Dev News: Netlify’s New Tool, JupyterLab 4.0 and MDN Gets AI
Web development platform Netlify acquired visual editing solution Stackbit on Thursday. Netlify said the acquisition expands its product portfolio, “empowering marketers and product managers with the ability to design, create and edit experiences within composable platforms.”
Stackbit can aggregate content from any headless CMS, content source or custom data source, “creating a streamlined singular workflow for frontend content changes.” It also enables updating objects within the UI and writing the changes automatically to the originating source, according to a company statement.
“Stackbit reduces bottlenecks and increases developer productivity by empowering all employees to make web changes quickly and easily,” Netlify said in the announcement. “The addition of Stackbit to the Netlify platform gives Netlify customers the tools they need to quickly update web copy and content, make edits, localize content, and execute faster.”
Among the capabilities Stackbit brings to Netlify are:
- No-code website editing by non-developers, who can easily make updates and edits to their websites in real-time, creating pages, dropping in components, and adjusting layouts and styles on demand.
- Teams can customize the design and layout of websites using themes or creating their own designs.
- Stackbit reads and writes content to and from CMS, database, e-commerce catalog, or custom data source.
- Stackbit works seamlessly with any CMS application, content source, headless frontend platform, and framework.
JupyterLab 4.0 a Major Release
JupyterLab 4.0 — heralded as a major release — is available this month. According to a blog post about the release, it includes an updated text editor called CodeMirror 6, a new extension manager, and improved document search; along with UI and accessibility improvements and updates to development tools.
This release also moved real-time collaboration into a separate package — jupyter_collaboration, version 1.0.0 — which is now available. It can support displaying multiple cursors and selections, and support for registration of new shared model types.
Accessibility improvements include improved focus and keyboard navigation in the file browser, more ARIA roles, and labels for UI elements. It also collapses the main menu into a hamburger menu when there isn’t enough space to display all menu items.
JupyterLab is a highly extensible, feature-rich notebook authoring application and editing environment, and, like Jupyter Notebooks, is a part of Project Jupyter, a large umbrella project that aims to provide tools (and standards) for interactive computing with computational notebooks.
Finally, there are more than 100 bug fixes resolved to add to JupyterLab’s stability and performance. The package is now available on PyPI and conda-forge. You can upgrade by running:
pip install --upgrade jupyterlab or conda install -c conda-forge jupyterlab.
AI Help Added to Mozilla Developer Network
Mozilla Developer Network (MDN) has added AI Help — an artificial intelligence tool designed to optimize searches — to its site. Users simply type a request into the search field and the AI “dives into our comprehensive repository of documentation, retrieves the most pertinent information, and presents it to you in a succinct summary,” the site said in a blog post about the new feature.
The blog post provides a “peak under the hood” to show you how the AI leverages two technologies. First, the site generated embedding for every section of its MDN articles. Embeddings was defined as “the transformation of high-dimensional data into a lower-dimensional space.”
“Each embedding, typically a vector, encapsulates certain characteristics or features of the original data,” the post stated. “These embeddings capture both semantic and syntactic relationships among words or sentences, positioning them within a continuous vector space. This placement allows the model to comprehend contextual and relational nuances among the words.”
The embeddings are stored in a PostgreSQL database, hosted by Supabase. It then adds generative AI to streamline the process of finding an answer.
“Instead of making the user read all the content and extract an answer themselves, we utilize OpenAI’s Chat API to craft a prompt,” the post explained. “This prompt includes several system-level instructions about the desired tone and other related considerations, along with more technical guidelines, such as the requirement to always reply in markdown. We then use the previously related content as a context for the Chat API and ask the user’s question on their behalf.”
This process enables a more efficient interaction by using AI to help users access the information they need, the post added.
