Bug 140418

Summary:	Automatically abort stalled / timed out DCC receive transfers
Product:	[Applications] konversation	Reporter:	Mook <mook.moz+sites.org.kde.bugs>
Component:	dcc	Assignee:	Konversation Bugs <konversation-bugs-null>
Status:	RESOLVED INTENTIONAL
Severity:	wishlist	CC:	b.buschinski, hein
Priority:	NOR
Version First Reported In:	unspecified
Target Milestone:	---
Platform:	openSUSE
OS:	Linux
Latest Commit:		Version Fixed/Implemented In:
Sentry Crash Report:

Description Mook 2007-01-22 02:50:09 UTC

Version:           1.0.1 (openSUSE-10.2/1.0.1-23) (using KDE KDE 3.5.5)
Installed from:    SuSE RPMs
OS:                Linux

When a DCC receive stalls, it just gets to 0 kB/s but doesn't abort for me.  This means that (having auto-resume set) the other party cannot just attempt a re-send to continue without having me manually click Abort first.

I would like to have some option of having a timeout such that, should no data arrive for that interval, the DCC receive is aborted automatically.  This way a retry can be initiated from the other end without me being around to look at the UI.

(Currently, Konversation appears to just queue the transfer...)

Comment 1 Bernd Buschinski 2012-04-21 14:07:27 UTC

This bug is really special, for normal situation, like loosing the connection, or not getting a connection at all, the dcc receive/send does get a timeout -> it gets aborted. But in some very special network cases, which are basically broken networks, we still a tcp-alive package, but never get any data.

As this only happens in some broken networks, I will not add an extra traffic watcher, as normally tcp would tell us if the connection is dead.