Bug 584870

Summary:	x11-terms/rxvt-unicode - enable/disable PRINTPIPE at compile time
Product:	Gentoo Linux	Reporter:	Renato Alves <simpledark>
Component:	Current packages	Assignee:	Jeroen Roovers (RETIRED) <jer>
Status:	RESOLVED UPSTREAM
Severity:	normal	CC:	wired
Priority:	Normal
Version:	unspecified
Hardware:	All
OS:	Linux
URL:	http://software.schmorp.de/pkg/rxvt-unicode.html
Whiteboard:
Package list:		Runtime testing required:	---
Attachments:	Exposes --enable-printpipe through ./configure

Description Renato Alves 2016-06-02 16:25:33 UTC

Created attachment 436212 [details, diff]
Exposes --enable-printpipe through ./configure

A couple of months ago I've started using rxvt-unicode after many years using xterm.

Today someone dropped a couple of pages on my desk containing random bits of output from one of my terminals. This person was able to trace the output back to me as among the output you could read my username. You can imagine my reaction...

Long story short, over the past weeks I have been sending multiple pages worth of text to our local printer. I was unaware as I rarely print documents.

It didn't take long to figure out how this happened.
https://wiki.gentoo.org/wiki/Rxvt-unicode#Printing

I have used rxvt in the past without problems, but on this keyboard the PrintScreen key is not in the usual location and accidentally hitting it is not so uncommon.

Attached is a patch to rxvt-unicode that adds a ./configure option (--(dis|en)able-printpipe) that should in turn allow customizing this behavior via USE flags.

Since this could be considered a security concern (there is no feedback that content was sent to lpr) I would suggest that the "printpipe" flag is disabled by default.

Comment 1 Jeroen Roovers (RETIRED) gentoo-dev

2016-06-06 07:49:38 UTC

We distribute rxvt-unicode, but we don't develop new features. Please talk to upstream if you want to have that added.

Comment 2 Renato Alves 2016-06-06 08:15:40 UTC

(In reply to Jeroen Roovers from comment #1)
> We distribute rxvt-unicode, but we don't develop new features. Please talk
> to upstream if you want to have that added.

Hi Jeroen,

I did talk with upstream. The patch was rejected on the grounds that it didn't add any new feature and that the behavior introduced by the patch could already be achieved by undefining PRINTPIPE in `features.h`. The recommendation was quite literally "gentoo developers can already use `sed` to perform the change if necessary".

In other words, disabling PRINTPIPE is already considered a feature.
The patch exposes this feature via configure making it easily accessible to portage via the "enable/disable" functionality.

If the patch is deemed unnecessary, I can provide a modified ebuild that implements the upstream recommended 'sed' and adds the printpipe use flag.

Is this reasonable?