DConf Online Third Edition

This talk is about compile-time and run-time reflection, template metaprogramming issues, and how strings just might be the way forward.

The calculation of gradients is a way to understand the universe. For the entire history of computing, people have been writing programs that calculate them. For many years these were written arduously (and often incorrectly) by hand, or in specialist languages dealing only in mathematics. Via the use of a set of techniques known as automatic differentiation, we can write programs which not only calculate gradients exactly but also without writing them manually—differentiable programming. I will summarise where the mathematics comes from, then walk through some simple code to implement both modes of automatic differentiation and apply them to some simple problems. Also included is a little historical discussion of some past techniques for specifying systems of derivatives to model systems, but which are not considered differentiable programming.

When building an application, at some point test data is required. Making up a list of somewhat realistic forenames and surnames isn't too much work, but then you need street names and postcodes for France. Faked can help you with that. This library translates the excellent faker.js library to D with help from the worst typescript transpiler known to man. This is its story.

This is a crash course on translating C code to D code, from small code snippets on stack overflow to entire C libraries. Since D has similar syntax and features to C, this is relatively easy, but there are some pitfalls, and it still requires tedious text editing. I'm going to present key differences you should be aware of and present the 'ctod' tool I made that does most of the tedious work for you.

DConf wouldn’t be DConf, online or otherwise, without the “Ask Us Anything!” tradition, where the language maintainers step onto the stage to face the assembled crowd, take their questions, and provide them with the answers they're looking for. So prepare your statements and compile your queries! Walter and Átila are very much looking forward to them.

Adam is back for a third DConf Online livecoding session. His previous two sessions were a lot of fun (DConf 2020 and DConf 2021). What will he code this time? Tune in to find out!

Please Note: this is not a pre-recorded talk. It takes place in a second livestream, separate from the DConf Online Q & A livestream. There is no pre-talk session, nor a post-talk Q & A session. Adam will be live and in color from the moment the stream kicks off, and will answer any questions posted in the livestream's chat box. A link to Adam's livestream will be posted here prior to the event.

Syntactic sugar can be a mystery. You know how pointers and values work, but how does that relate to syntactic sugar? The specification says what the sugar is supposed to do but not how it works. Knowing how it's implemented is key to making informed tradeoffs about when to make use of sugar. Personally, I am uncomfortable with using a programming language without knowing what is going on under the hood. After all, a race car driver that knows how his car works is going to get more out of it than a driver that only knows that stepping on the gas makes it go faster. I never understood virtual functions in C++ until I examined the output of Bjarne's C++ to C translator.

Here is the missing explanation of the secret inner workings of D's syntactic sugar.

As a long-time reader of the D forums, I have noticed that people new to the D language often ask what the best editor to use is. The usual answers are VScode, Vim, someone's other favorite app, or "check out the editor page at dlang.org". I will supply a more complete answer to this question in the hope that the beginning D language user may benefit from the information.

This talk is a continuation of the Dconf ‘22 talk on building a ray tracer in (less than) one weekend. What this talk will show is that D is a language built for software engineering and creating applications that scale. In this talk I will continue to take you through the journey of building a ray tracer, this time focusing on some of the key features of D. We’ll start by optimizing the previous ray tracer with static if and std.parallelism for example. Then I’ll show object-oriented programming in D and build a few data structures. Finally, we’ll display a final rendered image using the new and improved ray tracer with several new features.

This talk is about the modelling power of D, and why you should at least consider using structs or classes as models for your APIs. As D has features such as introspection and generative programming, building a model can be just as effective as using code to form your API. Plus a model can be a whole lot prettier in the end. Slap on some UDA magic and the code almost writes itself!

As I talked about in my 2019 talk on serialization, we want to use the D compiler to the best of its abilities. We’ll look at a few cases where modelling can shine as a way to express how your program should behave and look intuitive to those reading the code.

This talk is an introduction to my DConf Online ‘22 workshop, in which I walk viewers through the development of an 8080 assembler in D. Here, I'll talk about my motivations for the worskshop and what I learned when I published a blog post series on the same topic. I'll also opine on garbage collection and on ways we all can advocate for the D programming language.

Brian walks viewers through the creation of an Intel 8080 assembler in the D language. Note that this is a series of 15 short, prerecorded talks, of around 10 to 20 minutes running time each. It is not a livecoding session. The videos are not scheduled as premieres, but will be published following Brian's Q & A session at the end of his 19:00 talk. Starting at 20:15, we will have a Watch Party in the livestream: we'll broadcast each workshop video in the play list, back to back, in the livestream. Brian will be monitoring the livestream chat to answer any questions from viewers (note that there are no prizes for questions asked during this session).

Although the workshop videos will be available to watch and comment on any time, you'll only have a chance to get answers to your questions in real time during this livestream session. So stick around!