KDE way faster

Message

Mr Doolie · #1 Post by **Mr Doolie** » Wed 01 Mar 2006, 08:25

I've been playing around with different Linuxes and noticed that in Puppy KDE is **WAY** faster. Why? It's loading off the hard drive not RAM isn't it? Why is it 50 times faster than Mandriva's KDE or Ubuntu's KDE?

Is Puppy's KDE the "real" KDE? Just for fun I'd like to "bloat" it up with every KDE thing I can find. That's the fun of Puppy. You can back up your pup001 file and play around with a "disposable clone version" of your system.

Barry, you da man!

Lobster · #2 Post by **Lobster** » Wed 01 Mar 2006, 09:11

Not 50x but it certainly is much faster and it gives access to the wealth of KDE programs.

We have several versions of KDE
Kenny Revolutionists is the most complete and independent
Barry released a version (anyone got a URL) I think as a pupget
Nathan released a version as an .sfs (anyone got a URL) that worked OK with Puppy 1.0.6/1.0.7 (probably with 1.0.8

There was also some development from Bombayrockers and others - did this materialise - what happened?

KDE on Puppy - Sure! Faster by far
Gnome? On its way . . . (hopefully) any updates?

Is there anything that Puppy can't do?

bombayrockers · #3 Post by **bombayrockers** » Wed 01 Mar 2006, 09:31

I had released a version of KDE 3.4 for puppy (as usr_mode.sfs) which can be downloaded from
http://www.murga.org/~puppy/viewtopic.p ... highlight=
I am using puppy 1.0.7 and have not tested kde usr_more.sfs in 1.0.8, but I think it should work.

Nathan F repacakged this to work best with grafpup.

Kennys version of kde (kde 3.5.0) is available as an iso and is faster than mine, however it is far from complete as it is still under development and kenny is already working on puppy + kde 3.5.1. Kennys version is faster because everyting is loaded to the ram, so low ram users may not be able to use it.

I recently browsed slax repos and found that there is a 80 mb gnome module, so I have started working on this.

kenny is also working on gnome based distro.

Nathan F · #4 Post by **Nathan F** » Wed 01 Mar 2006, 14:38

Yes, my KDE squashfile was based off of Bombayrocker's work. It was also pretty Grafpup specific, so it's not exactly plug and play with Puppy (although it can be made to work with a little effort).

I'm actuall working a little towards Gnome myself, but wasn't intending for a full Gnome gui. I'm upgrading GTK and othe4r related libraries in Grafpup (the first step in my opinion) and compiling most of the core Gnome libraries. This ios actually just in order to provide easier accessability to certain Gnome based applications like Kino (video editing). My best guess is that attempts to get a Gnome desktop working in Puppy without first doing the GTK upgrade will be very difficult. Puppy's GTK2 is currently 2.4.0, while most current Gnome packages will require at least 2.8 and the current GTK is 2.8.13. Also, Cairo support is missing and this has to be compiled before Pango, which in turn must be compiled before GTK. You guys are welcome to prove me wrong but I think it will be very frustrating indeed.

As for why KDE is faster in Puppy? Even if KDE itself has to be loaded from the HD, the core of the operating system itself exisats in the RAMdisk, speeding everything up. OpenOffice runs a good deal faster too. The principals involved could really revolutionize computing if the rest of the world were to really take notice.

Nathan

Owl · #5 Post by **Owl** » Wed 01 Mar 2006, 15:04

Nathan F wrote: The principals involved could really revolutionize computing if the rest of the world were to really take notice.Nathan

This is what bugs my mind for a month or so now: everything is going sooo smooth in Puppy (e.g. OOo 2.0), why can't this be repeated in other distros ???

So far so good for Puppy, but I totally agree that Puppy represents a revolution in some ways, more then an evolution . . . !

However, I guess I am really off-topic here !

Regards,
Owl.

#6 Post by **Pizzasgood** » Thu 02 Mar 2006, 23:43

Maybe in a decade or so, the rest of the world will catch up. Then we will need 20GB of ram so we can load the big OSes into ram. Then they'll say, "Why don't we just skip the hd and go total ram?" They'll answer, "You can't save on ram." So, they'll make a newer, faster memory that is persistant AND faster than ram. And have the os use more of a save-state deal rather than a shut-down and startup. Push power, and you're back where you were when you turned it off like THAT. Other than a couple tasks like logging on to the network, it would be nearly instantaneous.

