VMD 1.8.5 Development

VMD Development Status

  • VMD 1.8.5 final release (August 25, 2006)
    • plugin page: Added a link to the CSCS "MultiMSMS" plugin page.
    • README: fix typo and remove stuff pertaining only to the unofficial builds
    • autopsf docs: made the autopsf window screen shot a more reasonable size
    • salt bridge docs: made the salt bridge plugin window image a more reasonable size
    • namdenergy docs: Added the units that output is given in
    • Add .i files to the source distribution
    • apbsrun: improved the load maps dialog giving three options for loading the output files.
    • final (hopefully) win32 installer updates
    • Turn off Python for Mac builds not using Apple's libraries
    • switchover for doing mac builds using our Tcl/Tk, last build of version using the Apple libraries is completed.
    • updated colorscalebar plugin docs
    • cleaned up the mutator doc page
    • updated apbsrun docs
    • Fixed a bad strcat() call in error handling within STAMP
    • apbsrun: Improved the uncharged molecule warning box. Added comment about buggy behavior when the top molecule is changed in between edits of an elec statement. This can occur when running autopsf or loading new molecules between edits, and should be addressed in a future revision.
    • molefacture: Fixed a bug in the protein builder that misplaced the resid of the new backbone nitrogens.
    • pulled over latest round of minor tweaks from the main multiseq CVS tree.
    • namdgui: fixed the behavior of the File | Reset All menu item, so it doesn't add duplicate copies of the parameter file each time, by correcting a line in the init_defaults proc.
    • biocorepubsync: Added check to make sure biocoreProject != -1, which should hopefully fix the BioCoRE errors. Also upped the required biocore package number.
    • biocore chat: Fixed help URL menu. (Had extra "-command" and other extra text.)
    • paratool: Got rid of namdserver use in charmm charges calculation. This feature is currently disabled until it is stable.
  • VMD 1.8.5rc6 (August 22, 2006)
    • biocore chat: Stored local version of current project, which should fix the bug of multiple project panels appearing in the chat window; Also fixed the help URL.
    • biocore: Added biocorechat image to help page.
    • biocore: http://www.ks.uiuc.edu/Research/vmd/plugins/ to [string trimright [vmdinfo www] /]/plugins/
    • biocore chat: a call in the biocore api that takes the listener id, but when switching projects, it was getting the BioCoRE session id.
  • VMD 1.8.5rc5 (August 21, 2006)
    • PMEPot patch to add a menubar similar to the one in irspecgui that allows to set charges, if they are not yet set. Also pmepot contained many 'error' messages, that are fine for textmode script, but in GUI mode it is better to have clean dialogs and abort the pending action. Patch by Axel Kohlmeyer.
    • clonereps: Patch to add support for symbolic names like 'top', 'active', 'all' and 'displayed' to the molecule selectors. Patch by Luis Gracia and Axel Kohlmeyer.
    • This patch (Axel) adds the correct expansion of the PYTHONPATH environment variable for python plugins shipped in subdirectories of the VMD plugins tree. python has a different strategy than tcl/tk, so it has to be implemented differently.
    • Added Alek's PMEpot reference paper to the docs.
    • updated docs with cone/cylinder default res and filled behavior.
    • added docs for 'mol reanalyze'
  • VMD 1.8.5rc4 (August 20, 2006)
    • Updated the README infolist.txt file in the win32 installer
    • Changed the MoleculeEvent to use an enum parameter, converted the code to use the new enums, and added a new enum to indicate when a structure has been re-analyzed.
    • Updated win32 installer
    • added 'mol reanalyze' to the README
    • If the molecule has no parts, it might later get some; for example, the script "mol new ; mol addfile 1atp" will initially generate a molecule with no parts, but the later addition of structure doesn't get noticed because the molecule is already registered. Justin Gullingsrud; 8/16/06
    • simpleedit no longer uses the external EDITOR environment variable, and now uses the VMD textview plugin for simple text editing purposes.
    • Multiseq now deletes any color bar it has created at shutdown time.
    • Removed redundant scoping, which causes problems with recent versions of g++.
    • Added a colorbar cleanup routine, so that the main multiseq window can remove it at closedown time.
    • Added a new "mol reanalyze" command to allow user to hack their structures and force VMD to re-analyze it on-the-fly.
    • added biocore remote job execution to the main exectool index page
    • added exectool plugin to master index
    • suppress bad structure warnings after they've seen 30 of them...
    • Added a dialog to tell the user if he tries to submit a new APBS run when and old one is already running, and gives him the opportunity to either keep waiting, or to forget the old one so that he can submit a new one.
    • namdenergy: If there's already a biocore job running, don't start up a new one. Print a dialog giving the user the option to either continue waiting for the old run, or forget about it so they can start a new one.
    • Added logic to nucleic ribbons per-fragment residue loop to abort rendering the entire fragment if we're unable to find a usable control point for a residue. Although we could probably hack the spline control point position using other methods, we have to pick a usable atom so that coloring and picking are done correctly. Added tests for terminal residue names inside the per-fragment residue loop.
    • namdenergy: If the job terminates abnormally, reset the job id to -1 so a new job can be started
    • improve comment
    • Changed DrawMolItem to duplicate cmdStr to prevent overlapping strcpy() buffer pointers.
    • Added clipping plane support to the clonerep plugin (AK)
    • Added links to the biocore and exectool/biocore help pages for the error dialogs.
    • Initial documentation for using the ExecTool plugin with BioCoRE
    • All of the video cards we care about going forward now support GLSL version 1.10 or later, and the old ifdef tests we did to deduce which types of bugs each implementation were of minimal value. Going forward it's better to use an all-or-nothing approach and maintain clean and fast shader code. Removing the crutches for old drivers that didn't previously support gl_FrontFacing is a necessary evil in order to get nice results with properly functioning cards. Those with old drivers can either copy in old shaders, or complain to the vendor to fix their GLSL compiler and drivers.
    • cranked version number
  • VMD 1.8.5rc2 (August 19, 2006)
    • Fixed record_file() FileSpec string generation code, as the FileSpec member name is "stride", but the Tcl command name is actually "step".
    • save FileSpec options used to load each file in the saved state.
    • save the file loading options in addition to the filename and type, so that saved states can restore the same frame range etc.
    • Added Axel's patch to allow the textviewer/editor to be opened with a given path. Fixed a few catch statements (didn't show, because the tk_Open/SaveFile dialogs catch them), now needed due to externally provided path. Have the window name contain the current file name.
    • added info about state saving improvements
    • save the state of all user defined clipping planes
    • make errors setting colors non-fatal, as we frequently run into the case where one molecule, residue, etc doesn't exist in the restored state, this way the code will grind along and set everything that's settable.
    • updated openURL.tcl for the new exectool namespace
    • Renamed "Glassy" to "Glossy" to make it distinct from the two transparent glass materials
    • namdenergy: Fixed an odd bug when multiple molecules are around in the command line invocation that I can't believe never caused problems before.
    • cranked version number
  • VMD 1.8.5rc1 (August 19, 2006)
    • fix the add-on extension/startup extension mechanism to deal with paths containing spaces
    • Renamed EXTRALIBS to VMDEXTRALIBS, in case users have an identical environment variable set for some reason. This is used for RPM bulding.
    • Updated the LICENSE file for the plugin tree with up to date information about included sources copyrighted by others
    • Updated apbsrun, autoimd, multiplot, namdenergy, namdgui, namdserver, paratool, resptool, and vmdmovie plugins to use the new exectool namespace, and to increment their respective version numbers so any interdependencies are properly handled, and fixed a few inconsistencies in package require/provide statement ordering here and there.
    • While we're revving exectool anyway, I'll fix the namespace so it matches the package, to reduce the potential for confusion...
    • namdenergy: updated package require for exectool and fixed order of package require statements
    • cranked version of exectool to 1.2, since 1.1 was self-inconsistent until Justin patched the version numbering mismatches.
    • Added documentation for recently added environment variables
    • Updated the README file with more information on updated/changed features and plugins.
    • Shorten "Data Set" label to "Vol", and added a "Range" label to the new isovalue slider min/max range control boxes.
    • release candidate mode
    • Added a doc subdir to the vmd multiseq cvs area and populated it with the latest from the ZS group.
    • enable SRPM builds more easily
    • Added Axel's plugin registration extension patch to allow RPMs to populate a (VMDDIR)/scripts/init.d directory with extra plugins, extensions, and random other routines at initialization time by sourcing all .tcl files found in that directory at startup time, immediately following the registration of the graphical extension plugins. Perhaps this ought to be moved into it's own proc?
    • migrated plugin and extension loading startup routines to their own file, and cleaned up vmdinit.tcl substantially.
    • Added cleanup of generated vmd.py to Makefile.swig per Axel's patch
    • Moved the login-checking code into exectool
    • Moved the login biocore window into exectool
    • Moved login-checking gui code into exectool, so the biocore login window pops up and gives the user the opportunity to log in and continue where he left off, or just cancel and return an error
    • Added a button to the not-logged-in error dialog to take the user to the Login Tk window if they aren't logged in.
    • Fixed missing MoleculeEvent callback for "mol delete all".
    • prevent integer wraparound (due to tiny pairdist) from causing a bad allocation, and a subsequent crash.
