- This topic has 78 replies, 14 voices, and was last updated Jan 1-8:23 am by Anonymous.
-
Posts
-
oops
“This article explains what multi-seat is, how it works, and…”
Multi-seat on Linux, demystified
…… Good explanation, so multi-seats behavior is a multi-users ability but not in the same time 😉
It’s that why I prefer Lightdm than Slim, only for this multi-seat feather.
Anonymous
Yes, multiple users, logged in at the same time.
Said differently, multiple logged in sessions (same user, or different user accounts) at the same time.https://wiki.debian.org/Multi_Seat_Debian_HOWTO
(page last modified 2020-04-24 21:29:24)Multi Seat refers to a PC that has more than 1 /videocard/monitor/X server/mouse/keyboard set so that more than 1 user can use the system at the same time. This is done locally with one computer, without the need for separate server hardware.
Modern computers are “multi user” in that when one person logs off another person can log in with their own unique profile, settings, and permissions-protected files. Multi-seat takes this to the next level and allows multiple users to use one computer concurrently, all with their own unique profiles and protected files.
olsztyn
skidoo wrote:
“This article explains what multi-seat is, how it works, and…”
Multi-seat on Linux, demystified
…A great, nicely written article on multi-seat architecture. It reminds me of my old times when we (users) had out terminals connected to mainframe computer and executing tasks independently and at the same time.
So in our modern times, when desktop computers and even some laptop can serve a role of such home ‘mainframe’ of the past, this multi-seat architecture can be implemented with cheap laptops (clients) connecting to the main desktop in multi-seat fashion using network protocols such as RDP, log in as remote desktop clients and have access to all the resources connected to the main computer, such as NAS servers, iSCSI drives, etc…
The restriction to being locally wired with keyboard, mouse and individual video card installed in the main computer (as in the original post) is no longer necessary, is it, with such modern-times architecture. This way all family using cheap laptops can enjoy all the power provided by central computer, along with all the software installed on it and resources connected to it such as printers, etc…
So if I am not misunderstanding the original objective:– What would I be missing with such architecture of remotely connected clients vs. being hard wired, with individual video cards as originally described?
– If antiX were to be the OS of the main computer, is lightdm and elogind still required in such scenario in place of Slim?Any insight would be appreciated…
Thanks and Regards…Live antiX Boot Options (Previously posted by Xecure):
https://antixlinuxfan.miraheze.org/wiki/Table_of_antiX_Boot_Parameters
Brian Masinick
Interesting! As you know, mainframe computer systems have been multi-user, multi-process for many years (decades).
Several classic minicomputer systems have also been multi-user, multi process systems for decades too. By the time I was using computers, most mainframe and minicomputer systems, except for special purpose real time systems allowed multiple login instances and also for users to run more than one thing at a time, though years ago, the actual number of processes that could be executed was limited by the physical memory and CPU resources.
Early PCs did NOT have this capability at all. At first I would say they were single user machines. A few may have allowed a few processes at a time, but that ability was limited. Later, personal computers allowed multiprocessing, but not until much later were any Microsoft operating systems available with multi user capability. In contrast, UNIX always had these capabilities, and so did even fairly early implementations of Linux. Today most systems, depending on the purpose of their OS, can certainly handle multiple processes, and many of them can be handle multiple concurrent users as well.
If you want some really cool history, read up on MULTICS. It was a research-based multi-user, multi-process system with parallel capabilities at virtually every level. The only thing even remotely close to it at the time was IBM’s mainframe operating system, but back in the sixties and seventies, though the mainframe CAN handle multiple users and multiple processes, it was hardly optimal; it was definitely designed for batch based processes, such as personnel and payroll systems with transactions written either into files or loaded in at run time from cards or tape storage.
These days a payroll system could conceivably run on most hardware we are familiar with; the main reason it’s still handled on mainframe computers is the physical and network security of those systems; while perhaps still not perfect (is anything truly perfect?) it is very mature (40-50 years of maturity, improvements), coupled with the significant physical security of a protected data processing center.
See Flynn’s taxonomy – https://en.wikipedia.org/wiki/Flynn%27s_taxonomy#Multiple_programs,_multiple_data_streams_(MPMD) for more information and links to additional information.
(By the way, I learned about these different organizations during my undergraduate Computer Science studies at Michigan Technological University, 1975-1979). Some of the things we were learning about did not have very many practical implementations at the time. UNIX had been invented, but it had not received wide recognition until the eighties. Sun Microsystems, Dell Computer, and companies similar in origins had not been formed. Hewlett Packard did exist; they did more work with instrumentation devices in the seventies, though they had a few (not very well-known) computer systems. Similarly, Texas Instruments and Intel were special purpose chip vendors, supplying the military and other special purposes. Both TI and HP had calculators that were more familiar to college students of that era than any computers they may have engineered.
