Depending on your location, and preparation. My home office room (which also doubled as a workshop and light storage room) used to be 10' x 12', and I was able to sit with my back towards a wall, and my desk facing space. And it was in Brooklyn, NY, not an area known for cheapest sq.ft. It mostly depends on how you plan the room.
Ehh, I find it difficult to distinguish between "taste" and "money". The shelving alone is a "contact us for pricing" situation. Premium items coupled with a too-clean-to-be-used work environment and natural light can do a significant lift in the "taste" department.
It reminds me a little of set dressing in movies. Every sophisticated character owns a chemex, but they use a french press to make coffee onscreen. Harks back to the days of Notting Hill when we had to believe that Hugh Grant ran a failing second hand bookstore while living in a well-decorated house in central London. Do we think the author uses his Teenage Engineering pocket operators, or are they window dressing? Do we need Godel, Escher and Bach as the backdrop for a completely unrelated photo?
People can be multi-dimensional. I’m a sysadmin/developer, yet I played in a symphony orchestra, and still play bass, take photos and read world classics, sci-fi and occasional philosophical books.
Why can’t he make music, read music history or biographies, or do other things?
Do all “software engineers” need to interface with a computer 7/24, Matrix style?
Of course you can, I think the author has taste, is clearly interested in design and I enjoyed looking through the images to see what I recognized. I should say that it obviously looks good, for the same reasons that movie sets look good and why we hire set designers/dressers.
It's also fair game to critique these photos from an artistic perspective. Some are clearly intentionally staged and I argue that the messaging is a little clumsy. Sure, it's hard to avoid if you've filled your space with expensive design objects. Another comparison is cooking blogs where the photographers add visual clutter that looks good on instagram, but is impractical and unrelated to the food being cooked. The space itself is very nice, though you've got to be absolutely anal about keeping clutter down.
I have GEB, TAoCP, Stevens, Crandall/Pomerance, Tannenbaum, Aho/Sethi/Ullman, Schneier, K&R and a bunch of other books on my shelves next to me. About 1000mm worth in total but I could probably trim it down to about 600mm if I stripped out the random extras related to old projects (Rails/JavaScript/Mysql/etc) or stuff you just don't need a book about (Git).
Putting them anywhere else in the house would either be more "showoff" or just less practical. It's true that I rarely ever pick them up but the few times I do I'm glad they're right next to my work desk.
Prices are on the linked page, or in a full price list PDF it links to.
(Though the fetishisation of this shelving seems weird. Maybe as I grew up in the UK, but I associate it with every single public and office building. Every library, every office, every school. It's not what I'd choose for home.)
It's great when you actually do want the flexibility. Not that you need anything ridiculously expensive, though. In a garage or workshop it's great because you can just put the brackets where you want and store long stuff like wood or pipes etc. But if you're just putting up shelves that you're never going to move it's less appealing. That said, I have used it in my study because I don't care how it looks and it's very strong.
A lot of what you pay for with these systems is them staying the same for decades at a time - a factory outfitted with shelving doesn't want to have to replace it all just because they're reconfiguring a part or adding on.
I know a person with a very good taste; his apartment is even emptier and cleaner than this. He's actually good at his job. Some people just find it actually comfortable. I'm not one of them, but these are real people, not posers.
There’s a chasm between them. I have seen people create great things with no money, and people who slaughtered spaces because they don’t have an eye for anything.
Slightly off topic, but what is the best way to build a cross platform GUI app these days, but something with good graphics, typography, etc. I mean a beautiful app. I would prefer to have a shared core in Go, and then something around it to give me the GUI. I know on MacOS it is straightforward to build something beautifully looking with their native Swift toolkit, but not sure on Linux and Windows. Is it better to just use a web view, or perhaps Flutter?
I keep coming back to this and not finding / choosing a solution. It sadly feels like people who are doing this are just going with Electron now. I'm mostly coding in Go now, so I seem to be looking for a similar solution to you.
This page has a lot of cross-platform GUI toolkits, but it focuses on C++:
I've been drawn to wxWidgets (actual native controls on each platform) or JUCE (most of the cross-platform commercial Windows/Mac/Linux audio effects I've bought are made in JUCE). But I've not had a chance to give either a proper try.
Years ago I used to all my cross-platform work with Xojo, a kind of cross-platform Visual Basic. That actually worked well for me, but then you're writing code in the Xojo Basic language, and it had some odd file formats for projects (not just raw text source files like C++ / Go would give me). Once LLMs hit I felt I probably needed to move on from Xojo to something less proprietary: https://www.xojo.com/
I'm actually working on that - it's called Hypen - (hypen.space).
You can build your core in Go or any other supported language, and write the UI in the Hypen DSL.
While desktop is still in the works and should be out in the next week or two, currently the alpha supports Native iOS, Android, Web and Web Canvas, and just like mobile, the Desktop will be _real_ native.
I like Flutter, it's actually not a webview unlike most other frameworks and likely the most mature out of them. I also am keeping an eye on Rust GUI frameworks, Dioxus Native in particular as they're making their own renderer.
The last one I did was using Fyne in Go, which is quite cross platform but its software drawn not native. Its targeting phones as well so its cross compatibility is very good but at the cost of giving you the full complexity of desktop applications, it does not have a highly capable table view for example. Since its written in Go this is what you will develop in.
Otherwise I think its QT and GTK on C/C++ as the other option across the desktop operating systems, neither is native on anything but Linux but they also look OK but I think a lot of people would rather avoid C nowadays for application development.
I have my eyes on that, looking forward to V3, maybe they manage to ship mobile support as well. That would be fantastic. For anyone that doesn't know, it's still a browser based stack.