    • added a comment about the extra NULL test and error message so I can return to this and work on it more later.
    • test for NULL subtree values when checking for recusion in atom selection macros, and complain about it. The yacc code also emits a syntax error message. More work is needed to clean up this corner case.
    • Patch apbsrun to replace many (but not yet all) fatal error messages that were sent to the console with dialogs.
    • hold back psfplugin changes until after release.
    • psfplugin: Added one extra column to the mass field, to deal with masses >= 1000 (useful for cg structures)
    • Changed autopsf to behave more intuitively if there are no automatically identified patches; it now goes ahead and creates the psf/pdb without forcing an extra button push.
    • namdenergy: GUI cleanup
    • biocore: more descriptive retry button text
    • apbsrun: Added a brief tutorial on coloring by potential.
    • timeline: Added Axel's patch to replace the text console input code for setting the scaling with a dialog box.
    • Added a test to rename an existing binary before copying in the new one, then subsequently deleting the other one, to avoid "text file busy" problems. Patch by Axel.
    • biocore: exported a couple of job commands
    • biocore login: added 'Return' to the password field as a way to submit the form. Move the code to execute upon submittal from the -command of the button to its own procedure so that it could be called from both places
    • Ghost is just barely visible in non-glsl mode, so I took off the GLSL tag in the name
    • fix clonerep error message box
    • added Jordi's glass material
    • corrected Schofield method names
    • added Jordi's glass material
    • fix string compare in palette tool
    • make plugin makefile permission finder target more self-explanatory
    • fixed another molecule name string compare in colorscalebar
    • Added Axel's glass material, still need to check out and see if the set is non-redundant prior to release.
    • irspecgui: button to copy the charge info from beta field.
    • namdplot: Updated namdplot interface with Axel's improvements
    • scriptlib mk3drama: updated with newest version
    • scriptlib webvmd: fix command to work well on windows
    • change the namespace usedin the colorscalebar plugin to avoid conflict with the older colorbar version still used by multiseq
    • string fix from Axel
    • Made the "RWB" (red->white->blue) color scale the default for VMD, as it's a more understandable default, and is frequently what people would immediately set it do anyway. Also added RGryB and BGryR color scales that use a light gray color as the middle value color, for better contrast between mid-range values and white backgrounds.
    • Fixed break in namdenergy_usage from missing quote
    • Elijah fixed the DB checking code for cases where the DB is updated after the initial run.
    • apbsrun: Added a button to open up the help page if login fails
    • exectool: Made some prompts more informative and tweaked the layout
    • updated multiseq info message with the final citation reference for the BMC bioinformatics paper. Online ref is: http://www.biomedcentral.com/1471-2105/7/382/abstract
    • Applied Justin's MACVMD patches so the main tree has the necessary mods to build with his native Mac GUI
    • cranked version number
  • VMD 1.8.5b9 (August 16, 2006)
    • added palettetool plugin
    • Added Axel's 3-D ramaplot extension to the ramaplot plugin, and cranked the plugin version.
    • Added Elijah's "MetallicPastel". Need to check on GLSL and versus other materials.
    • numerous multiseq improvements/fixes
    • Updated volmap ligand method with final journal citation.
    • Synced up plugins tree with master multiseq devel tree
    • exectool: Added returns so that remjob_set_default_projname/acctname don't print anything
    • apbsrun: Added a new APBSRun window for editing mobile ion concentrations and radii This initial implementation assumes that the user wants both positive and negative ions in equal concentration, with identical radii, as suggested by Nathan Baker. We can add a full ion list editor later to support more sophisticated calculations.
    • exectool: Added ::FinderTool::remjob_set_default_biocore_acctname and ::FinderTool::remjob_set_default_biocore_projname to allow users to define default values for job submission. Since these are likely to be used in the user's .vmdrc (before he's logged in to BioCoRE), no validity checking is done when they are called. However, later, when the exectool GUI gets built, if they are invalid, a console message is printed, and the first Project and Account in the user's list are selected.
    • namdenergy: Added several dialogs for error conditions Added "namdEnergy)" prefix to all console puts Changed the "Use BioCoRE" checkbox to a "Run locally/ Run remotely with BioCoRE" radio-button pair
    • exectool: Changed name of GUI button from "Continue" to "Start job"
    • Updated title bar to the apbsrun load maps window
    • apbsrun: set default fine grid size the same as the coarse grid size, at 1.5 times larger than the bounding box for the atoms in the system, so that the resulting potential grid and isosurface aren't truncated. Suggested by Nathan Baker.
    • volmap command-line: updated with latest (but not final) status of the implicit ligand sampling paper
    • readcharmmpar: Fixed error in comment.
    • paratool: Improvement and bugfixed for nonbonded refinement. Switched to faster angle measure routines Made Hessian GUI faster Taking signs of coupling constants into account Better tratment of bonds and angles in rings (this still has to be improved)
    • Replaced measure_angle and measure_dihed by much faster routines. (But it still would be nice to have a C++ based function provided by VMD.)
    • cranked version number
  • VMD 1.8.5b8 (August 14, 2006)
    • Print even more explicit messages about shader compile failures and driver bugs.
    • Made GLSL shader compile failure messages less intimidating, and put the blame clearly on the GPU/driver vendor where it belongs.
    • VMD will no longer print the GLSL shader compile/link error streams unless the user sets the VMDGLSLVERBOSE environment variable.
    • committed prototype font width/antialiasing toggling code, but in a disabled state so it's easier to experiment with until we make a decision.
    • Fix User's Guide typo found by Greg Sigalov
    • colorscalebar: set autoscale to off by default, since the plugin currently only tests the first rep, and has no means of allowing the user which rep is the one the color scale belongs with.
    • Added thin border lines around the color scale bar so it looks much nicer when displayed on a white background. Requested by Marcos.
    • Added another cleanup call for when the BioCoRE submission fails and the job is not submitted
    • If the user has no accounts, rather than printing an error, just return an empty list, and let the caller handle it
    • Updated win32 installer with fixed spectral density plugin, and updated VMD startup directory.
    • Changed initial parameter tags to match up with the new exectool naming convention of prefixing all BioCoRE field names to "biocore_*"
    • Renamed parameter tags to "biocore_*" (i.e. biocore_jobName instead of just jobName) to make the code clearer, and to make name collisions less likely. This affects all clients that need to pass in initial values to remjob_config_account or remjob_config_account_biocore.
    • prettied up the makefile for the irspecgui plugin
    • fixed typo in win32 portion of the spectral density back-end
    • added a delete button for the color scale bar, eliminated duplicated functionality, and made the deletion proc more robust.
    • molefacture: Fixed a number of minor bugs
    • namdenergy: Added a couple of dialogs to tell the user what's going on with the BioCoRE runs he's trying to submit
    • increase text line thickness to 2.0
    • namdenergy: Modified code that calls remjob_config_account for BioCoRE runs, to use the new exectool calling convention of passing single-element lists for parameter initial values, rather than just a string, to correctly handle spaces in parameter values
    • Fixed a typo, and updated code that reads initial account params to remjob_config_account, so that lists are expected rather than just a single value. By wrapping values in lists, initial values with spaces don't get broken up.
    • Yesterday's changes to exectool changed the calling conventions for initial values sent to remjob_config_account. Now, initial values should be a list with the first element the initial value, and subsequent elements being the flags. Note that the initial values must be encapsulated in lists, even if no flags are specified, to prevent initial values with spaces from being broken up.
    • irspecgui: fix typos caused by recycling an older gofr makefile
    • irspecgui: fix typo in the Win32 macro tests for DLL exports
    • exectool: Added error dialogs in the GUI for the odd cases where an account requires queue names, but then the connection to bicore is lost before the queue names can be retrieved, or for when no queue names are returned Added "exectool) " to all console output Removed an "after idle destroy" call for a frame in the gui that is being immediately redrawn. We're not sure if this is safe or not, but I suspect there might be problems if the after idle is left in, since the window may be immediately redrawn with the same name, causing the after idle to destroy the new window..
    • irspecgui: updated apbsrun package require version to match the shipping version number
    • cranked version of apbsrun to 1.2 since we've made non-trivial changes
    • reduced the size of the namdenergy doc pic
    • Updated apbsrun gui to be more explicit about where the job is running. Changed all console messages to identify themselves as coming from apbsrun, and made their format consistent with other plugins.
    • renamed biocore job submission button to make it clear what the purpose of the button is, for those that don't know what BioCoRE is.
    • apbsrun: added new map checking code that only allows the loader window to open if all of the expected output maps exist, are readable, and have non-zero file sizes.
    • exectool: Added 'after idle' for window destroy calls to avoid Tk crashes
    • updated exectool so that it won't find/accept directories with the execute bit set as a valid executable. (e.g. a directory named 'apbs' in the current working directory would be found rather than /usr/local/bin/apbs depending on search path ordering, without additional tests)
    • eliminated unused variables
    • Gave the caller of remjob_config_account the ability to declare some fields "readonly" or "hidden" so that they a) can't be modified by the user or b) don't show up at all in the GUI. Of course, they must be initialized to valid values by the caller code if the user can't set them.
