BackgroundColor strangeness

[Wilhelm Sanke]

On Thu, 20 Nov 2003 Alex Rice <alex at mindlube.com> wrote:

> > Question is, if there are so many problems with the Distribution
> > Builder, why don't you use the Metacard Standalone Builder?
>
> Because I'm not a Metacard user, and I there are workarounds for most
> of the aforementioned. And some are engine bugs not DBuilder bugs.


If you are not (yet) a Metacard user, it would be certainly instructive 
and helpful to have a look at the other half of the fraternal twins that 
are now both owned by RunRev. I think you - as a Revolution license 
holder - are even entitled to a free license key for Metacard under the 
new scheme.

Anyway, I could reproduce Graham Samuel's problem on my Windows computer:

> I have a stack with the backgroundColor set (via its Inspector) to a
> particular RGB color (0,50,25).I see this color in the inspector and
> in the stack itself. But if I write in the message box
>
>   put the backgroundColor of stack "myStack"
>
> I get a different result (0,128,64).
>
> It also appears that when I make a standalone with this stack in it,
> the second color is indeed used as a background, whether or not I
> choose to copy default colors.
>
> Am I going crazy?
>
> TIA for any explanation
>
> Graham



I followed the sequence of Graham's steps and got more or less the same 
results with Revolution 2.1.2. One thing differed as the message box 
returned the correct R,G,B color on Windows, but in the standalone the 
stack backcolor was different as with Graham's experience.

Again following the same procedure under Metacard, no such problem 
occured, the chosen backcolor of the stack also appeared in the standalone.

Third test: I transferred the Metacard Standalone Builder (along with 
the stack "windows version" as a substack) to the Revolution IDE and 
used it to build the standalone. Result: No difference between the 
original stack and the standalone concerning the backcolor.

So this is not an engine problem, but definitely connected to the 
Revolution Distribution Builder. I would be interested to learn what 
kind of similar problems could be engine bugs as you said.-

The Distribution Builder at present seems to be the weakest element of 
the Rev IDE. There are other problems, among them severe speed problems 
when building standalones with greater numbers of controls. The 
Distribution Builder can be 1000 (one thousand) times slower - this is 
not a bad joke, but a reproducible fact -, i.e. a build time of 1 second 
with the Metacard Standalone Builder corresponds to 16 minutes with the 
Rev Distribution Builder (tested for a stack of 300 K on a G4 Mac with 
10.3). Other stacks even needed a build time of longer than half an hour.

There was some discussion of these matters recently on the two other 
RunRev lists. No solutions to these speed problems were found and so far 
no feedback offered by the RunRev Team, who are most certainly very much 
engaged at the moment to find out what is wrong with the Distribution 
Builder - and probably will let us know soon in which way these 
deficiencies could be repaired.

In the meantime, it would be a good idea to add the Metacard Standalone 
Builder as an alternative tool  to the Revolution IDE.
In a broader sense  Metacard and Revolution could be seen as parts of 
the same Revolution project; the strengths of both twins should be 
explored and fully exploited. I imagine a "slim" or "power" version of 
Revolution - similar to the present Metacard IDE - that is indeed 
actively supported and developed as an alternative platform by the 
RunRev team (and , if possible, with the continuing active support by  
Scott Raney).

Regards,

Wilhelm Sanke