3 ms·
>when Google reverse-engineered the hack used against American diplomats in Uganda, they found an elegant, tiny piece of code that adapted software from 1990s X
by fouc 5y ago
>when Google reverse-engineered the hack used against American diplomats in Uganda, they found an elegant, tiny piece of code that adapted software from 1990s Xerox machines to fit a so-called Turing machine — essentially a complete computer — into a single GIF file.
LOL at describing PDFs as "adapted software from 1990s Xerox machines"
https://googleprojectzero.blogspot.com/2021/12/a-deep-dive-into-nso-zero-click.html https://googleprojectzero.blogspot.com/2021/12/a-deep-dive-i...
- oasisbob 5y agoHaving spent some time with OCR and scanning recently, I'd have to agree. A lot of news articles are describing JBIG2 as something archaic, when it seems to be as relevant and commonplace as ever. (see MRC, for a modern application)
- ciabattabread 5y agoThe article was originally written for Financial Times, whose audience will have a general knowledge of computing. Also, what the heck is MRC?
- klodolph 5y agoMixed raster content, I assume. Basically, you scan a paper document. It contains a mixture of line art, text, and photographs—“mixed” types of content. You can segment it and use different codecs to encode these different segments, and then combine the results in a PDF.
- oasisbob 5y agoNews articles, plural. Eg, here is another story by way of Ars: > Essentially, 1990's algorithms used in photocopying and scanning compression are still lurking in modern communication software, with all of the flaws and baggage that come with them. https://arstechnica.com/information-technology/2021/12/google-warns-that-nso-hacking-is-on-par-with-elite-nation-state-spies/ https://arstechnica.com/information-technology/2021/12/googl... Given than JBIG2 wasn't standardized until 2000, and that the Apple vulnerability was in their implementation (vs the standard) I just don't understand this characterization.
- _jal 5y agoThey were talking about JBIG2, not PDF. JBIG2 has been in the news periodically for a different sort of problem - you can't trust it to accurately represent what was scanned: https://www.theregister.com/2013/08/06/xerox_copier_flaw_means_dodgy_numbers_and_dangerous_designs/ https://www.theregister.com/2013/08/06/xerox_copier_flaw_mea...
- whizzter 5y agoThe NSO exploit started by pretending to send a GIF, that was sent down to some decoder that did automatic file-type detection based on data rather than filetype and "correctly" detected an PDF, the Apple PDF decoder in turn supports JBIG2 images where the actual exploit lives.