>NOT daily updates

This was an exaggeration that I did not think would be taken literally. We had updates as the case progressed. I would guess about once a month. These updates were on the corporate intranet home page which generally contained about 5 links to 'external news' or something.

>He WAS absolutely trying to steal code. Goldman high-frequency code isn't great...

You have written this in a weird way. I agree he was trying to steal code. The quality of what he stole is not relevant. The way you have written this, it could be read: "He tried to steal high-frequency code". This is very much debatable. I could easily argue that Log4J is part of Goldman's high-frequency code.

He did try to cover his tracks. What he did was wrong and he knew at the time it was wrong. Six years later, does the punishment fit the crime?

>He wasn't planning on selling it. He was planning on taking it to Teza

Yes. This is the interesting point. He wasn't planning on selling it because there is no market for it. By this, I mean that there is no place he could list the code for purchase because:

1. It would be illegal and other companies frown on that kind of thing (at least on paper).

2. The code would not be worth much without an understanding of what it does in the wider ecosystem.

Point two is key. Sergey understood the code. He was uploading it to help him remember the work he had done at Goldman's. There is very much a market for, "I did XYZ at Goldman's, I can do it for you too." I have no doubt that he would have been able to do it again without the source code, but he wanted his 'notes'.

GS didn't want him to apply his trade at another firm. Hence the 'vindictive employer' remark.

How long before another developer gets harassed by Goldman's because they printed off some technical documentation before leaving? What if the documentation is printed a month before leaving? ...a year? ...where is the line, exactly?

Do you have any proof, or do we have to believe the word of someone who belive's the word of someone else, who in his turn is unknown?

Haha your point is obviously valid. Without outing my friend, I can't prove it, so I hope you'll believe me.

In fact, even if you don't believe me, that's okay--Goldman Sachs deserves to be given some crap!

you're really equating the credibility of an online discussion forum with that of a court of justice?

>The thing that ultimately flagged him was that he kept trying to clear his .history file.

I'll assume you mean .bash_history This discussion has been had here before. There are plenty of good non-sketchy reasons to delete it, and at some places it is mandatory.

He said ".history" because he probably meant it.

There are many shells aside from bash. My systems don't even have bash installed.

Yeah, it was the .bash_history file. My bad.

He also encrypted the files, uploaded them to his remote computer, and then deleted the programs used to encrypt and upload them.

Oh boy, this thread again. see: https://news.ycombinator.com/item?id=9044805

  user@somemachine.gs.com:~/ >svn export --username Aleynikov --password Hunter2 \
  --non-interactive svn://subversion.ZOMG.think.of.the.children.com/plundered/exportFile
  user@somemachine.gs.com:~/ >rm .bash_history

It's also not unheard of to encrypt files for transit, and for storage on untrusted remote machines. I'm sure someone else around here could show you a few dozen ways to pipe that through gpg, a few of which might also require a password to be used in the command, necessitating the deletion of .bash_history

You may criticize him for using encryption but what would you have him use, plaintext and ftp?

Teza didn't use the same programming language as Goldman Sachs, so he definitely couldn't use it at Teza

The value in the code is not the code itself, but the models it encapsulates.

For those interested, Teza uses (single-threaded) Java and Hadoop (or at least they used to), and Goldman Sach's high-frequency code was written in C++.

He didn't steal the strats, only the code he was personally interested in.