Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.misc > #90485 > unrolled thread

Ancient History

Back to article view | Back to comp.os.linux.misc

Contents

  Ancient History rbowman <bowman@montana.com> - 2026-08-25 04:41 +0000
    Re: Ancient History c186282 <c186282@nnada.net> - 2026-08-25 03:15 -0400
      Re: Ancient History John Ames <commodorejohn@gmail.com> - 2026-08-25 08:19 -0700
        Re: Ancient History The Natural Philosopher <tnp@invalid.invalid> - 2026-08-25 19:57 +0100
    Re: Ancient History "Carlos E.R." <robin_listas@es.invalid> - 2026-08-25 11:41 +0200
      Re: Ancient History Nuno Silva <nunojsilva@invalid.invalid> - 2026-08-25 11:53 +0100
        Re: Ancient History The Natural Philosopher <tnp@invalid.invalid> - 2026-08-25 12:54 +0100
          Re: Ancient History rbowman <bowman@montana.com> - 2026-08-25 19:25 +0000
        Re: Ancient History "Carlos E.R." <robin_listas@es.invalid> - 2026-08-25 14:23 +0200
          Re: Ancient History c186282 <c186282@nnada.net> - 2026-08-26 03:28 -0400
      Re: Ancient History The Natural Philosopher <tnp@invalid.invalid> - 2026-08-25 12:31 +0100
        Re: Ancient History Chris Ahlstrom <OFeem1987@teleworm.us> - 2026-08-25 07:53 -0400
        Re: Ancient History Charlie Gibbs <cgibbs@kltpzyxm.invalid> - 2026-08-25 18:26 +0000
          Re: Ancient History rbowman <bowman@montana.com> - 2026-08-25 19:37 +0000
      Re: Ancient History rbowman <bowman@montana.com> - 2026-08-25 19:16 +0000
        Re: Ancient History The Natural Philosopher <tnp@invalid.invalid> - 2026-08-25 20:21 +0100
          Re: Ancient History rbowman <bowman@montana.com> - 2026-08-26 04:40 +0000
    Re: Ancient History The Natural Philosopher <tnp@invalid.invalid> - 2026-08-25 12:30 +0100
      Re: Ancient History c186282 <c186282@nnada.net> - 2026-08-26 02:49 -0400
    Re: Ancient History "Mr. Chang Man-wai" <toylet.toylet@gmail.com> - 2026-08-25 20:07 +0800
    Re: Ancient History Paul <nospam@needed.invalid> - 2026-08-25 11:20 -0400
      Re: Ancient History Robert Riches <spamtrap42@jacob21819.net> - 2026-08-25 17:44 +0000
      Re: Ancient History Lawrence D’Oliveiro <ldo@nz.invalid> - 2026-08-25 21:55 +0000

Page 2 of 2 — ← Prev page 1 [2]

#90540

On Tue, 8/25/2026 12:41 AM, rbowman wrote:
> In my OCD quest to rejuvenate old machines I finally dug out the tower 
> buried under other junk out in the shed. Turns out it's a Compaq Presario 
> 4550. It has a 233 MHz AMD K6 that was AMDs challenge to the Pentium II. 
> It was pin compatible and could be used in a Socket 7 board. It has 48 MB 
> of RAM (no, that's not a typo) and a 4 GB HDD. For your computing 
> convenient, it has a 56K modem. There are 2 USB ports although I think it 
> expects the mouse and keyboard to be plugged into the PS2 ports.
> 
> Two distros that claim to run in 48 MB are Tiny Core and Slax. slax 
> specifies text only unless you have more RAM. Tiny Core isn't very 
> specific.
> 
> I think I'll let that sleeping dog slumber on although I might see if it 
> boots at all. 
> 
> Now for the real puzzle. I was a fairly decent machine in 1997 and came 
> with Windows 95. Windows 95 claimed a minimum of 4 MB, recommended 8 MB, 
> and if you had 16 MB or more didn't bother creating a swap file. What the 
> flaming hell has happened to us?  Even by Windows 98 24 MB was 
> recommended. Now you can barely run a pared down Linux in 48 MB and I 
> doubt it's a very appealing UX.
> 

https://www.vogons.org/viewtopic.php?t=50935

    "I find working through all the "trouble" to be more rewarding than the result."

