The latest from the labs…
Welcome to the slightly revised BinOC/Labs. I have moved the forge over to The Labs domain and decided for, at least the time being, to NOT have a Projects and Experiments page but use the forge organizations and repositories them selves, directly. This is still accessible in the global header you have come to enjoy. Will this remain the case once something actually ships? Maybe.. Maybe not.
Besides, most people believe a Github Account or a Facebook Profile is a replacement for a website like I have been doing all millennium and indeed since 1996. The new Labs Proceedure is to not pre-emptively do ANYTHING, instead, wait for a positive result instead of show-casing what may materalize as if this was still the 2010s.
The only REAL remaining top-level Binary Outcast component left is the IRC server. Which you can always join and argue on trivial matters that have offended you based on something you read elsewhere. WHICH may not even include even a passing familarity with events that happened in the space-time continuum. Or one do something radical like say "Hello" and learn something but who has the time for that!
Until next update.
I am not interested in your obsolete pseudo-tactical horseshit or obvious fabricated intrest. It is not welcome here and neither are you. This place is for doing actual meaningful work that may one day help someone (which may even include myself). Which is not what any of you are about.
Keep your lolfunnymemes and your AI-mutated thoughts to yourselves. Any further random overtures by Moon-adjecent personnel will be ignored. Violators will be banned, repeat violators will be screamed at and banned again.
You have been warned.
Due to many factors including some beyond my control and despite all the work I have done thus far. Not just this year but in the past 12 years.. I am giving up on the collaborative open source nonsense. Which is totally politically infiltrated by ideologues and corporate top-down compliance gatekeepers and good old fasioned greed or dare I say the most moronic stupidity since before the Internet took off in any real way.
As a result, I have withdrawn the Donation button and am simplifying my one man operation to just BinOC/Labs and its projects and experiments. There is only XUL [dot org], for example, serves no purpose in an eco-sphere of Mozilla rejects (including myself, btw) who can't even agree that XUL technology.. The VERY DAMNED REASON we are rejects, needs to be preserved and advanced. Where Multiprocess is now considered fasionable, 10 years later, and WebExtensions desirable. Not to mention the entire Linux Ecosystem of Corporate Distributions with their SystemD, Wayland, and Rust AAAAAND increasingly Windows 10/11-style feature whoring.
I am simply done trying to convince anyone and pre-emptively setting up processes and systems that cater to a wider contributor base who do not actually exist in reality. I am also quite over waiting for someone ANYONE to give a shit as they virtue signal their social compliance out of manufactured fear and willful ignorance. It is simply enough. From this point forward, There is no Binary Outcast only Laboratories.
I make and resurrect software and ecosystems, the kind that are socially unacceptable to corporate governance. The kind that puts the newer systems to shame in everything but selfish profiteering and wasteful bloated hyperscales. We didn't need it before and we certainly don't need it now unless you are a greedy world wide company looking to profit off millions of REAL nerd's and geek's decades of hard work . Hard work that is now causally fed into an AI for the good of no one while core systems are endlessly refactored to use vertically integrated supply-chains. What we nerds used to call dependency hell.
If anyone wishes to assist me in redefining everything BACK to stable, predictable, knowable systems that anyone can just learn and use instead of performing for someone to throw them a fish or a dog treat. Do join our MY IRC server and the Repository Forge. Otherwise, enjoy your victory against sanity and sovereignty while you can because I shall prevail in the end even if it is in obscurity. For no other reason than I do the work. Like a fuckin' nerd.
Deal with it.
Due to the shifting nature of provisions in many new laws passed recently I have decided to outright restrict who can obtain and run a branded binary release of Snowball OS by U.S. State.
So far that brings us to residents, non-profits, and corporations residing or operating in the states of California and Illinois.
Which are hereby forbidden to obtain or execute any binary artifact associated with a branded Snowball OS release bundled or installed and must destroy any copies they may posess. This, however, is not a restriction on the source code that comprises Snowball OS but most entitlements on source code rely on legal obtainment of a distributed binary. Such binaries are permitted to be used according to grants based on terms to use my intellectual property in relation to that release. Those residents are also strictly forbidden to use Snowball OS branding or related intellectual property in association with anything they may produce them selves.
