[pLog-general] On php5...

[Oscar Renalias]

> > >>> 3 Add module support, and move admin class,view and etc to
> > >>> modules/admin£¬sumamry action,view and etc to modules/summary,other
> > >>> module can be added too. More generally, I think only plog core 
> > >>> classes
> > >>> should be putted into class directory, others should be treated as
> > >>> modules or plugins.
> > >>
> > >> I don't quite agree with this but I think we could discuss about it...
> > >
> > > If only considering blog system, moudle support maybe redundent. I want
> > > to add module support because:
> > >
> > > 1 New branch of plog will not only a blog system, but also an
> > > application framework for php to develope more complex system.
> > > 2 Module support is easy to implement:-)
> > >
> > > But, I have no idea which is suitable,module or plugin?
> > 
> > Sincerely, I don't like it.
> > 
> > Plugins is a concept that we're using to expand the functionality of 
> > plog itself by adding new menu entries (in 1.0) or exporting additional 
> > objects to the templates, etc. I also don't see how a "module" works or 
> > it is defined.
> > 
> > But if all you want to do is separate the base code of the framework 
> > from the code of plog, as a blogging tool, then we just need to 
> > rearrange the directory structure a bit.
> > 
> > Finally, there is already a framework that was spun off from the 
> > post-0.3 code of plog, but it hasn't gone quite public yet. I have to 
> > say that the idea didn't quite work out, and to tell you the truth, I 
> > don't feel like going through the same hassle again. Different people 
> > have different ideas about how things should work, etc, and plog was 
> > supposed to be the flagship application of the new framework but in the 
> > end we had to call the whole thing off (well, the framework is out 
> > there but plog was never integrated into it) because it was too 
> > difficult.
> > 
> > As a personal opinion, I have noticed that in order to build a powerful 
> > framework you have to make a lot of generalizations since that's the 
> > point of a framework: try to be the base for many applications as 
> > possible. And when you want to build an application, you have to ignore 
> > most of all generalizations and build concrete, precise things. It is 
> > indeed true that plog is built on top of its own framework, but it's a 
> > framework that has developed overtime and most important of all, 
> > focused exclusively on plog. I think plog's framework should stay that 
> > way, focused on plog and nothing else.
> > 
> > Also think about it, we also do not need to carry the burden of 
> > additional things that should be part of the framework when plog 
> > doesn't really need any of them, do we?
> > 
> > All in all, I would vote -1 against it...
> 
> I understand what you mean, and agree with you that a general framework
> is difficult to build, and also agree with you that we should first make
> 1.0 test works well and release 1.0 first, that's the most important
> work for you and for plog community, so talking about plog 2.0(?) is not
> very suitable at this time:-)
> 
> But how about let's do it eventually? Don't expect to make a good
> framework suddenly, but day after day, we'll make it better and better.
> 
> In fact, MVC archticture is a good start point, you have done it very
> well.  

Thanks :) But sometimes I'm not sure that what I've done so far would be useful 
for other applications... As I said, the first time we tried to build a 
framework off plog's code we realized that things in the current framework have 
been built to make things easier for plog, not for any other application, and 
the discussion started drifting towards how we should rebuild plog to make it 
fit into the framework.

I don't know, whatever we have now works well (all right, some parts could work 
better :)) I'd be more than happy to start building a powerful framework as 
long as it doesn't mean that plog has to change much but let's see what happens 
after 1.0!

Regarding getting plog 1.0, since that's the step before any discussion about 
architectures and frameworks, do you think you could give us a hand in getting 
1.0 work well with php5? Just working, we don't want to use any of the new 
fancy OOP features in php5 because the goal is to get plog work equally well 
under php4 and php5! Could you help there?
 
> I'd like to open another thread to talk about plog archticture.

Go ahead.
 
> 
> > Can't we try to fix log4php? Have you tried contacting the author? 
> > Perhaps he's got a new version coming out soon to solve all the 
> > problems with php5?
> 
> I have tried to contact with the author, but the mail was refused,
> sadly:-(.
> > 
> 
> > roadmap --> The 1.x releases (as many as needed) should be based on the 
> > current codebase with minor fixes to make it *just* work under php5. 
> > The 2.x branch should be based on the new development tree making use 
> > of the newest OOP features in php5.
> 
> Agree with you totally.
> 
> > 
> > But keep in mind that we haven't even released 1.0 yet and we're 
> > already talking about 2.0... :-)
> 
> Yes, but why not:-)
> 
> > 
> > new branch --> I say we call it "plog-php5", or something like that :-)
> > 
> > coding standards --> 
> > http://www.plogworld.org/wiki/index.php?pagename=Coding%20guidelines
> > 
> As to code style, my two cents:-)
> 
> 1 Name class with a uppper case word.

Aren't we already doing that?

> 2 Variables and functions name do not need a preceeding "_" when using
> php5,because we can use private to limit the access.

yeah but that's only for php5... For the time being, we'll have to live with 
that :)

Oscar