Project

General

Profile

Feature #8849

Bug #7552: Firefox extension to automatically verify the ISO checksum

Bug #6851: Fix ISO verification using checksum

Technical specifications for ISO verification extension

Added by sajolida over 4 years ago. Updated almost 4 years ago.

Status:
Resolved
Priority:
Elevated
Assignee:
-
Category:
Installation
Target version:
-
Start date:
01/28/2015
Due date:
% Done:

100%

QA Check:
Feature Branch:
Type of work:
Contributors documentation
Blueprint:
Starter:
Affected tool:
ISO Verification Extension

Subtasks

Feature #8850: Investigate feasibility of OpenPGP signature verification in JavaScriptRejected

Feature #8816: Investigate upload mechanisms and reactivity for browser app storesResolved

Feature #8855: Design data source for ISO verification extensionResolvedsajolida

Feature #8873: Decide which kind of verification would the ISO verification extension doResolved

Feature #9028: Check whether Tor Browser disables automatic updatesResolvedsajolida

Feature #9043: Check whether BitTorrent clients do proper hash verificationResolved

History

#1 Updated by sajolida about 4 years ago

  • Affected tool set to ISO Verification Extension

#2 Updated by sajolida about 4 years ago

  • Subject changed from Technical specifications for ISO verification extension to Security specifications for ISO verification extension

#3 Updated by sajolida about 4 years ago

  • Subject changed from Security specifications for ISO verification extension to Technical specifications for ISO verification extension

#4 Updated by intrigeri about 4 years ago

Might it be that Subresource Integrity, once supported by common browsers (no idea if that's the case yet) could replace the extension, and be a much cheaper way to get what we want?

#6 Updated by sajolida about 4 years ago

I read through the Chrome discussion. Current implementation status is only for <script> tag and secure origins. It apparently raised quite a hot debate about its utility in the first place (hTDUpMk_TV8) and was implemented to "to help answer some of the open questions in the WG" (355467). So I don't think that it's shipped in any official version.

Furthermore, what we want to do is not to validate a JavaScript script but 1GB of data downloaded by the user on their hard disk. Such a scenario (downloading something by the user) was never really discussed (566083003) and I bet that it would rise much different and more complex UX issues, for example about how to feedback that information to the user, whether it is validated or not, etc.

My conclusion from that is that, at least on the Chrome front, people are considering this firstly and mostly for sub-resource validation and we are very far from seeing whether this could help us, and when. Even if the W3C use case on "downloads" looks like what we want to do.

For the time being, the most widely available SRI technology seems to be this one: http://www.carbuyer.co.uk/reviews/vauxhall/corsa/sri.

#7 Updated by tchou about 4 years ago

Same feelings from the Mozilla front. The main contributor made a slideshare few weeks ago, oriented on SRI for Javascript : http://fr.slideshare.net/fmarier/integrity-protection-for-thirdparty-javascript

They made a website to promote SRI, and it's quite JS oriented too : https://srihash.org

But maybe we could ask them that's on the roadmap.

#8 Updated by intrigeri about 4 years ago

Furthermore, what we want to do is not to validate a JavaScript script but 1GB of
data downloaded by the user on their hard disk. Such a scenario (downloading
something by the user) was never really discussed (566083003) and I bet that it would
rise much different and more complex UX issues, for example about how to feedback
that information to the user, whether it is validated or not, etc.

Sure.

My conclusion from that is that, at least on the Chrome front, people are considering
this firstly and mostly for sub-resource validation and we are very far from seeing
whether this could help us, and when. Even if the W3C use case on "downloads" looks
like what we want to do.

OK, thanks for the research!

Shall I ask relevant people at Mozilla and Google about their plans and timeline for supporting the "download" use case?

#9 Updated by sajolida about 4 years ago

  • Target version set to Tails_1.4

#10 Updated by sajolida about 4 years ago

Sure, that's #9160 now.

#11 Updated by sajolida about 4 years ago

  • Target version deleted (Tails_1.4)

I'll send an email to Maone tomorrow with the detailed blueprint. So we're likely not going to have more precise specs before 1.4.

#12 Updated by sajolida almost 4 years ago

  • Status changed from Confirmed to Resolved
  • Assignee deleted (sajolida)

Maone is fine with the current blueprint and we'll refine the specs through the HTML code.

Also available in: Atom PDF