I am incredibly jealous of people for who this works for. Mine just become too unwieldy to manage or work with because they grow out in a crazy fashion.
My "productivity solution" is currently TriliumNotes with three work spaces as 1) Planner with sub notes for year, month, day 2) Brain Dump with subnotes for year and month 3) Projects with sub notes for each project. I manage tasks with Vikunja and then my time with Google Calendar.
It's an absolute mess, but it's the closest I've gotten to a solution that works the way my brain does.
Thank you for sharing. I feel similar to you; jealous this system works for others, sounds like a dream, but too overwhelming for me once it hits some point of no return. Your structure sounds interesting.
I'm genuinely curious how others do not get overwhelmed or sucked into yak-shaving some reorganization of a system like this.
I don't understand the rational for announcing that a vulnerability in project X was discovered before the patch is released. I read the project zero blogspot announcement but it doesn't make much sense to me. Google claims this is help downsteam users but that feels like a largely non-issue to me.
If you announce a vulnerability (unspecified) is found in a project before the patch is released doesn't that just incentivize bad actors to now direct their efforts at finding a vulnerability in that project?
The reason for this policy is that if you don’t keep a deadline upstream can just sit on the report forever while bad actors can find and exploit the vulnerabilities, which harms downstream users because they are left entirely unaware that the vulnerability even exists. The idea behind public disclosure is that downstream is now made aware of the bug and can take appropriate action on their side (for example, by avoiding the software, sponsoring a fix, etc.)
"Don't announce an unpatched vulnerability ever" used to be the norm. It caused a massive problem: most companies and organizations would never patch security vulnerabilities, so vulnerabilities would last years or sometimes decades being actively exploited with nobody knowing about it.
Changing the norm to "We don't announce unpatched vulnerabilities but there is a deadline" was a massive improvement.
Maybe for a small project? I think the difference here is rather minimal. Everybody "knows" code often has security bugs so this announcement wouldn't technically be new information. For a large project such as ffmpeg, I doubt there is a lack of effort in finding exploits in ffmpeg given how widely it is used.
I don't see why actors would suddenly reallocate large amounts of effort especially since a patch is now known to be coming for the issue that was found and thus the usefulness of the bug (even if found) is rather limited.
Echoing what everyone else has said here - awesome site, love how fast it was.
I did notice that when I put in a single book in a series (in my case Going Postal, Discworld #33) that tended to dominate the rest of the selection. That does make sense, but I don't want recommendations for a series I'm already well into.
Also noticed that a few books (Spycraft by Nadine Akkerman and Pete Langman, Tribalism is Dumb by Andrew Heaton) that I know are in goodreads and reviewed didn't show up in the search. I tried both author's name and the title of the book. Maybe they aren't in the dataset.
It did stumble with some books more niche books (The Complete Yes Minister). Trying the "Similar" button gave me more books that were _technically_ similar because they were novelizations of British comedy shows, but not what I was looking for.
For more common books though it lined up very well with books already on my wishlist!
Yes I would say the handling of series is probably the biggest problem. Once my test metrics got to a point I was happy with and my quality spot checks passed (can I follow the models recommendations from one generic history book to Steven Runciman, also making sure popular books don't always dominate the results), I was ready to release because I had been working on this project for so long. The solution is probably using the transformer model to generate 100-200 candidates and then having a reranker on top.
Not just series, but I seem to mostly get a list of other books from the same authors.
The recommendations from other authors are good, but as far as I can tell I’ve read every single one of them.
Continuing to aggressively add everything it recommends eventually does seem to result in some interesting books I wasn’t familiar with, but I also end up with more and more books that are of zero interest to me.
For what it’s worth, I started with:
Infinite Jest David Foster Wallace
Europe Central William T. Vollmann
Gravity’s Rainbow Thomas Pynchon
White Noise Don DeLillo
One Hundred Years of Solitude Gabriel García Márquez
It is possible that there simply aren’t many books like these in existence, so the pool of relevant recommendations gets exhausted fairly quickly. I’d guess trending towards unrelated popular books is also just a feature of the source data, that largely sums up my experience with goodreads anyway.
Very cool project though. I did end up ordering a couple of new books, so thank you very much.
Releasing is the right choice, well done with this it’s really cool.
I’ve only had a short play but a solution to this problem might be to show authors rather than books. Or select authors outside of the list the user has shared and then a top n (1,3,5) for each of those.
I feel like that’s how you’d recommend to someone else - type of book -> unknown author -> best matching few books from them.
After that the other side would be trying to find some diversity (if you think I’d like author X, personally you might suggest three different styles of book from them rather than three very similar books from them)
Going Postal is awesome. The flood of the mails and the test to be the post master where you would have a slide mail into the hole where a vicious dog is barking.
