufficient
> information. As a result, information will be extended for public
> modules only, because people start to try using them and ask for
> information on points they don't understand. Friend-only modules will
> not have any more support. So, many jewels will never be officially
> become usable by the community.
>
> 3. I don't like to make the available documentation a criteria of the
> friend-state. Just because I'd like to see more of the modules and even
> the start phase of NetBeans documented. It would be great, if the
> original authors would have done that, all the others need to look at
> the often poorly documented code and find out, what the original author
> wanted to do with it. In some cases this is far from being obvious. In
> this case, it doesn't help much, that the code is open-source - You'll
> also be unable to read some old texts, just because You cannot read the
> letters.
>
> 4. Yes, the main stability criterium for APIs for me is no changes for
> some time - it's usually an indicator, that it has been used
> sufficiently. But, of course, if I want to use some API, I'll probably
> find some problems.
>
> 5. Thinking about [4], the friend-state could probably be managed
> easier-to-use, as a first step, e.g. adding sth. like
> "personal.friends=module(s)-to-be-used-as-friend" in user.properties.
> This could take the burden of cloning and building the whole NetBeans
> for just trying some friend-only API. Probably, in this case don't build
> nbm files but only install locally (deploy in running IDE or run a
> separate IDE), so it cannot be pushed to plugin repository by accident.
> Just an idea.
>
> 6. The possibility to extend HTML syntax using html.editor.lib is only
> one of its possibilities, but that's only one of the affected modules,
> so if we want to discuss that, we should use another thread.
>
> 7. Same for CSL/GSF/etc. - my fault. Thought You could help me with
> these. Should have opened new thread.
>
> Kind regards
>
> Peter
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscr...@netbeans.incubator.apache.org
> For additional commands, e-mail: dev-h...@netbeans.incubator.apache.org
>
> For further information about the NetBeans mailing lists, visit:
> https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
>
>
>
>
