path: root/cad/doc/old_doc/CHANGENOTES.txt

 Copyright 2005-2007 Nanorex, Inc.  See LICENSE file for details. 
NanoEngineer-1 Change Log

$Id$


INSTRUCTIONS
------------

Developers should update this file after every commit that fixes a bug
or adds new functionality to NanoEngineer-1. Please include the following
information:

Name:
Date:
Files:
Summary:


CHANGE NOTES
------------

Name: Mark Sims
Date: 3/13/2005
Files: HistoryWidget
Summary: Changed startup height to 4 lines of text, requested by Damian.

Name: Mark Sims
Date: 3/13/2005
Files: ElementSelector
Summary: Changed button font from Times to default.

Name: Mark Sims
Date: 3/13/2005
Files: modifyMode
Summary: Fixed bug 13.  Removed "Separate" from context menu.

Name: Mark Sims
Date: 3/14/2005
Files: LightingTool
Summary: Fixed bug 432.

--------------------- Alpha 5 -------------------------------------

Name: Bruce Smith
Date: 3/16/2005
Files: lots
Summary: assembly/Part split -- clipboard items can be viewed or changed 
much like the main part.

----------------------------

Name: Bruce Smith
Date: 3/16/2005
Files: gadgets.py
Summary: fix the jig analog of bug 371

----------------------------

Name: Mark Sims
Date: 3/18/2005, 11:00AM

cookieMode: 
 - hide the modify toolbar while in cookie cutter tool

MainWindow:
 - added new "Select" menu (same options as Select toolbar)
 - added new Display menu items:
    - Reset Chunk Display
    - Show Invisible Atoms

MoleculeProp:
 - Changed reference to "Singlets:" to "Open Bonds:"
 - fixed bug 445, in which the 'Make invisible atoms visible' button reset the display mode for all atoms in the chunk.

movie:
 - changed GNUplot filename suffix from "-trace.plt" to "-plot.txt"

MWsemanitcs:
 - changed name of slot "dispResetMolColor" to "dispResetChunkColor"
 - added new slot "dispResetChunkDisplay"
 - added new slot "dispShowInvisAtoms"

part:
 - added new method "resetChunkDisplay"
 - added new method "showInvisibleAtoms"

PlotTool:
 - runGNUplot() verifies that the plotfile exists before attempting to plot it.

selectMode:
 - removed context menu items:
   - Connected
   - Doubly
   - Separate
 - added context menu items to "Right-Control Menu":
   - Reset Chunk Color
   - Reset Chunk Display
   - Show Invisible Atoms
   - Chunk Color...

WhatsThis:
 - Updated descriptive text for:
   - Select Chunks
   - Select Atoms
   - Move Chunks
   - Build Atoms Tool

----------------------------

Name: Mark Sims
Date: 3/18/2005, PM
File: Lots
Summary: Added a selection filter for Select Atoms.

----------------------------

Name: Bruce Smith
Date: 3/18/2005
Files: fileIO.py MWsemantics.py
Summary: exclude singlets from saved PDB files; warn user when this happens

----------------------------

Name: Mark Sims
Date: 3/19/2005
Files: MainWindow, depositMode, RotaryMotorDialog
Summary:

MainWindow: 
 - removed Element Selector from Modify menu/toolbar
 - Adding Phosphorus to Select Atoms filter combobox

depositMode:
 - Added Element Selector to Build Atoms dashboard

RotaryMotorDialog:
 - Changed torque units from "nN*nm" to "nN-nm"
 - Changed speed units from "gHz" to "GHz"

----------------------------

Name: Bruce Smith
Date: 3/22/2005
Files: fileIO.py
Summary: fix bug 433 (now ok to read empty pdb files)

----------------------------

Name: Mark Sims
Date: 3/24/2005
Files: LightingToolDialog, MainWindow
Summary: 
 - Fixed bug 473, added "Restore Defaults" button
 - Fixed bug 463, added "Sulfur" to SA Filter

----------------------------

Name: Mark Sims
Date: 3/25/2005
Files: MainWindow, MWsemanitcs, fusechunksMode, GLpane
Summary: Added support for new Fuse Chunks Tool

----------------------------

Name: Bruce Smith
Date: 4/07/2005 (and some earlier dates)
Files: runSim.py, part.py, movie.py, fileIO.py, chunk.py, chem.py, perhaps others
Summary: Minimize Selection (and some of bug 254 fix), and safer but perhaps slower movie playing

see email to cad for details.

