Release Notes for Gifa 5.2 ®

Main Novelties
- Large speed-up of the Inverse Laplace Transform code used in the processing of DOSY. Has been optimized for Velocity Engine® on Macintosh® G4 and G5. Processing time can now be nearly 10 times faster.

- redesigned user interface for the  processing of DOSY experiments

- enhance quality for the Macintosh version, using now the new JVM 1.4.1 from Apple (or higher)

- speed-up of the graphic display.

- New batch mode, permitting to run Gifa in a graphic-less environment



Release Notes for the 5.2005 technical release

New functionalties
New import filters:

- a new tool permits to import files issued by the SIMPSON program (available from the import dataset tool)
- a new tool permits to import text files with leading lines and several columns (available from the import dataset tool)
- the import from Jeol tool now handles real and integer generic files
Other:
- a new tool to compute buckets values, as used in metabonomics studies (available in the tools menu)
- clean up of the batch mode

Bug Corrections
In the Import commands
- import from Jeol was buggy in certain cases (when a different machines was used to run delta and Gifa), import was also accelerated
- import Varian was crashing Gifa in certain special cases, due to a buffer overflow in readv.

Elsewhere :
- when looking to vertical columns, the unit was displayed in a wrong unit
- the Store/Load macro utility was not working in the DOSY F2 processing tool
- Recompilation on MacOsX brings more stability and speed-up
- The display is reset after Fourier Transform if a FID was previously displayed


For the macro programer
- the comand int1d used for 1D integration has a global calibration problem, it now retains the initial data scale
- the new macro put_bruker_add.g permits to write new parameters into the Bruker parameter files
- the macros antislash2slash slash2antislash split and splitx have been improved and debugged
- the macro getdirfromname has been corrected
- the macro profile had a potentiel bug with long macro filename
- the fprintf macro was not functional
- the command READ_RAW in mode real8 swab was buggy, it now correctly swaps 8 bytes instead of 4
- the macro readv.g simulates the READV kernel command


Release Notes for the 5.2004 technical release

Bug Corrections
In DOSY mode :
- cleared a memory leak that caused Gifa to crash on long run (all plateforms)
- corrected the DOSY set-up, so now F2 processing and DOSY processing can be performed in two different runs.
- baseline points are now correctly reseted from one correction to the next
- DFACTOR calibration is saved into the 2D file, for later use
- DOSY processing can have a final size smaller than acquisition size
- spoil PFG (P19) is now interpreted in Bruker pulseprog
- molecular weight calculator is not calibrated on water anymore

Elsewhere :
- Recompilation on MacOsX corrected for random crashes - fully tested on 10.3
- batch mode corrected on MacOsX
- added lists for prefered sizes in the preferences
- filenames are now refreshed in processing dialog boxes when opening a new file.
- improved the export menu
- added a "run self-test" entry in the Option menu
- corrected default name in reading peak file

For the macro programer
- added a new "warning" command equivalent to  cprint 7
- added macros : slash2antislash getdirfromname


Release Notes for the 5.2 initial release
Bug Corrections
- estimate of time in progress bars
- improvement of the Bruker® import procecudure

For the macro programer
- printf frpintf sprintf commands have been implemented as macros
- the python() function can be used to call a python program