Alternative OS

OS market is market like any other market, it does not have any special properties.
Desktop segment already has big competitors in form of Windows, GNU/Linux and OSX.
If you want people to switch you have to provide them with superior alternative that will make switching to new platform worth it.
For example some developers switch from Windows to GNU/Linux because GNU provides them with superior development tools.
People use Android instead of GNU/Linux on mobile phones because Android provided them with superior UX on touch-only device. Some people tried to create GNU/Linux phone later but they did not have any "killer"/superior feature that makes people want to instantly switch to new platform.
Marketing and sales are 'force multipliers' for features on the market. Example of this would be Windows Server vs Linux where large companies take
As with many products that are successful both Windows and Linux appeared at the right time, when people wanted to use what they are offering and when competition for stuff that they are offering was low. Windows appeared at start of GUI era and offered GUI to non-Apple users. Linux appeared when there was no free (as in costs no money) UNIX based OS. They did not have to compete with existing products to take over market for services that they provided.
New/ alternative OS is more likely to be something specialized at the beginning, like smart-home OS or IoT OS and then spread to desktop/server after initial success in low competition environment.

I fucked up.
Example of this would be Windows Server vs Linux where large companies prefer WS just because MS has better marketing.

OP here. I ask because it seems that Windows and OSX are both getting worse the last 7 years, the entire Obama Presidency too. Also, when newer software comes onto the market older more popular products have real trouble adapting. For example, Microsoft and Apple seem to be trying to copy Google Docs success with collaboration. Are they succeeding? I don't use Office or iWorks to give an answer. It seems that an opening is being made for a new operating system to come onto the market. No one really wants Windows 10 and if a credible alternative exists then people might switch.
I use SmallTalk and Lisp because to compete against Microsoft will require a language that allows for higher productivity than C. What is the point of a new OS based on C?

I bet someone is working on an OS.js right now, we just don't know about it.

HaikuOS

Yes. There are more than a few OSs for real-time for instance: en.wikipedia.org/wiki/Comparison_of_real-time_operating_systems . But I doubt a different OS could get past niches and into main use unless it has some killer feature that the others can't copy (how?).

Unix is the correct answer to everything. It's just a matter of removing cancer and starting anew for something better.

Unix is cancer and the language used to implement it. I can't wait until the boomers are dead and the C cult can go with it.

They won't succeed, because anything remotely interesting is experimental in nature (TempleOS). For something to become a commercial success, it has to be dumbed-down and pumped full of bling. Also, it has to run all the latest commercial games and facebook apps. The two goals are diametrically opposed.
Even in the past, very good OS didn't make it in part because they just got bad publicity or marketing, or didn't run 100% of the software people used at work, where IBM & Microsoft had an monopoly.
And now the situations is more dire, because the hardware is overly complicated, so there are many more hurdles and time investment needed to try new ideas. Even on well-established OS, hardware drivers are still a problem unless you're Microsoft.

...

Goddamn, Holla Forums

That's what Plan 9 was supposed to be, a Unix with network, graphics and simplicity in mind. Unix was born simple but was corrupted, by the time that symbolic links were introduced, it was more bloated than Multics, thanks to HP, IBM, Berkeley and others.

Unix is a virus so powerful, it killed it's own decedents.

Are there any good books on applying Plan9? I've wanted to learn about it and its potential uses for a while now but don't know where to begin.

I just want to know all about the system itself and how to write programs on it that take advantage of the system and its capabilities.

I hear their C is a different dialect too, is it good?

I think it's pretty cool and better in some areas,

Windows 10 has great UX.

So does your mother.

cat-v.org

Plan 9's C is pretty much the same as C89, but with full UTF-8, can only be statically compiled (I can be wrong on this) and you can't go overboard with #ifdefs (which is good, ifdefs are harmful they make your program impossible to predict and hard to debug).

Do you mean like the integrated keylogger or the pre-configured adware?

Common Lisp really needs a clean-up & to extend some base tooling to the standard, CL21 is a good start on the first bit; The featureset of Roswell is decent on the latter.

