Posted by Bandersnatch on 04/04/2020 01:57:58:
Posted by John Baron on 03/04/2020 09:49:47:
If you look through the file you will see that it contains code sequences that are read and executed.
…
Yet multiple virus scanners find nothing wrong. Doesn't say much for the effectiveness of the scanners does it? And presumably the same scanners wouldn't find any problems with any other pdf's that we all frequently download.
…
A couple of inferences too far maybe?
Most security vulnerabilities have limited lives – once identified PDF readers, and other software, are updated to remove the threat. Once the software is fixed there's no point in reporting old problems!
Also, many, perhaps most, vulnerabilities depend on configuration. For example, it would be safe to open a PDF containing iffy Javascript if the reader didn't support Javascript. Similarly many vulnerabilities are blocked by the operating system by managing permissions.
AVMs know about this stuff. So a scanner or PDF reader might say to itself, 'I know about this dodgy code but it doesn't matter because this computer has up-to-date software and the environment is secure; as it's safe to display the content, I'll do without making a fuss'.
An important point about investigating security issues is the whole configuration matters. It's unwise to draw conclusions from the big picture. Version numbers and individual security settings matter enormously, making it difficult to compare your machine with mine. Windows-10 as updated 3 days ago has thousands of under the bonnet changes compared with the version released in 2015. Many of the differences are security fixes and improvements. Being unable to simply say 'Fred and I are identical because we both run Windows-10, Firefox, and McAffee', makes diagnosis tricky because version numbers, file permissions, extensions and registry settings matter too. Don't jump to conclusions too quickly!
I think 'Presumably the same scanners wouldn't find any problems with any other pdf's that we all frequently download', is surely too broad. Scanners find the problems they know about. Missing some is inevitable, but it doesn't mean all the others of the same type will be ignored too.
Dave
PS John's embedded example says: 'Thisiscode' in ASCII.