Hacker News new | past | comments | ask | show | jobs | submit login
A document that the NSA uses for teaching Python (archive.org)
91 points by enthd on Feb 16, 2020 | hide | past | favorite | 22 comments



Is this any better than other resources or just mildly interesting because it was made by the NSA?


Probably the latter. Anything that reveals how the NSA and other agencies operate is interesting.


I think the real question is - why does NSA make their own in-house training manuals for programming languages? Seems likely to get outdated fast. Do they have internal tutorials for Ruby, TCL, Lisp?


I adore governments' classification system which can state things are classified as 'unclassified', and by extension to hijack the meaning of 'classified' to mean 'top secret'.

On a topic relevant note, from the original twitter feed [0] which announced/released this (as part of FOIA) a helpful person also mentioned a lightweight download of the ?same document[1] (18MB vs 118MB).

[0] https://twitter.com/chris_swenson/status/1225836060938125313...

[1] https://deeb.io/wrdprs/wp-content/uploads/2020/02/comp3321_r...

Also, I notice in the top left corner there's a handy little Doc ID: 6689691. It'd be interesting if if it were possible to FOIA all those Doc ID's.


Top Secret is only one of several levels of classification.

You really need to mark things as unclassified in the government because other people need to be able to answer basic questions like:

- what computer am I allowed to view/edit/save/transmit this on?

- can I talk to my coworkers about this, and if so, which ones?

- what do I do with printed copies of this document?

Answering these questions wrong can land you in serious trouble, including jail time. So even as a courtesy to your peers, you want to mark them (U).


That's apparent to the guy you're responding to. He's pointing out that the so-called "unclassified" is actually the class "non-secret". A truly "unclassified" object would not have gone through the process of classification. It would not have a class.

Of course in English this is fairly common. The object of price zero is said to have no price. Yet clearly it has a price, it is just zero.

I have to say that's an amusing observation and sort of reveals the type errors that occur when we model real world things in type systems.


True, but there are documents that have information that is at multiple levels of sensitivity and/or classification. When you work with such information, having such granularity of classification is a good thing (no matter how awkward it looks).


The United Kingdom's previous system for marking school exams (GCSE) had a marking scheme from "A*/A" to "F", and then "U". Where "U" was Ungraded. Clearly it was graded, but as a "U".


What are they using now? I went through the system not that long ago when it was still A* to U.



Did they have E?


Don't forget U//FOUO.


It's not that great. You can find much better training materials for beginners all over the place.


Interesting, I don't disagree, but what better alternative would you suggest?


Here's the first place I send folks looking to learn Python:

http://greenteapress.com/thinkpython2/html/index.html


Automate Boring Stuff with Python is a great book to start


That book is an introduction to Python libraries, not programming and barely even automation.


Page 4 - the NSA uses Gitlab.



There’s probably nothing to this, but when I try to screenshot the pdf (https://ia802805.us.archive.org/7/items/comp3321/comp3321.pd...) on iOS Safari, my phone locks up for a moment with a black screen and spinner and then I have to unlock the phone. Doesn’t happen on other PDFs on archive.org or after I’ve saved it to iBooks. Can anyone replicate? Is this perhaps the result of some byzantine configuration state in the PDF spec that Apple has chosen to implement or that their implementation is choking on?


>my phone locks up for a moment with a black screen and spinner and then I have to unlock the phone

AFAIK that happens when springboard crashes. Maybe there's a 0day exploit in there?


When taking a screenshot in safari on iOS 13 it also tries to save a pdf of the entire page. If you open the little screenshot overlay in the corner there is an option to save as the full page as pdf instead of a screenshot.

Maybe the pdf is just too big to be re-rendered and saved as a new pdf and it crashes somehow.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: