Re: [Bug 756317] Re: Captive portals may corrupt apt translation indices

2014-03-25 Thread Bryan
Tim Hortons' captive portals break apt. Do they have Tim Horton's coffee shop in your neck of the woods? Bryan Harris, PE Research Engineer Structures and Materials Evaluation Group University of Dayton Research Institute bryan.har...@udri.udayton.edu http://www.udri.udayton.edu/ (937) 229-5561

Re: [Bug 756317] Re: Captive portals may corrupt apt translation indices

2014-03-23 Thread Bryan
The problem is that most people don't use apt at all. Their computer goes out to see if there are updates and breaks silently. It happens without user knowledge or intervention. On Mar 23, 2014 6:01 AM, "Monsta" <756...@bugs.launchpad.net> wrote: > Would be really nice if you'd fixed it before U

Re: [Bug 756317] Re: Captive portals may corrupt apt translation indices

2014-03-21 Thread Bryan
I see that there are http and https methods for downloading items. What if we just removed http download of files and defaulted to https? This would neatly guarantee that a captive portal (either broken or malicious) cannot get away with pretending to be a download server. Assuming https is impl

Re: [Bug 756317] Re: Captive portals may corrupt apt translation indices

2014-03-21 Thread Bryan
I'm willing to spend some time working on a patch. Does anybody have an idea where in the code a syntax check should go? Bryan Harris, PE Research Engineer Structures and Materials Evaluation Group University of Dayton Research Institute bryan.har...@udri.udayton.edu http://www.udri.udayton.edu/

Re: [Bug 756317] Re: Captive portals may corrupt apt translation indices

2014-03-21 Thread TJ
On 21/03/14 16:41, Monsta wrote: > It's problematic for one simple reason: APT ceases to function normally > until you remove the offending files and update the package index again. > Simply, it is a Denial of Service. Quite a neat way to create a window of opportunity for so-called zero- day vu