    • cranked version number
  • VMD 1.8.5b7 (August 13, 2006)
    • assume 10.4, and enable python etc for MacOS X powerpc builds
    • force OSX 10.4 builds to use FLTK 1.1.7
    • Use sydney and do 10.4 builds of the plugins using the Apple Tcl/Tk/Python
    • leave the performance-specific aspect of the forced stereo drawing code for a subsequent release. For now we'll stick with just fixing the Mac stereo bugs and leave performance tweak oriented stuff for later.
    • Added profuse commentary about the modification to Scene::draw() which works around the broken MacOS X stereo implementation.
    • Implemented a workaround for the MacOS X stereo bugs, by forcing draws to the left/right buffers when rendering in normal mode. Scene calls left() and right() with a non-stereo mode, which triggers drawing as normal(), but to GL_BACK_LEFT or GL_BACK_RIGHT.
    • Added a new variation to Scene::draw() which handles the case where a broken OpenGL stereo implementation requires separately drawing to the GL_BACK_LEFT and GL_BACK_RIGHT even when rendering monoscopically.
    • Added a new forced_stereo_draws() method to DisplayDevice, for use in implementing a workaround for broken MacOS X stereo drivers that don't correctly honor GL_BACK for drawing/clearing, and always require separately clearing/drawing to GL_BACK_LEFT and GL_BACK_RIGHT. Eventually this code path can be used to work around nVidia Linux drivers that have very bad performance, though they do honor GL_BACK.
    • Initial workaround for broken stereo drawing on MacOS X. Will also work as a solution for the 50x speed hit on nVidia Linux when drawing in mono.
    • Added new stereo flag to indicate a platform/driver/card that has a broken stereo implementation that requires drawing in stereo mode full-time when using a stereo visual for display, even when we'd prefer to draw in mono mode. The prime offenders here are the nVidia Quadro cards in general. On Linux these cards take a large performance hit when rendering monoscopically otherwise. On MacOS X, the drivers are actually broken and don't behave correctly when drawing to GL_BACK when a stereo visual is in use. When set, VMD will have to draw separately to GL_BACK_LEFT and GL_BACK_RIGHT even when operating in its normal monoscopic display mode.
    • set non-sensical normalization to 0 in PBC enabled cases for 'measure gofr'.
    • gofrgui: improve output file dialog options
    • irspecgui: catch a few more problem cases, and give more meaningful error messages
    • allow user to enable OpenGL extensions in MacOS X PowerPC at runtime, but leave them off by default since buggy drivers still appear to be the norm based on testing 10.4.7 on an NVidia Quadro 4500.
    • cranked gromacs plugin minor versions
    • gromacsplugin: Fix handling of non-orthogonal unit cells with corrected unit conversion in the pbc side/angle calculation.
    • PowerPC versions of OSX like the other stride code still
    • Enable stereo support for MacOS X builds done with FLTK 1.1.7 and later.
    • Wrote a workaround for bugs in the MacOS X stereo implementation.
    • fix typo in macosx ppc ifdef
    • added Makefile target to find bad file permissions in the plugin tree
    • added comment about apbsrun dependency
    • added comments about ff_parameter_init call during startup and reminders that this stuff should be migrated to a new plugin eventually.
    • fix pkgIndex.tcl for apbsrun, as it now needs access to it's directory immediately on startup in order to read the radii.dat parameters, so that other plugins can access this information.
    • Applied Justin's patch to make the solvent rep draw itself using the line and point array display commands rather than zillions of individual point/line commands.
    • Added several new parameters to control whether the color scale bar replaces any existing color scale bars, the screen location of the new bar, and made the code more robust.
    • Improve the default size and position of the color scale bar
    • Added in tick lines corresponding to each label
    • Added controls to allow the user to disable autoscaling and set the color scale bar range manually. Also added extra safety checks for the situation where no loaded molecules have data. Displays a tk error message dialog when a problem occurs.
    • Added a new radiobutton selector to select decimal or scientific notation for printing the color scale bar labels. The default is to print them with scientific notation, to 3 decimal places. (no point in displaying more than 3 decimal places since there are only 1024 distinct color scale colors anyway)
    • Fixed hard coded color scaling code in intersurf to work with new color ranges in VMD 1.8.5 (queries ranges dynamically now)
    • cranked version on the colorscalebar plugin to 1.2
    • Fixed the colorbar plugin to query the usable color range from VMD, rather than using hard coded color index and range values. I also fixed the quoting on the draw text command so that the labels are correctly right-justified. This plugin should be merged into the regular 'colorscalebar' plugin in a subsequent release.
    • eliminated a few more incorrectly hard-coded constants in the colorscalebar plugin
    • Corrected right justification of label text in the color scale bar plugin. Also fixed the color scale code to correctly query the min/max color scale color indices rather than hard coding 17/1042 as loop limits etc.
    • force the '-' and '+' characters to use consistent width to improve behavior of right and left justified numeric string printing.
    • Hacked the Hershey romans font used by VMD to yield constant widths for the numeric digit characters, and for the space ' ', and deciman point '.' characters. This allows right justified numeric strings to line up in columns when drawn with consistent y positions with 'draw text', for example.
    • correct comment typo
    • Added real irspecgui docs
    • Added support for loading atom charge map files into irspecgui
    • prevent custom window geometry code from running on MacOS X x86
    • applied Axel's update to improve the robustness of his irspecgui plugin
    • apbsrun now finds its install directory in such a way that it automatically tracks the plugin version and install location. the parameter info file reading code is now in a proc for easy reuse in other plugins.
    • Changed Ghost(GLSL) opacity to 0.10 to be more background neutral
    • Added Tim Isgro's Ghost and Steel materials
    • Added initialization proc to read default material list at startup.
    • Added new list of materials to make available by default
    • Make save state smarter about adding new materials, it now only adds them if they don't already exist. This eliminates the hack for handling the two default materials, and we can now include as many materials in VMD as we want with no penalty.
    • Make both MacOS X versions use the same stride code path
    • Enable renormalization full-time as a final cure for people remote displaying on machines that have broken rescale normal extensions, and in preparation for drawing ellipsoids etc. If we see a big speed hit, we can revert this change easily for the time being. If there's no longer a big performance penalty, then leaving renormalization on will make life easier on the buggy drivers. If performance suffers, we can revert to the previous behavior, and for drawing ellipsoids etc we can toggle between full renormalization and rescaling as necessary.
    • fix glprocaddress ifdef test for macosx versions
    • Updated comment about linux/macosx needing the static glTexImage3D()
    • Enable OpenGL extensions on the PowerPC builds (yes, we're playing with fire)
    • added a comment about the separate specular color issue with MacOS X, and simplified the GL version test
    • Added Axel's irspecgui plugin to the build
    • Updated placeholder irspecgui doc page
    • Added irspecgui plugin to the extensions menu
    • Added Axel's IR spectra plugin
    • updated window title for gofrgui
    • I added the initial code for running NAMD via BioCoRE. There isn't too much error checking yet, and there is no command-line way to run namdenergy yet. I'm pretty sure that if you don't set the "Use BioCoRE" checkbox, it will run as it did before.
    • Updated the default resolution of Licorice reps
    • Applied Axel's gofr gui updates for non-PBC handling etc.
    • updated the version of the clonerep plugin so it supercedes older versions released independently of VMD on the Bochum web site.
    • Bumped the Licorice Res to 10/10. Licorice looks much better when both bonds and spheres have the same res., also bonds at 6 made them looks artificially small vs. the spheres. The increase in res. is justifiable since licorice is by definition a "small number of atoms" and "zoomed in" type of representation.
    • biocore: copied the docs over from where they were to the docs/ subdirectory because John asked. Revised to the tcl code to point to the http://www.ks.uiuc.edu/Research/vmd/plugins/biocore/ instead of the old url.
    • paratool: Fixed 1-4 nonbonded interaction detection so that rings are treated correctly.
    • paratool: Changed behaviour of parameter editing GUI so that you can not manipulate parameters of types or bonded interactions that are defined in one of the loaded top/par files from the force field. This is more logical because otherwise you would break the force field. Nevertheless if you change the type names to a new value you can modify the parameters freely.
    • paratool: Fixed a few bugs in nonbonded refinement that I missed in the previous commit. A few GUI improvements.
    • namdserver: Decreased default debug output level.
    • paratool: Added GUI for refinement of nonbonded 1-4 parameters. Some bugfixes and housecleaning
    • paratool: Added field "Known" which is set to Yes if the atom type is defined in one of the loaded topology files.
    • paratool: housecleaning
    • optimization: Fixed bug that treated parameter bounds wrong in sim. annealing mode.
    • readcharmmtop: Fixed the function topology_contains_type.
    • Added chromatin example image, and vdna gui to vdna docs
    • Added minimal library dependency version information to the Doxygen Programmer's Guide as requested by Dan Strahs.
    • added gofrgui plugin to readme
    • BioCoRE produces a zero-length file when the user requests a file that ends up not getting produced by the submitted job. So I put a loop in to check for and delete zero-length output files. Unfortunately, the APBS GUI doesn't check for files that weren't actually produced for a particular run, and still tries to load them. (This will effect local runs as well as BioCoRE runs)
    • cranked version number
  • VMD 1.8.5b6 (August 9, 2006)
    • Added the plugins from the multiseq CVS tree to the build. These are kept in a separate list since they'll have to be synced up from the multiseq CVS on a periodic basis.
