Yeah it was the value historian comment that I was thinking of, Christopher. Thought it might point to some underlying QC bug. Point taken, sorry. DISCLAIMER: Did say it was re: something else entirely though ;)
2009/6/27 Christopher Wright <[email protected]> Backing off a bit... why is there conversion going on after the first >> conversion? (note attachment). Notice how in the example, if you keep a >> multiplexer on virtual it converts the boolean input works fine, and is >> converted to numeric on the output of the multiplexer. So even though there >> is conversion going on in the second "chain", it still works fine. Without >> seeing your exact scenario it's a bit hard to say what the problem is. >> > > Passing through virtual ports (via splitters, muxers, etc) isn't really > doing a type conversion (passing through mathematical expression, which only > uses number ports, or example, or using indexes via javascript, would > perform an actual type conversion). The only unusual boolean port quirk > that I can think of is how any non-zero number is true (and this is > completely correct in a programmer sense, just not necessarily intuitive to > non-programmers). I don't think I've said there are any problems with it > (contrary to initial post) that would be involved in this -- it's just their > internals with regard to some assumptions Value Historian makes that I was > referring to (100% our bug). > > with Virtual Ports, you get out exactly what you put in -- the port itself > is hands-off. Typed splitters will do conversions (it's impossible to get a > 2 out of a boolean port, as it'll convert the 2 into a True and call it a > day). > > On a side note, I strongly disagree that connections severing when >> changing type is a bug. I think that's extremely functional (... it can be >> irritating, but I think it's the "right" thing to have happen). >> > > > I see this breaking into 2 cases: When you change from non-convertible > types (to/from structure, image from number, to illustrate a few), I think > breaking the connections is "right" (because you can't possibly expect > anything meaningful to happen this way). Other conversions that are > convertible, however, (bool < -> index <-> number) have no negatives, and > shouldn't break (in my opinion) -- what happens will be a tiny bit > different, but nothing catastrophic. > > -- > [ christopher wright ] > [email protected] > http://kineme.net/ > >
_______________________________________________ Do not post admin requests to the list. They will be ignored. Quartzcomposer-dev mailing list ([email protected]) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/quartzcomposer-dev/archive%40mail-archive.com This email sent to [email protected]
