X Window System on a Floppy
pupngo.dk
pupngo.dk
* https://crackberry.com/heres-how-qnx-looked-1999-running-144...
I understand Ford Sync3 is based on it, but are there other consumer-visible products based on it?
And while Microgui is nice it purposely isn’t directly comparable to a compositing window manager.
Debian would just show the old curses-like installer, and pull the rest of the system from the repo, while QNX had a full GUI, network card/modem drivers, and a friggin web browser in 1.44 MB.
Out of curiosity, what are currently the most minimal, "graphic" (i.e. not Links-like), browsers available today? Any Open Source ones?
I wonder if anyone has tried to add optional javascript support to Dillo or something with Fabrice Bellard's QuickJS engine (full JS coverage in like 210 KiB of x86 code).
It really isn't too bad.
There's BrowseX, built in Tcl/Tk.
https://sourceforge.net/projects/brx/
The Suckless browser, "surf", is pretty light and featureful:
[0] https://www.floppix.com/ [1] http://menuetos.net/ [2] http://www.kolibrios.org/en/
I also like the idea that the media doesn't contain a computer that could be programmed to do something nefarious that you don't expect.
https://www.bleepingcomputer.com/news/security/heres-a-list-...
It's entirely possible for a floppy-disk to be an attack-vector.
After all, viruses were a thing in DOS days and they spread via floppy disks: https://www.trendmicro.com/vinfo/us/security/definition/boot...
But a modern Trojan-horse style computer-hidden-inside-an-innocuous-looking-peripheral in a floppy disk should be possible, considering what Sony fit into a floppy-disk's dimensions back in 2000: http://camera-wiki.org/wiki/Sony_Mavica_FD95
> It's entirely possible for a floppy-disk to be an attack-vector.
Yeah, but media with embedded computers adds more vectors on top of those that exist in dumb media. I'm not claiming that dumb media is absolutely secure, just that it has fewer attack vectors.
> But a modern Trojan-horse style computer-hidden-inside-an-innocuous-looking-peripheral in a floppy disk should be possible, considering what Sony fit into a floppy-disk's dimensions back in 2000: http://camera-wiki.org/wiki/Sony_Mavica_FD95
That's a novelty, though. I've never seen one of those, but that particular example could never be confused for a regular floppy disk, and I'm skeptical that a similar device that could be confused would even be possible to build.
Floppies are super light. To sneak a CPU into one would require power (which would not be provided directly by the drive), storage electronics, and something to interface magnetically with the heads. I think all that (especially the power and interface) would likely be noticeably heavier. I've had something might had some of the components (https://www.macrumors.com/2017/02/01/coin-shutdown-services-...), but I doubt the magnetic interface was sufficient for a trojan floppy.
But even if you could get the weight and balance right, inspecting whatever under the shutter would be a dead giveaway. And transparent floppies were a thing (https://commons.wikimedia.org/wiki/File:Transparent_floppy_d...) and using those would make it even harder to sneak electronics on board.
It should be possible, but I don't have the electronics to attempt it. Nor do I have the time.
One thing I can think of is people have done tricks with programmable storage like changing file contents after the first read (to bypass scanning). The link I posted above lists that as "7) RIT attack via USB mass storage - attack described in a research paper. It relies on changing the content of files while the USB mass storage device is connected to a victim's computer."
> it's not like usb or firewire where you have inappropriate levels of access to the system bus...
That's a good point. The main benefit of dumb media isn't so much that the storage is dumb, but that it connects to a far more restricted and inflexible interface. Though lacking a CPU does provide a smaller benefit (i.e. preventing that "RIT attack").
1.44 MB “floppies” were encased in hard plastic and not fragile
Basically we use a base Linux and modified Squeak to use SVGAlib instead of X-Windows.
Later on some friends created a whole OS in Squeak (e.g. you could browse the TCP/IP protocol as a class in Smalltalk): https://wiki.squeak.org/squeak/1762
https://github.com/nopsys/CogNOS
There is a conference paper about it, also from 2018:
I don't know if the project still has the same soul as it did then, but here it is:
Followed the Linux IP Masquerade HOWTO (http://www.e-infomax.com/ipmasq/howto/m-html/ipmasq-HOWTO-m....), and I soon had it running the network in my apartment building with a cable modem and 100mb SMC hub (also from a dumpster)
A couple weeks later I came home to find the internet "not working", it got hacked (I never applied updates), and I replaced it with Freesco and never looked back.
Buried deep into SCO's FTP server there still is a bootable single-floppy demo of SCO OpenServer intended for POS devices or so, but it's lacking any kind of GUI and most command line utilities...
