Its not too much to ask, just too much to expect.

_________________
Want to design a PCB for your project? I strongly recommend KiCad. Its free, its multiplatform, and its easy to learn!
Also, I maintain KiCad libraries of Retro Computing and Arduino components you might find useful.

BTW, Bruce Dawson writes great performance problem investigation reports on his blog:

Hey Synaptics, Can You Please Stop Polling!
What is Windows *doing* while hogging that lock
24-core CPU and I can’t move my mouse

Wikipedia: Apollo_Guidance_Computer
"The computer had 2048 words of erasable magnetic core memory and 36 kilowords of read-only core rope memory.
Both had cycle times of 11.72 microseconds."
The amount of computing power required to bring a man to the moon and back in 1969.

Wikipedia: Million_instructions_per_second
Intel Core i7 6950X -> 317,900 MIPS at 3.0 GHz
Would be interesting to know, what amount of computing power and RAM it takes nowaday (2017) on a modern PC
just to write one page of text about bringing a man to the moon and back...

A lot more than it does to write that text on a Commodore 128.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I booted with the one-floppy Unix called Tom's Root/Boot, and checked the size of /bin/cat - it was 7k. So, in a very weak sense, I could write a page of text on a modern PC using a 7k program...

... that said, the shell is 52k. The compressed kernel image is 240k. And the overall RAM footprint is about 9000k.

http://www.ibiblio.org/pub/Linux/system/recovery/
via
http://www.toms.net/rb/

Nine megs is really small compared to most major OSes. Windows 7 needs around 200 times that(2 gigs of ram, in my experience).

An absolutely minimal install of Win7 32-bit can run "OK" on 1gig, but I wouldn't recommend it. I remember reading an article about someone who investigated the absolute minimum RAM that Win7 would boot on; not necessarily be useful, but would boot - even if slowly. It turns out to be 64MiB, though it was slow as molasses because of the constant disk thrashing.

_________________
Want to design a PCB for your project? I strongly recommend KiCad. Its free, its multiplatform, and its easy to learn!
Also, I maintain KiCad libraries of Retro Computing and Arduino components you might find useful.

When we got our early Sun workstations, I can't honestly say how much memory they had. This was in the mid-80s. We had a Sun 2/160, and a Sun 3/50. They may well have been "maxed out" at 8MB and 4MB respectively (though I dunno about that 3/50).

We didn't have X-Windows, they ran SunTools, Suns internal GUI. The 3/50 was black and white, the 2/160 was color.

My NextStation, ca 1994, was a 25MHz 68040. It came with 8MB, but definitely ran better with 20MB. It maxed out at 32MB.

I'll say this again, the idea that "it's amazing" that a 1GB Rasperry Pi "can run Linux" is when I contrast it to what NextStep did with 8MB, much less with 16-20MB. How nice would it be to have NextStep ported to ARM and running on a Pi.

NextStep ran a full Display Postscript environment (not the cheapest, nor fastest way to generate graphics, I can tell you).

But, heck, I think XClock today compiles to several MB of code today.

The culprits today are really things like the high res displays, and the resources they need. I have icons on my Mac that are larger than the original 128K memory of the original Macintosh.

The other surprising sucker of resources (at least in the compiled executables, not necessarily RAM) is all of the multinational support. The plethora of extra languages and what not that come inherently with the runtime, it all has a cost.

I need to get my Pi fired up. Put NetBSD on it, and then Open CDE as a desktop. CDE was a desktop designed for machines with MB of RAM, not GBs. It should still load up a web browser (which will consume all resources until the heat death of the Universe). "Heat Death, thy name is Firefox!"

The Pi runs Linux just fine. It's just when you try to run a full-featured web browser or office package on it, it asphyxiates quickly.

_________________
WFDis Interactive 6502 Disassembler
AcheronVM: A Reconfigurable 16-bit Virtual CPU for the 6502 Microprocessor

...and to think the minis I worked with 40 years ago were lucky if they had 128K total RAM. Yet they could support 16 simultaneous users on character terminals. Needless to say, swap space on the disk got used a lot.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

An easy way to remember what was normal back in the day: emacs was nicknamed 'Eight Megs And Constantly Swapping' - but these days it counts as a lean and mean editor!

A well-chosen web browser on the Pi should be fine. Eclipse, on the other hand, I wouldn't attempt. (But note this: the original Pi had just 256M, with only half available to the CPU. The present Pi models have 1G. That's a factor of eight increase.)

Ubuntu recently ran a survey for users to select what application they would like to be the "default" app for different use cases.

There was a great post on /. with one users suggestions:

Well, yeah. Emacs is a great OS, but its editor sucks.

_________________
WFDis Interactive 6502 Disassembler
AcheronVM: A Reconfigurable 16-bit Virtual CPU for the 6502 Microprocessor

Hm.. the /. poster forgot Tetris. Works great in xemacs at least. Last time I checked.
I wouldn't say the editor sucks.. it's programmable, and I use it all the time (vim for /etc/hosts and such, just as the /. poster). It's a work horse. My favourite editor, together with TPU and PED (and most people don't know those two these days).

I did set up 8MB i386 Linux boxes as workstations/x-terminals for a company (b/c that's the hardware they had) back around '94. Worked fine.

Slightly OT: do you know of a version of PED available for current(ish) computers?

... Or maybe I should implement a PED mode for Emacs