Strip metadata in uploaded documents

Hi,

while it is nice to see how my two PirateBoxes are filled with data, I'm not happy with the metadata of the uploaded documents. Taking a random look over the documents most of them are filled with unneeded and potentially deanonyimisind data. Who saved the file, name of the machine the file was created and so on.

Do you see a way to implement "MAT: Metadata Anonymisation Toolkit" to the PirateBox image, running on a regular base over the uploaded documents or even right after a successfull upload. Perhaps not over all filetypes but for example PDF, pictures, Zip's and office documents ans stripping away the informatons which are a risk for privacy?

You are talking about that: [mat.boum.org] , aren't you?

Mh looking at the dependencies, yes on OpenWRT it should be possible... but as usual I'm only suggesting looking at the dependencies written down here: [gitweb.torproject.org]
AND I have no Idea how resource intensive this kind of media-data-parsing is... that can be a no-go issue

Do you travel with your box?
Is it an OpenWRT?

Ah, sorry, forget to add the link to mat.boum.org. Yes, OpenWRT. Precisely, an TP-Link MR3020.

And yes, I often carry it with me, thanks to a good battery pack. But the main purpose is to use it on Cryptopartys around FFM or other similar events. Traveling with Deutsche Bahn on the tracks FFM <-> Cologne and FFM <-> Hamburg is also funny with a running PirateBox

Happy to hear that I'm not the only one from FFM-region running around with PBx..

I totally understand your motivation... you can have a try but that is not easy.

When you install all packages, you have to install everything to the usb stick. For opkg, run with -d command

  opkg -d piratebox <package>

Then you may do some additional modifications... no idea what exactly

Ah, I see. I will try to get a bit time next weekend. If the election posters where on place.

There are some situations in which you don't want metadata stripped.

I'd suggest implimenting it with a tick-box on the upload page for 'strip metadata.' Ticked by default.

> Do you see a way to implement "MAT: Metadata
> Anonymisation Toolkit" to the PirateBox image,
> running on a regular base over the uploaded
> documents or even right after a successfull
> upload. Perhaps not over all filetypes but for
> example PDF, pictures, Zip's and office documents
> ans stripping away the informatons which are a
> risk for privacy?

Do you see a way to implement Metadata on a regular base over the uploaded documents or even right after a successful upload??

edit by Matthias: removed ads



Edited 2 time(s). Last edit at 10/15/2013 08:02PM by Matthias.

Stripping the metadata from a file isn't that resource intsensive. It only looks to a small portion of the file that is being stripped.
and the fields it looks for is also rather limitted in the case of MAT (but enough in most cases).