----------------------------

Name: Bruce Smith
Date: 4/07/2005
Files: runSim.py, movie.py
Summary: show warnings and errors from simulator during Minimize or Simulate

----------------------------

Name: Bruce Smith
Date: 4/08/2005
Files: GLPane.py
Summary: fix bug in which glpane or its edges would flicker to black during main-window resize

----------------------------

Name: Mark Sims
Date: 4/10/2005
Files: GLPane, MainWindowUI, modes, modifyMode, platform, translateMode (new)
Summary: Implemented new "translateMode", which adds support for sliding and/or rotating selected chunks 
along the x, y or z axis.  New buttons added to the Move Chunks dashboard give UI access to these features.
Users can also hold the X, Y, or Z keys while in Move Chunks mode to invoke these features.

----------------------------

Name: Mark Sims
Date: 4/11/2005
Files: MainWindowUI, MWsemantics, part, whatsthis
Summary: Changed "Weld Chunks" to "Merge Chunks" (bug 515).

----------------------------

Name: Mark Sims
Date: 4/12/2005
Files: MainWindowUI, MWsemantics, modifyMode, translateMode
Summary: Added "Move Delta" capability to Move Chunks mode.

----------------------------

Name: Mark Sims
Date: 4/12/2005
Files: MainWindowUI, MWsemantics, modes, modifyMode, platform
Summary: Changed/added UI and slot hooks for "Minimize Selection" and "Minimize All".

----------------------------

Name: Bruce Smith
Date: 4/12/2005
Files: runSim.py, others
Summary: Support for Minimize Selection vs Minimize All, and improved their history messages.

(This commit doesn't yet fix progress bar caption or text to say which command is running,
and doesn't yet turn singlets into H for Minimize All like it does for Minimize Selection.
Further improvements to these are planned, but won't be listed separately in this file.)

----------------------------

Name: Mark Sims
Date: 4/13/2005
Files: MainWindowUI, MWsemantics, modifyMode, part
Summary: Added "Modify > Invert" feature and moveDeltaPlus and moveDeltaMinus to modifyMode.  
Also improved the history message for the Stretch feature.

----------------------------

Name: Bruce Smith
Date: 4/13/2005
Files: GLPane.py
Summary: try to fix bug 530

This limits my bugfix from 4/8/05 (glpane or its edges would flicker to
black during main-window resize, which it turns out is a Mac-specific bug)
to the Mac, since that change also caused a new Windows-specific bug 530
(glpane flicker when covered by dialogs or menus, or maybe when they
disappear).

----------------------------

Name: Bruce Smith
Date: 4/13/2005
Files: GLPane.py
Summary: try to fix bug 507 (lighting prefs change works wrong unless in Home View)

----------------------------

Name: Bruce Smith
Date: 4/14/2005
Files: chem.py
Summary: fixed bug 498 (btw this is probably the same as bug 361)

----------------------------

Name: Bruce Smith
Date: 4/16/2005
Files: modes.py, depositMode.py
Summary: Use a "dynamic context menu" in Build mode; revise Set Hotspot for clipboard parts.

For details, see email to cad list, subject line:
  CHECKIN: dynamic context menus; Set Hotspot for clipboard parts

----------------------------

Name: Mark Sims
Date: 4/17/2005
Files: modifyMode
Summary: Implemented moveAbsolute() for one or more chunks.

----------------------------

Name: Mark Sims
Date: 4/20/2005
Files: modifyMode, fusechunksMode
Summary: Fixed bug 502-12. Implemented the Move Chunks modifier keys X, Y, and Z for Fuse Chunks to allow 
constrained translation and rotation around the 3 primary axes. I also added R and T modifier keys.