I really, really want a modern 'LispM-like' environment and/or OS; I tried for now 5 years & really hard for 3 to get 'used to' Nix ... just can't, just so shite when you realize how good Slime + Emacs is as a devel environment compared to anything else and how even shite that is compared to what could of been assuming it was turtles all the way down.

My first public opensource contribution was to Guix & within a year I lost hope. Guile & Guix are not designed to be independent of a Unix-like system & at this point, there is far too much traction for ALGOL tech that it's not worth forcing this weird hybrid approach.

The average consumer only cares about one thing:

"Can I use facebook?"

As long as an alternative OS is good enough to connect to the internet, load facebook, and also handle uploading pictures and videos and whatever bullshit you want to post up there, it will succeed.

Fuck SmallTalk, though I would love to see a modern SchemeOS (which is basically GuixOS, i guess).

The Common Lisp compiler is pretty fast already, and for Scheme there is Larceny which is pretty much on par with clisp (it's on github too).

Although this meme is hilarious, it's fairly accurate. It would be nice if the GuilEmacs variant could get some more people involved so Emacs can have a faster language behind it's inner workings.

Guix would be far more interesting if they were writing core utilities in Scheme. Core utilities in scheme has interested me for a while as it seems to be a fairly ideal way to look for performance issues between Scheme and C.

Speaking of which, are there any implementations of Lisp that don't rely on a garbage collector? Tuning the GC similar to how Golang does might not be out of the question to achieve some performance/latency improvements but if it wouldn't need to be there at all then this whole concept might go in a different direction.

That is not right, most modern consumers buy tech things so they could brag about them and feel superior to other people. Companies like Apple exploit this by increasing price to make their devices more 'exclusive' and people usually buy them as symbol for "I have large amount of money."

Op, yeah check out 'bone lisp' & 'carp'. Both pretty new and underdeveloped at this point -- but somewhat promising.

Regarding coreutils, talked to Ludo briefly a longwhiles back about this; He's expressed little to no interest about such things. I'm of two minds about this, but don't blame them how little their core team is still.

Huh.

Why are you a pasty diabetic with no friends?

Why are you a virgin in your 20s?

Wow. It really is the answer.

Spoken like a true Intel Itanium/Windows Vista astroturfer.

GNU stands for GNU Not Unix and Linux was never meant to be a Unix replacement, sure its similar but if you want a Unix system use BSD or Plan9

Emacs + Stumpwm + Slime is a decent environment though all & all (as a Tiling Desktop Environment). But I really want to be able to write system drivers in pure CL for example.

multics revival when

Yeah but the devs weren't able to implement graphics even to the standards of the 90s, so it was naturally going to fail. If someone could give a plan9-based system a pretty UI plus compatibility with certain graphical programs like firefox and gimp, it would develop a large and thriving community by alternative-alternative OS standards almost overnight.

First of all, you didn't give a definition of success. People still harp on about Linux being a failure because muh desktop, so what do you want? I could see a Lisp OS working really well for workstations, maybe servers, but there is no way in hell you'll get the current consumer retards to use it.

This is true for further reasons. x86 and ARM are terrible environments for Lisp and mismatches between the OS and the hardware become very costly very quickly; in the end, you cannot abstract completely over hardware.

This phrase is red flag because it is usually used by Modernization™ vermin, for who modernization means pointless replacement of old with new. I've taken a cursory look at CL21 and it smells exactly like that. What are some concrete gripes you have with Common Lisp?

Isn't that what Plan9Port essentially is? What can real Plan9 do that its userland port can not? I've wondered about this for some time, if real Plan9 really can do more then I'd want to look into it. Inferno also comes to mind.

Regardless is there any effort to improving the graphics libraries in P9?

In the early days, Linux really did function as a Unix replacement. Sure the libc and utilities were GNU, and the kernel wasn't derived from Unix roots, but it was close enough for me. I installed Slackware in 1995 (instead of "upgrading" DOS->Win95), and it seemed pretty damn close to my SunOS shell account in practice. I used basically the same tools on both.
But now Linux has become a lot more like Windows, with the focus on desktop shit (even to the point of moving away from X11) and crap like systemd.