    • Populated the VMD plugin CVS tree with the latest versions of the multiseq plugins, so people can build from the regular plugin source tree and get multiseq.
    • Added gofrgui to the extensions menu
    • Added Axel's updates for the clonerep plugin, with Tk dialogs for errors when running with Tk.
    • Added a dialog box that informs the user when APBSRun can't connect to BioCoRE, and directs him to the login option in the extensions menu
    • Added dialogs to report certain errors to the user in the GUI interface. If exectool can't connect to biocore, or the user is not in any projects or has not specified any job accounts, the error dialog directs him to the BioCoRE web site to fix the problem
    • Added gofrgui to the master plugin list
    • Added status update messages which are printed when gofr calculations take longer than 5 seconds
    • Added Axel's graphical front-end for the new 'measure gofr' command
    • minor bug fix to gofr parameter checking
    • measure gofr fix: must not add the 'edge' correction when we don't use pbc.
    • exectool: Fixed a return value, removed some debugging prints, and made the GUI call destroy the window after completion, so everything gets reinitialized correctly on a subsequent invocation
    • Commiting a big chunk of code to run APBS over BioCoRE. Main changes: Put files in a temp dir inside of temp, rather than just in temp Add a checkbox to the GUI to select BioCoRE or local mode Call the exectool biocore api to setup job, stage files, and run.
    • Moved plugin development info to the top of the plugins page.
    • vdna: Added new documentation text from Tom
    • saltbr: Applied Leo's update for the bugs/shortcomings found by Axel.
    • updated vasp plugin docs improving the formatting
    • Fixed Sung's email address line in the vasp plugin docs.
    • Added some initialization code to the GUI call that I forgot, and fixed some bad parameter tagnames.
    • cranked version number
  • VMD 1.8.5b5 (August 7, 2006)
    • Added Tom Bishops VDNA plugin to the Extensions menu in the visualization submenu.
    • Added vdna documentation to the master list
    • Added vdna to the master plugin makefile
    • Added Tom Bishop's virtual DNA viewer plugin
    • Replaced old lammps parsing code with Marco Kalweit's implementation.
    • exectool: This commit includes a revised API to the remjob_config_account_biocore so that it now takes a parameter list in the form of { key1 val1 ... } where the keys are the tags the particular job needs.
    • started integrating lammps plugin updates from Marco Kalweit
    • Fixed const specifiers and converted C++ style comments to proper C comments
    • Updated the XCrySDen plugin to recognize both the .axsf and .xsf filename extensions, and improved the pretty plugin name
    • updated axsfplugin docs
    • namdenergy: Changed a variable name in the namdserver related section for better clarity.
    • paratool: Fixed code for automatic adding of missing x-coordinates.
    • paratool: Added listbox to examine 1-4 nonbonded interactions. Improved angle refinement. Housecleaning
    • paratool: Minor bugfix in del_coordinate.
    • paratool: Automatically find rings after a new molecule is loaded.
    • paratool: Addedfunction to detect rings in molecules.
    • namdserver: Prettied up log output.
    • optimiztion: Added option to specify boundaries between that the parameters can only be varied during the optimization.
    • Added Axel's patch removing the dismiss button from the clonerep plugin
    • Added Axel's patch to cause the plugin build to early-exit if any step fails, and to fix the alphabetical order of cgtools in the list of targets
    • Switch the MacOS X plugin builds to use the Apple provided Tcl/Tk frameworks
    • Switch Mac Aqua builds to use the Apple-provided Tcl/Tk frameworks
    • removed old MacOS X 10.2-specific compile-time macros
    • fix typo in frame selector comments
    • re-enable python and numpy since they aren't the cause of the 10.4 build crash at shutdown.
    • regress the 10.4 specific PowerPC build until I track down a weird crash
    • Fixed problem in handling of empty selections pointed out by jhenin
    • Added docs for VASP plugins
    • crank versions of all of the VASP plugins by on minor rev so these get used rather than any older ones on the user's system.
    • Added Sung Sakong's VASP plugins to the build
    • VASP plugins: rewrote the rotation matrix initialization to please the SGI IRIX compilers.
    • Replaced VASP plugin header acosf() and sqrtf() calls with plain sqrt() and acos() calls until we have appropriate compatibility macros added for the plugin builds like we do for VMD itself. Changed the type and initialization of the 3x3 r[][] matrix to something accepted by more compilers.
    • vaspchgcarplugin: replaced use of snprintf() with plain sprintf() so the plugin builds on Win32. This should be safe since the variable part of the format string is only a single character.
    • Added Sung Sakong's VASP plugins to the source tree. A few modifications are needed to integrate these and fix portability, but the changes will be pretty minor.
    • Fix for problems with non-gui packages requiring paratool (still might want to think more about this organization longer-term, possibly making a separate toolkit for common functions)
    • Added a 10.4 based PowerPC MacOS X build that includes Python.
    • Added Python to the default MacOS X x86 builds.
    • MacOS X x86 builds now use the system-provided Python framework, with a manually added Numeric 24.2.
    • Set MacOS X PowerPC builds to use the system provided Python framework, with a manually added Numeric 24.2.
    • Mac builds will now use Python frameworks rather than the plain old Unix libraries, as this is the only way to get modules to work correctly with the latest versions IIRC.
    • cranked version number
  • VMD 1.8.5b4 (August 3, 2006)
    • Added the clonereps and navfly plugins to the main plugins web page
    • ClipTool: removed the confusing clip plane activation logic that occurs when the user selects a new plane number. Also made "Top Molecule" the new default target for all clipplane operations.
    • Patched qmtool to keep it's arrow procedure in its local namespace rather than adding it to the global namespace.
    • Added Axel Kohlmeyer's clonerep plugin
    • Applied Justin's patch to update Mutator's alchemify C extension by requiring it as a package rather than manually loading the shared library
    • Applied Justin's patch to eliminate the main routine in hesstrans
    • molefacture: A little more house cleaning prior to release, plus fixed a bug if you get a negative valence value
    • molefacture: Got rid of a lot of debug spam
    • Added Justin's patch to cause vmdinit.tcl to more gracefully deal with a missing plugin directory.
    • Added Justin's patch to query current light colors from Scene
    • biocore: Replaced call to "getRecentChats" to "getRecentMessages" to get all recent messages.
    • eliminated AtomRep queries from MSMS and Surf rep code.
    • eliminated AtomRep query from newcartoon code
    • eliminate AtomRep queries from ribbon drawing routines, fix a typo in the cpk code.
    • Updated the trace rep to use parameters rather than querying AtomRep directly.
    • Added help menu to BioCoRE chat window.
    • Eliminated unneeded ifdefs from initial implementations of code that is now solidly part of the standard builds. Changed many rep routines to accept explicit parameters rather than having them query the AtomRep structure for themselves on-the-fly.
    • eliminated old VMDISOSURFACE compile time define
    • renamed multiseq menu item as requested by the ZS group
    • Started adding README details for this release now that we're getting close..
    • Added pick points to the Beads rep.
    • updated cvs logs with new cvs changelog builder statistics.
    • Updated 'measure gofr' docs.
    • Improved gofr implementation returns information on the number of frames used/skipped/total, calculates default box sizes better, handles single frame runs using the current atom selection frames appropriately, and improved parameter checking for rmax and delta.
    • measure gofr: added proper handling of empty selections, misc cleanup.
    • Force the label form to update itself when molecules are deleted. This requires catching MOL_DEL commands, and forcing the geometry list to fully update itself prior to regenerating the label browser.
    • autoimd: added dependency on the pdf docs for the distrib target
    • fixed typo in textview example in docs
    • mol2plugin: fix minor column spacing issue when writing mol2 files
    • Updated AMBER plugins to convert to and from elementary charge units
    • correct 'measure gofr' argument checking, and eliminate unused variables.
    • saltbr: updated docs to fix an error
    • cranked version number
  • VMD 1.8.5b3 (July 24, 2006)
    • updated multiseq plugin
    • psfgen: Fix bad count of guessed atoms due to continued guessing after success.
    • added cgtools and saltbr plugins to the main plugin doc page
    • Added saltbr plugin to the extensions menu
    • Added tk callback for salt bridge plugin
    • added salt bridge plugin by Leo and Elizabeth
    • added saltbr docs
    • added docs for new "measure gofr" command
    • eliminated the old measure gofr code now that all key functions are performed in the back-end.
    • added migrated the new gofr code into the back-end, and added support for updating the atom selections each frame.
    • moved gofr coordinate buffer memory allocation entirely out of the frame loop
    • updated the new version of the gofr code with proper handling of PBC data, pre-checking the trajectory PBC cell for orthogonality etc.
    • eliminated more unnecessary code before merging into the back-end.
    • Added Axel's new gofr implementation into TclMeasure, prior to merging into the back end Measure code.
    • move the core of the gofr code from the Tcl interface and use the new AtomSel::timestep() method to get pbc info from sel1.
    • Added timestep access function for selections, to make it easy to access the periodic cell info for a selection.
    • renamed CommandQueue::check() to CommandQueue::check_events() to avoid conflicts with Apple's Cocoa/Carbon frameworks.
