Also relevant (though mostly for entertainment and to spark curiosity): https://www.youtube.com/watch?v=kPRA0W1kECg

You can include Jira and Pivotal in GitLab. I haven't used them so far, but seen the integration option while setting up my GitLab server.

I wish there would be a standard/protocol for interpreting issue-tracking/milestones and wikis from the repo, such that I could write them as markdown files in a specific directory/file structure, commit them to git and they would be recognized by services like GitLab. In turn, when creating wiki pages etc in the webui GitLab would automatically also commit those to your repo. It would prevent the lock-in to one service and make migration much easier.

I've recently set up my own GitLab server in order to have control over my data without missing the immensely useful features of the webui such as issues, milestones etc. The biggest problem for me was the idea that there is data that may be created for my projects that is not checked into the git repo and only hosted on the service, eg. issues, milestones and the wiki. Luckily GitLab is pretty damn good and open source, so I was able to set up my own server on some old hardware I found in a closet.

They talk about it being an anti-pattern there, but I think the idea of keeping software components loosely coupled but related is a pretty good thing.

Absolutely, it's part of the core UNIX philosophy and makes a lot of sense for maintainability and flexibility.

The problem is on layer 8.

Spaghetti - code that is difficult to maintain because it executes through unorganized and duplicated portions of business logic.

And the lesser known OO equivalent:

Lasagne Code: When the data you seek is buried so deep under layers and layers of abstraction and matryoshka doll-like data structures, that it takes hours to get a mental image of where that one fucking string, that you need, is coming from and who is overwriting it with garbage. Most commonly observed in enterprise® solutions™ written in Java.

• vim-plug to manage your plugins (I have tried several plugin managers and that one is my favourite by far)
• fugitive for excellent git integration
• ctrlp for quick access to files and buffers

Don't worry, I'll manage. I always ever do one thing at a time and write all the others on a list to work on when done with my current focus.

Well damn it, I already don't have enough time for my own side projects, and now you drop this bomb on me? Well, guess I have to learn me some Haskell for great good then, right after the exams and after learning Python, which I already started, and after finnishing my current pet project in C and my other pet project in Lua.

It's really not as much fun discussing something while constantly being insulted. You may notice that I did not do that in my post, there is no hostility from my side. The trait that you are painting on my back, that I would be irrational, opinion based and of a "wake up sheeple" mentality is what you are demonstrating yourself by the means of your expression. Please, if you want me to accept that I only throw around opinions while your perspective is the factually correct one, do show some evidence. Because so far we have both only voiced opinions or conclusions.

You are wrong if you assume me to be too ignorant to admit fault or see the truth when presented to me, I am willing to continue the discussion; but only under the condition that you refrain from defaming me or insinuating things you can't know about me.

According to those present at the conference, Windows developers attending the conference did not deny that the "NSA" key was built into their software. But they refused to talk about what the key did, or why it had been put there without users' knowledge.

If you have closed source software that contains hardcoded keys, one of which is labelled as NSAKEY in the debug-build, coming from a company that is factually known to cooperate with the US government under the obligation of NSLs and give at least the FBI and NSA access to their users data. And if those agencies are determined to "collect it all" and have demonstrated disregard to the law they operate under and have claimed in internal documents to be very successful in breaking either crypto or implementation of SSL, TLS and SSH; and if those agencies are part of a government that has been prohibiting export of strong crypto outside the US since the 90s while having interfered with the definition of security standards by standardizing weakened cyphers and curves through NIST. And if there is no explanation or comment from developers of that company in regards to why those keys are hardcoded into a central part of the crypto library of the Windows operating system and there is no good reason for those keys to be there in the first place; then assuming all is fine and you can trust Microsofts implementation of SSH is unreasonable and being sceptical and refusing towards it is not spreading FUD.

Hold it right there, criminal scum! will be the welcoming message of my SSH-server for clients connecting from the PowerShell or with keys generated by Windows - followed by a forced logout.

Why? Because there is a backdoor in the central crypto DLL of all versions of Windows newer than early Win95. A backdoor consisting of two hardcoded keys, one confirmed to be labelled NSAKEY in the debug-symbols. Allowing a closed source SSH-client from Microsoft and under Windows to connect to your server means compromising security.

I generally don't swear, but: Fuck no! Microsoft offering SSH in Windows is not an increase in security, it is precisely the opposite.

The last demo that left me this speechless was We Come in Peace by Elude, which is running on the Amiga AGA and has real time lighting with shadows.

Either call it awesome or include OpenSSL.

Well, first you have to draft out the general idea behind the language, then you would design the usage by creating the semantics and syntax and defining the behaviour. This can take years, the result is a standard that can be implemented in a compiler infrastructure.

In this case you can watch how it's done very closely as Jonathan Blow is doing it openly and documenting the progress.

And here is the donation page, for the lazy ones

I don't even know how to react apropriately to the SSH thing, this is like a death sentence to secure system administration. There are not many details yet on how it works and it looks like it's no 100% solution, but they seem to be able to read SSH connections with a certain probability.

Here are the documents referenced in the talk

Casey Muratori is generally worth watching. If you like what you see there, you might also be interested in the videos of Jonathan Blow (a friend and collegue of Casey) regarding the new programming language he is designing.

This started with a few talks [1,2] about the design of a new programming language for game programming (that is naturally also useful for other tasks). Blows motivation for this is related to the increasing complexity (and ugliness) of C++, which he explains in detail. About a month after the first talks a third video appeared demoing a first implementation of some of his ideas [3]. This video was followed by two live coding sessions, the first of which is linked in the OP, the second one can be found in [4].

I have often seen people asking for actual unedited coding videos from experienced programmers and thought some of you might find this interesting.

[1] https://www.youtube.com/watch?v=TH9VCN6UkyQ&list=UUCuoqzrsHlwv1YyPKLuMDUQ

[2] https://www.youtube.com/watch?v=5Nc68IdNKdg&list=UUCuoqzrsHlwv1YyPKLuMDUQ&spfreload=10

[3] https://www.youtube.com/watch?v=UTqZNujQOlA&list=UUCuoqzrsHlwv1YyPKLuMDUQ

[4] https://www.youtube.com/watch?v=GJhvmNgJFdQ&list=UUCuoqzrsHlwv1YyPKLuMDUQ&spfreload=10