A lot of software you could run on such a machine, would have an "SSE2 requirement",
so you may have to stick with software from the era.

At one time, packed data structures were "quite popular". Today, you can have
an individual write an image utility, where instead of using 24 bits for
a pixel, "each color is represented by a 64 bit number" or "we're using a floating
point number for R, for G, and for B". Proud claims of piggish design :-)
The reason I note this, is one day, I attempted to open an image,
sure, the image was a large one, the machine returns "out of RAM".
And that's when I started checking for what exactly they were
doing inside. You do an eye roll. You're yelling "why would you do this?".
And so on :-)

   Paul

[toc] | [prev] | [next] | [standalone]

#90548

On 2026-08-25, Paul <nospam@needed.invalid> wrote:
> On Tue, 8/25/2026 12:41 AM, rbowman wrote:
>> In my OCD quest to rejuvenate old machines I finally dug out the tower 
>> buried under other junk out in the shed. Turns out it's a Compaq Presario 
>> 4550. It has a 233 MHz AMD K6 that was AMDs challenge to the Pentium II. 
>> It was pin compatible and could be used in a Socket 7 board. It has 48 MB 
>> of RAM (no, that's not a typo) and a 4 GB HDD. For your computing 
>> convenient, it has a 56K modem. There are 2 USB ports although I think it 
>> expects the mouse and keyboard to be plugged into the PS2 ports.
>> 
>> Two distros that claim to run in 48 MB are Tiny Core and Slax. slax 
>> specifies text only unless you have more RAM. Tiny Core isn't very 
>> specific.
>> 
>> I think I'll let that sleeping dog slumber on although I might see if it 
>> boots at all. 
>> 
>> Now for the real puzzle. I was a fairly decent machine in 1997 and came 
>> with Windows 95. Windows 95 claimed a minimum of 4 MB, recommended 8 MB, 
>> and if you had 16 MB or more didn't bother creating a swap file. What the 
>> flaming hell has happened to us?  Even by Windows 98 24 MB was 
>> recommended. Now you can barely run a pared down Linux in 48 MB and I 
>> doubt it's a very appealing UX.
>> 
>
> https://www.vogons.org/viewtopic.php?t=50935
>
>     "I find working through all the "trouble" to be more rewarding than the result."
>
> A lot of software you could run on such a machine, would have an "SSE2 requirement",
> so you may have to stick with software from the era.
>
> At one time, packed data structures were "quite popular". Today, you can have
> an individual write an image utility, where instead of using 24 bits for
> a pixel, "each color is represented by a 64 bit number" or "we're using a floating
> point number for R, for G, and for B". Proud claims of piggish design :-)
> The reason I note this, is one day, I attempted to open an image,
> sure, the image was a large one, the machine returns "out of RAM".
> And that's when I started checking for what exactly they were
> doing inside. You do an eye roll. You're yelling "why would you do this?".
> And so on :-)
>
>    Paul

What image [mis-]handling software was being that wasteful of
RAM?

ImageMagick also has a limit on the size of a "pixel cache" with
a default of 2GiB.  That limit can be exceeded when working with
fairly high resolution scans of full-page paper documents.
Fortunately, it is fairly easy to specify a larger limit in
/etc/ImageMagick-7/policy.xml

-- 
Robert Riches
spamtrap42@jacob21819.net
(Yes, that is one of my email addresses.)

[toc] | [prev] | [next] | [standalone]

#90571

On Tue, 25 Aug 2026 11:20:03 -0400, Paul wrote:

> At one time, packed data structures were "quite popular". Today, you
> can have an individual write an image utility, where instead of
> using 24 bits for a pixel, "each color is represented by a 64 bit
> number" or "we're using a floating point number for R, for G, and
> for B". Proud claims of piggish design :-) The reason I note this,
> is one day, I attempted to open an image, sure, the image was a
> large one, the machine returns "out of RAM". And that's when I
> started checking for what exactly they were doing inside. You do an
> eye roll.

Sounds like OpenEXR. Quite popular among the CG content-creation
crowd. Also incidentally something that Adobe isn’t quite au fait
with.

<https://openexr.com/>.

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | comp.os.linux.misc

csiph-web