Re: ISSUE: what does getStyleProperty return?

Hearing no objections, I implemented the proposed clarification in the 
for-publication draft, [1] and [2].

[1] 
http://jigedit.w3.org/lhenders/WWW/Graphics/WebCGM/drafts/current-editor-21/WebCGM21-DOM.html#getStyleProperty
[2] 
http://jigedit.w3.org/lhenders/WWW/Graphics/WebCGM/drafts/current-editor-21/WebCGM21-DOM.html#getStyleProperty-APS

-Lofton.

At 04:11 PM 1/20/2009 -0700, Lofton Henderson wrote:
>It sounds like everyone agrees, and current implementations have done the 
>same thing.
>
>If there are no objections, I propose to treat this as a non-substantive 
>editorial clarification.  (Maybe roll it into the pending publication.)
>
>For the getStyleProperty entries in both the Picture [1] and AppStructure 
>interfaces:
>
>In the description of getStyleProperty "Return value" change:
>"the Style Property value as a string, or the empty string if that 
>attribute does not have an explicitly set value (see the inheritance model 
>for further related discusion)."
>to:
>"the Style Property value as a string, or the empty string if that Style 
>Property has not been explicitly set by a setStyleValue call (or 
>equivalent XCF).  For Style Properties that may be set in different modes 
>-- e.g., NVDC or percentage -- the return value shall be in the mode in 
>which it was set.  For example, if stroke-weight were most recently set to 
>a percentage value by a setStyleValue call, then getStyleValue shall 
>return a percentage value.  (See the inheritance model for further related 
>discussion)."
>
>Any objections or suggestions?
>
>-Lofton.
>
>[1] 
>http://www.w3.org/Graphics/WebCGM/drafts/current-editor-21/WebCGM21-DOM.html#getStyleProperty
>
>At 09:58 AM 1/20/2009 -0800, David Cruikshank wrote:
>>I think WebCGM 2.1 already states that if a SP is not explicitly set, it 
>>returns a null string, so that shouldn't be a problem.  I interpret 
>>"explicitly set" to mean explicitly set by a call to SetStyleProperty 
>>(not set by a CGM attribute or control element)
>>
>>Upon reflecting on the issues involved, I guess I would propose that 
>>GetStyleProperty should return the value that was passed with the 
>>SetStyleProperty.
>>
>>Dave
>>
>>On Mon, Jan 19, 2009 at 11:51 PM, Weidenbrueck, Dieter 
>><<mailto:dweidenbrueck@ptc.com>dweidenbrueck@ptc.com> wrote:
>>>
>>>All,
>>>
>>>
>>>
>>>apparently a small but tricky question.
>>>
>>>
>>>
>>>getStyleProperty could return a variety of things:
>>>
>>>
>>>
>>>1.      The value used by setStyleProperty
>>>This would ensure that a script writer could retrieve values set via the 
>>>DOM, and change them. There is an analogy to setTransform/getTransform 
>>>here IMO.
>>>Open: what should the call return if there was no prior setStyleProperty 
>>>call, should it be null or an error?
>>>
>>>2.      The computed resulting value on the object
>>>This is also an interesting value, because it would tell the script 
>>>writer about the original state of an object (in case there was no prior 
>>>setStyleProperty call) and the resulting value.
>>>Issue: if you get resulting values back there is no way to detect 
>>>whether this value was in the file, or whether it had been manipulated 
>>>by script.
>>>
>>>
>>>
>>>I personally had always expected 1) because of the analogy to 
>>>transforms. Whatever you set you can get back. So the term "style 
>>>property" to me refers to the value set by the call, and not to detailed 
>>>CGM elements. In fact, it was always my intent to eliminate references 
>>>to specific CGM elements as much as possible, because no script writer 
>>>has a clue how CGM elements work. Example: there is no straight 
>>>equivalent to stroke-width in a CGM file, it affects a variety of things.
>>>
>>>
>>>
>>>A small issue is the return value in cases where there is no common 