    • Added dynamic bonds to the other ball/stick reps, patch by Justin
    • Misc updates to the new 'measure gofr' code to prepare for implementation of version with built-in normalization, and to move more of the brains out of the Tcl-specific part of the code.
    • cranked version number
  • VMD 1.8.5b2 (July 20, 2006)
    • Updated win32 installer with new plugins
    • pbcbox: new version from Olaf Lenz
    • pbctools: version 1.1 from Olaf Lenz
    • pbcwrap: version 1.3 from Olaf Lenz
    • multiseq: new versionof multiseq
    • New "measure gofr" which calculates the unnormalized histogram of g(r) for two selections.
    • biocore: handle the case where a user logs out of biocore from another method (webpage or another control panel). now it converts the chat window to a login screen when that occurs.
    • added cgtools plugin to the build
    • added placeholder docs for cgtools plugin
    • Added makefile for cgtools
    • fix pkgIndex.tcl for cgtools plugin
    • Initial cgtools import
    • Added a new "VMDSHEARSTEREO" environment variable to enable the new perspective projection code that uses the shear matrix rather than eye rotation to generate stereo views. Orthographic projections are still done using eye rotation.
    • message about logins/logouts/open control panels/closing of control panels is now displayed (to the AUTP, assuming the user is in it, or to another project if not)
    • biocore: changes some GETs to POSTs in biocore.tcl Made biocorechat so that it handles private messages to and from the user, handles emotes, and handles private messages from the server.
    • biocore: fixed a bug in verify that was causing the procedure to not hit the server because [clock clicks -millisecond] had started returning a negative value. Now it handles that.
    • TkCon: Changed TkCon window name to ".tkcon" (used to be ".", which is a really bad thing!) Removed the "exec wish" call since it does not seem necessary
    • cranked version number
  • VMD 1.8.5b1 (July 17, 2006)
    • save label text offset to saved state file.
    • Added Justin's patch for atom label X/Y window offsets, and atom label format substitutions for resname/resid/charge/etc.
    • Added a display command for offsetting text relative to X/Y window coordinates
    • Added JString::to_camel() conversion to camel case
    • Updated win32 installer
    • Updated the Win32 build of VMD to link against VRPN 7.03 now, in order to support the new HDAPI haptic interface.
    • Updated win32 installer with latest plugins etc.
    • molefacture: Gave amino acids proper handedness (...oops)
    • molefacture: Disable undo, which no longer works due to some more important feature additions
    • tweaked biocore.tcl to fix a typo in listener registration which was causing control panels opened via api/vmd to not register themselves properly with the server. Now they are getting registered, which means that other people running real control panels can now see that a user has opened a control panel, even if via vmd. This has a side effect of also causing older messages to automatically be sent to the control panel, just like it does for java. So, the code invokation to manually retrieve these older messages has been commented out, which seems to speed up startup by a couple of seconds at least.
    • cranked version number
  • VMD 1.8.5a9 (July 13, 2006)
    • don't draw the closing sphere on the ribbon-end of the nucleotide cylinder, as the rep looks much nicer without them most of the time.
    • fixed a window creation problem that was occuring iff a biocore session file existed, but had bad data in it.
    • revised biocore::verify so that it only hits the server at most once per second. The 'redraw' proc in biocorepubsync was calling verify, and caused the redraw call to occur 39 times every time the pub sync window opens (once per widget, maybe? haven't investigated..) This change causes the window to appear quite a bit faster since it saves about 38 hits to the biocore server. This 'redraw' issue needs to be revisited at some point in the future when we have more time.
    • Added a special right-eye X-axis reflection stereo mode needed to support the new variety of LCD panel-based Immersadesks developed by Tom Defanti's people up at UIC.
    • biocorechat: Don't send empty chat messages.
    • molefacture: Updated docs to add new building features
    • pmepot: Expanded docs.
    • re-enabled the snapshot feature using the new component of the api that allows you to attach files to notebook messages. Now, the tcl code just writes a special tag into the saved state that says to 'put a link to the file here' when needed. Tested on both linux and windows. more testing is really needed and will be done over the next few days.
    • Limit notebook chats to VMD messages; also scroll to the bottom after adding a VMD message.
    • updated to version 1.22 of biocore.tcl
    • autopsf: updated docs
    • Scroll to the bottom of a list when a new chat arrives.
    • biocorechat: vmd end of fix for the equals sign. the biocore server needs to be updated in conjunction with the release of this code to the wild.
    • namdenergy: Fixed writing of xsc file
    • Added free energy perturbation info to mutator plugin docs.
    • Added a cache to the getUserName call. Should greatly reduce the number of calls to the server and speed up the user experience as well. Added functionality to the API to allow a file to be attached to a notebook message. Next step is to have VMDs pub/sync use that to attach snapshots to the notebook messages themselves. New getRequiredJobParameters routine that returns a list of field names that are needed to run a job of this particular type with the specified account.
    • biocorechat: Commented out extra "puts" statement and changed a call to "new_error_window" to a tk_messageBox.
    • molefacture: Fixed a geometry calculation bug
    • applied Warren Delano's patch for TRJ reading to the Gromacs plugin.
    • Added Jerome's mutator error/warning patch
    • autopsf: Minor gui fix
    • updated mutator docs with new text and image from Marcos
    • molefacture: Couple small bug fixes
    • molefacture: added some fragments to the "Replace hydrogen with *" menu
    • molefacture: New fragments
    • Added simple 'makemovie' proc stub which can be used to build a movie from a script in text mode.
    • molefacture: Fixed the bonds (I mean it this time)
    • pretty up configure file list
    • molefacture: Fixed bond/angle confusion upon deleting atoms
    • molefacture: Fixed a bit of weirdness creating new molecules, including the bug pointed out by jhenin
    • added ring data structure used for carbohydrate rendering
    • molefacture: A decent set of base fragments. Now I need to do the editing fragments.
    • cranked version number
  • VMD 1.8.5a8 (July 6, 2006)
    • Fixed parameter order in the call to find_subfragments() for the proteins.
    • Added a new "measure dipole" command for calculating the dipole moment of an atom selection.
    • Added first rev of ring finding structure analysis code by Simon Cross and Michelle Kuttel. This is the first step towards doing something better with carbohydrates and polysaccharide structures.
    • paratool: Code optimizations (brackets around expr expressions). New code to compute force constants from Hessian. Before it was using the diagonal elements of the internal Hessian thus omitting any couplings. Since many coordinates (in natual iternal coord basis) are linear depedent, e.g. the angles in a trigonal planar or a tetrahedral moiety (you cant change one angle without changing others, too) we are better off honouring the couplings between these coordinates. Thus our potential is lot closer to the scanned potential.
    • qmtool: Some changes so that paratool can use the new method for computing force constants and target potentials.
    • Updated the backbone atom classification code to find nucleic acid terminal residues, reduced the minimum nucleic backbone atom count required for residue classification down to 4 rather than 8, and added several additional tests for nucleic acid terminal atoms H5T and H3T. Together with updates to the nucleic ribbon drawing code, DNA/RNA should look much better now. We still need better handling of various modified bases and non-standard residues, which can probably be done with additional logic.
    • readcharmmtop: Fixed bad ACEH entry.
    • Make the nucleic acid ribbons code do some extra checks on the terminal residues, and use the O1P/O2P from the following residue if necessary.
    • Added "backbonetype" and "residuetype" atom selection and 'get' keywords for use in internal testing of BaseMolecule::analyze() and its many helper routines.
    • fix typo in build.csh
    • Disable test of selection of the nucleotide cylinder start atom, so that we can continue to use "backbone" and "not backbone" to select the ribbon backbone and the nucleotide cylinder now that the C3 atoms are used as the spline control points for the ribbon.
    • Oops, can't add anything beyond expected nucleic backbone atom names or it throws things off, we may have to add another stage of testing in order to solve issues with nonstandard residues etc.
    • Added hybrid topology file to the distrib target
    • Topology file containing hybrid amino acid residues (all non-standard)
    • moved the hybrid topology file to the readcharmmtop plugin where the other topology files live.
    • Got rid of a nasty hydrogen addition bug
    • Added additional BaseMolecule tests for nucleic atom names: O4*, O4', C2*, C2', O2*, O2', C1*, C1'
    • Added Marcos' new mutator gui with the FEP stuff.
    • Added Justin's patch to simplify the OpenGL header inclusion ifdefs on MacOS X.
    • Added Justin's NameList patch to eliminate warnings about not returning a value from a non-void function.
    • biocorechat: Fixed some window naming errors, added a check for no network connection in redraw.
    • mutator: Added Jerome's alchemify shared library to the mutator build, and got things bubble gummed and taped together for the first rev. Need to fix up libalchemify to load as a separate package etc like we do for autopsf, fix the internal module name, etc.
    • mutator: added makefile to update /w swig when necessary
    • added Jerome Henin's alchemify source files to the mutator plugin
    • Fix mutator plugin to work with filenames containing spaces
    • namdenergy: minor documentation fix
    • psfgen: Added "(duplicate?)" hint to segment error message.
    • Changed some window naming conventions.
    • biocorechat: tweaks to get \, {, and } working as they should for chat messages that are sent.
    • Fixed biocorechat XML parsing
    • fixed psfcheck package loading to follow the standard convention, as pointed out by Justin.
