Running ACI/Oceanlog on Linux/Macos via Wine

[bperrybap]

Doug,
Would it be possible to steer the PC s/w development team
to start considering updating/modifying
their Windows PC s/w code so that it can run under Wine?

This would allow it to run on Linux and MacOS using wine.
I believe that it would offer the quickest & easiest way
for Pelagic to pick up support for those operating systems rather
than continue to leave this as a void in their product offerings.


To give you an idea of where things currently are with respect to running
under wine.
Currently:

AUD 2.0.5 (which is based on the 1.x code) seems
to load install and run ok and be able to display dives from the sample diver logbook.
There are a few font issues in a few places
but nothing major in the GUI. I can't actually try a download as for some
reason, even on native windows version it doesn't seem to work
with the AtmosAI.


ACI 1.0.7 will install and run under wine - same as AUD.
You can even bring up the "Personal Logs" and view the Elite dives of the
Sample Diver.
I haven't successfully gotten a download to work yet but I haven't really
tried yet. I am assuming that once the wine registry is fudged up properly
and the FTDI driver is up and working, that downloads will also work.


ACI 2.0.4 will install and run under wine on Ubuntu. However, some of the
initial database API calls fail so it fails to open the logbook.

ACI 2.2.4 will not install as it is using some API calls that are not yet supported
by wine.

==================================================

My guess is that it wouldn't be too difficult to either get wine updated
to support some of the needed calls or modify ACI/Oceanlog to do things differently
so that the windows PC install package could install and run under wine.


--- bill

 

[AndyDragon]

The problem will not be with the software itself, likely the install issue is some bizarre Vista-specific bug. The problem is the driver support for the cables.

 

[bperrybap]

I'm not worried about FTDI USB drivers. FTDI drivers are available
and given the simple needs of the download protocol, I don't think getting
serial data flowing along with a few control signals will be difficult.

My first attempt will be to install the virtual comm driver and fudge up the pseudo
registry to try to get ACI/Oceanlog to open the comm port directly.

If that fails,
I'm thinking of writing a simple wrapper application that creates a PTY then fudges up
the pseudo registry to trick ACI/OceanLog into opening the PTY instead of any
sort of "real" comm port. That way the wrapper application/daemon can do whatever
is necessary behind the scenes and hide the actual real interface.

The wrapper application can also fire up wine to bring up ACI/OceanLog as well.

So my big concern is installing the s/w and accessing the database.
If ACI/Oceanlog won't install or come up because they can't access the database,
it doesn't matter if the serial data portion is functioning.

For the time being I'll be able to debug everything using my downloader application
then test it with ACI 1.x. I think it should all work ok,
but eventually I'd like to get ACI/Oceanlog 2.x working and that
will more than likely require some minor changes to the application itself,
which requires support from Pelagic.

--- bill

 

[bperrybap]

So a bit of an update.
It is good news mostly.....

The stock FTDI driver in Ubuntu 8.10 can talk to the Pelagic USB data cables,
once you patch in a new udev entry for the proper USB vendor/product ID.
This is extremely easy to do as it is done by simply copying a
pre-made .rules file to the proper location.
(This can even be automated with a simple post install script)


So I have successfully fully downloaded an Aeris Atmos AI on linux
using my downloader application.

Getting the existing windows OceanLog/ACI code to run under wine is not possible without
a modification to wine itself due to the way the Pelagic code uses the registry
to locate the USB data cable.

The current version of wine does not populate the windows registry DEVICEMAP
with any entrys in the SERIALCOMM hive. There is currently wine work going on that
will populate the SERIALCOMM hive with real serial ports but there will need to
be an additional modification for virtual comm ports that use a USB interface.
Not difficult code to write but would require updating wine itself.


While this could easily be done and would allow AUD/ACI/OceanLog 1.x s/w
to run, it won't allow the newer 2.x s/w to run under wine.

Older 2.x will install but aborts when trying to access the database.
Newer/latest 2.x release won't install under wine.

Given that ACI/OceanLog 2.x needs a few modifications to work under wine anyway,
it might be easier to make some simple additional modifications to the
ACI/OceanLog code to allow an alternate method of port location that is
compatible with the existing wine.

In the mean time, I'll create a patch for wine that will populate
the VCP0 value that the ACI 1.x s/w and potentially other windows s/w
is looking for which should allow 1.x s/w to work under wine.
(I'll also see if this can be put into the main line wine trunk so that all wine users
can pick it up as part of the normal wine updates)

Doug, any thoughts on tweaking the ACI/OceanLog 2.x code to allow
it to run under wine?

--- bill