Hi Josch,

Thanks for taking the time to give that explanation.

In my testing, I observed exactly what you said - using '-d' sets the
Distribution field in the .changes file.  Not using '-d' causes the
Distribution to be populated with the distribution value from the
d/changelog entry

But I feel like there's still an issue, or I'm still not understanding
something correctly.

When sbuild calls lintian, it passes the filename of the .changes file.

If my understanding is correct, using or not using '-d' has already
affected the contents of the .changes file at that point.  Then linitan
runs against that .changes file, and gets all of its information from

So shouldn't running lintian outside of sbuilt using the *exact same*
.changes file produce the same output?

I would think yes, unless-

1) the .changes file passed to lintian is different than the .changes
file written to the log and left on disk after the build


2) the output from lintian is somehow altered/modified by sbuild.

Neither of those seem to be the case, but I'm not 100% sure.  However,
either of those could explain the discrepancy.

> Do you see any way sbuild could improve to be less confusing?

I think most of the docs are pretty clear.  But I guess that depends on if
there's really an issue as I suspect, or if I'm just greatly
misunderstanding thing.

I guess I'll let you know once we figure out which one :-)

Best regards,

Reply via email to