I created this small utility because doing a lot of talks recently with live coding sessions is usually terrible, in terms of experience. On retina display when mirroring to SXGA, WXGA or even lower, coding is very hard. For this reason I created this small utility that gives the opportunity to select just a small portion of the main screen, that is later available to be fullscreened on a secondary screen.

It's not about just talking, but what about Lavabit and all the problems the guy faced when he denied to give the encryption keys away:

http://www.rt.com/usa/lavabit-owner-fears-surveillance-arrest-595/

Wondering if Swift could ramp in this top 10 chart up after being available for Linux as well later this year...

Yes, some builds are still there, but the code has been turned down to avoid troubles. I hope he will be safe, even removing doesn't mean he will be safe in the future.

Probably one of the best lighting talks around the web!

Great answer. Just few things to add:

The observer pattern is very much based on mutation (because you have to keep the list of observers) and therefore doesn't really work all that well with lazyness.

Exactly and there's one more important thing to remember when dealing with the GoF's definition of the Observer pattern: it's imperative. This means that the value of a variable at the time of evaluation is important, in Reactive Programming this side can be ignored, because state is not anymore a component. The outcome of this is that, when dealing with asynchronous computation, the observer pattern will create what we call "Spaghetti Code" and will crate a lot of troubles. :)

Yes! Another one that looks very promising is Sodium that correctly has a denotational semantics.

It honestly took some time to write that post, I read a lot of papers to truly understand the origin of FRP and what was the definition of a Reactive System in the 90s'. It was very fun, a lot of math and papers, but worthy.

I complete disagree with the path we are taking right now as developers as community.

Code of conducts, gender-only conferences and all other related things are turning the developer's world into a sort of unstable community. If I say that one doesn't know how-to-code, later figuring out that is a her, I would probably be banned and burned by the SJW.

I also completely disagree with the necessity to have groups like "Girls Who Code", "Rails Girls", "Women Who Code", etc... why they don't come to the all other conferences instead of creating one women-only? This is a kind-of discrimination.

At this point, I would actually want to see if somebody will ever come up with a "Blacks that Code" or "Hispanic that Code", both of them are minorities in the IT world and no one gives a heck about them. Code of Conduct will drive us into the hell of not being anymore able to have a serious confrontation. Without confrontations the community will never evolve anymore, we will lose smart developers driving open source projects because an opinion offended someone and they got kicked out or, even worser, banned from the platform. This is going to be the future of Github.

I would say Bitbucket. All my private repos are there since a year or so. I am also considering in porting all my public repos used in the book about Reactive Programming in Swift that I am working on.

Considering that Bitbucket is owned by Atlassian that is a B2B oriented-company, I would say yes. They can't introduce this kind of things having businesses with companies that don't give a **** about these rules.

I played with Mongo for few weeks, I never really liked new technologies when dealing with data storage, I still have to find a good solution to databases like Postgres, MySQL, Cassandra and Elastic Search... All these 4 have a long history and are widely used in different domains, I tried to port some web apps using a SQL oriented database to MongoDB, the result was absolutely horrendum! Tuned PostgreSQL outperformed tuned MongoDB by almost an order of magnitude for some queries, and after few days started to freak out due to the amount of data to handle.

Even if I am relative young as developer, I never felt the necessity to find a new technology when dealing with data storage, old proved solutions are still the best choice, at least they don't lose data... :)

For sure NASA has hardware constraints to deal with. I can't agree with people quoting NASA code guidelines or CERN's one when they are implementing a REST webservice delivering the current weather. Every technology has some guidelines, as well programming languages. At the bottom we have the project context that is important, in this case NASA uses C.

I had to write a geo-localization software for a satellite during my bachelor, it was the start point of the actual one in use right now, I know how hard it can be having so big constraints... :)

Very interesting! I kind of like the fact some concepts have become mainstream after a decade or more of their invention, like OOP has been implemented in Smalltalk, but gained attention and became mainstream only 20 year later with Java.

Quick note: Smalltalk is wrong, first appearance is dated 1972 (after 3 years of development), not 1980. :)

++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++.

C is a key language, if you want to understand memory management and resource handling it's the way to go. K&R book is still a good resource, I wouldn't go after another book before finishing it.

C is a very basic language in terms of syntax, there's almost no complexity if compare to languages like Clojure, Haskell, Lisp or Scala, but on the other side is a wider world, you can do a lot of stuff with many potential solutions and the only way to handle them is to try. C, more than other languages, requires a lot of time to code and try different solutions, mainly when asynchronous computation is introduced. Keep going!

It always depends on the language and technologies used. Code clarity and maintainability is a key component in most softwares. Writing C code is way different from writing Scala code, static typing helps a lot in writing better code, rules are way different between the 2 languages.

I honestly would not use C to code a life-or-death system, ever! In fact, the most used programming language in fields where failure could drive to tragedies is Ada. The list of the Ada applications in fields like aviation, air traffic management, power stations, medical industry, etc... is very long. Code guideline is important, but it's more important to pick the right technology to do the right job.