I'm just going to touch on a few points. The TL;DR is: simplicity is subjective, and people should use the tools that work for them.
On Thu, Jun 5, 2025 at 9:40 AM Daniel Maslowski via 9fans <[email protected]> wrote: > [snip] > Let me explain. I am using machines to assist me with what I do, and if one > doesn't fit, I will change it. If the system doesn't offer what I need, I > will pick a different one. I have worked with all the major ones, be it > Windows, macOS, Linux or FreeBSD, and they all have their ups and downs. I > want mostly the ups, and I anticipate everyone does. As you should, IMHO. Computers are tools. In other areas of life, I use a lot of tools, and I own a lot of tools. Some of those tools are very high quality (I own some really very nice chisels for woodworking, for example); some are cheap junk. But something that I've learned is that the most effective tools are those that _let you do some job that you consider important_. If plan 9 does that for you, great! If it doesn't, but some other system does and that other system is accessible to you, then that's also great! I'm no longer going to judge someone because they use a different text editor than I do or whatever; do what you need to do. But that said, someone needs to build the tools. Someone who can do so skillfully is filling a very important role; much as I might want to, I am simply not equipped to forge my own chisel, but I'm very happy to consume someone else's work here. We should be willing to ask questions: "is this the best tool for the job? How can it be improved?" What constitutes an "improvement" is subjective, of course, and I think that that, along with what dimension the improvement is made, are the sources of much of the impedance mismatch between plan9 and other systems. But still, it's valuable to have people asking those questions, experimenting, making prototypes, evaluating them, and reporting the results. That is, doing research. Without that, we run into orthodoxies around our tools. > [snip] > Now you may know that I've got a decade of experience in web development, a > field that exploded over the years, and I also acknowledge that many of you > simply hate it. Why though? Does user experience design discomfort you? I do > understand that a lot of the development is not perfect, and I also see how a > lot of the web has turned into ad platforms. We can easily agree that those > aspects need improving. So back to the OS. I think some people are turned off by the entire model of the web; I confess, I'm one of those. When it first came about, it struck me that, because the protocol is fundamentally stateless and contained no notion of a "session", let alone granularity of transfer less than that of a "resource" (e.g., a full page), it felt like a huge step backwards from the sorts of rich interactive graphical applications people were building at the time. HTML struck me as both overly verbose and also decidedly anemic; it couldn't even usefully represent mathematical notation! And that SGML-inspired syntax was baroque and wasteful.... The web was so-so for publishing basic information, and while hyperlinks were a nifty idea, it all felt very one-way/batch style. Honestly, using a web browser in the early 1990s, before Netscape 2.0, reminded me of using a 3270 terminal. Over the years, the situation has improved, of course, and now we've got very rich apps running on the web. But those are implemented using layers, and layers, and layers of abstractions and code that are meant to paper-over the original deficiencies, rather than fundamentally address them. So the web is now undeniably useful, but if you're already approaching it from the mindset of exploring a system that's built on the premise of questioning a lot of base assumptions about how systems _should_ be built, I can see where it would remain off-putting. Is anyone rethinking its fundamentals, in the way that plan9 rethought Unix's? If not, why not? Who's job would it be, anyway? > [snip] > And here goes the idea of "simplicity": It isn't simple nor easy to *develop* > those things, but the primitives are simple. On the other hand, it is the > developers' burden to deliver simplicity to the end user. Let's keep that in > mind: Missing out on a decent user experience creates tons of complexity on > the side of the user. Like, say, having tons of abbreviations and little use > of colors and such in 2025, in which we have 8k screens, terabytes of > storage, gigabytes of RAM, touch input, and tons of gadgets in everyone's > hands - that can change. This is the part where, I think, it becomes highly subjective. True story: when Go was created, Rob Pike gave a talk about it internally and said it was a "simple" language, syntactically and semantically. I didn't really think so: it seemed kind of complex to me, so I sent him an email and said something like, "if I want simple syntax, it's kind of hard to beat Lisp. He wrote me back, and I forget exactly what he said, but it was along the lines of, "common lisp is very complex." It is true that some aspects of, e.g., Common Lisp are extremely complex (reader macros, arbitrary properties on symbols, the MOP, CLOS, anything related to path names, etc) but syntactically, it's hard to argue that it is more complex than Go. So my takeaway is that what one considers simple really does vary from person to person. I like the spartan interface provided by plan9; it matches my sense of aesthetics. But people can certainly differ there, and that's fine. I get that some people won't appreciate it, or will find it lacking. Beauty is in the eye of the beholder and all that; some people might even have a different favor. I think what some people are flummoxed by when it comes to these things are folks who discard the rest of the system because, say, they don't like the color scheme in the text editor. Perhaps I've said this before, but several years ago, back at Google, I have a demonstration of using plan9. One of the people who watched was a fairly important person in the Linux community, and I thought he'd appreciate namespaces, the regularity of the interface, and all of that; but after the demo, his only comment was, "the window system looks like it's from 1991." Ok, perhaps, but that wasn't the point. Anyway, I continue to think that it's important to have folks who are asking fundamental questions about the most basic layers of software, and that we've gone way too far down the road towards a homogenous orthodoxy. But not everyone needs to do that, and some folks really do just need to get things done, and that's ok. - Dan C. ------------------------------------------ 9fans: 9fans Permalink: https://9fans.topicbox.com/groups/9fans/Tf84d656c78bbda91-Me87cd34e24d2f807db0f56e3 Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