----------------------------

Name: Bruce Smith
Date: 4/21/2005
Files: Utility.py chunk.py files_mmp.py jigs.py
Summary:  fix bug 406: save leaf node hidden state in mmp file

----------------------------

Name: Bruce Smith
Date: 4/21/2005
Files: Utility.py files_mmp.py
Summary:  save group open state in mmp file

----------------------------

Name: Bruce Smith
Date: 4/25/2005
Files: files_mmp.py jigs.py node_indices.py Utility.py
Summary: "new jig mmpfile behaviors seem to work in simple tests"

The new jig behaviors deserve mention in release notes, as one aspect
of the assembly/Part split. 

They have no effect unless you drag jigs (or the chunks containing the atoms
they attach to) from one Part to another in the model tree, or unless
you use the new MT context menu command to disable a jig. And if you
do these things, it's designed with the intention that you can figure
out what's going on.

For details: I summarized the plan in recent mail to cad list (late last
week), subject "full list of jig/clipboard requirements for Alpha 5".

----------------------------

Name: Bruce Smith
Date: 4/25/2005
Files: movie.py
Summary:  fixed bugs 519, 384-6 

Warning: these fixes are not only "band-aids" (instead of the required
major surgery), but they only cover the wounds when you look at them from
certain directions, which happen to include the ones in those bug reports.

----------------------------

Name: Bruce Smith
Date: 4/27/2005
Files: part.py MWsemantics.py modelTree.py
Summary:  fixed bug 566 [not sure if this bug predated last release (A4) or not]

----------------------------

Name: Bruce Smith
Date: 4/28/2005
Files: chem.py moviefile.py movie.py movieMode.py
Summary: preliminary new movie code for Alpha5

Excerpt from my CHECKIN mail; for many more details see that mail:

I committed a major rewrite of the lowest-level movie code, as an
attempt to fix or mitigate a lot of its bugs (new and old) for Alpha5....

The basic change is that if you change the Part and then re-play the movie,
or change it while you're still in movieMode playing the movie, then
rather than crashing (tracebacks), it does its best to "do what you're
asking for" (whether or not it's sensible), safely. That is, it keeps
track of which atoms the movie is about, and moves them (except for deleted
ones), and doesn't move the newly added ones.
....

----------------------------

Name: Bruce Smith
Date: 4/29/2005
Files: assembly.py Utility.py MWsemantics.py
Summary:  fixed bug 413; revised wording of unsaved changes message

------------------------------------------ Alpha 6 --------------------------------------------

### something by Huaicai about cookie cutter snap to grid? I don't know
if he needs to add that here, I just wanted to make a place for it;
Huaicai can remove this comment [bruce 050502]

----------------------------

Name: Bruce Smith
Date: 5/2/2005
Files: bonds.py chem.py chunk.py depositMode.py files_mmp.py
Summary: split bonds.py out of chem.py; very partial support for bond2 etc, including in mmp files

For details (about how preliminary and partial it is) see email to cad list.

----------------------------

Name: Bruce Smith
Date: 5/4/2005
Files: TreeWidget.py modelTree.py
Summary: added model tree context menu command to select all atoms of any Jig or set of Jigs

----------------------------

Name: Bruce Smith
Date: 5/5/2005
Files: Utility.py jigs.py modelTree.py
Summary: fix Alpha5 bug 593: disabling/enabling Jigs now marks file as changed

----------------------------

Name: Bruce Smith
Date: 5/5/2005
Files: modelTree.py
Summary: commands in MT empty-space context menu: new clipboard item; delete all clipboard items.

----------------------------

Name: Bruce Smith
Date: 5/5/2005
Files: files_mmp.py Utility.py chunk.py
Summary: chunk color now saved in mmp file.

----------------------------

Name: Bruce Smith
Date: 5/10/2005
Files: chem.py
Summary: fixed bug 466: Transmute now sets file-modified flag.

