To fight the threat of ChatGPT and other competing machine learning technologies based on GPT-3, Google is doubling down on AI. The first announcements under the new stategy are expected at Google I/O 2023.
I may be old school, but I'm not looking forward to these upcoming Google products and technologies.
As a user, my experience with the Google Assistant has been unimpressive. The features I tried work half of the times and waste my time in failed attempts at getting understood. When such localized features are rolled out at all to my country years later, that is.
So I'm not excited at the prospect of, say, an AI-enhanced Google Search experience.
I don't want a nice, perfectly idiomatic conversation with a chatbot that ends up recommending the usual shallow, keyword-stuffed, SEO-optimized crap of top search results. Maybe created by some ChatGPT-powered bot.
I want products and features created with traditional, boring software and hardware technologies. Something that works all of the times and consistently makes a difference in my productivity and ability to carry out tasks.
Here are a few Google product features that would make me jump with joy:
Markdown in Gmail
native Android support for full app data syncronization in the cloud, so seamless and compelling developers just can't say no
enforcing the requirement that all new Android apps must work also on chromeOS
true, substantial (at least 8X) optical zoom in Pixel phones
native Docker support in chromeOS
full Android app mirroring on chromeOS, not just messaging
Between Google chasing the latest hype, and advanced users always getting a back seat in product decisions, something tells me I'm not going to see any of these features anytime soon.
Larry Masinter, one of the original Medley Interlisp designers and current maintainers, invited me to join the project. I'm really honored and happy to have accepted, as I'm meaning to explore Medley Interlisp and use it as my main development environment. Also, I hope to give something back to this incredible retrocomputing project.
My first pull request was merged. It's a trivial change that adds a new section, written by another maintainer, to a page of the project website. But I'm pleased as it helped me familiarize with the development process, the code base, and the toolchain.
I'll initially focus on documentation and outreach, a byproduct of my learning journey. In the long term I may venture into contributing some fixes or additional Lisp code.
When I researched a printer for my first Chromebox in 2015, going with a Brother HL-2340DW seemed like a no brainer.
It was one of the few affordable, wireless laser units with support for Google Cloud Print, then the only way of printing from chromeOS. Print quality is okay. But I never liked the HL-2340DW much as it's slow, and it often went into a deep sleep mode from which it was difficult to wake up.
When Google discontinued Cloud Print in 2020 I could still use the HL-2340DW from my new Chromebox via the IPP protocol, but the unit would often ignore print jobs. When printing from the Chromebox the HL-2340DW would acknowledge receiving the data, then just ignore the jobs and return to the ready state. No troubleshooting or combination of restarting chromeOS, the Chromebox, or the printer would help.
One of the times, exasperated, I initiated printing the Google Drive files I wanted from my Pixel 4 XL Android phone instead of the Chromebox. The printer worked flawlessly, and it has been from Android since then.
My first Medley Interlisp blog post was shared on Hacker News, got over a hundred upvotes, and ended up on the front page, where it still is a day later after climbing up to number 5. So far my post received over 24K views, and counting.
I'm really glad Medley Interlisp is gaining some very well deserved attention.
Imagine someone let you into an alien spaceship they landed in your backyard, sat you at the controls, and encouraged you to fly the ship. This is the opportunity Medley Interlisp offers.
The project
Medley Interlisp is a project aiming to restore the Interlisp-D software environment of the Lisp Machines Xerox produced since the early 1980s, and rehost it on modern operating systems and computers. It's unique in the retrocomputing space in that many of the original designers and implementors of major parts of the system are participating in the effort.
The project started slightly over three years ago and is little known, so I stumbled upon it by chance. I was immediately drawn to Medley Interlisp for two reasons.
My experience with Lisp
First, I have a soft spot for Lisp. Since the early 1990s, for a couple of decades I extensively used several Lisp dialects and systems such as Scheme, Common Lisp, and Emacs Lisp.
I read a lot about Lisp Machines and their lost wonders but never got a chance to use one.
Although I always wanted to try such a system, hardware solutions are impractical and there are only a handful of Lisp Machine emulators, usually limited or difficult to set up. So the second reason Medley Interlisp caught my attention is, unlike similar efforts, it offers a complete environment that's easy to install and run. It can even run in the cloud and be accessed in a browser, here it is on my Chromebox:
Why Medley Interlisp is different
The Medley Interlisp project is significant because it provides access to the full Interlisp-D software system. Working with a Lisp Machine environment feels like an encounter with an advanced civilization, so the spaceship metaphor is fitting.
A number of features make the image-based environments of Lisp Machines unique and still largely unparalleled by traditional file-based development envi
