[Envisage-dev] chaco errorbar change

hi folks,

i'm planning on changing how the errorbar renderer looks for data in its 
plotvalues.  this change is to support some new stuff in the rheology project. 
  this change will also not be backwards compatible, so PLEASE let me know if 
you are using errorbars or know of a place where they are used so i can 
discuss these changes with you before i make them.

if i don't hear back from anyone by mid-afternoon or so, i'll go ahead and 
start working on these changes.

thanks,
peter

Re: [Envisage-dev] Refence cycles with TraitLists

>>>>> "Dave" == David C Morrill <dmorrill@...> writes:

    Dave> Prabhu Ramachandran wrote:
    >> Maybe this should be done by default in TraitListObject?  Some
    >> of my TVTK tests started failing as soon as I added a List
    >> trait to a class, which is why I noticed this!
    >>
    >> I guess the same problem will exist for the TraitDictObject
    >> also.

    Dave> I had to back out the changes I made earlier to use
    Dave> weakref.proxy objects since I found that it caused subtle
    Dave> semantic changes in the app we are developing. Until I have
    Dave> time to figure out what those changes are and eliminate
    Dave> them, I'm going to leave the proxies out.

No problem.  I just used Tuple instead...  :)

cheers,
prabhu

[Envisage-dev] Mathematica-style notebooks

Recently, two students Toni Alatalo and Tzanko Matev were selected to 
participate in the Google Summer of Code to work on a Mathematica-style 
notebook for ipython. I am helping Fernando as a co-mentor on this project.

The goal is to develop a GUI component that represents a rich-text 
document with embedded interactive "cells" that serve as the input 
prompt and other cells representing output and graphics. For example, a 
milestone that we want to achieve by the end of August is to have the 
Scipy tutorial in this notebook format such that the user can open the 
notebook, edit one of the example prompts to change, for example, some 
input variables and thus change one of the plots.

Although Envisage is not being targetted at this early stage (these guys 
have enough to deal with ), we do hope that the component will be 
clean enough that it could be plugged in without too much pain.

I have already written some code to read and write an XML-based notebook 
format. I can also export those notebooks to reST and HTML. Toni 
primarily wants to work on document generation, so he will be tweaking 
the format and writing more export code (I did a fair bit of his job for 
him. Oops). Tzanko will be working more on the GUI and ipython 
integration. His first milestone will be to take a notebook file, render 
it unstyled in a wx.StyledTextCtrl, allow one to edit, delete, rearrange 
and otherwise manipulate the input and output cells and the surrounding 
text, and finally save the results to disk.

If you have any user stories you would like to add, please do so on the 
Wiki. We would appreciate your input on any aspect of the project.

http://www.scipy.org/wikis/featurerequests/NoteBook

Re: [Envisage-dev] Mathematica-style notebooks

Robert Kern wrote:
> Although Envisage is not being targetted at this early stage (these guys 
> have enough to deal with ), we do hope that the component will be 
> clean enough that it could be plugged in without too much pain.

I'd be very interested in plugging it into PanGalactic's
wxPython-based gui, PanGalaxian (for info about
PanGalactic, see <http://ned.gsfc.nasa.gov>) ... I'll
follow this work with great interest.

Hope you are using wxPython 2.5 or (preferably) 2.6, but
I guess that you might be using 2.4 to be consistent with
Envisage, as the last I heard they had not migrated.  Even
then, it won't be too difficult for me to modify the code
to conform to the 2.5/2.6 api's if necessary.

Cheers,
Steve

Re: [Envisage-dev] Mathematica-style notebooks

Stephen Waterbury wrote:
> Robert Kern wrote:
> 
>> Although Envisage is not being targetted at this early stage (these 
>> guys have enough to deal with ), we do hope that the component will 
>> be clean enough that it could be plugged in without too much pain.
> 
> I'd be very interested in plugging it into PanGalactic's
> wxPython-based gui, PanGalaxian (for info about
> PanGalactic, see <http://ned.gsfc.nasa.gov>) ... I'll
> follow this work with great interest.

Man, you guys take buzzword-compliance to a whole new level, don't you? 
The last time I saw so many acronyms was when I worked on an Army base.  

> Hope you are using wxPython 2.5 or (preferably) 2.6, but
> I guess that you might be using 2.4 to be consistent with
> Envisage, as the last I heard they had not migrated.  Even
> then, it won't be too difficult for me to modify the code
> to conform to the 2.5/2.6 api's if necessary.

We will be targetting 2.6. Envisage *will* catch up eventually, so 
writing code for 2.4 doesn't really make sense unless if you're directly 
developing for Envisage. Which we're not, yet.

