Google has added a fresh round of features to Project IDX, its cloud-based development environment.
As may be surmised from a name that includes the term “project,” Google’s work on Project IDX isn’t entirely done.
Take Project Starline, Google’s video telepresence gambit, which surfaced in 2021 and is still being refined. Or Project Maven, a drone initiative with the military that was abandoned in 2018. Or Project Dragonfly, a censored search engine Google planned to roll out in China until Mountain View gave up on the idea in 2019.
Project IDX was announced in August and initially the AI-enhanced, cloud-based integrated development environment (IDE) was accessible only by invitation. It’s still in public preview, which means those who submit an email to the waitlist at some point should be admitted to test the nascent web-based workspace.
Based on Codey, a foundational coding model built on Google’s PaLM 2 large language model, and on Gemini, Project IDX aspires to offer the sort of AI-assisted developer experience available to those using Microsoft Visual Studio Code in the cloud with AWS Toolkit extension (for Code Whisperer), or the GitHub Copilot extension if it were available.
Another similar offering would be GitHub’s Codespaces, with Copilot, although the VS Code comparison is more apt because Project IDX is built atop Code OSS, the open source foundation of VS Code.
(A reminder that neural-network-generated source may lower the quality of your codebase if ingested. Generative models all too confidently offer broken code from time to time. YMMV.)
Those admitted to Club IDX can expect code completion that is allegedly smart, a chat interface that’s been getting some praise online, and on-demand technical exegesis through the addition of comment prompts like “add comments” or “explain this code.”
IDX environments are Debian VMs, with full command line access, built using Google’s Cloud Workstations service.
The latest additions to Project IDX include an iOS simulator and an Android emulator that have been built directly into the browser. Thus, if you’re writing, say, a Flutter app using the Dart programming language and you want to preview how it would look on iPhone and Android hardware, you can launch your code in the appropriate simulator or emulator without exiting your workspace.
According to the IDX team, this capability is still experimental, meaning buggier than usual. But that caveat applies to the entire service, which is not intended for use in a production environment.
“This is an experimental, pre-production release of Project IDX,” the documentation says. “Project IDX is likely to change and is not subject to any service level agreement (SLA) or deprecation policy. The implementation is subject to change without notice and in future releases.”
Given Google’s history as a serial killer of products, such warnings should be taken seriously.
In any event, there are also several new templates for spinning up new projects without the usual command line arcana. Those looking to create Astro, Go, Python/Flask, Qwik, Lit, Preact, Solid.js, and Node.js applications with a few clicks should be pleased.
There’s of course support for importing repos from GitHub and for environmental configuration and customization using Nix, a package management and system configuration tool.
IDX workspaces now can auto-detect network ports when applications need to make firewall adjustments. They also allow command-line tools, scripts, and utilities to be run in workspaces without local installation. And they make it easier to work with Docker containers and images via Docker config settings in the dev.nix file.
As for AI code help beyond the US, the IDX team says this has been made available in 15 regions – India, Australia, Israel, Brazil, Mexico, Colombia, Argentina, Peru, Chile, Singapore, Bangladesh, Pakistan, Canada, Japan, and South Korea – with more to come.
Keep in mind that AI code suggestions also come with a disclaimer: “Our generative code features are still experimental and you’re responsible for your use of suggested code or coding explanations. So you should use discretion and carefully test and review all code for errors, bugs, and vulnerabilities before relying on it.”
In other words, you have been warned. ®
![](https://techcodex.com/wp-content/uploads/2023/11/Eugen-designstyle-birthday-m.png)
Eugen Boglaru is an AI aficionado covering the fascinating and rapidly advancing field of Artificial Intelligence. From machine learning breakthroughs to ethical considerations, Eugen provides readers with a deep dive into the world of AI, demystifying complex concepts and exploring the transformative impact of intelligent technologies.