The purpose of Zoof is to lower the barriers for (scientific) computing by providing a simple and intuitive IDE that promotes a powerful interactive workflow to use dynamic languages like Python. Getting started is easy, since Zoof can take care of installing the programming language and additional packages.
We believe that making (scientific) computing easier accessible promotes freedom and independence, and is an important step to make science more reproducible.
Since most dynamic languages are easy to learn, they are well-suited for education. Similarly, their interactive workflow provides many benefits for data analysis. This is propably why dynamic programming languages are becoming increasingly popular . For instance the Scientific Python ecosystem is blooming and provides powerful tools for science and data exploration. Exciting development like LLVM, Pypy, Julia, Kivy, ASM.js demonstrate that dynamic languages can be nearly as fast as compiled languages.
All these tools are open source and free for anyone to use. So, whether you're a high-school kid who wants to create a game, a scientist who wants to process some data, or anything in between, the tools you need are probably there ...
However, newcomers often don't know where to start, and the development environments that most developers use are not well suited for beginners. Zoof aims to reduce these obstacles by making it as easy as possible to get access to the tools, and provides an intuitive environment to use them.
We believe that the art of writing software can (and should) be a much more common skill. Many exciting tools already exist; with Zoof we bring them to a wide audience.
(These are the features that Zoof will have, see status below)
Zoof is just getting started ...
Zoof can be seen as the successor of IEP and Pyzo. Although quite successful, it's time to move things to the next level. Zoof essentially combines the two projects in as single application. Further, we want to open up to other dynamic programming languages, and we want to make the user interface more intuitive and modern.
We start with a clean sheet, allowing us to lay a strong foundation with a modular design, a well-documented public API, unit-tests, and approporate coding style. From there we include the pieces from IEP and Pyzo, while adapting them to the Zoof's design principles. Along the way, some functionality will be replaced by modern features (see below). With this approach we think that we can get Zoof ready to roll in a relative short amount of time, while achieving a healthy code-base for future development. You can see the progress at our github repo.
Since Januari 2015 we are actively developing Zoof. The first step is developing a web-based GUI toolkit, which will form the basis for Zoof's UI.
If you would like to participate in any way, please contact me.
Here is the main idea of Zoof: Download Zoof and start it up, set up an environment with the language of your choice, and start coding. An intuitive interface will guide novice users through the process, and will provide tutorial materials if necessary. At the same time, the interface does not get in the way and is powerful enough to support the workflow of advanced programmers.
Zoof supports multiple progamming languages, focusing on dynamic languages (which support the interactve work flow of Zoof). At first we will focus on Python and Julia, and other languages can be added via an extension mechanism. Over time the languages that Zoof natively supports will be increased.
Like IEP, Zoof will be available in multiple (natural) languages. Let us know if you want to help with the translations!
Zoof will probably support two ways to provide kernels. On the one hand we will piggy-back on the kernel mechanisms of project Jupyter (formerly IPython). On the other hand, we are developing protocols that are aimed to be easy to implement in any dynamic language, and have few dependencies. Therefore, making Zoof work with a new language should be quite simple. Imagine building a toy language with LLVM and being able to target it from Zoof by spending one day implementinging a few simple protocols.
Zoof uses Conda to spawn and manage new environments. Conda already provides full-fledged support for several version of the Python programming language. Via Binstar it will be straightforward to allow Zoof to spawn other languages as well.
Some encouraging testimonials from IEP and Pyzo:
This is fantastic IDE - exactly what I was looking for!For quite a while now I’ve been looking for a way to switch from Matlab to Python. The last remaining problem was a lack of descent IDE for interactive work. I do a lot of short calculations, so a ‘cell’ execution mode is crucial for me. Thanks again for a great project!
I don’t even remember how I found IEP, so confused and desperate I was testing [all available IDE’s] I could find. Installed it, and 10 minutes later I was coding, connecting with a glade created user interface, and everything is running so well. I ain’t gonna change now. IEP is my choice. Thank you very much, and keep up this amazing work.
What I like especially of iep is the plain and simple design of the UI [...] while providing really powerful functionality. Please, keep this project running.
It’s performing beautifully, the UI is very intuitive and stylish (to my eyes anyway). [...] This is a very impressive app, on OSX close to alpha dog status (I’ve used ALL the OSX alternatives).
There are some features that we plan on incorporating (or at least experiment with). The details need to be worked out, and much is still up for discussion:
- Once Zoof has started, most interaction with the file system will be done via a proxy that runs in a dedicated thread, so that Zoof remains responsive even if your hard drive is busy. This is already implemented in IEP's file browser tool. This approach can later be extended to transparently map remote file systems (e.g. cloud computing).
- No more tabs, only one file is "open" at a time. Begone with the old-fashioned concept of unsaved files. Just edit each file directly. The state of each file (cursor position, undo, etc.) is preserved once you re-open it. Open/select a file by typing in a text field, which will smartly suggest recent files and files that match the part of the file name that you wrote. The idea is similar to Sublime's "open anything" and Zed's method for opening files. Don't worry, we will also include IEP's powerful file browser that has support for filename filters and search.
- No more menu bar. Each component of Zoof has its own small menu that only lists functionality specific to that component. Instead of searching for functionality in the global menu, you simply select the menu of the component to which the required functionality applied.
- A generic "tip" mechanism to guide novel users in using Zoof, that is easily put out of the way when (no longer) needed. Also give hints about keyboard shortcuts to help users adopt an effective workflow. The same system provides extra information on menu items and user-settings. This should help keep the UI simple and clutter-free while making it easy to get more information when needed.
- Source structure tool (like IEP has) is shown in the editor itself and is only shown when ALT (or another key?) is pressed down. This should reduce clutter while still making it easy to get an overview of your code. Will need to see how this works out in practice though.
- Concept of "pages". Much functionality is provided via pages that are displayed in the area where your code is normally displayed. E.g. configuration options can be accessed by selectig "zoof:config" as a "file name". This reduces the amount of dialogs popping up, and makes Zoof very easy to extend. Common pages can easily be loaded via a menu option.
- The start page (which can be accessed using a single click) provides a quick way to common tasks, like launching and managing environments.
- Zoof will get an extension mechanism to make it easy to register extenstions that provide new pages, tools, and other functionality to customize Zoof to your needs.
.. toctree:: :maxdepth: 2