summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorBruce Smith <bruce@nanorex.com>2009-02-25 16:35:40 +0000
committerBruce Smith <bruce@nanorex.com>2009-02-25 16:35:40 +0000
commite11258994f63313d26aa2f324e377b36d2be961d (patch)
tree83c47fcb35c3add4c3831c2a9f09446aa13918df
parentb5942f91e26d9e9f33a31d375e0fe951905ffac5 (diff)
downloadnanoengineer-e11258994f63313d26aa2f324e377b36d2be961d.tar.gz
nanoengineer-e11258994f63313d26aa2f324e377b36d2be961d.zip
remove files there is no time to finish for this release
-rw-r--r--cad/src/graphics/model_drawing/changesetscratch.py73
-rw-r--r--cad/src/scratch/TransformNode.py (renamed from cad/src/model/TransformNode.py)15
-rw-r--r--cad/src/scratch/TransformState.py (renamed from cad/src/model/TransformState.py)3
3 files changed, 16 insertions, 75 deletions
diff --git a/cad/src/graphics/model_drawing/changesetscratch.py b/cad/src/graphics/model_drawing/changesetscratch.py
deleted file mode 100644
index bb27189b4..000000000
--- a/cad/src/graphics/model_drawing/changesetscratch.py
+++ /dev/null
@@ -1,73 +0,0 @@
- def _C_distorted_things(self):
- """
- Make sure the side effects associated with self.distorted_things are up to date,
- so that can be iterated over, and anything subscribing to its changes has them.
- """
- make sure input changesets are processed by asking for them in some sense
- (this also subs to their invals for next time)
-
- for t in self._xxx: # possibly distorted things, got added to some changeset by inval of motions they care about
- t.actually_distorted(): # _since_when?
- output_set[t] = t
-
- not sure if we need to subscribe to something more here?
-
- return output_set # not accurate, we return something that stands for an incremental set....
-
-
-
-
-
-
- """
- Turn the possibly distorted bridges into the actually distorted ones.
- Do this by noticing which bridges touch transformnodes whose actual transform data has changed
- (counting only the data in the dynamic transforms on them, I think, not the static ones),
- and for each one, check whether all its transformnodes have the same dynamic transform or not.
- Put the distorted bridges into our output set.
- """
- def xxx(self): # _C_ or something more special? or use a decorator?
- for bridge in input_set: ## get this from where? argument?
- d = bridge.is_dynamically_distorted() # "is"??? seems correct but also seems wrong in principle...
- # also this ought to be constant during the drag... so can't we just notice whether the symbolic transform changed??
- # not quite, since when a rigid motion occurs, the boundary sets dont change but they do get distorted.
- # but can't we track that by having an object just for that boundary,
- # which will be a single object that stands for all the bridges that cross that particular combo of drag-motion-kinds ==
- # symbolic drag transforms? that object tracks the actual data, changes in which distort its members
- # (for the bridging members, it's trivial -- any input change distorts them)
- # (for the non-bridging members, an input change to one dynamic transform symbol only moves them all rigidly --
- # it has no change-consequence at all except to invalidate their abs atom positions, important when next drawing
- # their external bonds)
-
- # ok, we're noticing symbolic transforms instead, on an actual bridge (eg EBSet).
-
- if d:
- put bridge in output_set
- return something
-
-
-
-
-
-### REVIEW / DISCUSS: can we use separate "banks" so that we don't care about sharing TCs across banks?
-# That is, the limit is only how many we use within one bank?
-# Or is this handled for us (by existing banking code for just that purpose)...
-# btw it means we have to keep track of which bank a CSDL is in, and it causes
-# trouble if we move it across bank bounaries -- TC merging might be required at that time! maybe not detected until we draw
-# (esp if we try to not worry if all affected nodes are hidden, etc).
-# anyway we have to tell the allocator about the banks... or give it a callback to tell us about overloads, now or later.
-# or do 2pass draw -- 1 draw-prep (report problems we need to solve, like too many TCs, with details of which ones),
-# 1b fix them, 2 actually draw.
-
-
-# If Russ can give me more TCs by doing the merging with identity himself (ie retransforming the vertices)...
-# but don't I have to find out and respond somehow? not sure...
-
-# In that case I'd end up fragmenting my usage of them, indefinitely... how would he know which ones to implement in the harder way?
-# Decide each frame, implem the least used ones slower??
-
-# but he also has to pass the tc ids in an attr array... not great if they keep changing.
-
-
-
-
diff --git a/cad/src/model/TransformNode.py b/cad/src/scratch/TransformNode.py
index 3c48e6571..17c83392a 100644
--- a/cad/src/model/TransformNode.py
+++ b/cad/src/scratch/TransformNode.py
@@ -2,7 +2,20 @@
"""
TransformNode.py -- mutable transform classes, shared by TransformNodes
-NOT YET USED as of 090204
+NOT YET USED as of 090204... declared obsolete as of bruce 090225:
+- much of the specific code is obsolete, since we abandoned the idea of a
+ number-capped TransformControl object (no need to merge TCs when too many
+ are used)
+- the concept of a separated dynamic and static transform, and letting
+ them be pointers to separate objects, potentially shared, is a good one,
+ which will be used in some form if we ever optimize rigid drag of
+ multiple chunks bridges by external bonds (as I hope we will),
+ but there is no longer time to implement it in this form (new superclass
+ for Chunk), before the next release, since it impacts too much else about
+ a Chunk. If we implement that soon, it will be in some more klugy form
+ which modifies Chunk to a lesser degree.
+
+Therefore, this file (and TransformState which it uses) are now scratch files.
@author: Bruce
@version: $Id$
diff --git a/cad/src/model/TransformState.py b/cad/src/scratch/TransformState.py
index 97c50bb47..7294e426b 100644
--- a/cad/src/model/TransformState.py
+++ b/cad/src/scratch/TransformState.py
@@ -2,7 +2,8 @@
"""
TransformState.py -- mutable transform classes, shared by TransformNodes
-NOT YET USED as of 090204
+NOT YET USED as of 090204... declared a scratch file, bruce 090225;
+see comments in TransformNode docstring for why.
@author: Bruce
@version: $Id$