RCA, Honeywell, and GE made computers, and at one time or another, their systems ran this cool MULTICS system.
At AT&T’s Bell System Laboratories of the sixties and seventies, they were aware of MULTICS, and they probably had access to one or more of them at some point in time; both the hardware and the associated software, due to their advanced nature and the much higher cost of physical hardware at that time were prohibitively expensive. That’s why Ken Thompson took a relatively inexpensive piece of hardware, designed and wrote a smaller operating system called UNIX- a clear “play on words” – Eunuch – eu·nuch
/ˈyo͞onək/
Learn to pronounce
noun
noun: eunuch; plural noun: eunuchsa man who has been castrated, especially (in the past) one employed to guard the women’s living areas at an oriental court.
Linux, which was a complete rewrite, somewhat similar operating system, was written explicitly to run on mid to high end personal computer systems from the late 1980s and early 1990s, started in 1991 by Linus Torvalds.
“What is the difference between Unix and Linux?
You may have heard of Unix, which is an operating system developed in the 1970s at Bell Labs by Ken Thompson, Dennis Ritchie, and others. Unix and Linux are similar in many ways, and in fact, Linux was originally created to be indistinguishable from Unix. Both have similar tools for interfacing with the system, programming tools, filesystem layouts, and other key components. However, not all Unices are free and open source.
Over the years, a number of different operating systems have been created that attempted to be “unix-like” or “unix-compatible,” but Linux has been the most successful, far surpassing its predecessors in popularity.”
Source: https://opensource.com/resources/linux
https://en.wikipedia.org/wiki/LinuxI hope this historical digression is actually helpful in appreciating and understanding the topic at hand, regarding “Multi-seat” and what it really means, and what it can do.
- This reply was modified 2 years, 8 months ago by Brian Masinick.
--
Brian MasinickThanks for this extensive historical digression…
Both TI and HP had calculators that were more familiar to college students of that era than any computers they may have engineered.
Texas Instruments… My first personal ‘computer’ was actually TI-59 programmable calculator, on which I developed programs in 1979 (mostly scientific) and wrote them on that calculator to magnetic cards for subsequent their execution…
Over the years, a number of different operating systems have been created that attempted to be “unix-like” or “unix-compatible,” but Linux has been the most successful, far surpassing its predecessors in popularity.”
Yes, however the lack of standards and unified strategy made Linux fragmented into hundreds of competing individual distros and multitude of immature pieces apps doing often the same thing, developed just for the purpose of being different.
This lack of ‘Linux strategy’ killed any chance of Linux becoming one popular desktop system, as Windows had become… I will probably get a lot of flame for saying this but Linux developers do not seem to care about Linux success, instead just about showing off their ‘fantastic’ distro as different from their neighbor’s distro… Even virtually the same file managers are being renamed for different distros to obfuscate the fact they are the same… Such immaturity and selfishness has been abundant and fragmentation prevented acceptance of Linux as a popular desktop system. This is unlike use for servers, where there was more control…
Thanks again and regards…Live antiX Boot Options (Previously posted by Xecure):
https://antixlinuxfan.miraheze.org/wiki/Table_of_antiX_Boot_ParametersBrian Masinick
I can understand the variations between GNOME, KDE, and Xfce – three of the original major desktop environments.
The first versions of these started under THREE completely different licenses. KDE and Xfce actually got a slight start before GNOME really got off the ground, and so neither of them initially used the GPL, the GNU General Public License (all three now have provisions through that license. Xfce switched midway through its third major release to use Gtk, the same toolkit that GNOME uses, but it does things differently, a bit simpler and lighter than GNOME.Both GNOME and KDE use a reasonably full-featured object-oriented architecture – at least the originals did – and the current versions are quite different than the very first ones. KDE got really big for a while, and it is still fairly large, but the most recent implementation does not “consume” all resources unless they are used by the objects and programs that are invoked. I’ve really not looked at GNOME at all in recent years. I’ve generally used either a light implementation of Xfce, and often I do without that; simple cut and paste is enough “value” for me, so I can use light stuff like IceWM.
Interestingly enough, I am playing with KDE today – first with a recent release of KDE Neon, and now with the recent MX Linux KDE 19.2. With KDE and the Chromium Web browser and two tabs open, plus konsole to display htop, I’m consuming 1.37 GB, a whopping amount, given my modest needs, and that’s why I don’t use these powerful desktop environments much. Though I have 8 GB of available memory that does not mean I have to try to use it every time I’m logged in, only if I am doing memory intensive or CPU intensive work do I ever “tax” my system with my resource usage.