I really dislike that it has to come to this but these laws are being challenged in court as well as being discussed on the Federal Level. Though, world-wide scope doesn't have a direct path to affect Snowball OS in any way but I must warn anyone outside the United States that you may have provisions of your own.
Rest assured, that if the law prevents even a limited scope distribution of Linux by an individual with no pretense of modern provisions.. One can always build it them selves. Though, beware my branding restrictions. It may just save the project.
It has been brought to my attention that since I've been pouring my self into the Labs past few months that developments have indeed developed. Old news now but stellar none the less. Thanks to the Colorado effort by System76, California has loosened their restrictions when it comes to free and open source software. However, it was California's Silicon Vally that created the modern preditory practices AND the so called solutions negotiated with lawmakers to them that has set the framework and precident for these laws in the first place. So my ban on California stands. They are not permitted to use this OS. Colorado and System76, however, are to be commended.
Doesn't matter how much rpmbuild, or lorax, or kiwi-ng you learn. No matter how much you practice or simulate. You can't make a remix of a distribution you can't get onto the disk via its own installer.
No. I can use other means to conjure EL10 onto disk. Then I would need to figure out WHY there is an issue and then figure out HOW it happened and WHY and then not submit it to upstream cause they aren't into that if you don't 100% buy everything they say (Or a RHEL Subscription, either works). Or bypass anaconda by creating a different means of redistribution and installation. But the REAL question is.. HOW much is Enterprise Linux 10 worth to power through these unique difficulties that do not present in the reletive Fedora versions pulled from?
Frankly, not a lot. The work to try and make what was supposed to be a simple stress-free remix/rebrand with some extras for independent Mozilla developers to have a ready made system is greatly exceeding the value any specific user would get out of it. I should have anticpated this, and I did, but I hoped that by .1 or .2 the disaster of deeming a Beta as Final almost immediately after releasing said Beta would have been largely resolved or at least bugs like mine would be more known.
Value wise, EL8 is the most valuable but a keep-alive project is very different from a remix. So I am just.. Done with RedHat, Fedora, all of it.. While it may be some time if ever for Snowball OS to be considered "Enterprise Grade" but given how that grade is barely skating a D (in american grading systems) at this point.. I will just have to let the hats simply.. Wear it without interference.. I have other plans.
Not every idea is a winner and most winning ideas are time sensitive. The moment simply slipped past the window.
One of the Labs' most exciting new areas of research and production is in the area of Linux. Yes, that penguin that seems to run 90% of the post-20th century infrastructure on this planet. Remember, any android device IS a linux device.. Just quite a specialized one. Like the addition of Linux to cars, Android in cars IS linux in cars after all and distros are specifically edging into "Automotive Computing". But I am not here today to talk about car-linux. Though we need to have a discussion about it, eventually. No, I am here to talk about just THE Linux. The one that runs on devices still shaped like computers. Specifically, our pending Linux OS experiment. Snowball.
While it is true, for There is only XUL, that we are working on a low-impact remix of an industry-grade OS in the form of Fez Enterprising Linux but this is just a means to an end. Largely, for independent Mozilla developers. Which is great and all.. But it doesn't really SOLVE the Linux problem.. Just helps with the Mozilla problem. To work towards solving the Linux problem, we need to BUILD Linux not just rebrand it. That is where Snowball OS comes in.
Snowball OS intends to be a "Millennial-style Workstation" which basically means it is following Traditional lines. But unlike most rebel distros and forks this does not mean that only the old is paramount. No, it is about capturing that conceptual methodology that built the foundations for our modern world and allowed everyone to become so .. NOT FOUNDATIONAL and forget that without the foundation (that is being ripped to shreds and oxidized) Vibing will come to a sudden and spectacular end. To hopefully rediscover enough to provide a contrasting example to counter the continued Dark Future of Tech I was on about in 2011.
