I finally compiled and expanded on all my various blog posts, tutorials and other Python goodness into a book: Working with Python. It is available as a free pdf download at: https://mkaz.blog/working-with-python/
It's grown over a dozen or so years and when I finally decide to compile into a book, everyone now uses AI and no longer read and learn from books but instead through LLMs.
Fantastic. I wish I'd started on writing something like this years ago (although I'd wanted to teach explicitly rather than having a collection of how-tos).
> when I finally decide to compile into a book, everyone now uses AI
This is part of what discourages me from starting now, sadly. That, and having more concepts for actual Python projects than I know what to do with.
Great book! I already use python for some simple projects and your book is in the perfect level of practicality that I need.
Thank you!
Suggestion: create an epub version as well. It would be awesome to read it on a kindle or other e-ink devices.
> everyone now uses AI and no longer read and learn from books
Not me, I read the shit out of documentation and also books like yours which distill knowledge from professionals down to a bunch of useful points. I have never not learned something (even if I knew and forgot it) from reading a good book about "Working with X".
Thanks for your hard work, and for giving it away to others gratis.
Agree, it seems like a mismatch of features and audience.
As someone who doesn't know much about working out or what exercises to do this sounds like a good app. I need help, but picking based on muscles is off. My thought and goals are not by muscle group, but losing weight or getting more toned.
Conversely, someone who knows what muscle groups they want to target, probably already has some sense of the exercises to target and thus less likely to need the app.
For someone like me, who have had an accident (dislocated kneecap) and need to focus on special muscle groups to compensate it makes sense to search for exercises based on muscle groups.
Interesting, although I wouldn't say that's the audience that the author says he's targeting.
Also - for most people who had accidents they'd probably rather click on "Dislocated Kneecap" and then have the software suggest exercises to help with that condition - vs needing to bring that knowledge to the app.
great feedback and I agree that (i forgot tbh..) most beginners don't think in muscle groups, but in goals like "lose weight" "feel better", "beginner friendly", etc...
The goal is to make the app more welcoming by offering goal-based (or filters,let's see) entry points like "fat loss" "beginner full-body" or "3x/week routine" and not require anatomy knowledge to get started.
The muscle filter will just be one of many ways to browse, not a gatekeeper i guess. Thanks a lot for highlighting this!
You can also create a writing space in your browser using a data URL, paste the following into your address bar and boom! instant editor. You can bookmark it and will always be available.
```
data:text/html, <html contenteditable>
```
Plus bonus ephemeral, close the tab and your notes are gone with them.
Here's an old post with examples using data URL adding style: https://news.ycombinator.com/item?id=6005295
That works too, for sure. I’ve paid special attention to certain features like how task creation works in Markdown with Ephe. Would love for you to give it a try.
A basic command-line tool to track your books read, show some basic stats and charts. All stored in a sqlite3 database and you can import from a Goodreads export CSV.
I used mostly Claude 3.7 but recently been trying out Gemini 2.5 Pro which works really well too. AI really makes projects like this easy, it's a basic tool and doing simple tasks of converting or selecting data.
I believe they are saying for the rest of the web, where all the information and millions of web pages. If no one goes to those web pages anymore and digests all the info via chat interfaces (ChatGPT, Google instant answers, etc...) then what's the point of having web sites.
I think pointing at HTML and CSS is odd, having complete individual sites that no one visits is the impact
I see you point. My line of thinking was, SEO has become such an optimization game that websites have become less usable in order to rank higher in many cases.
So, it would be logical to think over time, content friendly to AI bots will be favored over anything optimized for humans.
Now, I don't know what incentives will be put in place for content creators by AI companies, because if no one is viewing pages, then no one is viewing ads so what pays for all this content... no idea?
Maybe AI companies will need to adopt a Spotify type of payment model where authors are paid cents each time their content is referenced.
Ah. In that case, I reject the premise. Chat-based UX is kinda awful, and while some companies are pushing it, I don't know many people who actually thrive in such apps. I doubt such UX will result in people no longer visiting web pages.
I absolutely agree that the chat UI is horrible, and somehow chat = AI to do many people right now.
I hope it will change, but I have been observing more and more how people, out of share convenience, just go to chatgpt for anything now, they don't even bother with a Google search.
Obsidian is free for individual use. The $50/yr is a commercial license.
They also have a $4/month sync product (sync across devices with e2e encryption), but you can use icloud, google drive, etc too.
Thank you, an excellent tool and works well so far. It was a nice easy setup, you already convinced me to purchase the lifetime pro to say thanks. Gotta support indie developers in an ecosystem I use.
A minor bug I noticed, you don't have media controls on the demo video on your website. I had to right-click and select play, clicking the video didn't start it either.
My photos linked above are licensed under creative commons, so feel free to use any. I'd be interested in helping with the project too and can supply higher res if needed.
Also, look for Silicon Valley '07 by Gabriele Basilico, it's a photo book - not a huge coffee table book but pretty good with about 80-100 photos from 2007
I really love this. It makes me feel weirdly nostalgic, even though I've never been to California, or indeed the US. As a tech-obsessed teenager in the 90s, some of the older companies here felt almost mythical at the time. Plus lots of California generally would show up in so much pop culture, films, music videos and so on, and the sort of vintage-y vibe I'm getting here feels really reminiscent of that.
Nice work. That Sun Microsystems one should be the same as the Facebook one, but the back of that sign ;) (as I see you've included in the FB writeup. Good show)
Thanks! The majority of them are digital, most with a Fujifilm camera either X-T1 or X-PRO2 depending on when they were taken. I do shoot a fair amount of film, so a couple will be Kodak Portra. I prefer the film aesthetic so my digital presets aim for that look.
I got a X-T1 about 6 months ago from eBay for around 250€. It does a wonderful job of films simulation.
Still very surprised that a 10year old camera can take such wonderful pictures.
It's grown over a dozen or so years and when I finally decide to compile into a book, everyone now uses AI and no longer read and learn from books but instead through LLMs.