Only reasonable way is shared core with thin UI layer on top. For Rust there is Crux, don’t know for other languages. Everything else is just compromise, like all Flutter apps I know on iOS are just atrocious.
I like this approach, it's what I had in mind, but Crux doesn't seem to support desktop targets. I know on MacOS you can get nice looking apps with their native toolkit, on Linux you have GTK4 which can be decent looking, but not amazing, and on Windows, I truly don't know. Native apps on Windows look crap to me (without even mentioning the advanced fragmentation in UI toolkits in Windows). Maybe someone has some good examples for Windows and Linux, using native SDKs.
Only partially true: macOS is supported, and one can fall back to the web. But you're right in that native Windows and Linux are still missing.
> Linux
Problem with Linux of course is that it's almost as fragmented as Windows, with Qt and GTK being the main toolkits, but a dozen more if you ask the wrong people :D
I personally don't like GTK, to me it (well, mainly Gnome) looks and feels like trying to copy macOS without understanding what makes it great, but Qt is a toolkit I can get behind…
I see a lot of negative feedback here, but I don't agree with it. This is really fantastic what you have built, especially for longer running agents that are used repeatedly, in which case the initial investment of giving only the permissions it needs is worth the effort. To that end, ability to combine several agents which have different roles, which are narrowly scoped in terms of permissions, would be a very useful feature. Perhaps you could even have an agent or UI overlay driven by AI, which can quickly scope the permissions for a new agent, so that users don't need to do it manually.
There's a very dark scenario where for some reason or another (all out nuclear war or asteroid hit) sunlight is blocked, in which case having stable base load energy production from nuclear would be very useful. I know this is an unlikely scenario and hopefully it never happens, but it's always good to think about tail risks like these.
If sunlight is blocked the amount of people who die due to starvation from crop failures will probably more than make up for the difference in lost solar power energy. That is to say, we'll have much larger issues than a stable power grid to contend with.
If we directed worldwide LED production all into artificial light for farming, and grew whatever was most calorie-efficient, I think we could theoretically feed every human alive if the sun was blocked out tomorrow.
Obviously that isn't what would happen. The poor would starve whilst the rich still fed cows to eat steak.
It doesn't necessarily take a full blockage of sunlight. Extreme weather conditions that create multi week collapse in solar outputs is enough to create grid stress, if one is totally dependent on solar and 24hr batteries.
Volcanic winters are far more frequent than catastrophic asteroid blasts. Disregarding a volcanic winter possibility and its impact is like disregarding the possibility of a pandemic.
True. But if you're working in public policy in a vaguely-democratic country, and trying to get anything useful done - then the public feels vastly more familiar with "giant asteroid wiped out the dinosaurs" than with volcanic winters. So, just like "Zombie Apocalypse (wink)" disaster prep - you go with a "close enough" scenario which lets you achieve some actual preparation.
Surely you must realize that the fuel for nuclear power plants is not more freely available than sunlight. In the event of "all out nuclear war or asteroid hit", you're not getting those shipments from Kazakhstan.
Now that Zed supports remote development, I really hope they can release it for tablets (iOS/Android) so that we can use it as a client for a remote development machine. That would be delightful!
Wondering how gpt 5.5 is doing in your test. Happy to hear that DeepSeek has good performance in your test, because my experience seems to correlate with yours, for the coding problems I am working on. Claude doesn't seem to be so good if you stray away from writing http handlers (the modern web app stack in its various incarnations).
Very cool to hear there is agreement with (probably quite challenging?) coding problems as well.
Just ran a couple of them through GPT 5.5, but this is a single attempt, so take any of this with a grain of salt. I'm on the Plus tier with memory off so each chat should have no memory of any other attempt (same goes for other models too).
It seems to be getting more of the impressive insights that Gemini got and doing so much faster, but I'm having a really hard time getting it to spit out a proper lengthy proof in a single prompt, as it loves its "summaries". For the random matrix theory problems, it also doesn't seem to adhere to the notation used in the documents I give it, which is a bit weird. My general impression at the moment is that it is probably on par with Gemini for the important stuff, and both are a bit better than DeepSeek.
I can't stress how much better these three models are than everything else though (at least in my type of math problems). Claude can't get anything nontrivial on any of the problems within ten (!!) minutes of thinking, so I have to shut it off before I run into usage limits. I have colleagues who love using Claude for tiny lemmas and things, so your mileage may vary, but it seems pretty bad at the hard stuff. Kimi and GLM are so vague as to be useless.
My work is on a p2p database with quite weird constraints and complex and emergent interactions between peers. So it's more a system design problem than coding. Chatgpt 5.x has been helping me close the loop slowly while opus did help me initially a lot but later was missing many of the important details, leading to going in circles to some degree. Still remains to be seen if this whole endeavour will be successful with the current class of models.
Do you an idea of how well these models perform on set theory problems or more niche fields in mathematics? So the model would have to both understand a paper that’s not in its training data, and use this to write proofs.
This is all fairly niche stuff I'm trying it on (well, the first three problems anyway), so yes, it needs me to give it several papers that are not in its training data and use them to write proofs. I would expect my experiences to transfer to set theory problems as well.
There are several things going on even now, 1 hour after your comment. But I appreciate that they list them. That hopefully means that they have a good culture of honesty, and they can improve.
I looked through the issues and basically only ongoing thing is that backup power is not working in one of the data centers (could be a problem). The rest are warnings about planned shutdown of some services and speed limitation of object storage in one location.
I am sure it's luck but we have few hetzner VPSes in both German locations and in last 5 years afaik they've never been down. On our http monitor service they have 100s of days uptime only because we restarted them ourselves.
reply