More directly to the topic you raised, yes, it may be worth having 3-5, or perhaps even ten varieties of browsers, desktop or window managers, and a few variations to allow experimentation. We see that we actually have four current variations of antiX at the packaging level: net, core, base, and full, plus we offer variations with Sid, Testing, and the current release. It’s not difficult to see how we end up with so much variation across distributions, but you do make a point: with a bit of coordination, perhaps we could have three or four desktop variations, three or four window managers, three or four packaging styles, and then collaborate a bit so that, yes, we could have hundreds of variations and permutations, just with a few of these in different combinations.
I doubt that we could cut the number of distributions WAY down, but it would be nice if more development teams either contributed to one or more of these areas: application software, system configuration tools, desktop widgets or window management infrastructure, or packaging. I’m sure that there is some sharing that already takes place – we receive code from Debian and Devuan at the distro level, and we have used a variety of applications from various sources over the years. I hope that we have shared feedback, defect reports, and possibly bug fixes or new modules of code. Maybe there is more of that going on than the average person realizes. Most distributions are 90% packaging as far as each team’s effort, but an equal 90% of the stuff going into the software is common code that hundreds of distributions already share – such as Firefox or Chromium code, Thunar or Rox file manager, etc.
So there IS a lot of application and infrastructure sharing, and the kernel code is most definitely shared too.
--
Brian MasinickBobC
I am creating a new Slim Theme, and I have everything working except the Welcome message doesn’t appear to tell you to Press F1.
Does anyone know how to fix that?
( edit: removed misinfo posted earlier )
grep welcome /etc/slim.conf
^—v
# welcome message. Available variables: %host, %domain
welcome_msg (Press F1 to toggle sessions)On local machines, I’ve changed it to:
welcome_msg (Press F1 repeatedly to toggle your choice of available xsessions)Yes, I didn’t change the welcome_msg line, then I tried changing it a little, but it had no effect either way.
Here is a link to the file… https://www.mediafire.com/file/j3su0r7sms5nous/antiXMagic.zip/file
PS: The reason I moved the input to the you left is because in a Multi-Display setup, the login is half on the left screen and half on the right, and you have a 50/50 chance that the 2 sides will be reversed unless you move the monitors around to be correct. This way it always fits on the far left screen.
- This reply was modified 2 years, 4 months ago by BobC.
- This reply was modified 2 years, 4 months ago by BobC.
Attachments:
> the welcome_msg line, then I tried changing it a little, but it had no effect either way.
Possibly, as a byproduct of prior edits, your slim.conf contains multiple lines declaring welcome_msg.
If so, and you noticed and edited only the topmost of those redundant lines…
…the latter (last parsed), unedited line may be superceding the freshly-edited line.Another approach for you to consider would be to edit the theme’s background (or panel) imagefile and emblazon the text upon (within the pixels of) the image.
I can’t assist further b/c I’m off-the-reservation (using my forked version of SLiM, out of touch with upstream/mainstream changes).
ps:
https://gitlab.com/skidoo/slim-themes
contains some antiX-branded panel images you might care to repurpose in your themeOk, thanks for the idea s. Here is a pic done that was. It looks pretty good I think.
I do wish I had a clearer picture of the logo.
The 2nd pic shows what it looks like when booted with both screens connected. The login part normally comes up split between the two screens.
- This reply was modified 2 years, 4 months ago by BobC.
- This reply was modified 2 years, 4 months ago by BobC.
I got it working reasonably. I built the F1 prompt into the panel picture as suggested. I had to really stretch the limits of space to accommodate both low resolution screens and multiple HiDPI screens with the same pictures and theme. I finished testing today a couple times and finally got it to what I think is good. The auditorium pic is great. I’ll say no more there. If anyone wants to try what I created, feedback would be appreciated.
the welcome_msg line, then I tried changing it a little, but it had no effect either way.
Bear in mind that SLiM is decades old.
That “welome_msg”, we never see it ~~ it’s a blip.
Back in the slow ol’ days, it served as a vehicle for MOTD (message of the day) for graphical session users.
It isn’t displayed until the user has successfully authenticated.If you run the theme preview option
(can do it while you’re logged into a desktop session, it is intended for testing while tweaking themes)
slim -p /usr/share/sllim/themes/<mytheme>
you can enter uname+password and view the rendered welcome_msg onscreen for a few seconds until the preview ends.
marcelocripe
Hi BobC,
Will we be able to have Slim translated into several languages, with the possibility of correcting the inserted text, or of returning in the “Username” field after the cursor is in the “Password” field?
marcelocripe
(Original text in Brazilian Portuguese)———-
Olá BobC,
Será que conseguiremos ter o Slim traduzido para vários idiomas, com a possibilidade de corrigir o texto inserido, ou de retornar no campo “Username” após o cursor estar no campo “Password”?
marcelocripe
(Texto original em idioma Português do Brasil)marcelocripe, please download the zipfile attached to BobC post #48160
View the “background.jpg” and “panel.png” imagefiles.
Understand that SLiM has no ability to display dynamically inserted text translations.
- You must be logged in to reply to this topic.