>>>value. This may always happen, no matter what getStyleProperty will 
>>>return. Typical for these cases is to return an error or a special value 
>>>indicating that there is no common value.
>>>
>>>
>>>
>>>Regards,
>>>
>>>Dieter
>>>
>>>
>>>
>>>
>>>
>>>From: 
>>><mailto:public-webcgm-wg-request@w3.org>public-webcgm-wg-request@w3.org 
>>>[mailto:public-webcgm-wg-request@w3.org] On Behalf Of Lofton Henderson
>>>Sent: Dienstag, 20. Januar 2009 01:17
>>>To: David Cruikshank
>>>Cc: WebCGM WG
>>>Subject: Re: ISSUE: what does getStyleProperty return?
>>>
>>>
>>>
>>>At 02:49 PM 1/19/2009 -0800, David Cruikshank wrote:
>>>
>>>On the other hand there isn't any analogous value for raster 
>>>intensity...so it has to return the intensity value that was set.
>>>
>>>
>>>Right.  But I don't see that as a problem, since there is only one mode 
>>>for setting raster intensity.
>>>
>>>Responding to your initial idea, I think there is some logic to 
>>>returning the value in the mode that corresponds to how the CGM:1999 
>>>attributes are expressed.  But wait!  It is not as simple as it looks:
>>>
>>>stroke-weight SP:  NVDC or % [as a multiplier on current line/edge widths]
>>>LINE/EDGE WIDTH cgm attribute:  VDC, or MM, or ScaleFactor (times 
>>>impl-dependent "nominal"), or % (of VDC extent).
>>>
>>>The APS content in the metafile could have the line/edge width in one of 
>>>4 Specification Modes, and it would seem that you would have to select 
>>>VDC/NVDC arbitrarily as the "canonical" mode.
>>>
>>>But the stroke-weight % option might present a wrinkle that determines 
>>>the answer of this issue.  Consider this example.
>>>
>>>BegAPS;
>>>line width 5mm;
>>>polyline;
>>>line width 10mm;
>>>polyline;
>>>EndAPS
>>>
>>>If a DOM call sets stroke-weight to 10mm, then both polylines are 
>>>10mm.  But if it sets stroke-weight to 200%, then the first polyline is 
>>>10mm and the second is 20mm.
>>>
>>>In the latter case, getSP('stroke-weight') could NOT return an NVDC 
>>>value that was meaningful, right?  So I think the answer to the issue 
>>>needs to be:  return it in the mode in which it was set (or "empty 
>>>string" if the SP was never set).
>>>
>>>Does anyone see a way around this conclusion, that does not involve a 
>>>bunch of weird rules and tedious calculations?
>>>
>>>Regards,
>>>-Lofton.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>On Mon, Jan 19, 2009 at 12:38 PM, David Cruikshank 
>>><<mailto:dvdcruikshank@gmail.com>dvdcruikshank@gmail.com> wrote:
>>>
>>>Just shooting from the hip here, but I would expect the return value to 
>>>correspond to the common value of the attribute in CGM.  We introduced 
>>>these "intensity" or percentage values WebCGM, but for simplicity I 
>>>think the return should probably be the value that makes sense when you 
>>>consider the same attribute value in CGM.
>>>
>>>Just my thoughts.
>>>
>>>Dave
>>>
>>>
>>>On Mon, Jan 19, 2009 at 9:10 AM, Lofton Henderson 
>>><<mailto:lofton@rockynet.com>lofton@rockynet.com> wrote:
>>>
>>>ISSUE:  does getStyleProperty have to return the SP in the same form in 
>>>which it was set?
>>>
>>>Ref: 
>>><http://lists.w3.org/Archives/Public/public-webcgm-wg/2009Jan/0005.html>http://lists.w3.org/Archives/Public/public-webcgm-wg/2009Jan/0005.html
>>>
>>>Discussion:  Benoit asks the simple question, which we apparently have 
>>>never addressed...
>>>
>>>At 10:59 AM 1/7/2009 -0500, Bezaire, Benoit wrote:
>>>
>>>[....] I wonder if the implementation has to return the exact same 
>>>string that was used to set the attribute.
>>>
>>>Example for stroke-weight: setStyleProperty("stroke-weight", "200%"), 
>>>getStyleProperty("stroke-weight") returns 0.5 (assuming original was 0.25).
>>>
>>>
>>>Question 1:  do any of our past minutes or email address this?
>>>
>>>Question 2:  has anyone implemented yet?  What have you implemented?
>>>
>>>RECOMMENDATION:  (none yet.)
>>>
>>>### end ###