    • Fixed psfcheck so it doesn't have a main() by default, as pointed out by Justin.
    • Fixed VMD TkCon to set it's directory in the same way we do for the other plugins, rather than hard coding it.
    • Fixed a small issue where computing a ligand map over one frame without using -allframes left some inf's in the map. This is not very consequential since ligand maps should not be computed that way anyways. Also added some comments.
    • doc: Added a few more details about the volmap ligand stuff
    • Fixed molefacture_gui to be able to have no selection. Minor updates to protein building routines. Several bugfixes to get rid of bad assumptions on the existence of certain atoms. Fixed leucine and proline. Properly allocate new hydrogens when needed (ie, when you've added more than 100 atoms at a time). Made bond/angle list updates more frequent to fix a number of potential problems. Added an "add from sequence" section to the protein builder. Yes Virginia, we still need the fragment libraries. Soon.
    • Protein builder is a lot more fleshed out now. Still need to add fragment libraries and fix the dummy hydrogen issue
    • Fixed protein secondary structure
    • Updated win32 installer
    • only reset the view when the first frame is loaded.
    • Updated the volumetric axes to track the axes color category items rather than using hard coded colors. Updated DrawMolItem to regenerate reps when the axes, and display fore/background are updated as well.
    • volmapgui: fix typo in variable name pointed out by Thomas Huber
    • added display foreground to the save state script
    • Added "Foreground" color item to the "Display" color category
    • Added 16 more colors to the VMD fixed named color table The new colors are in a color wheel arrangement so it's easier to find complementary colors. We may want to consider further color table reorganizations to make the whole thing work better.
    • eliminated out of date comments
    • Added missing files to the molefacture fragment library
    • cranked version number
  • VMD 1.8.5a7 (June 28, 2006)
    • exectool: Fixed a problem with determining the bicore project id given only the project name
    • biocorechat: commented out a puts
    • Fixed X index loop control bug in VolMap::smooth(), reported by Thomas Huber.
    • biocorechat: fixed the once-per-second updating. Moved the 1000 back to 300000 and uncommented the fileevent line that monitors the socket.
    • biocorechat: Added error message if the user tries to send a chat with no internet connection.
    • biocore: handled some boundary conditions that were showing up on multiple window openings and invalid users
    • added a menu bar.
    • Added Paul's navfly plugin next to Jordi's and added (Keys) and (Mouse) to differentiate them.
    • biocore: added another button to the 'utilities' menu for biocore. 'Information' just returns simple stuff about you.. your id, your name, your login time, the biocore server.. More can be added as we think of new things to put there
    • navfly: added docs
    • molefacture: Fixed up the fragments a bit
    • molefacture: Generalized amino acid addition works now
    • biocore chat: Added a dialog window and close the chat window if the user logs out of BioCoRE after opening the chat window.
    • biocore: Some changes to help with publishing from a link in the Chat window.
    • biocore: User registration now allows the user to register a full user (as well as the partial user, which was possible before). The BioCoRE server will be restarted this afternoon to add support for this capability.
    • Added extra macros so VMD can be built against both the old and new revs of VRPN.
    • molefacture: Fragment files for protein builder
    • molefacture: Added the very first version of the protein builder. Still has bugs, and still a lot of features to add.
    • Fixed problem in VDW-1-4 parameter support.
    • paratool: When parameters of selected internal coordinates are edited manually then the changes are now committed to all equivalent internal coordinates.
    • autopsf: Added interface for dealing with patches Fixed a minor bug with chain deletions
    • autopsf: Fixed problem pointed out by Peter. It was not due to disulfide bonds but due to multiple frame PDBs.
    • namdenergy: Added PME and management of periodic boundary conditions
    • fix a bug with greater than's in chat parsing. Had to move the regex's around a little bit.
    • Updated win32 installer for VMD 1.8.5
    • started updating docs for 1.8.5 release..
    • updated announce mesg in prep for upcoming release
    • Updated the Windows registry key used for VMD 1.8.5
    • Changed the behavior of the nucleotide cylinder code and the nucleic ribbon drawing code to use the C3' or C3* atoms as the ribbon spline control points, which makes the nucleotide cylinders always match up with the ribbon, with the added bonus that coloring by residue/resname/etc visually makes more sense now. Still need to sort out a few issues with drawing the terminal residues, but half of that problem actually lies in the BaseMolecule class.
    • Corrected some inconsistent package version numbers reported by Justin.
    • Fix package tk callback proc names as pointed out by Justin.
    • Revised redraw(), fixed various other things that should hopefully alleviate the problems others were having with the chat window not opening correctly.
    • added more comments to nucleotide code
    • make the nucleotide thickness ratio more sensible.
    • Added pick points for the nucleotide cylinders, and cleaned up a few bits of the code.
  • VMD 1.8.5a6 (June 23, 2006)
    • biocore: user registration now includes a checkbox for 'Automatically join the VMD Public project'. Assuming the prospective user leaves the box checked, they will be automatically added to the vmd public project upon completing the registration form. Since the vmd public project is specific to the biocore server here at ks.uiuc.edu, the value passed in via biocorelogin would need to be changed or removed for other biocore servers. This shouldn't be much of an issue in the near term.
    • biocoreutil: minor tweak to text box to make it non editable
    • Changed DrawMolecule::append_frame() to check for bond updates whenever at least one frame exist. This also required updating MolFilePlugin::read_optional_structure() to avoid re-analyzing the structure except when additional bonds are added.
    • fixed typo in example of user-defined movie procedures
    • Added TODO item for the atom selection error that can occur if a noob tries to run autoimd on a molecule with no coordinates (e.g. load a PSF with no PDB or DCD.)
    • changed the help url to about.html#utilities
    • tweaked the description field to use a text area instead of a label. Gives more space.
    • Added in the ability to join projects. From the utilities menu you can choose the 'join project' button, and you will be shown a list of projects that you are eligible to join (public projects and private ones to which you have been invited). You can choose one of them, click 'join' and you are in it.
    • Updated populate_user_list to take a projectId parameter, don't display names in AUTP, display chats from other Control Panels in the proper chat window, not just the one that is active.
    • Added deregisterListener{} proc, modified getUsersLoggedIn{} to take a projectId parameter, modified getNotebookEntryListFull to take an optional projectId parameter.
    • Added old changes.
    • added the biocore utilities plugin to the extensions menu
    • Fixed a typo in the 'pyrimidine' atom selection macro (URA resname typo..)
    • biocoreutil: tweaks to get the window opening working as it should. The login window now pops up properly after adding the deiconify and lets the user reattempt a connection. Eventually I think I'll add a way for biocorelogin to call a callback which can automatically redraw the invoking window once the login is correct.
    • molefacture: Fixed namespace silliness (fixme from previous commit) Changed makefile to properly distribute fragment libraries
    • don't let bignum export/import * into the global namespace
    • added a biocore utilities file. The utilities window currently has a single option (change password), but more will be added (join project, etc). For change password, give it the old and a couple of new, and it changes the password on the server. I still need to do some tweaks for the option to work right if the user hasn't logged in yet. (works right the first time, window doesn't redraw subsequent times.. the user won't typically see it multiple times, though) Updated Makefile, pkgIndex to reflect the new file. Need a line to add the entry to the plugins menu. I'm using: install_extension biocoreutil biocoreutil "BioCoRE/Utilities"
    • latest version of biocore api
    • added a 'forgot your password' link that will reset the password to a random value which is then sent via email to the user.
    • ClipTool: The plane-selector controls are very ugly and unusable in this version of Tk. Updated them to look half-decent.
    • Cranked version number
  • VMD 1.8.5a5 (June 19, 2006)
    • fixed molefacture make distrib target, which was missing new source files
    • biocorechat: Updated "pack" commands for window resizing.
    • biocorechat: Reversed order of recent chats (most recent on the bottom).
    • Updated all but a couple of the remaining plugins to use the more readable registration style for the plugin entry points. This will also make them insensitive to future changes in the order of entry points in the structures, since the values are initialized explicitly by member name rather than implicitly by order.
    • Updated 11 more plugins to use the more readable style for initializing the plugin entry points.
    • biocorechat: Changed MessageBoardMessage to NotebookMessage in parseListenerMessages{}.
    • multiplot: fixed typo pointed out by Jerome Henin
    • biocorechat: Added initial version of adding saved states to the window. You can click on them, and they should load the state. Also bound "Return" in the chat entry to submit the message.
    • Changed "::biocore::BiofsDirectoryEntry configure..." calls in BiofsDirectoryEntry_scanFile to "$self", which is what scanDir does, and seems to work better.
    • Added in the "new molecule from base fragment" feature TODO: Gather some real fragment libraries, and fix the stupid kludge on fragment file names
    • molefacture: Added the ability to perform a "replace hydrogen with fragment" operation. Now we just need to build up the database.
    • molefacture: added methyl, benzene fragments for use in builder
    • molefacture: Reorganized molefacture to make it somewhat maintainable Also, improved autotyping
    • exectool: Removed many diagnostic prints. Changed file transfer behavior so that if no output files are specified, it just says the transfer is done, rather than indicating an error. Changed project selection so that if an id is specified and valid, use that. Otherwise, if a name is specified and valid, use that. If neither is specified, run the text-based selector. No checking is done to insure that the Name and ID match; the ID has priority. Project name is no longer cached. It is requested shortly before use, so if the project is renamed, it will be correct. Added calls to get the project list, get the account list, and to get a project id, given the project name. Project name from id is easily done by turning the project list into an array and referencing it by id. Same for account id to name.
