27 Jul
2017
27 Jul
'17
8:09 a.m.
Hi,
On Jul 27, 2017, at 1:21 AM, Alexandre Bergel
<alexandre.bergel(a)me.com> wrote:
I understand your request, but I have no easy to fix this as the semantics is not clear
to me. It seems that the source:connectFrom:… looks for any elements, and not only the one
defined in the current scope.
The issue with the creation of edges is not difficult: instead of traversing only the
children, we simply traverse deeply all children of the current scope. This worked in two
versions of Mondrian.
Nevertheless, the main issue I reported related to the application of layouts inside the
nested objects.
Using the semantics you describes, what the following
script should display?
-=-=-=-=-=-=-=-=-=-=
b := RTMondrian new.
b nodes: #(1 2) forEach: [ :i |
b nodes: #(1 2).
b layout verticalLine ].
b shape line
color: Color blue;
withShorterDistanceAttachPoint.
b edges connectFrom: 1 to: 2.
b
-=-=-=-=-=-=-=-=-=-=
Exactly what it does right now. You start looking in the scope you defined edges in, and
find the nodes.
I feel we are reaching the limit of what Mondrian is
aiming to solve. In the script above, I would expect the behavior we have currently. If
"b edges connectFrom: 1 to: 2.” would somehow inject some edges within the outer
nodes, then this feel really magic and rather counter intuitive.
It really is not magic at all. It is declarative :). For example, in the same realm, you
also do not get an error when one of the nodes does not exist because the intention is to
provide a declarative way of building graphs. But, the main issue at hand is how to apply
layouts, given that layouts are eager and they get applied when we encounter the layout
instruction, and not at the end. Any ideas here?
Cheers,
Doru
Alexandre
--
www.tudorgirba.com
www.feenk.com
"We are all great at making mistakes."
On Jul 22, 2017, at 3:03 AM, Tudor Girba
<tudor(a)tudorgirba.com> wrote:
Hi,
I think there is a misunderstanding.
I want to be able to have hierarchies inside each container node, but by defining the
edges once for the overall graph. So, the graph would look almost like in the picture
below, but without the shifting:
<Playground-mondrian-issue.png>
For example, I want to show classes grouped in packages, and show inheritance, too. Right
now, I have to do it like this:
packages := { 'Glamour-Core' asPackage . 'Glamour-Presentations'
asPackage }.
view := RTMondrian new.
view shape rectangle fillColor: Color yellow.
nodes := view nodes: packages forEach: [ :pack |
view nodes: pack definedClasses.
view edges source: pack definedClasses connectFrom: #superclass to: #yourself.
view layout tree ].
classesWithInheritanceBetweenPackages := (packages flatCollect: #definedClasses)
select: [ :each | each isClass and: [each package ~= each superclass package] ]
thenCollect: [ :each | each -> each superclass ].
view edges source: classesWithInheritanceBetweenPackages connectFrom: [ :x | x value ]
to: [ :x | x key ].
view
This means that I explicitly have to pick the edges that I want to draw in the different
contexts: once to draw only the local edges, and once to draw only the non-local edges. If
layouts would be applied lazily, I could simply do this:
packages := { 'Glamour-Core' asPackage . 'Glamour-Presentations'
asPackage }.
view := RTMondrian new.
view shape rectangle fillColor: Color yellow.
nodes := view nodes: packages forEach: [ :pack |
view nodes: pack definedClasses.
view layout tree ].
view edges source: (packages flatCollect: #definedClasses) connectFrom: [ :x | x
superclass ] to: [ :x | x ].
view
As this is not possible now, I was thinking of still constructing the graph in the
simplified form by defining edges once, but then we apply the layout later inside each
container node, which is still cumbersome.
Cheers,
Doru
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
_______________________________________________
Moose-dev mailing list
Moose-dev(a)list.inf.unibe.ch
https://www.list.inf.unibe.ch/listinfo/moose-dev On Jul 21, 2017, at 8:53 PM, Alexandre Bergel
<alexandre.bergel(a)me.com> wrote:
Using nodes that belongs to different parent nodes?
Well… I do not know how to do this.
Alexandre
--
_,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
Alexandre Bergel http://www.bergel.eu
^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
--
www.tudorgirba.com
www.feenk.com
"Value is always contextual."
_______________________________________________
Moose-dev mailing list
Moose-dev(a)list.inf.unibe.ch
https://www.list.inf.unibe.ch/listinfo/moose-dev On Jul 21, 2017, at 2:41 PM, Tudor Girba
<tudor(a)tudorgirba.com> wrote:
Hi,
I expect to see a “tree" inside each top node.
Doru
> On Jul 21, 2017, at 8:40 PM, Alexandre Bergel <alexandre.bergel(a)me.com> wrote:
>
> Hi Doru,
>
> Nesting is indeed a weak point of currently version of Roassal.
>
> Future version of Roassal will ease this.
