Three months into the project, I want to take a brief pause.
The low-hanging fruit has already been picked. Every subsequent improvement now demands exponentially more resources. Naturally, I’m not talking about human hours—no single person could pull off a project like this alone—but about AI usage. Agents deliver an impressive price-to-performance ratio, but even so, the cost per task is becoming quite steep.
To be specific: I spend roughly $300 a month on agent subscriptions. That is the hard ceiling of what I can justify spending on a hobby project. To speed things up, I would either have to spend more or downgrade to cheaper, less capable models—which is practically a gamble. The other option, of course, is simply to wait longer; that choice is always on the table. The only difference is that instead of sharing a couple of cool features every week, updates will slow down to one or two a month.
This makes it a great milestone to showcase the project to a wider audience.
For legal and ethical reasons, I cannot publish the entire codebase publicly. Those same concerns prevent me from using the Opera brand name.
Therefore:
otium:about, though legacy identifiers are retained for backward compatibility). Mentions of Opera and Opera Software ASA copyright notices have been preserved where appropriate.branding.ini configuration file has been introduced, allowing anyone to customize the naming for all identifiers.I am fully aware that this won’t offer absolute protection if the project catches the eye of the copyright holders. My strategy is simply to avoid provoking them—by ensuring this work never interferes with how they make their living.
I’m still hesitant to open up a public issue tracker, as I’d quickly drown in bug reports. The primary bottleneck right now isn’t an ignorance of existing bugs, but a lack of bandwidth to handle user support alongside everything else.
Still, if you’d like to follow the project’s progress or help out with development and testing, feel free to join the Telegram channel.