Re: Known MacOS programmer about WO/EOF

[David Ayers]

Dennis Leeuw wrote:

> David Ayers wrote:
>
> > What we need is developers that already know how EOF/WebObjects and 
> > can spend time working on it.  There are a few that are trying to get 
> > into it.  But they often lack the experience and resources (reference 
> > implementations) to really get the correct implementation.  I'm sad to 
> > say, that I simply don't always have the time or even the know-how to 
> > help them out, even though I try when I can.  Another problem I see, 
> > is that some of the work to squeeze out more efficiency has made the 
> > code much harder to read, making it even harder for new folks to help 
> > out. I'm still hoping to find a way to remedy that, but to do that 
> > I'll need to find some time to do some real benchmarking myself to 
> > make sure we don't loose measurable performance.
>
>
> On the otherhand releasing let's say a 0.1 version might create 
> attention which then might attracked more developers. I don't think any 
> developer expects a 0.1 version to be a version to base a production 
> application upon.

I've had often had this discussion.  I also often had discussions on IRC 
about how much someone has been wasting their time with GDL2 because 
they expected this or that to work or this or the other class to be a 
fundamental part of GDL2.  I have the feeling that frustrating the folks 
which are willing to take a look by premature releases is worse than the 
status quo.  On the other hand, I do not want to seem like I'm hijacking 
any project here.  If someone wants to make a GDL2/GSWeb release, please 
go ahead.  Personally I think it would be a disservice but that's just 
my personal opinion and feel free to prove me wrong.

In my view, the issue is not so much that GDL2/GSWeb are hobbyist 
projects.  If they were hobbyist projects, I think the implementation 
would be much "cleaner" (as in easy to learn from and extend).  But 
instead they are a foundation for certain production implementations. 
This means that certain hacks remain to make sure production code 
doesn't break.  It means that it's filled with logs to identify issues 
in when debugging in production scenarios which obfuscate the actual 
code.  It means that there is a lot of dead code that can presumably be 
easily reactivated if deemed necessary in a production environment.

Personally, I think what probably needs to be done to making both 
projects more accessible to others, is to remove dead code, the a lot of 
the logging and any hacks that are papering over application issues. 
But to do this without jeopardizing production code, we also need an 
ObjC based testing suite.  I've worked on top of Alex's scripts and 
tried to integrate it into -make but it's still has some ugly hacks and 
caveats, but I'll clean it up an float it around to see if someone can 
pick it up.  If it gets integrated, I can commit the tests I have for 
GDL2 and from there we may be able to build a test suite that we can use 
to verify the functionality and rely less on logging.  Then we need to 
get into real profiling so that we can insure performance without the 
global IMP caching we currently started.  Of course we also need 
documentation.  Well all this is just the vague direction I'm looking 
at, since the time I can spend on it is limited.  Manual, could float 
these ideas to see if we can agree on a general plan?

Cheers,
David