    • silence iRetVal mesgs in the chat window
    • biocorechat: Changed update time from 10 seconds to 1 second.
    • Silence all of the biocore chat status messages that an end-user isn't going to understand. Some of these are probably going to make it into the GUI (biofs notificatons?) but until they're in there, it's best not to spam the user's console with them.
    • Wrapped some variables in quotes to protect against spaces. Actually I think most of these are extraneous, but its working, so I'll leave them in
    • Eliminated chatty debug messages in ::biocore::getUsersLoggedIn
    • silenced debug messages from the chat plugin
    • added Mike to the cvs commit mesgs
    • cleaned up biocorechat mesgs
    • Added Robert to the commit mail logs
    • I added checks to make sure the specified output files can be written in remjob_config_output_file, remjob_config_stdout_file, and remjob_config_stderr_file, before the job is run, and also in remjob_get_file, after the file is run.
    • Misplaced comment caused an error in remjob_start_transfer.
    • Previously, if a user re-used a job name, it created an exception which got reported on the server. I tried to implement nicer checking for this in the biocore api, but some functionality it depends on is not currently working. So, I instead added a 4-digit number to the end of the file staging directory on BioCoRE, so the server won't complain. Note that on the 1 in 10000 chance that there is a duplication, the server will still complain but everything will work as far as exectool is concerned.
    • Added functions that produce text prompts for account and project info for biocore if the user passes null values into remjob_config_account_biocore
    • Re-populate user list when switching projects.
    • Added the biocore chat to the package index
    • added biocorechat.tcl to the makefile, deleted the old vmdchat file
    • customized the search box to search only the plugins subtree of the web site
    • updated name of the chat extension in the plugin registration code
    • Renamed vmdchat.tcl to biocorechat.tcl, and changed references to "VmdChat" to "BiocoreChat".
    • Make symbol_data::convert() much more bulletproof, even when NULL pointers get passed for string data.
    • Added biocore chat plugin to the Extensions menu
    • pretty
    • Force initialization of string pointers to NULL in the ParseTree code.
    • Added first revision of the new remote job submission code that uses the BioCoRE APIs to define a job, define files to be staged, define commands to be run, stage input files, run the jobs, poll job status, and stage output files back, etc. This first version is the bare minimum, and some functionality exists as stub routines that don't really do anything yet. Future revs will revise the APIs slightly, fill in some stub routines, and implement more of the needed interactive GUI functionality.
    • Added getUsersLoggedIn() to get a list of the users logged into a project. This is the function that could be revised to use funtionality already in the API. Also made some modifications to getNotebookEntryListFull.
    • Revised opening the login window.
    • Revised open_new_login_window to handle opening from a different parent window.
    • Initial version of vmdchat.tcl. It will require a biocore restart until some methods are revised.
    • -BioCoRE plugin now allows logging in without a Control Panel. You can login and share states with collaborators without ever hitting a webpage. Rev'd VMVERSION to 1.1. Will leave it here until release of this version of VMD. Updated API to newest version, which uses more proper packaging. Everything is actually in ::biocore:: now. Added new features. User can login straight from VMD, as well as create a new guest account. Reduced functionality: saving a snapshot image has been temporarily disabled until a security hole can be closed (before VMD release, most likely) related to users being able to arbitrarily add any files they want to server space on the BioCoRE server. There are a few boundary conditions that aren't yet being handled. Most of them deal with users who login to multiple BioCoRE servers. This basically affects the BioCoRE devs who have their own dev servers. There are probably other currently untested routes into the windows that might cause problems. Additional testing is underway. Todo before VMD release: forgot password change password join project (VMD public project, most likely)
    • updated the navfly plugin with Paul's latest code.
    • updated movie maker docs, moving example script into a separate file.
    • Added the simplest movie making example as a separate file.
    • updated docs to link to example movie script
    • Added a simple example of a user-defined movie script which fades in a material and rotates a structure.
    • membrane: Fix for the W100 bug noted by Axel, which caused a clash between the segnames and the name of a segment in the template. Output will now be unchanged; I twiddled the name of the segments in the template.
    • qmtool: A few minor improvements.
    • paratool: All sorts of little updates and changes. Added new file paratool_hessian.tcl which contains a routine that converts internal hessians much better into force constants (including treatment of coupling constants). This is now just in CVS and will be merged into paratool later.
    • Prevent subsequently loaded PDB (and similar files) with CONECT records specifying only partial bonding information for special bonds from replacing previously determined fully specified connectivity information gleaned from other files, or automatic bond search. The "mol addfile" command now honors the "filebonds" flag in the same way that "mol new" does.
    • intersurf: fix info printing
    • Raise maximum bond count to 12 so that face centered cubic lattices of gold and other metals used in Alek's nano-bio structures don't encounter problems.
    • Added bignum package. Clean target clears out the compile directory every time.
    • Added bignum makefile and package index.
    • added pure Tcl version of the 'bignum' arbitrary precision arithmentic package
    • mac builds are going back to cupertino, while sydney gets reinstalled
    • updated mol2 plugin docs
    • Volmap docs: fixed small typo regarding the volmap ligand "-dihomo" option (which was wrongly stated as -dimono)
    • plugin test main: updated call to read_bonds() to pass NULL for the new optional bond order parameter.
    • namdenergy: Added in ability to use periodic cell and PME. This is command line only and undocumented; both of those shortcomings will be fixed shortly
    • Added highlighting of selected chain, and some better water handling
    • xyzplugin: Fixed a minor bug that caused xyz files to fail to load when driven by CatDCD.
    • catdcd: Added dependency on libmolfile_plugin.a/.lib so that catdcd gets rebuilt anytime plugins are re-linked into the static lib.
    • Cranked version number
  • VMD 1.8.5a3 (May 24, 2006)
    • Documented the volmap -checkpoint option Fixed a lot of typos throughout the tcl-commands docs
    • Added Justin's isovalue control range patch, though there's some GUI work to be done here still...
    • VolMap: Added a "-checkpoint " options, which will write out intermediate volmaps every frames for long trajectories. The default is every 500 frames. A value of zero means no checkpoints.
    • multiseq: fix colorbar makefile This plugin is probably greatly superceded by the newest "colorscalebar" plugin in the main VMD plugin tree, so we should consolidate the two together and eliminate the redundancy here.
    • multiseq: Fix find syntax the "-or" flag isn't portable, but "-o" is.
    • multiseq: fixed return type for main() in the cluster plugin
    • Fixed faulty reading of atom coordinates when using molecular symmetries other than C1. This fixes the problems reported by "mashaojienuaa" on 05/21/06 10:12 pm.
    • Updated rmsdtt with Luis' latest version which fixes a minor bug.
    • Applied Justin's patch to improve isosurface extraction performance, with a few additional minor changes. The ResizeArray template doesn't have a 'resize' method presently, so that bit was left out.
    • pqrplugin: fix up various lameness and eliminate bogus error messages.
    • pqrplugin: Added spaces between atom coords and added an END record, as suggested by Justin.
    • paratool: Fixed deg2rad scaling problem in potential fitting.
    • qmtool: Don't generate a new frame when coordinates are modified in the ModRedundant section. Replace current frame instead. Leads to the correct number of frameswhen reading logfiles.
    • paratool: Combed out another few bugs related to potential scanning.
    • paratool: Fixed atom scrambling bug in refinement, too.
    • paratool: Fixed a bug that would scramble atom order in angle definitions of scanned internal coordinates.
    • namdgui: Fixed bug that prevented proper unitcell handling.
    • qmtool: Major overhaul of potential scanning code. Rigid and relaxed potential scanning supported. TODO: Plotting of the scanned potential (can be taken from paratool).
    • paratool: Major overhaul of the QM potential scanning feature. Supports rigid and relaxed potential scanning now. cleaned up code. Lots of small bugfixes.
    • VolMap: ARghh! the "average" combination mode was _still_ broken due to a missing "break". Now the bug is fixed, tested, and done with for good!
    • paratool: Working towards better and faster refinement
    • Cranked version number
  • VMD 1.8.5a2 (May 9, 2006)
    • changed measure_sumweights() to use double precision floating point for the accumulator, since a surprising amount of error can accumulate for million atom selections otherwise. We may need to scrutinize the other measure commands for similar weaknesses.
    • add handling of unrecognized nucleotide residue names, and draw them if we can find atoms for both endpoints.
    • Use the sphere array primitive more efficiently when drawing nucleotide cylinders.
    • cap nucleotide cylinders with spheres
    • Scale the nucleotide cylinder lengths so they meet up more closely with the ribbon.
    • cleaned up nucleotide cylinder drawing code some more
    • nucleotide cylinders now use C3' and C3* atoms rather than phosphates by default, and they'll fall back to C1'/C1* or P atoms if necessary.
    • Added recognition of URA residues, and single-character residue names for nucleic acids used in the nucleotide cylinder drawing code.
