September 2018 Summaries

In September 2018, Google Fonts launched support for Japanese web fonts, following the introduction of Korean fonts earlier in the year, by optimizing their font slicing system to handle the large and complex character sets typical of these languages. This optimization reduced data transfer by over 30% for Korean users, and a similar approach for Japanese fonts involves analyzing character frequency from millions of webpages to efficiently slice and deliver only the necessary font parts. By adopting unicode-range, woff2, and HTTP/2 technologies, Google Fonts enables finer-grained slicing and concurrent delivery of small files, significantly reducing the byte load for users. For Korean, segmenting the top 2,000 characters into 20 slices resulted in a 38% reduction in data transfer, while segmenting the top 3,000 Japanese characters into 20 slices led to an 80% reduction. These advancements allow for faster user experiences and pave the way for future collaborations with the W3C to develop new web standards, while also indicating Google's ongoing commitment to further optimize and expand their font offerings.

Google has announced updates to its augmented reality platform ARCore and the 3D rendering library Sceneform, aimed at enhancing app performance and developer flexibility. Key improvements include algorithm enhancements for reduced memory and CPU usage, and a new API in Sceneform that supports runtime glTF model loading, eliminating the need for pre-conversion to SFB format. Additionally, the release of the Sceneform UX Library's source code allows developers to customize AR elements such as plane detection and object transformation, while a new ARCore API assigns unique IDs to point cloud dots, aiding in the identification of stable structures across frames. ARCore's compatibility has also expanded to include more devices, notably the Acer Chromebook Tab 10, broadening the reach of AR experiences. More information and resources are available on Google’s developer website and their GitHub projects.

Google Cloud Platform (GCP) and G Suite can be integrated to enhance data analysis and presentation capabilities, as demonstrated in a video tutorial by Wesley Chun. By utilizing Google Apps Script, a serverless development platform, users can access G Suite APIs and some GCP tools like BigQuery to perform data queries. The process involves running a query to extract data from BigQuery, storing the results in Google Sheets, generating a chart from the data, and creating a presentation in Google Slides. This workflow allows users to transform big data analysis into an executive-ready presentation, showcasing interoperability between GCP and G Suite. The tutorial highlights the potential of using both platforms together, offering a step-by-step guide through a codelab and a GitHub repository for further exploration and inspiration.

Google's AR/VR team announced new experimental features for the Daydream platform, specifically for the Lenovo Mirage Solo headset, to enhance virtual reality experiences. These features include support for 6DoF (six degrees of freedom) controllers, which allow for more natural hand movements in VR without the need for expensive external hardware, utilizing machine learning and optical tracking. Additionally, a new see-through mode is introduced, leveraging WorldSense technology to integrate physical and digital worlds, enabling applications such as augmented reality prototypes. Finally, the capability to open smartphone Android apps in VR is presented, allowing developers to adapt existing 2D applications for a virtual reality environment without extensive redevelopment. These innovations aim to make advanced VR experiences more accessible and versatile for both developers and users.

Flutter Release Preview 2, announced at Google Developer Days in Shanghai, marks the final major milestone before the highly anticipated Flutter 1.0 release, focusing on creating pixel-perfect iOS applications. This release greatly enhances support for Cupertino-themed controls, offering a comprehensive library of widgets and classes to facilitate the development of apps adhering to Apple's interface guidelines. Key updates include new widgets like CupertinoApp, CupertinoTimerPicker, and improvements to existing ones such as CupertinoNavigationBar, which now supports parallax transitions. The release also introduces support for executing Dart code in the background and achieves a notable reduction in application package size, with a minimal Flutter app on Android now at 4.7MB. Flutter's growing momentum is evidenced by its status as one of the top 50 active software repositories on GitHub and its use in major applications by companies like Alibaba and Tencent. The strong community support is highlighted by a high satisfaction rate among developers, with ongoing contributions encouraged. For those looking to upgrade to Release Preview 2, the process is straightforward, and resources are available for beginners and experienced developers alike to explore the Flutter ecosystem further.

Google announced the general availability of the Google Photos Library API, designed to enhance photo and video experiences across various apps and devices by allowing integration with Google Photos features. The API, which follows a successful developer preview, supports mobile, web, and backend development, providing a REST interface and client libraries in multiple languages to facilitate rapid implementation. Users can authorize API requests, ensuring they maintain control over their data while developers can offer features like photo organization, direct uploads, and shared albums. The API also enables the use of machine learning for smart filtering, allowing users to efficiently locate specific content. Developers interested in extensive integration can join the Google Photos partner program, and those transitioning from the Picasa Web Albums API can use the provided migration guide.

Cathy Pearl, Head of Conversation Design Outreach at Google, emphasizes the importance of writing sample dialogs as a fundamental step in designing conversational systems for Actions on Google. Sample dialogs, resembling film scripts, guide the development of user interactions by mapping out potential conversational paths before any coding begins. Pearl notes that many developers skip this step, mistakenly thinking that conversation design only pertains to the surface-level words users see or hear. However, she explains that it influences the entire structure of an experience, including backend integration and error handling. By drafting sample dialogs, developers can identify issues such as overwhelming users with information, failing to invite user input, and inadequate error prompts. Writing sample dialogs helps refine the interaction flow, ensures better user experience, and reduces the likelihood of future design flaws. Google provides guidelines for creating effective conversational designs, encouraging designers to read dialogs aloud, avoid branching paths, and iterate based on feedback.

Launchpad Studio, an acceleration program by Google, has announced a new cohort focused on finance startups utilizing applied machine learning for various innovative solutions. This cohort includes startups like Celo, Starling Bank, and Go-Jek, among others, from across nine countries. These startups aim to tackle challenges in financial inclusion, stable currencies, and identity verification, leveraging machine learning and blockchain technologies. Participants receive 1:1 mentoring and access to Google's resources, including its product managers and ML research teams, to enhance their product development and drive global economic growth. Since its inception in 2015, Google Developers Launchpad has fostered a global network aimed at solving significant world challenges through innovation, and this initiative continues that mission by supporting diverse startups in creating inclusive financial products and services.