[kepler-dev] diva fork

Chad Berkley berkley at nceas.ucsb.edu
Tue Aug 10 14:00:59 PDT 2004


Thanks for the tips.  See my comments below.

Stephen Andrew Neuendorffer wrote:
> Looks like a good start, but this seems to be really heavyweight.  I 
> think we should
> strive for something more transparent.  For this to work well, updating 
> the BRM
> should be something that a programmer does not need to do anything 
> 'extra' for.
> 
> setTransform should not have a method with a flag... it should always 
> update the BRM.
> I'm guessing you have this to avoid the infinite loop?  I think we'll 
> run into problems
> coordinating the panner and the scrollbars this way?

I removed the flag.  It was just an artifact left over from 
experimentation.  You shouldn't have to do anything "extra" for the bars 
now.  I'm going to add this as a configuration option (it only works 
with the kepler BasicGraphFrame class right now).

> 
> _horizRangeModel and _vertRangeModel are not words....  In fact, they're 
> probably not even necessary,
> just do _canvas.getHorizonalRangeModel().

done

> 
> setRangeModel is not a "setter" method..  change it to _updateRangeModel()?
done.  I haven't seen any other private methods in diva that start with 
   and underscore.  Is that the way you want it?  I thought it was only 
private variables.

> 
> FigureLayer needs to call _updateRangeModel when a figure is moved out 
> of range.
I'll take a look at this.

> 
> _updateRangeModel should only do something when the canvasPane is the 
> toplevel pane i.e.: _canvas.getCanvasPane() == this
done

> 
> _getViewSize() and getVisibleSize() might be useful as public methods of 
> _canvas, instead of in this class?
done

chad
> 
> Steve.
> 
> At 11:53 AM 8/10/2004, you wrote:
> 
>> Hi Stephen,
>>
>> I actually just checked them in.  I saw Christopher's vacation note so 
>> I figured I'd just do the checkin and back out anything you guys don't 
>> like.  See diva/canvas/JCanvas.java and diva/canvas/CanvasPane.java
>>
>> chad
>>
>> Stephen Andrew Neuendorffer wrote:
>>
>>> Could you send me the changes?  I'm curious what you needed...
>>> At 11:29 AM 8/10/2004, you wrote:
>>>
>>>> Hi Christopher,
>>>>
>>>> I've integrated our diva changes into the ptII source tree.
>>>> Everything still seems to work fine.  We only changed two files 
>>>> (JCanvas.java and CanvasPane.java) and the changes are actually 
>>>> pretty minimal.  I also wrote a build file for diva that I'd like to 
>>>> commit.
>>>> I just wanted to check with you first before I did so.  Let me know 
>>>> if these commits are ok, then I'll be able to get rid of the kepler 
>>>> diva fork.
>>>>
>>>> thanks,
>>>> chad
>>>>
>>>> Christopher Brooks wrote:
>>>>
>>>>> Ok, sounds good.
>>>>> I'll jump through many hoops to avoid code duplication, it turns out
>>>>> that we've wasted lots of time fixing bugs in duplicated code.  It
>>>>> drives Edward and I nutz when this happens :-)
>>>>> If you want, you could create a branch in the current ptII/diva tree
>>>>> and work from there.  I have no problem with that.
>>>>> Diva is an interesting case for us development-wise, since it is a
>>>>> package we use, but we are not the primary developer, though Steve and
>>>>> Edward have made lots of contributions.  We were not very happy to do
>>>>> the split, but diva was not our package, and it seemed for the best.
>>>>> Unfortunately, it seems like not much is happening with the Diva
>>>>> project, there has been only one check in to the master diva tree on
>>>>> embedded since 1/13/04.  There was a bunch of activity October
>>>>> through January though.
>>>>> -Christopher
>>>>> --------
>>>>>     Hi Christopher,
>>>>>
>>>>>     Our plan is definitely to merge these changes with your tree.  
>>>>> I just     need to get the code cleaned up a bit first.  We want to 
>>>>> make the     scroll bar changes a  configurable option so different 
>>>>> configs can     choose whether to use them or not.  I'll hopefully 
>>>>> get this cleaned up     by the end of the week then I'll talk to 
>>>>> you about merging it in with     your tree.
>>>>>
>>>>>     chad
>>>>>
>>>>>     On Aug 9, 2004, at 11:39 AM, Christopher Brooks wrote:
>>>>>
>>>>>     > It looks like you are in the process of forking a copy of diva?
>>>>>     >
>>>>>     > Is there anyway we can work from the diva tree that is in the 
>>>>> ptII
>>>>>     > repository?
>>>>>     >
>>>>>     > We already forked once from the primary diva tree.  The ptII 
>>>>> diva tree
>>>>>     > is a subset of
>>>>>     > http://embedded.eecs.berkeley.edu/diva/
>>>>>     >
>>>>>     > We split the ptII version of diva off because of development 
>>>>> tensions
>>>>>     > between ptII and diva.  Basically, the original author of 
>>>>> diva wanted
>>>>>     > to do development of diva, and ptII needed a more stable 
>>>>> platform.
>>>>>     >
>>>>>     > When I did the split, I did lots of code cleanup and pared 
>>>>> away parts
>>>>>     > of diva that ptII did not need.
>>>>>     >
>>>>>     > It would be great if we could stick with one tree that would 
>>>>> support
>>>>>     > the scrollbar work you are doing as well as the current 
>>>>> Ptolemy II
>>>>>     > work.
>>>>>     >
>>>>>     > -Christopher
>>>>>     >
>>>>>     > _______________________________________________
>>>>>     > kepler-dev mailing list
>>>>>     > kepler-dev at ecoinformatics.org
>>>>>     > http://www.ecoinformatics.org/mailman/listinfo/kepler-dev
>>>>>
>>>>>     _______________________________________________
>>>>>     kepler-dev mailing list
>>>>>     kepler-dev at ecoinformatics.org
>>>>>     http://www.ecoinformatics.org/mailman/listinfo/kepler-dev
>>>>> --------
>>>>
>>>>
>>>> _______________________________________________
>>>> kepler-dev mailing list
>>>> kepler-dev at ecoinformatics.org
>>>> http://www.ecoinformatics.org/mailman/listinfo/kepler-dev
> 
> 



More information about the Kepler-dev mailing list