Found the cia nigger!

P9P is just userland programs like acme and mothra ported to *nix. They still have to operate under Linux or a Unix kernel and GNU or a Unix userland. Acme won't give you the neat networking capabilities of Plan9. The Linux kernel doesn't support dumb terminals, for example, like the Plan9 kernel does. P9P is basically like a nice leather interior on an old rustbucket; you're still using a 50+ year old car with an inefficient engine too big for its case.
Plan9 from Bell Labs is no longer being developed. The devs just up and abandoned it when they couldn't make it take off. The most acting P9* community today is 9front, and they've long since stopped bothering with graphics.

This is stupid and could only be true in the mind of a Holla Forums faggot.
There are already fast as fuck Scheme compilers.

Have you LISPniks ever used a Burroughs or any of the other computer architectures or operating systems made for ALGOL?

LISPniks destroy cultures and then wipe out their history, all to get more AI welfare money.

It's already been happening, you dumbfuck. If you followed the *BSD mailing lists, you would have already noticed an influx of Linux refugees.

Do you have your head in the sand?

Samsung bought Joyent and plans to use their datacenter structure for their business, that means Illumnos.

BSDs already exist in larger ecosystems like CA so their platform can't get taken out by one OS hack.

This website runs on FreeBSD, Toyota has been found running OpenBSD code, which is also what Bionic is based on.

BSDs are more imbued than you think and the server market is already showing more support for them, Digital Ocean and VULTR deciding to support BSD support this.

I'm pretty sure every Sony product since the 80's has used some form of Unix and then later BSD variants in modern products like their televisions etc.

A few years ago I could not go to any company without seeing a SUN machine so I'm not surprised to hear Illumos is still popular. I hope things work out well for the Joyent team and it's not some Oracle 2.0 situation.

Tons of routers still run OpenBSD from what I can tell.

I'm curious what the BSD percentage is at Microsoft these days, I know it went down when they were pushing Azure but I wonder if it's entirely phased out yet.

I have to imagine BSD is still used for movie production backends. I think there was a recent video of Pixar or Dreamworks (I can't tell them apart anymore) using a lot of Linux for thin clients connected to a BSD render cluster.

*BSD on servers are nothing new, though. It literally has absolutely nothing to do with systemd. In fact, many sysadmins agree systemd isn't bad (or even beneficial) on a technical level when you have the right configuration.

Basically, people who are getting freaked out about systemd probably wouldn't have even touched systemd in the first place.

I thought 8chin used CentOS.

Back in the day, it used Lunix. CentOS, if I recall correctly, but I also remember hearing something about Arch. Ever since 8ch was sold to 2ch, FreeBSD became the OS used in the servers, mainly because one of the employees also happens to be a FreeBSD developer.

OP here, not even sure what you are getting at; I'm saying that there is to much technical traction to try to bring back 'proper' LispMs (as an architecture) for anything besides maybe a hobby market when low-grade chip fabrication becomes a more semi-trivial feat.

At this point you have to work with the given arch regardless, & Also I'm not convinced that if we even have home fabrication -- that the amount of technical debt in LispChips would ever be comparable to modern processors with millions of dollars and decades of years in R&D.

I don't even really care about low-level all too much anyways -- I'm still wanting a more modern higher-level stack ... which CL is SLOWLY working that way as of the past ~5 or-so years.

Nobody said anything about being representative, you faggot. There is an influx, period.

You are not OP.

Totally am... but okay.

To be clear, not 'OP OP' but regarding the OP who wants a modernization effort for CL & is disheartened by the prospect of GuixSD being some sort of a savior in this space.

Quit pretending shill.

Ok.

Windows dominates the market too much nowadays but they fucked up big time with W10. I work in a bank and we just upgraded to W7 from XP last year. The reason for this is the world wide banking compliance that all banks must be at least W7 or Server 2008. We didn't choose 8 or 10 because there's too much vurnerability in them and they're unfit for use in a financial sector.

