In Europe, television remote controls had a magic text button. Years before the Internet entered homes, pressing that button brought up digital information resources with hundreds of regularly updated pages. Living in Ireland in the 1980s and 1990s, my family accessed the national telephone text service—Aertel—several times a day for weather and news alerts, as well as things like TV program guides and airport flight arrival updates.
It was a great system: fast, low bandwidth, not affected by the user load, and delivers readable text even on analog television screens. So when I recently saw it was the 40th anniversary of Aertel’s test transfer, it revived a thought that had been bouncing around in my head for years. Can I do a ham-radio version of teletext?
What is Teletext?
First developed in the United Kingdom and broadcast to the public by the BBC under the name Ceefax, teletext took advantage of the quirk of analog television signals. These signals transmit video frames as lines of light and color, as well as some additional blank lines that were not displayed. Teletext piggybacked the digital signal on these backups, transmitting a carousel of pages over time. Using their remotes, viewers typed in a three-digit code for the page they wanted. Usually within a few seconds, the carousel would rotate and display the desired page.
Teletext created text that was unusually readable in the 8-bit era by expanding alphanumeric characters and translating new pixels by looking at existing pixels that touch diagonally, and adding white space between characters. The letters of the pictures were not translated, and included blocky pieces known as sixels in their 2-by-3 arrangement. My modern entertainment uses the open source font Bedstead, which replicates the look of teletext, including graphic characters. James Provost
Teletext is composed of letters that can be one of eight colors. Control the codes in the distribution of characters, choose colors and can reproduce effects such as overlapping text and double-height characters. Text quality was better than most computers could handle at the time, thanks to the SAA5050 character-generating chip at the heart of teletext. Although the characters are stored inside the chip in 6-by-10-pixel cells—fewer pixels than the typical 8×8 pixel cell used in home computers in the 1980s—the SAA5050 interprets the extra pixels of alphanumeric characters on the fly, making the effective resolution 10 by 18 pixels. The trade-off is very low-resolution images, including characters that use the set of 2-by-3 block pixels.
Teletext screens use a 40-by-24 character grid. This means that a kilobyte of memory can store a full page of multicolored text, a fraction of the memory required for the same amount of text, for example, the Commodore 64. The BBC Microcomputer took advantage of this by putting the SAA5050 on its motherboard, which can be found in one of the computer’s graphics systems. Besides raw graphics, some educational games used this mode, especially Grandma’s Gardenwhich filled a similar cultural space among British school children that The Oregon Trail for their US counterparts.
By the 2010s, most teletext services had stopped broadcasting. But teletext is still remembered by many, and those who like to keep it alive, recover and archive old content, use Internet-based services for current news feeds, and develop programs that make it possible to create and display teletext on modern TVs.
Bringing Teletext Back On Air
I wanted to do something different. Inspired by the way BBC Micro chose teletext for its purposes, I thought it would make a great radio protocol. In particular I thought it would be a digital counterpart to slow scan television (SSTV).
SSTV is an analog method of transmitting images, usually consisting of banners with ham-radio call signs and other messages. SSTV is fun, but, as its name suggests, it’s slow—the most popular protocols take less than 2 minutes to send an image—and it can be difficult to get a perfect image with readable text. For that reason, SSTV footage is often broadcast multiple times.
Teletext is still fondly remembered by many.
I decided to send text over the phone using the AX.25 protocol, which encodes ones and zeros as audible tones. With VHF and UHF transmissions at a baud rate of 1,200, it can take 11 seconds to send one text screen. For HF bands, AX.25 data is usually sent at 300 baud, which can result in an unacceptable 44 seconds per screen. When a teletext page is sent repeatedly, any missed or corrupted lines are filled in with new ones. So in a little over 2 minutes, I could send the screen three times over HF, and the receiver would automatically integrate the data. I also wanted to build a system in Python for portability, with a page editor, an AX.25 encoder and decoder, and a monitor to display the acquired images.
The reason why I haven’t done this before is because it requires digesting the AX.25 standard and teletext’s spec, and then translating them into the software, which I never seem to have time to do. So I tried an experiment within an experiment, and I turned to vibe writing.
Despite the popularity of developer vibe writing, I have a reservation. Even if concerns about AI slop, environment, and memory retention were off the table, I would still worry about relying on centralized systems that vibe coding brings. The whole point of a DIY project is to, well, do it yourself. A DIY project allows you to do things for your own purposes, not just to work within someone else’s profit margins and policies.
However, criticizing technology from afar is not fair, so I directed Anthropic’s Claude towards the AX.25 and the teletext specifications and told him what I wanted. After about 250,000 to 300,000 tokens and many nights of going back and forth about bugs and features, I had a complete system that worked without writing a single line of code. To be honest with myself, I doubt that this program—which I call Spectel—would have ever happened without vibe coding.
But I didn’t read anything new about how teletext works, and more about AX.25. Updates are dependent on my payment of Anthropic fees. So I’m always deeply ambivalent about the vibe code. And the last test remains in any case: trying Spectel on the HF bands. Of course, that means I’ll need willing partners outside of the ether. So if you’re a ham who’d like to help, let me know in the comments below!
From Your Site Locations
Related Topics on the Web
