does not report incompatibility between deb and host architectures (32vs64)

[Sandeep]

Linux menzoberranzan 2.6.31-11-generic #36-Ubuntu SMP Fri Sep 25 06:37:23 UTC 2009 x86_64 GNU/Linux

Distributor ID: Ubuntu
Description: Ubuntu 9.10
Release: 9.10
Codename: karmic

gdebi 0.5.9

take a i386 .deb package and double click on it in an AMD64 host.

The error reported is :
Could not open <something>
The package might be corrupted or you are not allowed to open the file. Check the permissions of the file.

while "sudo dpkg -i" gives :
dpkg: error processing <something>.deb (--install):
 package architecture (i386) does not match system (amd64)

the error message from gdebi is highly misleading and incorrect.

[Sandeep (sandys-gmail)]

I have some more information on this - the deb I had gotten was corrupted. However, the fact still remains that cmdline "dpkg" was able to tell me incorrect architecture and gdebi was not.

The issue can be duplicated by taking a good i386 .deb and corrupting it (by opening it in an editor and changing some values). dpkg will still report incompatibility between architectures, while gdebi thinks it's a permissions related issue.

Ideally, I would appreciate some corruption detection - however at the very least to be identical to dpkg's error reporting.

[esscrow (esscrow)]

I have the same bug in ubuntu remix (karmic koala 9.10). dpkg -i does not install .deb package giving back information about errors. However, gdebi install such package and it does not work correctly.

[Michael Vogt (mvo)]

Thanks for your bugreport.

I just tried this on my amd64 machine and it looks like I do not get this error:

$ uname -m
x86_64
$ gdebi /tmp/4g8_1.0-3_i386.deb
Reading package lists: Done
Reading state information: Done
Reading state information: Done
Reading state information: Done
This package is uninstallable
Wrong architecture 'i386'

Same same for gdebi-gtk and double-clicking. Could you please provide me with
steps how to reproduce?

[Sandeep (sandys-gmail)]

@Michael - the issue is not reproducible any more. Probably behavior got changed, for the better, along the way.

I guess it should be fixed for @esscrow as well.

Please feel free to close the bug.

Thanks!

-Sandeep

[Luca Falavigna (dktrkranz)]

Closing as per user request, feel free to reopen if problem persist, thanks!