As noted in the other review this should solve the problem we're having.

Thanks, I think that should do the trick.

Thanks for the feedback. You're right, I didn't think about in flight packets from the previous connection. I see also RFC 6191 references it as a potential problem. (And apparently the ISN generation isn't as random as I for some reason though.)

I think something like this should work if using TCP timestamps instead. Would this be preferable?

In D56307#1288534, @tuexen wrote:

I would like to understand why this change is proposed?

The output looks god to me.

I think you can do something as simple as this to get an extra white space it the plain text output, since it's outside the "uptime-human" format string it won't be included in the encoded outputs.

The json and xml output looks good to me, but in the plain text output it's missing a space between "up" and the time.

Add braces after if/else where the following expression is split over multiple lines

Fix typo in LICENSE_FILE variable name.

Update format of diff and bump port revison.

Just ignore my previous question, it's because of libxml2. Looks good to me.