Hello,
The latest releases (FormatConverter 1.0.14 and Analysis 1.0.11) have just
gone 'live'. (There was a last-minute hitch yesterday to do with
backwards compatiblity code for DNA data, but that should be ok now.)
These are available in the usual place:
http://www.ccpn.ac.uk/downloads/downloads.html
The installCode.py script has been updated further to hopefully cope
better with 64-bit machines, but as always, this is a complicated script
so something might well be wrong somewhere or other (so let me know, as
always, and I'll email patches).
A list of changes we could remember is given at the bottom. Because of the
slight delay to the release, Tim managed to get in one extra feature this
morning (not listed below), namely that pulldown menus now pack into
multiple columns if there are more than 20 items on the list, so this
should fix the problem that people had when the pulldown menus were longer
than the screen.
The most important structural change in Analysis is that the 'update'
feature now allows you to update from an old release to the current
release. What this does is download from the ftp site, and install the
code using your existing environment.txt file (which is hopefully still
valid, otherwise this procedure will fail). This should make it much
easier to upgrade to new releases. (Otherwise the 'update' facility
remains as now, i.e. you can get patches only if you have the current
release.)
Because none of you have this new update feature in your existing code
base, I'm attaching a gzipped tar file with this code in it. Unpack this
code in your existing top-level CCPN directory, the one that is probably
called ccpnmr (unless you renamed it) and contains the sub-directory
ccpnmr1.0. (So the files in the gzipped tar file begin
ccpnmr1.0/python/...) Then run the update in your existing release and
(hopefully) it will notice that there is a new release and install it for
you. We have tested this a bit here but of course since this is
complicated and new something or other might go wrong (if so, just upgrade
the old-fashioned way, but let us know what went wrong).
The next release will involve some structural changes to the data model
itself. This means that the incremental patch system will fail at that
point (unless we post a heck of a lot of files, which would get rather
tricky to manage). So my guess is that you will get incremental patches
for the next week or two for the obvious bugs you find straight away, and
then there will be a blackout period for a couple of weeks while we
upgrade the code with the changes, then we will probably do another
release soon after that. That one might prove to be buggier than normal
because of those structural changes, so be warned in advance. (And there
will be further structural changes in the next few months to the API.)
Wayne
**************************************************************
Changes from previous release:
Added browseAtomsMacro - Goes in for the 'b' key as a system macro.
Allow ambiguous peakDim anotations e.g. "(4Ala/6Ala)CA([123]/[456])"
if there are two contribs
Added function to find constraints given selected peaks in
ConstraintBrowser.
Added option to remove resonances from peaks after deassigning
resonance inResonanceBrowser.
Merge botton now does Merge/Split depending on selection, in
ResonanceBrowser.
Added in NOE dist function manipulation for constraint generation
including network anchoring.
Added storage of preferred NOE intensity type by spectrum.
Aria1.* export uses constraint serial rather than peak serial.
Testing shift matching distance constraints now lists peaks with poor
merit
Added getNoeDistance() in ConstraintBasic to do the d = k / r^6
Added attempt to resolve prochirals (not ensembles yet) for violation
estimation
CalcAmbigDistConstraints now works with 4D and higher NOESYs. Bonded
dims no longer passed-in - now deduced from experiment type.
Fix for loading structures with imperfect residue mappings, e.g.
smaller -> larger num residues.
Added atom type check (from BMRB data) when assigning resonances.
First version of Network anchoring module. Not for anonymous
assignments at present.
Added options and functions to find symmetry related peaks.
First versions of assignment report and quality control apparatus.
Added CSI v2 file reading.
Prochiral distance measurements now non zero.
Changed default display of strip midpoints and numbers to on.
Added "2 to 4 residues" option in AssignmentGraph.
Added option to read NmrView data files (courtesy of Gary Thompson of
the University of Leeds) and Bruker data files.
Changed orthogonal dimension j/k shortcuts so that they work on
strip where cursor is rather than on active one.
Using ctrl drag middle mouse button you can now zoom into a specified
region.
Put extra padding of half a point in orthogonal dimensions for peak
viewing and selection.
Fixed bug in rates analysis error estimates (some variables were not
being initialised to 0).
In Rates Analysis popup add option to use bootstrap method for
calculation of the TC error estimate (the default) or to use the previous
(pre 1.0.9 Analysis) method.
|