----------------------------

Name: Bruce Smith
Date: 5/11/2005
Files: ops_atoms.py
Summary: fix bug 376 (only tested for no traceback, not correctness)

----------------------------

Name: Bruce Smith
Date: 5/11/2005
Files: lots
Summary: introduced "atom types" (very incomplete); part of the new bonds ontology

(see mail to cad for details: CHECKIN mail has list of files, prior mail has explanation)

----------------------------

Name: Bruce Smith
Date: 5/12/2005
Files: Utility.py
Summary: fixed bug 614 - Hide/Unhide now mark file as needing save (probably too minor for release notes)

----------------------------

Name: Bruce Smith
Date: 5/12/2005
Files: TreeView.py
Summary: fixed bug 620 - exceptions in MT when clicked while waiting for long op to complete

----------------------------

Name: Bruce Smith
Date: 5/13/2005 [and several days thereafter]
Files: 28 files
Summary: optimized mmp read, VdW draw, movie play; '==' -> 'is' in many files

For details see email: "CHECKIN: optimized readmmp, draw VdW, movies (I think)".

Basically: some ops on large parts are faster now, e.g. movie playing, file opening,
selecting atoms (due to the redraw speedup).

----------------------------

Name: Bruce Smith
Date: 5/19/2005
Files: Utility.py VQT.py drawer.py jigs.py
Summary: rotary motor circular arrow rotates with motor during movies

(It also rotates in Move Chunks mode, but only because we haven't yet fixed the
bug or NFR about wanting the motor to move along with the chunk. During this kind
of motion there can be glitches (unjustified sudden changes) in the displayed angle,
since the spokes get far from their nominal arrangement. These glitches are harmless.)

----------------------------

Name: Bruce Smith
Date: 5/19/2005
Files:  assembly.py TreeWidget.py modelTree.py ops_select.py

Summary: fix bug 500, revise selwhat vs selmode code, related behavior changes

See email to cad, "selection issues summary".

----------------------------

Name: Bruce Smith
Date: 5/19/2005
Files: jigs.py modelTree.py
Summary: added MT cmenu command for rotary motors: Recenter on atoms

This mitigates the bug/NFR asking for the motor to move automatically 
when the atoms do (if they are all moved together, in Move Chunks mode).
It doesn't really fix that, but it's not fully obvious how to fix that
(e.g. when only some of the atoms move). Even so, I assume we'll fix
that sometime, so this is a temporary "holdover" until the real fix.

----------------------------

Name: Bruce Smith
Date: 5/20/2005
Files: op_select_doubly.py ops_connected.py
Summary: New code for Select Doubly (fixes bug 582; makes this op linear time).

The old code was wrong (at least since Alpha1) and slow. The new code is
not obviously wrong on the test cases I tried, and fast. (It is also,
in theory, compatible with Selection Filter, so that should be changed
to no longer disable this operation, but that hasn't yet been done.)

Warning: the new code (like the old) does use recursion rather than iteration
and might get an exception for big parts due to hitting the recursion limit.
If this starts to happen, it can be considered a new bug, and it can be fixed
w/o too much trouble (maybe another 1/2 day of work).

----------------------------

Name: Bruce Smith
Date: 5/26/2005
Files: lots
Summary: support for copying selected atoms, jigs, Groups

(and the next day, did some bugfixes and extended the change to DND copy in MT)

For details, see mail to cad list, not yet sent, hopefully to be sent
within a few days.