Heh. Why not? No, but seriousl-- Actually I am serious.. WHY NOT?!
Not quite. This distribution would be aimed at the tinkerer but focusing on bringing back to the forefront the most open and diverse era of expression in Linux and technology in general. To empower the user to not simply be a free customer but the ability to make their own system from scratch or use our base as a starting point. Something between the hefty freedom-taxes of following the Linux from Scratch book (which has discontinued non-SystemD) or trying to rice a managed-integrated-ecosystem just as inflexible as any corporate-adjacent distribution. Just shifting the cost of production to your hardware for their system.
That is the question isn't it. How? Well Linux from Scratch WAS a solution. It still is but even it is starting to show an ideology in its design beyond teaching you how a Linux system works. Since they dropped non-systemd, which also makes more advanced non-systemd init systems harder to swap in, it simply isn't a universal starting point any longer. How this changes my original plans is subtle.
Originally, I was seeking to solve the RPM-Bootstrap paradox and the solution is simple. Don't. You solve it by admitting that it is less compute time to build a reduced sysroot with enough to build and run rpmbuild. Which is like 75% of the LFS Book anyway. Then you simply provide a few fake-packages to get you owning the filesystem and providing bash then switch to a final build process of building Ch. 8 RPMs installing as needed until you finish the book. Afterwhich all you do is rpm root install the new packages to a new root. Then continue from there on post-baseos packages.
It was a good plan. I got past the initial problems of it as well. But then, they cancelled the Sysvinit version of the book. While it continues in development-only format largely mechanically continued not actually tested it threw a huge wrench in my simple scheme. Now I have to own the baseos package SELECTION as well. So Snowball OS as a result is merely LFS inspired not LFS powered.
Well that is the thing. I am not completely sure yet. I see two possible paths. One path has us do a FreeBSD thing (which they are actually un-doing) where we ship a monolithic base and use packages as add-ons. It would be a VERY familiar setup for the Labs to pull off. The second is control the package selection also but follow the original plan just with added work. Of course, the former does not preclude the latter but the reverse isn't true. In any event, the NEW first step is to re-establish a reliable build procedure and selection to set the base. Which is being done with the help of the currently experimental Axion Build Environment.
From there I am leaning more to the FreeBSD model with RPM packages acting as an add-ons system (or I guess ports?). Again, we can always become fully rpm-managed later if it proves advantageous but when one considers that outside of specialized chroots and containers, everyone always installs the same 300-800 base packages it starts to make sense. At least for now.
Oh a marketing blurb.. Sure thing! "Snowball OS is a Millennial-style Workstation featuring SysVInit, X11, PulseAudio, and is either RPM-powered or extended by a "Reliable Ports Manager" (with rpm)." There is no console-only version, no server version, no hyperscale version. Just a traditional workstation that has just enough to build more. FOR COMPUTERS, and you!
The Linux Problem can be summed up like this: "Is a free and open source system still free and open source in practice if only elite specialists and server clusters can build and provide it?".. Who controls the means of production? Because in open source it should be .. uhh.. EVERYONE!
I can't believe Moonchild reached his goal of truly emulating copying Mozilla in almost completely reducing our platform to an over-glorified browser library and has yet to work out that the extension ecosystem is that special sauce that makes XUL useful even when what XUL was primarily abused for, web browsers (and sometimes e-mail), has long since had no use for it.. Or XML for that matter. But let's detail some of the inescapable conclusions using my trusty rusty UXP Coordinator skills.
First off we will ignore the canonized archived record AND version 7 of the MCP fan-fic re-telling of the original drama born from intentionally botched removals and PM30 sabotage along with the fall out from removing my server from MCP's fleet and focus ONLY on the tangible results since I have been gone (in more ways than one, admittedly).
