Firefox profiles are still janky and I had a lot of issue switching and managing two profiles effectively. Specifically the biggest problem I had was that clicking on work links would not open them in my last used profile (work profile). They would always open in a default profile (non-work).
I ended up keeping chrome just for work, and using firefox for personal life.
Then I grabbed Browser Tamer and set up an AHK v2 script that, when I click on and focus any browser window, it executes browser tamer's CLI to update the default browser. Thus I get the behavior of "open links in last used browser", which is the correct browser for whatever link I click 99.9% of the time.
OK, that mental overhead of "last used browser" would drive me nuts.
My solution is: The system has a default browser that opens default links. On my work machine I have a different browser where I am for example logged in to my private github account I just never want to open a clicked link there anyway. Copy/pasting 2 times per day is fine.
so youre just manually doing what i automated but that 1% of the time when it opens in the wrong browser is too much mental overhead? if youre at work and you tap a link you were almost certainly last using your work profile so its nearly always right and its not something i ever have to think about
When I want that sort of separation, which I do between work and play, I run browsers (and everything else) as different users. That works with any browser and I don't even have to worry about bugs in profile or container separation, and it reduces (though of course didn't remove) the chance of idiot here using the wrong instance for the wrong use. Heck, where possible I even use a separate machine. DayJob provide a PC on the office that I remote into (via VPN+RDC) for work purposes, so the contact point between that work and everything else is minimal (in fact my main desktop is a VM I "remote" into, I only use the base metal when I take have too which is usually things unhappy running that way (Bambu Studio and games, which do not like the lack of faf free access to the GPU)). You can still access everything from one machine, or even have the different users instances on the same desktop (this does reduce the barriers a touch though).
The only real cost is that running things this way eats more memory, but I've not experienced OOM issues for years away from deliberately small VMs (for testing or small sever tasks) that turned out to be too small.
IIRC you can use the -p flag to open that menu on launch. It also opens when you haven't set a default profile. And it's possible to access other profiles via about:profiles.
I recently tried Vibe (https://github.com/thewh1teagle/vibe) from a recording of a meeting taken on one side. It was able to identify the speakers. As Speaker 1, 2, etc. But still useful to see.
I fall into the 25 year of experience category. Probably a few more. For me, this agentic coding couldn’t have come at a better time. I still love thinking about solutions to problems and creating those solutions. I’m becoming less and less interested in the implementation details of those solutions.
I tend to use Claude Code in 2 scenarios. YOLO where I don’t care what it looks like. One shot stuff I’ll never maintain.
Or a replacement for my real hands on coding. And in many cases I can’t tell the difference after a few days if I wrote it or AI did. Of course I have well established patterns and years of creating requirements for junior devs.
We are in a similar length of experience, but weirdly as I got older, it's the opposite for me: I got more particular about clarity, readability, especially in the context of handling edge cases. The 10% of situations that require 90% of effort. My new hobby is a codebase that can read as a business rulebook.
Both of my wife's parents died with cancer in their 60s before MAID. These days we're seeing the parents of others, and even our peers choosing MAID. It breaks my wife's heart every time that MAID wasn't available to her parents.
I work on a fully open project that sustains itself… And we really push for a ticket before any major work is done just to make sure it fits our scope.
At the end of the day most contributions come from people as part of their job, and they move on. We’re left to maintain it.
I don't want to hijack this thread, but what does Django do?
I ask, cause I have a service that is a Rust backend (kind of specialized in what it does) + a SolidJS frontend... But I've seen a potential area for growth, I didn't imagine at first, and that would be stuff like user profiles, authentication, and managing stuff. And I really don't want to cut all this more "CRUD" like stuff in Rust and an SPA. I don't know Ruby, but have done some rather large Python projects in the past, just never ventured into Django land.
Out of the box Django offers: authentication, database management and migration, a web server, a template engine, URL routing, and probably more - pretty much everything an app needs in a single framework.
Django also has some GIS tools builtin if you need to do anything with geographic data. It has a solid library of error handling, messages, and web security tooling you will probably need, and, the third party ecosystem, both Djano specific modules and the wider Python ecosystem are very powerful.
It comes with a built in admin portal that lets you work with all your data in list views and detail views, with sorting, filtering, and searching. For a lot of scenarios you can get away with just using the admin portal (like we have accounting people log in there and import CSV files, etc).
It’s great for managing a data model, setting up a database schema and managing the migrations for you. And once that is setup it’s pretty trivial to enable API endpoints to do all the CRUD-style stuff, if needed.
Nothing against Django, just suggesting another option: Laravel offers the same "out of the box" tools that Django does. Django is built with Python, Laravel with PHP. Someone will be along in a moment to start the flame wars about why one or the other or both totally suck, but for your purposes it's most important to choose the language and approach that will be most comfortable to you.
Laravel's first-party ecosystem, both paid and free, on top of the framework is quite widespread. This also brings a lot of people to the party on the third party side with excellent tools like Filament and Ploi. I don't see this nearly as much for Django. Have a look under the Ecosystem tab on the Laravel homepage. Within a few `composer install`s, you can get:
- Local dev environment (i.e. actually running the app): Herd, Sail
- Stripe/Paddle support: Cashier, Spark
- Starter kit scaffolds including all the usual auth goodies (e.g. 2FA): Breeze, Jetstream
- View layer alternatives to plain old HTML views: Livewire, Inertia
- API and Social logins: Passport, Sanctum (for SPAs), Socialite
- Fulltext search: Scout
- Robust APM: Pulse, and Telescope on dev
- Admin panels: Nova, and 3rd party Filament
- Websockets: Echo, Reverb (coming soon)
- Infrastructure, both traditional and serverless: Forge, Envoyer, Vapor, Ploi
- Completely alternative PHP runners to replace php-fpm: Octane with Roadrunner, Swoole, or FrankenPHP
- Automated browser testing: Dusk
You can find Django packages for these things too, but I struggle to see many of them coming from the actual Django team. I also think the Django Admin solution is very neat for getting off the ground, but doesn't seem as robust as Nova or Filament.
Lastly, call it my bias, but Python's overall lacklustre developer experience with package and environment management (compared to Composer and Ruby gems) also affects Django as a result.
Also PHP has Laravel Spark[0]. They basically bootstrapped a SaaS but for a price. Not sure if it's worth it, but it's from the guys who made Laravel, and everyone only has good things to say about that so...
Kopia supports zstd compression, I found it to be within +10% to -5% of the size of my Borg repo.
It also has extensive support for ignoring stuff and it works very well.
I still use Borg because its policy of expiring older snapshots is more useful for me, but Kopia is extremely solid and I would use it any day if I didn't care that it doesn't actually keep one monthly backup for the last 3 months as Borg does (it decides which older snapshots to keep with another algorithm; it's documented on their website).