    • paratool: Working towards proper Urey-Bradley term handling in potscan and refinement.
    • paratool: Reorganized code.
    • paratool: Changed order of parameters in zmat
    • readcharmmpar: Added some comment into the code.
    • qmtool: Changed order of parameters in zmat.
    • paratool: Added a new file for refinement
    • Remove hard-coded package require of resptool
    • Fixed unchecked package require of resptool.
    • Check atom selection state for nucleotide cylinders before drawing
    • updated with more links to STL format info
    • continue prettying up the molfile plugin main page
    • Added .xmol to recognized file types for xyz plugin
    • doc: Added namdserver and optimization plugins to the plugin page
    • qmtool: updated documentation
    • resptool: Using Findertool to find the binary. Fixed typo preventing proper closing of window.
    • resptool: Fixed circular dependecy with paratool.
    • paratool: Changed a few function names
    • paratool: Fixed bug preventing proper loading of QM single point calculations.
    • qmtool: Added comments at the beginning of qmtool.tcl Removed debugging output.
    • paratool: Renamed paratool_charges to paratool_respcharges.tcl
    • paratool: Renamed paratool_editcoor.tcl to paratool_intcoor.tcl
    • updated X-PLOR and CNS doc links
    • updated links to charmm, namd, xplor docs
    • Added '.spi' to recognized extensions for the spider plugin
    • Added '.sit' to recognized filetypes for situs plugin
    • Added '.cub' to recognized file extensions for Guassian cube file reader
    • Added '.cns' and '.xplor' to recognized file extensions for X-PLOR map reader
    • Added '.omap' extension to dsn6 plugin
    • Updated versions and information about various plugins, and improved web formatting slightly.
    • Accept .dsn6 as another recognized file extension in addition to .ds6
    • resptool: got rid of .paratool references
    • paratool: Added fuction to find equivalent internal coordinates/parameters. Symmetrizing of parameters is now based on the current selection of internal coordinates.
    • paratool: Renamed variable segrestrans to tmsegrestrans.
    • paratool: Enclosed argument of 'expr' in braces for performance.
    • namdenergy: Fixed a few bugs in the servermode code. Added flag -highprec for sending high precision energies from the callback through the socket.
    • namdserver by default reads energies from the stdout log that was piped back into VMD. Now you can also specify 'variable highprec 1' in order to use NAMD's callback mechanism which provides energies with higher precision. When the next NAMD build is out with the fixed callback function then this will become the default.
    • Fix memory leak in atomsel_moveby(), reported by Marcos
    • Pretty up atomselect_move() and atomselect_moveby() before fixing moveby memory leak
    • Avoid merging residues with different segnames, as we already do for residues with different chains
    • Cranked version number
  • VMD 1.8.5a1 (May 5, 2006)
    • fix namdgui package index
    • Added paratool_energies.tcl to the distrib target
    • updated colorscalebar version
    • fix pkgIndex.tcl to set NAMDGUIDIR prior to sourcing the main script since the main script then uses the environment variable when sourcing the namdgui_tclforces.tcl script.
    • Enable the experimental DNA nucleotide rendering code as part of the NewCartoon representation to start getting user feedback and find/fix bugs.
    • added graphics picking demo script
    • added lubos' color_coord script.
    • added Lubos' hbond-based residue coloring script
    • Added Dan's kiosk script to the script library
    • Updated parm7 plugin to use current flags
    • Added new safety checks to test whether plugins are correctly setting the optflags optional data flags parameter when their read_structure() entry points are called. The dlpoly history reader plugin wasn't doing this, and it was causing random crashes since VMD would get all kinds of bogus flags and invalid data as a result. It had gone unnoticed because VMD wasn't initializing the optflags parameter before sending it into the plugins, and didn't do any sort of sanity check on the returned flags. The new code sets the optflags parameter to a special MOLFILE_BADOPTIONS sentinel value which should never happen otherwise. If the parameter is still set that way on return, the file load is aborted and an error message is printed indicating that the plugin failed to set the optional data flags correctly. In general we can't trust plugins to do the right thing, we have to double check on everything they do since they are typically less rigorously tested than the main VMD code.
    • Added a new MOLFILE_BADOPTIONS flag which should never be returned by a molfile plugin. Before calling the read_structure() plugin entry point, the optflags parameter is initialized to MOLFILE_BADOPTIONS. On return, if the optflags parameter is still set to MOLFILE_BADOPTIONS, then the caller should abort any use of the returned data, since the plugin failed to intialize optflags correctly.
    • eliminated unnecessary data from dlpoly struct, and added proper initialization of the optflags parameter which was previously unset.
    • cranked version of dlpoly plugin
    • updated dlpoly plugin to use explicit initializers
    • Added comment marking likely cause of the weird behavior when loading many PDBs into VMD by scripts, as a means of creating a trajectory. (symptom is that autobonds are dropped once subsequent PDBs with CONECT records are loaded). Also added a comment in the code where we'll likely want to add the calls to read in angles/dihedrals/impropers/cross-terms, etc.
    • Added rough draft implementation of new MolFilePlugin methods for loading QM data and writing volumetric data. Lots to be done still..
    • Further improvements to the color scale bar plugin to make it more bulletproof and to use a selector for the text color since that's easier for people to use, despite being slightly less flexible than an arbitrary color index.
    • Added Alex Saladino's text color patch for the color scale bar plugin
    • doc: fix out of date text regarding behavior of $sel frame for out-of-range values.
    • VolMap: Added a missing line of code which caused all volmap maps that used frame averaging (i.e. the "-combine avg" option) to be completely wrong. This bug only affected "avg".
    • Added navfly to plugin distrib target
    • Applied Justin's patch which adds a new "pickpoint" graphics/draw command, and a corresponding vmd_pick_graphics Tcl callback. We'll need to add the necessary docs, Python bindings, etc still.
    • Added the navfly plugin developed by Paul McCreary with help from John
    • paratool: paratool_energies.tcl now contains all the code responsible for retrieving energies from namdserver.
    • optimization: Improved simulated annealing.
    • paratool: Improved potential scanning and refinement code. Added refinement GUI.
    • paratool: Formatting of floats is now done only for the displayed internal coordinate listbox and not for the underlying data lists.
    • started eliminating old static initializers in molfile plugins and replacing them with explicit initializers which are not sensitive to the order of structure members, etc.
    • Added in tests for new read write routines for angles, volumetric
    • Added test for new qm related routines
    • improved the text describing the qm timestep reading behavior
    • Added new read_timestep() routine with QM support, which will eventually supercede the old read_next_timestep() routine.
    • Implemented some of the code needed to move the data read-in by the plugin into the data structures now provided by VMD. Currently completely untested.
    • molfile plugin interface headers: added comment that the per-timestep QM stuff is going to need to be done along with atom coordinates in order to keep our sanity. We'll just add parameters to read/write timestep entry points. But before we do that I'm going to clean things up to make this change simpler.
    • paratool: added support for plugins namdserver and optimization. Other minor updates and bugfixes.
    • fix minor issues with new plugins
    • plugin makefiles: Added optimization and namdserver plugins
    • Updated QM data structures with more fixes.
    • namdenergy: some modifications necessary for NAMDserver.
    • Yet another useful plugin: NAMDserver Provides an interface for starting NAMD in "server mode". NAMD is started externally and its output is piped back into VMD such that the energy output can be utilized. You can generate a NAMD input file that contains code such that NAMD starts up, opens a socket and goes into waiting mode (event-loop, waiting for socket to be connected). The NMADserver plugin polls the server and as soon as it's up it will connect a client to it. In the following every line that the client sends to the server will be evaluated in NAMD as a command (TCL and NAMD commands). The results will be sent back to the client through the socket. Thus you can interactively control NAMD. More documentation included...
    • namdgui: Fixed failure to find par files if there are spaces in the readcharmmpar path
    • added docs for optimzation plugin
    • A new plugin for multidimensional optimization! You can choose between the downhill simplex method or simulated annealing. It's all TCL and thus slow, but I needed it for Paratool and the slowest part is the function evaluation anyway.
    • autopsf: Fixed problems with water sometimes causing an explosion in the number of segments
    • fix a few typos in the new QM data structures
    • updated plugin header copyright dates
    • eliminated some redundant information from the QM data structures and split the general QM info into per-trajectory timestep items, and constant items.
    • First draft of new qm related data structures and plugin entry points, along with a volumetric data writng callback, and angle/dihed/imprp/cmap read/write entry points.
    • enabled the ribbon width modulation hack so that users can play with this feature until there're proper GUI and text interfaces for it.
    • doc: Fixed some latex transgressions in several of the big keyword tables where text was overflowing the table boundary
    • dlpolyplugin: Applied Anti Liivat's bug fix for an unchecked ts pointer in the periodic cell reading code.
    • updated delaunay script for new web site formatting.
    • updated links to Hull code with patched version by Axel Kohlmeyer
    • fix cliptool so it doesn't freak out when calculating minmax of molecules that don't have atoms
    • readcharmmpar: Corrected error in an error message;-)
    • paratool: Fixed problems from previous 2 commits
    • paratool: Fixed problem caused by a bad commit
    • cranked version
  • VMD 1.8.4 final release (April 16, 2006)

Please email any questions to vmd@ks.uiuc.edu.