--

-- 
Robert Kern
rkern@...

"In the fields of hell where the grass grows high

Re: [Envisage-dev] Mathematica-style notebooks

Robert Kern wrote:
> Stephen Waterbury wrote:
> 
>> Robert Kern wrote:
>>
>>> Although Envisage is not being targetted at this early stage (these 
>>> guys have enough to deal with ), we do hope that the component 
>>> will be clean enough that it could be plugged in without too much pain.
>>
>> I'd be very interested in plugging it into PanGalactic's
>> wxPython-based gui, PanGalaxian (for info about
>> PanGalactic, see <http://ned.gsfc.nasa.gov>) ... I'll
>> follow this work with great interest.
> 
> Man, you guys take buzzword-compliance to a whole new level, don't you? 
> The last time I saw so many acronyms was when I worked on an Army base.  
> 

Yeah, sorry about that -- the slides are a bit over the top! ;)
The sad part is I'm actually serious about compliance
with most of them ... in the engineering data integration
world, standards are our friend (er, sort of).  OTOH, "the
great thing about standards is there are so many to choose
from" ... :)

> We will be targetting 2.6. ...

Excellent.

> ... Envisage *will* catch up eventually, so 

Re: [Envisage-dev] Mathematica-style notebooks

Hi folks,

We actually have got Envisage up and running on wx 2.6, although we 
haven't officially made the switch yet 8^(  I would hope that the 
*real* move to wx 2.6 and Python 2.4 will occur sometime this month...

Martin

> Stephen Waterbury wrote:
>> Robert Kern wrote:
>>> Although Envisage is not being targetted at this early stage (these 
>>> guys have enough to deal with ), we do hope that the component 
>>> will be clean enough that it could be plugged in without too much 
>>> pain.
>> I'd be very interested in plugging it into PanGalactic's
>> wxPython-based gui, PanGalaxian (for info about
>> PanGalactic, see <http://ned.gsfc.nasa.gov>) ... I'll
>> follow this work with great interest.
>
> Man, you guys take buzzword-compliance to a whole new level, don't 
> you? The last time I saw so many acronyms was when I worked on an Army 
> base.  
>
>> Hope you are using wxPython 2.5 or (preferably) 2.6, but
>> I guess that you might be using 2.4 to be consistent with
>> Envisage, as the last I heard they had not migrated.  Even
>> then, it won't be too difficult for me to modify the code
>> to conform to the 2.5/2.6 api's if necessary.
>
> We will be targetting 2.6. Envisage *will* catch up eventually, so 

[Envisage-dev] Bug in tree editor

Hi Dave,

The recent checkins break the tree editor.  For example just run the
example in traits/ui/tests/tree_editor_test.py

Attached is a patch to fix the problem.

cheers,
prabhu

Index: wx/tree_editor.py
===================================================================
--- wx/tree_editor.py	(revision 4625)
+++ wx/tree_editor.py	(working copy)
 <at>  <at>  -42,10 +42,10  <at>  <at> 
 from enthought.traits.ui         import View, Item, TreeNode, MultiTreeNode

 from enthought.traits.ui.undo    import ListUndoItem

 from enthought.traits.ui.menu    import Menu, Action, Separator

+from enthought.traits.ui.helper  import Orientation

 

 from editor                      import Editor

 from editor_factory              import EditorFactory

-from helper                      import Orientation

 

 #-------------------------------------------------------------------------------

 #  Global data:

Index: helper.py
===================================================================

[Envisage-dev] Range trait cannot handle longs

Hi Dave,

I ran into a problem using a long with Traits.  The following does not
work:

from enthought.traits import *
class A(HasTraits):
    l = Range(0L, 10000000L)

I get this error:

 TraitError: TraitRange can only be use for int or float values, but a
 value of type <type 'long'> was specified.

Could this be fixed?  Thanks.

cheers,
prabhu

[Envisage-dev] minor change to Enable graphics context

Hi folks,

I just removed the set_alpha() function from the Enable graphics context.  (It 
was overriding the behavior in Kiva's graphics context.)  The proper function 
to call in its stead is EnableGraphicsContext.alpha().

I think all the places that really care about GCs and alphas are doing the 
right thing, but this is a heads-up to folks that I've made a change to the 
lower-level graphics stuff, and if you see weirdness, specifically, things 
being transparent when they shouldn't be, or things being opaque when they 
should be invisible, etc., please shoot me an email.  (Sorry for the 
overabundance of commas in the last sentence.)

Thanks,
Peter