[fdo-internals] RE: FDO raster image data model does not provide the actual number of bits used by certain image data formats (Re: FdoRfc1 & FdoRfc2)

[Greg Boone]

One more point,

Properties 'bool force' & 'bool approx_ok' should have default values
assigned...

I propose...

    FDO_API virtual bool GetMinMax( FdoDouble &min, 
						FdoDouble &max, 
                                    bool force = false, 
						bool approx_ok = true);


-----Original Message-----
From: Greg Boone 
Sent: Tuesday, April 03, 2007 3:01 PM
To: FDO Internals Mail List
Subject: RE: FDO raster image data model does not provide the actual
number of bits used by certain image data formats (Re: FdoRfc1 &
FdoRfc2)

Hi Frank,

Ok. Let's try and proceed with just the min/max property. 

Lets put a hold on RFC1 until RFC2 is agreed on by the PSC.

I will rework RFC2 to have the new properties placed on FdoIRaster

Can we rename GetMinMax() -> GetMinMaxValues() ?

On the technical side, if min/max values are not available from the
underlying raster library, but BitsPerPixel is available, how will
min/max values be determined from BitsPerPixel? 

Greg

-----Original Message-----
From: Frank Warmerdam (External) 
Sent: Friday, March 23, 2007 1:00 PM
To: Sarat Venugopal
Cc: Greg Boone; Jeff Childers; Orest Halustchak; FDO Internals Mail List
Subject: Re: FDO raster image data model does not provide the actual
number of bits used by certain image data formats (Re: FdoRfc1 &
FdoRfc2)

Sarat Venugopal wrote:
>> OK, please go ahead and make a case for doing both.  Off hand, seems
>> like duplication and unnecessary complexity.
> 
> Formats such as Ikonos distinguish themselves by the actual number of
> bits used, so I think including this field in the data model helps API
> clients with various things without performing additional processing.
> Besides, this information could be used to arrive at the approximate
> minimum and maximum values which is superior to a method that only
> involves sampling the image.

Sarat,

But I don't see how this provides something that wouldn't be provided
by the min/max handling since for an Ikonos image you can just return
0/2047 for the min/max.

> I certainly don't see any complexity in adding this additional field.

There is extra complexity in every single property, or method added to
an interface.  It wouldn't take much value to justify this very small
increment of complexity, but it is extra complexity.

The multiple approaches to scaling will certainly complicate future
provider and application code.  If I write application code I will be
faced with deciding which information to use, and have to start second
guessing which providers implement which well.

Best regards,
-- 
---------------------------------------+--------------------------------
------
I set the clouds in motion - turn up   | Frank Warmerdam,
warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo,
http://osgeo.org