With 7 support ending by 2018, Windows has basically up until next year to unfuck W10 or we'll be moving to Linux lol.

How do you plan on running Excel on Linux? I NEVER seen a finance person NOT use Excel.

I'm pretty sure even Gnumeric would be good enough for these people. LibreOffice is pretty good, and Excel runs on Wine anyway.

I doubt they'd switch

Wanna know a dirty secret? A huge number of core business operations are nigger rigged in excel scripts, it is the maintenence job of many of my friends in the industry.

People are too retarded to use anything else.

These people who use those horrible scripts are retarded, adverse to change, and your boss values their opinion more than yours.

We had a large company in my town switch to Ubuntu, it was practically seamless but the moment the senior IT admin left the MS suits came in, sweet talked the bosses with some wine and BAM they were back to Windows.

The existence of MS and Oracle is owed to salesmen like those people and retarded bosses.

well its Linux or we stay in W7. Even though W10 is free for upgrade right now, W10 is still unfit due to the many holes security holes in it. From what I've heard from our developer teams as well, W10 is hard to make programs for as well unlike the xp - 7 transition.

Teach those men some Perl!

washingtonpost.com/news/wonk/wp/2016/08/26/an-alarming-number-of-scientific-papers-contain-excel-errors/

I'm always surprised considering support contracts for non-Windows should be cheaper, not to mention licensing costs. You'd think business people would care about recurring costs.

Are you talking about the univeral windows platform? I thought the point of that was lock in into the micrsoft store.

what's the point? do you all still use unix software?

Have you guys heard about ReactOS? It's basically open-source windows clone. What do you think?

It needs NT 6 compatibility before it can be even considered useful.

All you need is a niche that allows you to have a monopoly in an area
Your project is viable as long as you have a monopoly in area that generates enough revenue to pay the bills or motivates enough people to contribute time and energy.

It lacks a decent editor tho

There are OS made in Lisp as Movitz (a toolkit to build an OS actually), Mezzano and Yalo. It is just that no ones seriously cares about them and their development really lags.

I will also mention Harvey OS, a community developed descentant of Plan 9.

Someone can explain what the fuck Urbit is. I found it when the domain was still pointing to a blogspot post and to this day I have no idea what it is about.

google is trying that with Fuchshia

Most everyday computer users don't give a shit what OS they run. They will use whatever is preinstalled on their machines.
For those who actually install their OS themselves, there is two very important factors to consider, both of which will hinder the success of any new alternative OS nowadays:
1) Drivers / compatibility
The greatest OS in the world doesn't do you any good if it doesn't work on your machine.
No drivers = no users
No users = no one to code drivers / no reason for manufacturers to port their drivers
2) Applications
An OS can only succeed if there are applications for it. Windows isn't just successful because it comes pre-installed with most PCs, it's also because muh Word, muh Photoshit and MUH GAMES.
Again, it's the chicken and egg paradox, no applications mean no one is going to use the OS; no one using the OS means no one will bother to write applications for it. See: Firefox Mobile, Ubuntu Phone, Windows RT, Windows Phone

Urbit is like peer 2 peer distributed internet plus facebook plus monarchist propaganda.

It looks interesting but their site still doesn't make much sense. Or I'm just an idiot as it seems they successfully raised $200k from 1k people.

"Wants A 'Modern' CL" OP here.

Some FPGA tooling and/or a reimplementation for Movitz on RISC-V would be some kind of saving grace for the so-called 'lispm' community me thinks at this point; The trouble is the obviously absurd amounts of technical debt one would have to claw out of to form such foundations. We're kinda seeing the community work backwards in tangent with such goals via various means, mcclim & paulownia are great starts (sans the latter seemingly currently havivng a hard-depend goal to be built on QT) in making interfaces so generic where "eventually^tm" one would be able to write such a thing. Fragmentation is and always seeminjly has been a problem, especially in lisp, I assume we all know.

Regarding Harvey, is that the current 'most development here' fork? I was looking at 9front awhile ago and not much seemed to be happening there.