I don't know what happened around that SHA-1 collision thing to get so many people to visit it, but I'm sure sad that it's this mostly down-to-earth answer that got the most votes out of all of my previous answers T.T I mean, please, my Garbled Circuit one took certainly ten times more time to write.

@Lery I know that feel. My highest voted answers on Sec.SE are simple things about DDoS, the definition of certain hacker jargon, why wiping encryption keys is faster than wiping an entire disk, whether or not physical access to a sensitive system is dangerous... But then ones where I actually spend hours figuring out the answer for a fascinating question only gets a few upvotes.

It is normal, the more basic and the more common the more people will vote for it.

Yeah, it's just unfortunate.

@bdegnan What makes SIMON/SPECK better than PRINTCipher? iacr.org/archive/ches2010/62250016/62250016.pdf

Random question: How would I go about calculating the effort required to perform logjam-type precomputation attack against a 1536-bit DH modulus?

Meh, my highest votes go to explain CER / DER and such and of course explaining that PKCS#5 based padding is essentially the same as PKCS#7 based padding for 8 byte block sizes.

My XMas special answer (32 bit calculations on Java Card) took some days to figure out. I could find examples in C but those were for unsigned variables, and I had to figure out which algorithm was most efficient (hint: Knuth knows).

Oh wow, I had no idea you had so much rep on SO.

Yeah, I'm kinda in the lead on the cryptography and encryption tags :P And of course Java Card.

I'm not a fan of SO. Although it has some good answers, most of the really good ones are pre-2013. The average quality of questions and answers there is far lower than on many of the other sites (excluding certain specific sites which I shall not name which I believe are almost 100% full of it).

I also had some rep on Expert Sexchange. But that's kind of history now. It got flooded with payed subscribers mainly from India who thought their company had payed for the privilege and you had to deliver.

In the pre-hyphen days?

Yeah.

My then-company actually blocked it due to "sex" being in the server name.

Also, I could not visit zallmanusa.com for my cooling solutions :P

The Scunthorp problem never ceases to amaze.

Which shows that you cannot use a Squid proxy with a too simple regex as your filter.

Oh, I thought I spelled that wrong, it is zalmanusa.com They must have dropped an L somewhere in their history.

There are so many databases of "naughty sites", too, which would be easier to use (and probably lighter on the firewall).

They were a bunch of numbnuts mostly to be honest. I told them that using an FTP server + scripting was not a good idea to receive files.

heh

Then, later, they came with a file for which the signature verification failed. Can you guess why?

Someone uploaded porn?

Nah, it was from the intended sender...

Who placed the right signature...

Oh, not an anonymous FTP server? I dunno, just regular corruption then?

Well, if you use a separate FTP server, how do you know that you received all of the file?

I just check the file size.

The server may now the file size as the protocol tells it to. But the receiving file system certainly doesn't know if all of the file has been received.

So they used a timeout, and if nothing was received for some time then the file must be complete, right?

I am against using any kind of scripting for these kind of things. I also explained that using separate utilities for generating XML and then signing it was a bad idea. Then later they received more data and they found out that the IO speed sucked. Well yes, duh.

ouch

That's so strange, because there are so many good ways of transferring files.

OK, that's fine for HNQ if you ask me, don't know if anybody will be interested in that but alright.

It's better than many of the HNQs we get.

@forest You can remove questions from the HNQ by visiting it and hitting the mod button.

I was removing questions from HNQ even before I was a mod. :P

(also sorry lol)

I was asking how, but I probably don't wanna know.

Oh, just adding ${}$ into the title.

MathJax doesn't play well with HNQ, and whenever they go together, MathJax wins and the question gets bumped off.

The SHa-1 question is not hit the chat. There might be some problem with the bot.

@forest I'was doing the reverse, if I feel that the Q is a good HNQ, I replace the MathJax.

Man, consider the math sites. Almost no HNQ.

They could fix it by rendering it server-side then having it display as HTML+CSS.

It will cost the extra CPU time on the servers, no way! How do they earn the money?