>
> Try this script:
> -=-=-=-=-=-=-=-=-=-=-=-=
> view := RTMondrian new.
> view shape rectangle fillColor: Color yellow.
> nodes := view nodes: { { 1 . 2 } . { 11 . 12 } } forEach: [ :col |
> view nodes: col ].
> view edges source: { 1 . 11 } connectFrom: [ :x | x ] to: [ :x | x + 1 ].
> view edges source: { 1 . 2 } connectFrom: [ :x | x ] to: [ :x | x + 10 ].
> nodes do: [ :each |
> RTNest new
> layout: RTTreeLayout new;
> centerOn: each elements: each nestedElements ].
> view
> -=-=-=-=-=-=-=-=-=-=-=-=
>
> I am not completely sure to understand what the result should be. What do you
expect?
>
> Alexandre
> --
> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
> Alexandre Bergel http://www.bergel.eu
> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
>
>
>
>> On Jul 21, 2017, at 6:45 AM, Tudor Girba <tudor(a)tudorgirba.com> wrote:
>>
>> Hi,
>>
>> Actually, my script does not work because the nested nodes are shifted
internally.
>>
>> So, now I do not know any solution for my issue. Any ideas?
>>
>> The issue is visible when we give the root elements a color:
>>
>> view := RTMondrian new.
>> view shape rectangle fillColor: Color yellow.
>> nodes := view nodes: { { 1 . 2 } . { 11 . 12 } } forEach: [ :col |
>> view nodes: col ].
>> view edges source: { 1 . 11 } connectFrom: [ :x | x ] to: [ :x | x + 1 ].
>> view edges source: { 1 . 2 } connectFrom: [ :x | x ] to: [ :x | x + 10 ].
>> nodes do: [ :each |
>> RTNest new
>> layout: RTTreeLayout new;
>> on: each nest: each nestedElements ].
>> view
>>
>> <Playground-mondrian-issue.png>
>>
>> Cheers,
>> Doru
>>
>>
>>> On Jul 21, 2017, at 12:42 PM, Tudor Girba <tudor(a)tudorgirba.com>
wrote:
>>>
>>> Hi,
>>>
>>> I am trying to use Mondrian for a case in which I have nested elements and
crosscutting edges. The issue is that I only found a cumbersome way to specify the
layout.
>>>
>>> Here is an example:
>>>
>>> view := RTMondrian new.
>>> view shape rectangle fillColor: Color transparent.
>>> nodes := view nodes: { { 1 . 2 } . { 11 . 12 } } forEach: [ :col |
>>> view nodes: col ].
>>> view edges source: { 1 . 11 } connectFrom: [ :x | x ] to: [ :x | x + 1 ].
>>> view edges source: { 1 . 2 } connectFrom: [ :x | x ] to: [ :x | x + 10 ].
>>> nodes do: [ :each |
>>> RTNest new
>>> layout: RTTreeLayout new;
>>> on: each nest: each nestedElements ].
>>> view
>>>
>>> With bold I have showed the code necessary for applying the layout. In
previous versions of Roassal/Mondrian, layouts were applied lazily, after the graph was
constructed. This allowed us to write something like this:
>>>
>>> view := RTMondrian new.
>>> view shape rectangle fillColor: Color transparent.
>>> view nodes: { { 1 . 2 } . { 11 . 12 } } forEach: [ :col |
>>> view nodes: col.
>>> view layout tree ].
>>> view edges source: { 1 . 11 } connectFrom: [ :x | x ] to: [ :x | x + 1 ].
>>> view edges source: { 1 . 2 } connectFrom: [ :x | x ] to: [ :x | x + 10 ].
>>> view
>>>
>>> Did I miss something or is this the only way supported currently?
>>>
>>> Cheers,
>>> Doru
>>>
>>>
>>> --
>>> www.tudorgirba.com
>>> www.feenk.com
>>>
>>> "To utilize feedback, you first have to acquire it."
>>>
>>
>> --
>> www.tudorgirba.com
>> www.feenk.com
>>
>> "Yesterday is a fact.
>> Tomorrow is a possibility.
>> Today is a challenge."
>>
>>
>>
>>
>> _______________________________________________
>> Moose-dev mailing list
>> Moose-dev(a)list.inf.unibe.ch
>> https://www.list.inf.unibe.ch/listinfo/moose-dev
>
> _______________________________________________
> Moose-dev mailing list
> Moose-dev(a)list.inf.unibe.ch
> https://www.list.inf.unibe.ch/listinfo/moose-dev
--
www.tudorgirba.com
www.feenk.com
"Innovation comes in the least expected form.
That is, if it is expected, it already happened."
_______________________________________________
Moose-dev mailing list
Moose-dev(a)list.inf.unibe.ch
https://www.list.inf.unibe.ch/listinfo/moose-dev
_______________________________________________
Moose-dev mailing list
Moose-dev(a)list.inf.unibe.ch
https://www.list.inf.unibe.ch/listinfo/moose-dev