first commit

This commit is contained in:
libroot 2026-11-19 04:39:11 +02:00
commit 0d90363116
Signed by: libroot
GPG key ID: 2AC68C029EBE9ACE
6510 changed files with 679146 additions and 0 deletions

View file

@ -0,0 +1,65 @@
DYNAMIC PAGE -- HIGHEST POSSIBLE CLASSIFICATION IS
TOP SECRET // SI / TK // REL TO USA AUS CAN GBR NZL
SIDtoday
(U) Raul Responds
FROM: Raul
A DNI Analyst
Run Date: 01/13/2005
(U) Ooooh, weeeeee!!! Dog my cats and corn my pone! My last article obviously hit a few
nerves. So I'll provide a little feedback to the feedback .
(U) Even as old and ancient as Old Raul is, he can still learn a few things. First, in all future
articles I'll be absolutely sure to specify things so as not to allow folks to read into what my
point is. Also, I have to be careful in my choice of words.
(U) In my last article I used the word "considered" meaning a well thought out response by both
the technical and managerial sides of the house to the PDF processing problem. I did not simply
mean whether anyone, while eating a hamburger or chewing a stick of gum, had happened to
consider any of the issues I mentioned in the article. Obviously many folks, including myself,
had. Whether anyone had "considered" the issue(s) at all was not my point. Mea culpa. Having
said that, here we go.
(U) Before we start skinning the chipmunks, let me spell out a few of the requirements one, or
at least Raul, would surely think should have gone into a "considered" solution. Here they are:
1. There are two areas to be addressed: processing for selection and filtering and
processing for display. If we can come up with one solution adequate for both, fine. If
not, then we'll have to have multiple solutions.
2. In both cases above, the processes have to be capable of handling damaged data. No
exception here.
3. Whatever the solution is, it has to be readily available to the analyst (i.e. available via
the standard processing & display tools and compatible with the operating system the
analyst uses). If I have to call ten people, mail or ftp the thing and then wait days, weeks
or months for a response, and then perhaps go to a special computer to view the output,
well, that's just not going to cut it. You'd be amazed at how many people around here
think that is a perfectly reasonable way of doing business.
4. Since we are the National Security Agency, the process has to be able to handle foreign
language material and if used for selection and filtering, must output all text and images
in standard formats that are compatible with standard processing and display tools.
5. The appropriate resources have to be made available to ensure success. Success doesn't
mean perfection. Let's say it's that famous 80% solution. 80% though does not mean it
does 4 of 5 things and doesn't do the 5th at all.
(U/FOUO) Now, having said all that, let me note that not one alternative sent as feedback fit the
bill. Either the solution couldn't handle damaged data or foreign text or wasn't available to the
masses or was no better and perhaps worse than just trying to slog along with Adobe Acrobat
Reader 4. Bin Laden is still safe and we, to the best of my knowledge, still have no reasonable
solution to the PDF problem. But in case I'm wrong, and you have a working solution that meets
the requirements above, please let me know.
Happy New Year!
"(U//FOUO) SIDtoday articles may not be republished or reposted outside NSANet
without the consent of S0121 (DL sid_comms)."
DYNAMIC PAGE -- HIGHEST POSSIBLE CLASSIFICATION IS
TOP SECRET // SI / TK // REL TO USA AUS CAN GBR NZL
DERIVED FROM: NSA/CSSM 1-52, DATED 08 JAN 2007 DECLASSIFY ON: 20320108

View file

@ -0,0 +1,42 @@
{
"title": "Raul Responds",
"type": "SIDTODAY",
"description": "Apparently, Raul received several angry emails in response to his previous about PDFs, in which he stated that Osama bin Laden could use PDFs written in Arabic and NSA would never discover those PDFs by searching for Arabic words. Raul responds to the criticism with a list of requirements for how to solve the foreign language PDF problem.",
"long_description": "In this SIDtoday column, an analyst using the pseudonym Raul responds to critical feedback he received after a prior article about the difficulty of processing PDF documents, particularly foreign-language ones. He clarifies his earlier remarks and lays out five requirements a 'considered' solution to the PDF processing problem should meet, covering selection/filtering, display, handling of damaged data, availability to analysts, and foreign language support. He concludes that none of the alternatives suggested in the feedback met these requirements.",
"tags": [
"pdf processing",
"foreign language",
"data analysis",
"sigint tools",
"reader feedback",
"adobe acrobat"
],
"document": {
"creator": "NSA",
"date_created": "2005-01-13",
"classification": [
"UNCLASSIFIED"
],
"dissemination": [
"FOUO"
],
"banner": "DYNAMIC PAGE -- HIGHEST POSSIBLE CLASSIFICATION IS TOP SECRET // SI / TK // REL TO USA AUS CAN GBR NZL",
"genre": "newsletter",
"programs": [],
"countries": [],
"people": [
"bin Laden, Osama"
]
},
"publication": [
{
"publisher": "The Intercept",
"date_published": "2017-04-24",
"link": "https://theintercept.com/snowden-sidtoday/3676182-raul-responds/",
"doc_source": "https://theintercept.com/snowden-sidtoday/3676182-raul-responds/",
"doc_url": "https://s3.documentcloud.org/documents/3676182/Raul-Responds.pdf"
}
],
"pdf_sha256": "4445d2e3a067c0e7728f1f4bec9ea64a2470ac5b01c683c65cc0eab46134b863",
"pages": 2
}