Known Bugs

home page

This page cites all the bugs that the author is aware of in all the Gepasi 3.xx versions. The version that contains a fix is indicated (alpha and beta versions available from author on request). Bugs marked severe can affect results of simulations (contact the author for further details).

Most recent bug is on top.

Description Severe Date Versions affected Version with fix
The calculation of Jacobians of systems with mass conservation is wrong. yes November 2004 3.00-3.30 none yet
The predefined Allosteric inhibition kinetic function is not being properly calculated, and the help file is also wrong. yes November 2001 3.00-3.21 3.30
When compartments have non-unit volumes Gepasi fails to calculate several kinetics properly. yes November 2000 3.20-3.21 3.30
Gepasi crashes if an optimisation or fit are attempted without adjustable parameters. no July 2000 3.20-3.21 3.30
The predefined Ping Pong Bi Bi kinetic function is not being properly calculated. yes 25/9/1998 3.00-3.20 3.21
The differential equations with mass action steps were incorrectly set up when the volumes involved were not unity yes June 1998 3.10-3.20 3.21
Gepasi allows empty user-defined functions to be entered. These cause the program to crash when running. no 26/3/1998 3.00-3.10 3.20
If any volume of a compartment was not unity, the differential equations were not built correctly. yes 29/8/1997 3.00-3.03 3.10
An error in the file saving routine caused loss of precision on numbers smaller than 1e-16. no 29/8/1997 3.00-3.03 3.10
Perturbations on concentrations of a member of a conservation relation and their effect were not correctly calculated. yes 29/8/1997 3.00-3.03 3.10
When parameters are linked to user-functions (rather than other parameters) Gepasi does not calculate the correct value for the linked parameters. yes 4/06/1997 3.00-3.03 none yet
Gepasi crashes when some reactions are added to models with a specific configuration of branches. no 18/03/1997 3.00-3.02 3.03
Gepasi crashes if any of the output files are opened in exclusive mode by other programs (eg Excel) when the simulation runs. no 20/01/1997 3.00-3.01 3.02
Under Windows 95 and Windows NT 4.0 (but not 3.51) the list boxes do not have a horizontal track. no 10/01/1997 3.00-3.02 none yet
Divisions by zero are not trapped by the Intel executable. This may later on result in error -3 in the integration routine (LSODA). This is only an incomplete error message (Gepasi should report a "division by zero"), not a bug. no 05/01/1997 3.00-3.01 3.02
Help for Ordered Uni Bi kinetic type is not available. no 10/12/1996 3.00 3.01
The user-defined kinetic type parser points all substrates to the first substrate and all products to the first product; the results of any model with user-defined kinetic types with more than one substrate/product are wrong. yes 10/12/1996 3.00 3.01
If the steady state and time course filenames are equal the program crashes. no 10/12/1996 3.00 3.01

Last modified: November 16, 2004 Copyright © 1996-2004 Pedro Mendes
Gepasi - biochemical simulation