{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Concluding Remarks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Julia: Interdisciplinary science perspective\n",
    "- For software development in interdisciplinary fields like materials science, quantum chemistry, biology, physics and countless others, one problem is to bundle the efforts from all involved research communities.\n",
    "- One issue is that each field has different methods and approaches, resulting in deviating demands, for example:\n",
    "    - **Mathematicians** often consider reduced models and edge cases\n",
    "    - **Scientists** want to develop scientifically sound models and not worry about numerics\n",
    "    - **Computer scientists** want to gain the best performance for a given hardware\n",
    "    - **Practicioners** want to use the software to understand molecules or materials,\n",
    "      so software should be easy-to-use, reliable and black-box\n",
    "- As a result existing software is often only suitable for the research of some involved communities, but fails to provide a general platform. One example from materials science is that the methods suggested by mathematicians (working on the their toy codes) sometimes don't make it into the packages used by practicioners.\n",
    "- I think Julia is well-suited to overcome these challenges of interdisciplinary software development. Both high-level and low-level developments can be jointly underdone and at the same time one is able to design interfaces, which can be used by people less oriented towards programming."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Julia: My wishlist\n",
    "- Better support for more advanced unit tests (test tagging, test selection, fixtures, plugin system for tests)\n",
    "- Better support for static code analysis\n",
    "- Better support for code completion and refactoring in editors (at least in VIM it's still only mäh to this point)\n",
    "- Package maturity and quality not uniformly distributed amoungst fields, but ... that's not such a big problem (see below)\n",
    "- **Important:** I think these aspects are mainly a question of time ... the basics are there, but compared e.g. to Python things have not yet developed so much into a stable common ecosystem."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Julia: User perspective\n",
    "- Rich and growing package landscape\n",
    "- Bonus on top: Keep using all your favourite python, R, C, C++ snippets\n",
    "- Open and welcoming community with plenty of places to ask for help\n",
    "- Advantageous language design:\n",
    "    - Code generic by nature\n",
    "    - Multiple dispatch decouples packages (since e.g. it avoids the binary operator problem)\n",
    "    - Risk of \"dependency hell\" well reduced\n",
    "- Surprisingly small effort to use cutting-edge features (automatic differentiation, mixed precision, GPU)\n",
    "- Often combining niche packages, which have certainly never been used jointly, just works out of the box ...\n",
    "    - ... or it takes 1 or 2 simple patches with one-line fixes to get it there"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Julia: Developer perspective\n",
    "- High-level language, good for rapid prototyping\n",
    "- The first **prototype has adequate speed** to go beyond simplistic test cases\n",
    "- Tools to inspect, benchmark and profile (`BenchmarkTools`, `Profile`, `@code_llvm`, `@code_native`, ...)\n",
    "- Tricks to selectively optimise performance (`@threads`, `@inbounds`, `@views`, `@simd`, ...)\n",
    "- **Philosophy:** First get it to work, then **make it fast**\n",
    "\n",
    "\n",
    "\n",
    "- Combination of strong type system and just-in-time compilation:\n",
    "    - Code can be written generically\n",
    "    - User can pass domain-specific knowledge about his problem\n",
    "      to the compiler in the form of types (e.g. symmetries, positive-definiteness ...)\n",
    "    - Enables short-cuts and dynamic switching to better optimisations\n",
    "    - Dead code and branches may be eliminated\n",
    "- Machine-specific optimisations can be used ... even on inhomogeneous clusters with nodes of different processor generations\n",
    "- **Philosophy:** Write code once, but **reuse** it many times"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Julia 1.3.0",
   "language": "julia",
   "name": "julia-1.3"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.3.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}