Re: [mh] I2CS code testing

[Michael S. <mi...@st...>]

On Jan 14, 2013 1:46PM, Kevin wrote,
> 
> >> Can you explain how it is useful for i1 devices assuming that "sync links" and "sync
> >> all links" will install the same link?  I'm considering removing it but would never
do
> >> that if you or others say it is useful.
> >
> > We're talking about 3 things:
> >
> > 1) link to interface -> this is useful for I1/I2 devices to save walking between the
> > PLM and the device to link them so that the PLM gets status updates when said devices
are
> > actuated.
> >
> > 2) sync linkss for a single device is somewhat useful when you have bus problems and
> > sync all links is going to do a lot more than you want (talking to other devices that
are not
> > responding well).
> > It tells MH to only bother syncing links for that one device
> >
> > 3) sync all links does what you know.
> >
> > You could argue that #2 is not super useful, but I've used it in the past.
> > That said if it were to stop working, it would not be the end of the world since if
> > #3 is reliable, than #2 isn't that useful anymore (assuming your home network is
> > reliable, which mine isn't, but will be when I switch everything to dual band)
> 
> I think Michael is saying that sync_links (for a single device) and
> link_to_interface perform the same basic functions.  Yes sync_links
> performs more, but Michael's question is when do you only want to link
> a device to MH but not sync all of the other links defined in MH?

Exactly.  One thing to note is that "sync links" only syncs the links on a single device.
 
> I tend to think Michael is right.  But I suppose in a rough
> environment (few devices, lots of noise, ...) someone may have a
> difficult time communicating with the device.  In that instance trying
> to push a bunch of links in addition to the PLM link may be annoying.
> I tend to think this instance is rare or non-existant.  If you have a
> bunch of links to push, you should have sufficient devices to overcome
> the communication problems.

If you don't want to sync all the links on "that" device then you could delete the links
for that device from the .mht file and just leave the object definition.  Then "sync
links" would be identical to "link to interface".  If your environment is too unreliable
to sync the device links in your .mht file for a single device then I doubt that you can
effectively use Insteon.  If the underlying issue is bugs in the sync process then lets
debug those and I'll fix them.
 
> I dunno, for conciseness it may be easier to remove.  Removing it may
> cut down on the confusion encountered by new users as well since
> adding a new device would simply be sync_links.

That's exactly why I want to remove it.  It confused me and I don't think it is a useful
command.