----------------------------

Name: Bruce Smith
Date: 6/14/2005
Files: chem.py
Summary: fixed bug 661

----------------------------

Name: Bruce Smith
Date: 6/14/2005
Files: chunk.py depositMode.py
Summary: fixed bug 703, and related older bug requiring mouse motion to see new-hotspot color

----------------------------

Name: Bruce Smith
Date: 6/15/2005
Files: depositMode.py
Summary: fixed bug 264

----------------------------

Name: Bruce Smith
Date: 6/18/2005
Files: files_mmp.py chunk.py jigs.py Utility.py
Summary: permit right paren in node names in mmp files (fixes part of bug 474)

----------------------------

Name: Bruce Smith
Date: 6/29/2005
Files: ops_connected.py
Summary: fix bug 714 by making Select Connected non-recursive to avoid Python stack overflow

----------------------------

Name: Bruce Smith
Date: 6/29/2005
Files: MWsemantics.py
Summary: permit Select Connected and Select Doubly to be used when Selection Filter is active

----------------------------

Name: Bruce Smith
Date: 6/29/2005
Files: op_select_doubly.py
Summary: fix bug 725 by making Select Doubly non-recursive (analogous to bug 714)

----------------------------

Name: Bruce Smith
Date: 7/4/2005
Files: jig_Gamess.py jigs.py modelTree.py
Summary: Duplicate for Gamess jigs; Copy/Cut in MT cmenu for selections in clipboard

The Duplicate command needs no separate mention in release notes, but needs to be
mentioned somewhere in the Gamess documentation (if we write any before that entire
interface gets redesigned).

----------------------------

Name: Bruce Smith
Date: 7/5/2005
Files: jigs.py
Summary: implemented "Recenter on atoms" for Linear Motor

----------------------------

Name: Will Ware
Date: 11/3/2005
Files: NanotubeGenerator.py
Summary: Some nanotube chiralities can produce funky valences.

	 I ended up leaving open bonds on the ends, graphitic bonds
	 elsewhere, and all the carbons are sp2. When you hydrogenate
	 the result, you sometimes get the funky valence messages, and
	 I don't know a good fix for that. We should at least
	 release-note the funny valences. Maybe our users will be good
	 enough chemists to understand them and patch them manually.
	 We can't protect users from ever learning any chemistry at
	 all.

This happens when a carbon has a single graphitic bond to one other
carbon, so when you hydrogenate, it picks up two hydrogens, then it
whines about not having enough bonds for its hybridization.

On this topic, Damian wrote:

	If you're looking for an easy fix, just delete all carbon
	atoms with two single bonds on them (the sp2 carbons with the
	two H's on'em). The nanotube is defined by the pi-connectivity
	between carbons and the graphitic (in-tube) or aromatic (edge
	of tube) bonding of all the carbon atoms, not the absolute
	length and the need for all atoms to fill that space to be
	there. Unless your intention is to sew pairs of carbons into
	the tubes to connect each "single" carbon to the nearest
	neighbor aromatic carbon, which would be fine, too.

Here is what I think Damian is talking about with the "sewing" thing.
Make a (5,2) nanotube, 5 angstroms long. Hydrogenate it. There will be
two of the under-valenced carbons. Each one has a hydrogen that's
unrealistically close to another hydrogen. If you popped out those two
hydrogens and replaced them with another carbon, maybe that would work
better? I tried it, using sp2 carbons to replace the four hydrogens,
and it didn't seem to help.

I think the real solution is for the user to make the nanotube a bit
longer than he wants, and then hand-trim the ends of the tube. Either
that or come up with some brilliant algorithmic method to generate
clean tube ends. Then we might be able to implement Bruce's idea of
growing the end of the tube by one carbon layer at a time, though for
many chiralities it's more like growing spirally around the tube.

----------------------------

Name: Bruce Smith
Date: 12/1/05
Files: wiki_help.py and several others
Summary: wiki help feature

Extended Will's prototype of Eric D's idea of a wiki page per program 
feature or UI element, reachable from the program. The module docstring
describes the idea like this:

  wiki_help.py -- associate webpages (typically in a wiki) with program features,
  and provide access to them. Pages typically contain feature-specific help info,
  FAQ, forum, etc.

and also has this Terminology note:

  We use "web help" rather than "wiki help" in menu command text,
  since more users will know what it means,
  and since nothing in principle forces the web pages accessed this way to be wiki pages.

  But we use "wiki help" in history messages, since we want people to think of the wiki
  (rather than a rarely-changing web page) as an integral part of the idea.

----------------------------