>>7481934 lb
>take a credit card advance to have enough cash under your mattress to last for 6 months
Imagine thinking this is a good idea.
bakers cant pin anything, all the baker can do is make this a notable
kel<
probably not, global notables are for board admin related stuff
who elected this assclown blumenthal?
oh wait probably not the living, lawful citizens of this country
A revised version of the report was released yesterday. Pic related, page 3 of 480 from https://www.justice.gov/storage/120919-examination.pdf
Presumably this includes running the OCR software on the whole thing again, which would change the errors in the text. From a quick look there are about 3300 lines of text that changed in some way between the two versions of the report.
I have the original version of the report from the 9th but it is too large to post here.
Date fail, the revised version was released today not yesterday.
Okay so this is basically fuckery 100% confirmed, light softwarefag and mathfag analysis follows.
If the "James 8. Corney" bullshit was just random OCR errors then we would expect to see some instances of C-o-m-e-y and C-o-r-n-e-y in both the original report and the revised version, and we should expect to see some of these instances changing between C-o-m-e-y and C-o-r-n-e-y as the document is OCRed again.
This is not what we see.
There are 0 instances of the word C-o-m-e-y in the text transcript of the initial or the revised document. However, I do see multiple instances of C-o-r-n-e-y changing to C-a-m-e-y or C-o-r-n (space) e-y between the initial and the revised versions of the report.
The only conclusion I can draw from this is the software the government uses to OCR their documents is compromised to avoid or even completely prevent recognizing certain names or keywords accurately.
As others have mentioned, this would be very effective at thwarting attempts to search for information via FOIA requests.
I'm not set up to look for all possible misspellings but here is what I found in a quick search:
Tru mp: 2 instances (3 in revised report)
Trump: 678 instances (683 in revised report)
Much more importantly, it is very clear to me that every single mention of Comey was forced into being one OCR error another, and what's more, the specific error changed between the original and the revised version of the report.
This is impossible - an OCR software that was actually doing straight OCR would have had errors, but it also would have identified at least some of these instances correctly each time it was run.
This is damning evidence that has me 100% convinced that there is massive fuckery afoot to thwart FOIA requests.
>to thwart FOIA requests
I should add: this technique would also be effective in thwarting any internal government investigations. Government moves slowly on the best of days so this probably explains some of the delays we've been seeing with… everything.
Absolutely, unbelievably desperate.
Here's hoping the white hats have software to adequately correct for this fuckery. In any case I hear nothing can stop what is coming.
Q confirmation would be cool but not needed. It is mathematically impossible for this to be anything other than what I've described.
Wasn't expecting to effortpost in this thread kek, but I might as well put together a notable line.
Baker, submitting this for your consideration, I think I've showed conclusively that there is no chance our OCR errors are coincidental.
>>7482321, >>7482371, >>7482464, >>7482476, >>7482552, >>7482563, >>7482585, >>7482617, >>7482642, >>7482676 Dig on government OCR and evidence of systematic fuckery
Sure seems that way. Every day I learn a bit more about the magnitude of the job that Trump has taken on.
Baker
This format (including all posts in the notable rather than linking to a bun) also works better once the notable gets onto wearethene.ws