With that in mind, having an antique version of X running from a floppy isn't that impressive. It's still really cool though. :D
See the Wikipedia section [1] on nomenclature for details.
[1]: https://en.wikipedia.org/wiki/X_Window_System#Nomenclature
Cringing intensifies.
2. The name is not "X Window".
Citation: See the June 1988 "XNextEvent" newsletter, The Official Newsletter of XUG, the X User’s Group, "Things That Happen When You Say ‘X Windows’", for the definitive list of its official names, and the dire consequences of getting it wrong:
https://donhopkins.medium.com/the-x-windows-disaster-128d398...
>Things That Happen When You Say ‘X Windows’
>I was digging through some old papers, and ran across a 15 [now 33] year old “XNextEvent” newsletter, "The Official Newsletter of XUG, the X User’s Group", Volume 1 Number 2, from June 1988. Here’s an article that illustrates how far the usage of the term "X Windows" has evolved over the past 15 [33] years. (Too bad The Window System Improperly Known as X Windows itself [still] hasn’t evolved.)
>Someone on slashdot asks, “Why is it still called X-Windows?”. Predictably, the first reply says: “It isn’t. It’s called ‘The X Window System.’ Or simply ‘X’. ‘X Windows’ is a misnomer.”
>He didn’t ask why it is “X-Windows”. He asked why it’s called “X-Windows”. You’re wrong that it isn’t called “X-Windows”. It is! It’s just that it isn’t “X-Windows”. Being something is independent of being called something.
>The answer to the question ‘Why is it still called X-Windows?’ is: It’s still called X-Windows in order to annoy the X-Windows Fanatics, who take it upon themselves to correct you every time you call it X-Windows. That’s why it’s called X-Windows.
>The following definitive guide to the consequences of saying “X Windows” is from the June 1988 “XNextEvent” newsletter, “The Official Newsletter of XUG, the X User’s Group”, Volume 1 Number 2:
>Things That Happen When You Say ‘X Windows’
>THE OFFICIAL NAMES
>The official names of the software described herein are:
>- X
>- X Window System
>- X Version 11
>- X Window System, Version 11
>- X11
>Note that the phrases X.11, X-11, X Windows or any permutation thereof, are explicitly excluded from this list and should not be used to describe the X Window System (window system should be thought of as one word).
>The above should be enough to scare anyone into using the proper terminology, but sadly enough, it’s not. Recently, certain people, lacking sufficient motivation to change their speech patterns, have fallen victim to various ‘accidents’, or ‘misfortune’. I’ve compiled a short list of happenings, some of which I have witnessed, others which remain heresay. I’m not claiming any direct connection between their speech habits and the reported incidents, but you be the judge… And woe betide any who set the cursed phrase into print!
>- You are forced to explain toolkit programming to X neophytes.
>- Bob Schiefler says, “You should know better than that!”
>- The Power Supply (and unknown boards) on your workstation mysteriously give up the ghost.
>- Ditto for the controller board for the disk on your new Sun.
>- Your hair falls out.
>- xmh refuses to come up in a useful size, no matter what you fiddle.
>- You inexplicitly lose both of your complete Ultrix Doc sets.
>- R2 won’t build.
>- Bob Schiefler says “Type ‘man X’”.
>- Your nifty new X screen saver just won’t go away.
>- The window you’re working in loses input focus. Permanently.
>[...] Don wrote the chapter on the X-Windows Disaster. (To annoy X fanatics, Don specifically asked that we include the hyphen after the letter “X,”, as well as the plural of the word “Windows,” in his chapter title.)
In terms of feature, using it through SSH always impressed me and allowed some quick tricks that would be much more complicated otherwise. It also allowed advanced usages like in Compiz, which I personally got involved and used to impress my friends who did not use Linux.
The remaining problems were more complicated to fix. Old code base, old unused extensions, lack of isolation and sandboxing, different DPI and refresh rates on devices, no guarantees of when something was drawn and single thread architecture that introduces unneeded latency. That's why we got Wayland.
I'm nostalgic to see Xorg go, but I'm happy to know it will be replaced by something better.
Perhaps it was "2-Disk Xwindow Linux": https://web.archive.org/web/20020408075857/http://www.mungki...
It makes me sad how programmers can keep up with the insane improvement of the hardware, only to make software slower and bigger than ever, while adding very little useful functionality.
However, to lament programmers not eking out every megabyte of payload size is nonsensical - in my opinion. Why waste a scarce resource (labour hours) optimising the use of a cheap and plentiful resource (bandwidth/storage).
Imagine a world where fusion energy was the most commonly employed energy source and the cost of electricity in both environmental and economic terms was negligible. Should we still have energy-star rated refrigerators? Why? What purpose would it serve?