When I left, Moonchild had trouble restarting services but once they did they have been in 100% control at full operational capacity since let's be generous.. May 2022. They lost an E-mail client which was quickly and haphazardly replaced by epyrus. Also, Basilisk was restarted after it was determined kronosschoty wasn't allowed to buy it because he was hanging out at BinOC's IRC channel, of course. So DJames1 now owns and develops it. So that is Pale Moon, Basilisk, and Epyrus.. Epyrus doesn't get active development or regular releases so it is worse off than FossaMail was before Interlink. They restored the Cross-Reference (albeit an inferior unoptimised version) and Phoebus and stole the Binary Outcast Template that was attached to Phoebus. So they have had it all and make no mistake they had everything from me by the 23rd of March in 2022 and every previous backup as well. So UXP without Tobin. Cool great.. sure.. And the result?
What are results?
Results are outcomes of events, usually actions, direct or indirect that form the unfolding causal nexus of spacetime. Wait, that should be "What are THE results".. Silly me.
What are the results?
In short, we lost the browser war. That's correct. We lost. We lost because we couldn't get along and have nice things. But why do I say this? Because UXP is no longer the buzz or bane of anyone anymore (unless they venture to the Pale Moon forums). No one cares. Basilisk is a parody of itself and Epyrus does not get any real development. But the real issue is Pale Moon.
A very long time ago I said the following words "Pale Moon is not Firefox and never will be again". This, turns out to not be the case in the end. Pale Moon and Moonchild and most of the remaining contributors, that I personally recruited have become effectively Mozilla. The wrong blasted one, the Firefox First one. The kind of Mozilla that takes your extensions and choices away while telling you it is the most customizable version yet. This is what I have seen in the tree when you ignore the people-drama and the schemes and the renewed focus on marketing fluff and strategic business alliances and non-disclosure agreements and backroom deals with Cloudflare. Which all sounds like MozCo to me. But the state of the tree…
On a general application framework and cross-platform toolkit AND runtime environment basis, UXP is not fit for purpose. It is slow, stutter-y, its XUL primitives are more unstable than ever and interfaces are being culled. Interfaces used LARGELY by extensions if not internally by more than a few call sites. Versus my original parameters and vision this code base can only RELIABLY produce a Firefox-class Web Browser and only an e-mail client still because Mozilla continues to produce one and I used to. It is maddening and very very sad all at the same time. A great waste of potential and investment. Rendered as useless to NEW application development as any of the 40s and 50s were from Mozilla upstream. Worse actually because years of careful orchestration and planning was thrown out the window by 2024 and negated completely by 2025.
There are no real options for "modern" UXP.
As the creator of the Unified XUL Platform, Take 2 the one that worked, and coiner of the name its self.. I declare the platform as a dream completely over. It simply isn't capable of it any longer due to project mismanagement over time.
It's TRUE! Check it out. It is even better than before!
The experiments page is now up and along with the Projects page made more clear. However, a more complex structure will be needed especially for projects.
Rest assured there is more happening that will be brought to the forefront in short order.
Until next update…
So Meta's new AI Scraper was abusing the There is only XUL Community Cross-Reference with 69 seperate IP addresses in the 57.141.4/24 block. They can also bypass the Anubis AI Firewall. How is anyone supposed to get any SCIENCE done with AI destroying people's servers and stealing their content?
In any event, it is mitigated for now and the experiment continues!
The new server is operational with the code Cross-Reference which has been refit with the latest in Metropolis Utilities and the Advanced Phoebus Model.
Labs will be migrating to that server along with the rest of the xulorg. There is a high chance for Gogs in the nearterm forecast!
Until next update.
The projects page now exists ready for project pages them selves to be written. The new BinOC large server will be acquired tomorrow. At which time the real experiment begins! Also, these early posts are sure to be removed once more structured operations are in full swing.
BinOC/Labs is now under the runtime authority of the Advanced Phoebus Model as provided by the Metropolis Utilities Anti-Framework v3.0.0a1.
This page will contain all the exciting new updates to projects and experiments as well as status updates for on-going BinOC activities as they related to software and infrastructure development and field use!
For now, stay tuned!
The Binary Outcast name, associated brands, styles, designs, artwork, and original content are Copyright © 2001-2026 Matt A. Tobin - All rights reserved.
All material is the intellectual property (such as copyright or trademark) of its respective owners.