Relatedly there's a Map Men video on why north is up. [0] I don't buy the whole top is 'good' and lower is 'bad'. I think the bias is just a lot of the groups that made maps were located north(ish) and traveling roughly southward which made it a convenient orientation, especially during the age of sail.
The discomfort visible in many commenter's reactions is telling, even if it isn't a good vs bad dichotomy.
The fact of the matter is that any data visualization brings with it some advantages and drawbacks. This can be projection, orientation or centering related. Acknowledging these drawbacks can be useful, and so is trying out alternative representations of the data every so often.
One thing I think can be acknowledged is that the poor job traditional maps do of representing Africa has affected policy towards countries in the continent for the worse. For instance misguided infrastructure projects.
And looking at the map, it would be hard for those map makers not to be north(ish) since the South is mostly ocean. Not too many civilizations that have sprung up in the ocean.
This is kind of one of the points Jonathan Rauch made in his book "Cross Purposes"[0]. He talks about how the common zeitgeist went from being christian and conservative to being christian because you were a conservative and because of that people are treating politics with the same fervor that they would have treated religion in the past.
I also want to shout out The Sprits which serves as a book club for cocktails. Very good if you're just exploring. Each week you get a cocktail and a themed playlist to go with it, plus some other random musings.
It's not just information density but rather intended use design. A lot of engineering/manufacturing parts suppliers tend to have good information dense websites that are really catered to their customers for finding parts.
Take mouser.com, digikey.com, grainger.com rockauto.com or mcmaster.com. They all have a bit of a "landing page" but once you go to search for parts you've got something that was really designed to be an intuitive parts search. Compare that with jameco.com which competes with mouser/digikey but has a more classic webshop search system. It’s a bit more frustrating to use.
Some news sites also do a great job of presenting headlines and highlights well in a small area. I think semafor.com is probably my current favorite, but I'll readily admit that it's not the most information dense.
CAD software also tends to be good at this, but that might be just because the UI has chugged along since the 90's. AutoCAD/Inventor/Solidworks/SolidEdge/KiCAD/Altium/Virtuoso are all great examples where if you've got prior experience with them (or even similar software) you can sit down and quickly get up to speed on a project and see what's been done. I think the distinction is that a lot of software/websites are designed to keep the average user focused on a single aspect and so they are designed to either remove or hide the complexity but for more “professional” level tools you need all that data and information. You can probably blame (for better or for worse) material UI for a lot of this spaced-out thing. In my mind that was the first mobile first UI scheme that really took off and it's basically influenced everything that's come sense then. Computer first software might be your best bet to get some examples. Because a lot of the web is mobile first/mobile forward now you probably aren't going to find a lot of examples on that. I would love to see examples of information dense mobile first sites.
A few other examples I just wanted to brain dump:
- labgopher.com
- tld-list.com
- The Bloomberg Terminal
- Ghidra
- Most plane cockpits, especially modern fighter planes if you ever get to see/sit on one.
- A lot of “professional level creative software” – Reaper, Affinity
The Complete Yes Minister, a novelization of the TV show, but probably more approachable. A kind of "high brow" skewering of politics and government. There's a lot of interplay between politicians and civil servants that mirrors some play between politicians and people, especially in the case of "systemic lag".
Erasure by Percival Everett. A book on racial conformity and expectations. A weird case where the movie (American Fiction) might be better than the book. Pretty easy and quick read however. I don't know what it was but this book has stuck with me ever since I've read it.
The Code Book by Simon Singh. This is the book that got me into cryptography. It's a bit old and outdated now (published in 1999) but it was responsible for forming a lifelong interest in me.
Arriving Today: From Factory to Front Door by Christopher Mims. The premise was supposed to be tracking a product from production to consumer, but then the COVID happened. The book turns into an exploration of how just in time production and supply lanes work today.
The Dictator's Handbook by Bruce Bueno de Mesquita. I had a professor who was friends with BBM, so when we were discussing selectorate theory we actually got to meet him. At it's core this is a cynical book about realpolitik, talking about how leaders get in power, stay in power, get money and foreign aid, and deal with revolutions and war. It is very political focused but the theory can be abstracted out to most big organizations. It fundamentally changed the way I look at interactions between countries. This is 100% a more mass market appeal book than the original paper (and imo a bit dumbed down) but everyone I've recommended it to has come back appreciative about it.
My "productivity solution" is currently TriliumNotes with three work spaces as 1) Planner with sub notes for year, month, day 2) Brain Dump with subnotes for year and month 3) Projects with sub notes for each project. I manage tasks with Vikunja and then my time with Google Calendar.
It's an absolute mess, but it's the closest I've gotten to a solution that works the way my brain does.