[Bug 47592] HFSS antenna modelling software running but erroneous output !

[wine-bugs at winehq.org]

https://bugs.winehq.org/show_bug.cgi?id=47592

--- Comment #9 from Jean-Michel Friedt <friedtj at free.fr> ---
Thank you for your reply. Indeed I was not clear about the 3D radiation
pattern issue which I try to describe better below:

> 1. one or more calculations done by the program produces slightly inaccurate
> results under Wine,
> 
> 2. the radiation pattern plot doesn't appear.
> 
> (1) is tricky, because Wine does little to no math itself, instead relying
> on the host C library. Presumably the code is calling into some C library,
> and there are some floating point inaccuracies? The first thing I'd try is
> to replace that C library with the native version to see if that is indeed
> the problem, but first we need to know which one it is. Can you attach a log
> when running the program (and doing whatever actions lead to this inaccuracy
> being visible) with WINEDEBUG=+loaddll?

Please find attached screenlog_nodebug.0 the output of running HFSS simulation
with NO debug information (just program output with mostly genuine messages but
a few Wine warning it seems) and screenlog_withdebug.0 the full output with
loaddll option enabled, from starting the software to displaying the result.
The
log file is annotated with 3 breakpoints manually inserted and prefixed with
## JMF:
1/ loading dynamic libraries all the way to displaying an empty graphical
user interface at "JMF: END OF OPENING SOFTWARE GUI". Then I File -> Open
the model which hardly leads to additional messages up to "JMF: END OF LOADING
MODEL". I then run the simulation all the way to "JMF: END OF COMPUTATION". I 
then plot the reflection coefficient (the curve that is not wrong but slightly
different from the MS-Windows output) all the way to "JMF: END OF PLOTTING 2D
S11". Finally I compute the 3D radiation pattern which does provide a graphical
window but no relevant content (see below) all the way to the end of the log.
I stopped the log before quitting HFSS.

> (2) I may need more information for. Is this potentially a graphical
> failure? e.g. is the plot blank, or are there axes or something similar but
> no data? Screenshots may be helpful here.

My mistake in describing the issue: the graphical interface opens fine and
displays a chart axis and labels, but no curve. My understand is that the lack
of curve is due to erroneous computations which this time are not "slightly" 
wrong but so wrong that the displayed parameters are all 0. Attached is the
antenna parameters (the output which would be plotted on the 3D graph) computed
from MS-Windows (radiation_report_MSWindows.png) with reasonable values
(radiation efficiency close to 1, gain close to 1.8 dB) as opposed to the
Wine outptu (radiation_report_Wine.png) which shows an efficiency of 0 and
a gain of 0 (stating that the antenna is radiating no energy at all, hence
the inexistent radiation pattern in the plot). The same information is provided
as exportparams*.txt file although I don't think the detailed content is
relevant to the bug
report.

Thank you, Jean-Michel

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.