The future of teaching programming is on mobile devices

Nikolai Tillmann, Michal Moskal, Jonathan de Halleux, Manuel Fähndrich, Judith Bishop, Arjmand Samuel, Tao XieITiCSE 2012 · DOI

TL;DR

This paper argues that mobile devices, not traditional PCs, are the future of programming education. They introduce TouchDevelop, a novel mobile-first programming environment and language designed for touchscreen input, enabling students to learn and create applications directly on their smartphones.

What Problem Does This Paper Address?

The authors identify a significant technological shift: the increasing prevalence and power of mobile devices, particularly smartphones, as the primary computing platform for many individuals. They observe that while technology has historically influenced teaching methods, programming education has lagged behind, still largely relying on traditional PC-based development environments. This creates a disconnect, as students are more likely to own and use mobile devices than PCs for everyday tasks. The paper addresses the challenge of adapting programming education to this mobile-centric reality, proposing that learning to program should happen on the devices students use daily, without the need for a separate PC for development and testing.

Methods

The core of the paper's contribution lies in the TouchDevelop environment. This is not just a port of existing programming languages to mobile but a re-imagining of the entire development experience. Key aspects of their approach include:

  • Mobile-First Language Design: TouchDevelop features a typed, structured programming language designed from the ground up for touchscreen input. It aims to simplify syntax and interaction, moving away from character-by-character input.
  • Semi-Structured Code Editor: Instead of a traditional text editor, TouchDevelop uses a visual, semi-structured editor. This guides users by presenting lists of statement kinds and expression tokens, significantly reducing the possibility of syntax errors and making navigation easier on a small screen.
  • Inferred Types and Predictive Completion: The editor leverages inferred types and mines existing programs to provide highly predictive auto-completion suggestions, allowing users to select from lists rather than typing everything.
  • Social Programming Features: TouchDevelop incorporates social elements, enabling users to publish, download, and discuss programs, fostering a community around shared creations.
  • Cloud Integration: The environment is designed to seamlessly integrate with cloud services, allowing for distributed state management and access to remote functionalities, addressing the limitations of mobile device storage and processing power.

Datasets

Not discussed in the paper. The paper focuses on the development of the TouchDevelop environment and its pedagogical implications, rather than on training or evaluating models on specific datasets.

Key Insights and Ideas

  • Instant Gratification is Key: Allowing students to program directly on their phones means they can immediately see their creations in action, share them with friends, and experience the satisfaction of building something tangible. This is a powerful motivator.
  • Leveraging Personal Content: Programming on personal devices opens up opportunities to use students' own photos, music, and data as input for their programs, making the learning process more relevant and engaging.
  • Touchscreen Input Demands New Paradigms: The limitations of finger-based input on touchscreens necessitate a departure from traditional keyboard-centric programming languages and editors. TouchDevelop's approach to structured editing and predictive input is a direct response to this.
  • Cloud as an Extension of the Device: The cloud is not just for storage but an integral part of the programming model, allowing mobile devices to access rich functionalities and manage state seamlessly, overcoming hardware limitations.
  • Democratizing Programming: By removing the PC barrier, TouchDevelop has the potential to make programming education accessible to a much wider audience, especially in regions where mobile devices are more prevalent than computers.

Key Results

The paper reports on experiences with students at various levels:

  • High School Students (9 weeks): Two students with prior PC programming experience developed 19 games and applications using TouchDevelop, with some creations rivaling commercial apps.
  • High School Students (90 minutes): In a computer science class, 30 students were able to write mobile applications within 45 minutes of a brief introduction, demonstrating the environment's accessibility.
  • Eighth Grade Students (2 hours): 90 students, with no prior programming experience, were able to create unique applications incorporating phone sensors and media libraries.
  • Overall Usage: After six months, over 3000 games and applications were published by users, indicating significant engagement and productivity.

Limitations and Future Work

  • Platform Heterogeneity: The authors acknowledge that the current mobile landscape is heterogeneous, with different platforms supporting different programming environments. Standardization or cross-platform support is crucial for widespread adoption.
  • Teacher Preparedness: Educators are not yet fully prepared for this shift, and existing course materials are largely tailored to PC-based development. Adapting materials and training teachers is a significant undertaking.
  • Ecosystem Development: While TouchDevelop aims to be self-contained, the broader ecosystem of development tools and practices for mobile-first programming is still nascent.
  • Future Work: The authors are actively creating course materials (books, slide decks) for introductory programming with TouchDevelop. They also highlight the potential for "seamless learning" where students can learn and practice programming anytime, anywhere, on their mobile devices. The integration of cloud services and the evolution of the language and editor to better support mobile development practices are ongoing areas of development.

Summary updated

BibTeX
@inproceedings{tillmann2012future,
  author = {Nikolai Tillmann and Michal Moskal and Jonathan de Halleux and Manuel Fähndrich and Judith Bishop and Arjmand Samuel and Tao Xie},
  title = {{The future of teaching programming is on mobile devices}},
  booktitle = {ITiCSE},
  year = {2012},
  doi = {10.1145/2325296.2325336},
}
Figure