1. we save a couple of labour hours
2. make every phone and computer a lot slower and more annoying to use for hundreds of millions of people
I am a software engineer but I can't even imagine how a flashlight app can be several megabytes. What is inside it?
I'll give a counterpoint to this example; MineSweeper on the Windows Store.
Last time I took a look, the installation size of this app was more than a 'functional' install of the entire Windows 98 OS, including a slightly-less-colorful version of the same game.
Or, for a more 'Business' example. Autocad 2013+; it became bloated, frustrating to work with, and expensive.
> Imagine a world where fusion energy was the most commonly employed energy source and the cost of electricity in both environmental and economic terms was negligible. Should we still have energy-star rated refrigerators? Why? What purpose would it serve?
In a 'happy path' fusion scenario, the best refrigerator design would be one with the lowest ozone depleting and/or greenhouse gas emitting setup, so long as the energy consumption was overall sustainable on the grid a lot of them were hooked up to.
It has perfect functionality but doesn't waste your time by loading bloatware for 10 seconds.
Even 25 years ago, on Slashdot I could tell it not to show comments with less than a certain number of upvotes, and thus filter out a lot of the dross from busy conversations.
I can't do that on HN. I can do something like it on Reddit, although I know of no "expand all" functionality, but on HN it's all or nothing.
That is imperfect to say the least.
Bandwidth and storage are not always plentiful. I might have a lot of other things to store and not enough room for a half a gig deck of cards. And maybe my local tower is down and I've got an intermittent signal from the next one over.
> Imagine a world where fusion energy was the most commonly employed energy source and the cost of electricity in both environmental and economic terms was negligible. Should we still have energy-star rated refrigerators? Why? What purpose would it serve?
A more efficient fridge dumps less heat into the kitchen (or wherever), which means temperature fluctuation and more comfort, even if A/C is effectively free, it makes noise and what not. The other comment about environmental impact is also worth considering, although if energy is free and plentiful, I expect most power plants to switch and for syngas to takeover where liquid fuel still makes sense, which might make worrying about emissions obsolete (although who knows, anyway fusion seems capital intensive for the forseable future, so I'm not going to plan for free and plentiful energy)
Today's apps are meant for things like 4K screens and have graphical resources to match.
Then, everyone loves to bundle in third-party frameworks and libraries, which bloat individual apps and cannot be shared.
If you look at the actual application logic code, it has grown much more reasonably in proportion to apps 20 years ago. It's just that there's a lot more junk tagging along with it. If you look at typical Linux packages, where library reuse is still the norm (though perhaps not for long, with containers...), you'll find fairly sensibly sized binaries.
Modern Linux kernels don't fit on floppies any more, but if you look at the actual code in there... there's good reasons for it to exist (and of course you can compile time disable a lot of it, though probably not enough to get it small enough for a floppy any more). Complexity is unfortunately necessary for things to work and perform well on modern systems (e.g. think of how much more complicated things got once SMP/multicore became the norm).
[Cries in Zircon static linking]
Eg, take calculating the length of a string. That's just "add 1 to count until reaching \0", right?
Not in the modern day, where you have UTF8, and what qualifies as a character is a lot more complicated than it used to be. A modern program, wanting to accurately calculate the number of characters in a string then brings up a bunch of UTF8 machinery with it. Then there's a good deal more than that to UTF8, so that also brings up a bunch of other code that is needed for other purposes, but for now may not be used.
But these days we mostly use GUIs, so instead of "length of a string in characters", you probably want "length of a string in pixels". In comes the TrueType library you need to parse the font and implement the virtual machine (!), because these days a font file contains code that helps the system produce the best possible result.
And so on, and so forth. Doing the most basic version possible in English on a fixed width text display with fixed width fonts is just a few bytes of code. Doing the modern version that takes into account things like that Japan exists quickly expands the problem to megabytes worth of code.
Just tried this out for fun on my 486DX2 PC ( https://i.imgur.com/NOeR1Ad.jpg ). It boots, but obviously the minimal kernel only has included support for a small set of hardware, which does not include serial mouse, S3 805 VGA (works, but graphical artifacts), and 3C509 ISA network card. Will have to try out the provided toolchain and scripts to try and see if I can rebuild with better hardware support for my 486. Why bother, you ask? Why not, I say!
Linux 2.2.26 is kind of hilariously old. Linux 2.2 was released in 1999.
Back in the day, Damn Small Linux (DSL) made the controversial decision to stake themselves to Linux 2.4 because 2.6 was just too big and bloated, they said.
uses 5.13.0-rc2.
Very cool.