[I just had 1H or so I spent to write Emails back to you this weekend,
but now I'll be back at work with long hours, so I won't have much more
time from Email]
On Mon, Sep 03, 2012 at 09:21:06PM -0700, George Farris wrote:
> > I'm not sure if I agree. I only wrote limited amounts of new code, but I
> > rewrote most of the hai rcxx driver, and wrote the CM26a driver using the
> > w800 driver as an example, and I didn't find writing drivers was hard.
> Great so then would you be able to provide a layout of what code goes
> where and what one has to modify to let MH know there is a new driver?
I just took the W800 driver and modified it. Open the file and read it,
or for that matter open my CM26a driver which is even shorter.
You'll see the callback in there.
> Sorry I didn't mean the main svn branch I meant a copy of the main
> branch and call it test or next version or whatever handle people want.
> Obviously we wouldn't modify trunk that was a slip on my part.
Understood, although to be honest for now I'd rather not see new
branches in svn until there is a concrete plan and you have code to show
If this were git, you could just do this at home anyway. Sorry that with
svn it's just a bit more awkward, but ultimately even with svn you can fork and
develop your idea at home for now.
> > Granted I do use ia5, but others may use other ones. How about you just
> > delete this on your tree at home if it's really in the way?
> Is there not a need for an updated WEB view to Misterhouse? I thought
> that is one of the things people were after. I was just putting ideas
> out to try get some interest flowing.
It's not about the need, it's about the people writing the code.
Deleting old iaXX directories does not magically write new frontend code
in other directories :)
> Well Insteon isn't the only thing I'm interested in. I have code for an
> Ardunio board coming along to replace a Weeder board, which is where I
> started looking into adding a driver.
> So since you have done this could you provide some details about how you
> did your driver and what files you had to touch. It's way easier than
> reading 300,000 lines of perl and trying to figure it out solo, not to
> mention all the rest of the files.
Read lib/X10_CMxx.pm or really most files in lib that are drivers.
> > If you can finish up the insteon code to a point that we agree that it's
> > worth switching the default to that, you'll then have earned enough trust
> > points to make other suggestions and big changes to the code base :)
> Well not to be blunt but if I finish up the Insteon code it will be for
> myself, because I want to use it. And "we" can take it or leave it, it
> won't matter to me. Who is "we" anyway?
Sorry if the wording was imperfect but what I'm trying to say that if
you or anyone else finishes the insteon code, it will still take a
general agreement from people here that the new code is finished enough
to make it the new default that everyone should switch to.
There is nothing worse than switching people to non finished code.
Arguably, the current insteon code is very close to being finished,
several of us use it successfully enough (it's not perfect, but better
than the old code for me).
> Don't get me wrong Marc I've seen you reply to messages on this list for
> a long time, years, and I think you are doing an awesome job because you
> do respond. I think you deserve much credit for that. I do however get
> a slight impression that there is some resistance to any change in MH.
> I'd enjoy working with you and others to move MH forward.
I'm warry of sweeping changes from people with good intentions before
seeing that they make sense :)
No offense to you, I've just seen this too often in too many projects
and the damage it has caused.
Anyone can make any suggestion here, but ultimately to earn the right to
make big changes to the code, which is what you're suggesting, you or
someone else has to have earned some community trust and show and he/she
is able to write a bunch of code and understand the major bit of the
Just to be clear, I barely consider myself qualified for this, and
understand that I get scared when I read "I'm not too sure how all this
works yet, so let's delete a lot of code to make it easier" :)
Note, you can totally do that at home until you become more confident.
If you make drastic changes, you are actually encouraged to do a full
fork of mh and get people to try it out. When people generally agree
your tree is better than the current one, the old one gets deprecated
and yours becomes the new official one.
If that's not quite where you get, you can still make big changes, and
find that a few are better than the original tree and work to merge them
This is how open source works, and sometimes you will have a better idea
but may not be able to convince people that it is, however it is by
doing it on your side and publishing it, that you convince people that
you were right.
The linux kernel people always say "show me the code" :)
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
.... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/