But for now, we have both Puppy and bragging rights

Mr Doolie · #7 Post by **Mr Doolie** » Fri 03 Mar 2006, 00:52

Pizzasgood wrote:Maybe in a decade or so, the rest of the world will catch up. Then we will need 20GB of ram so we can load the big OSes into ram

Would you settle for a 4Gig Ram board that looks like a HDD?
http://www.techreport.com/reviews/2006q ... dex.x?pg=1

Now THAT would be great for testing distros and fooling around. If you farkle it no harm done whatsoever.

It has some limitations of course (don't leave your system unplugged for more than 10 hours) but it wouldn't be difficult to write a script to backup/restore the entire thing. Perhaps another model has a better battery backup or something.

Barry. Puppy V3.0 with RAMHDD support?

#8 Post by **Pizzasgood** » Fri 03 Mar 2006, 03:02

That's what I'm talking about. Now, get a bigger, better, cheaper version without that 150mbs bottleneck, and we're in business.

It does seem like everybody focuses on processors and graphics cards (just another processor) when they could make something like this even better. Sure, processors will get things processing faster, but the starting times won't change. Sitting there waiting for it to load things is the process that irks me the most, not waiting for it to get done thinking. And that's on a 450MHz system. So if I had a 3GHz system (like my prospective college thinks I should have), it would be more than fast enough. But I'd still be sitting there waiting for stuff to start. Except I won't, because I'll have Puppy.

Lobster · #9 Post by **Lobster** » Fri 03 Mar 2006, 08:25

Owl wrote: This is what bugs my mind for a month or so now: everything is going sooo smooth in Puppy (e.g. OOo 2.0), why can't this be repeated in other distros ???

So far so good for Puppy, but I totally agree that Puppy represents a revolution in some ways, more then an evolution . . . !

However, I guess I am really off-topic here !

Other distros (as was pointed out to me recently) do different things. They have a different history.
Just as I am not impressed with MS hype, the same is true of Linux excuses.
The Linux mantras, "Do not run as rute", "If it moves mount it". "C is the Linux language", "Choice is good" and so on are all embedded memes hardwired in a penguins distrobox of a brain.
Well Puppy can be made to run as spot but mostly it runs as rute and like most users I have no complaints. None. I am rute. So?
If I tell a CD draw to open it is because I want it to open [no more clues computer]
If I move data to a disk it is because I want it copied there.
Nobody has explained mounting to my satisfaction. It is a relic. A way to ensure I really really REALLY am sure I wan to commit hardware. Relic.
People still write in C because it has no successor. I blame Sun, they had java and did what? Hardwire it? Optimise it? No they waited for MS to rebrand it as .NEt
LOL - Choice is not good until you have generic commonality

. . . and now to Puppy. Linux never had an advantage on the desktop over Windows or Apple - servers were more reliable yes yes yes - I know that. Desktops are now very impressive. Linux is now on a par and will soon overtake (year or two perhaps - perhaps).
Puppy does have an advantage. It works beautifully. It is interesting for example that (I am sorry I forget where) in 1.0.8r1 some of the rather developorish wizards are being replaced by a few lines of polish. Most people equate outer polish with inner gleam. tsk tsk . . .

I don't yet understand puppy2 but as primarily a user I don't have to. If XP or Redhat or any combination works for you - use it.
The trend at the moment is for online Operating Systems (server side) - temp move - yet to emerge fully.
Long term the distinctions between OS and data will become blurred.

. . . OK enough ranting . . .
Don't believe the hype.
Use Puppy and be happy.

I think I will go on
for an emergency Puppy rant here:

http://tmxxine.com/forum/

Sage · #10 Post by **Sage** » Fri 03 Mar 2006, 09:56

PG makes some good points. The future (as was the past!) lies in embedded OSes, possibly, if using Puppy, embedded apps. EEPROMs are not only cheap, but can be flashed - they just need to be more/bigger. As so many have observed, once the ~600MHz cpu speed was breached there has been little objective value in further increases. It's just that AMD & Intel are the market drivers! Memory keeps getting reinvented - 45nm (IBM 39.5!) DDR4 will arrive long before that line width in cpu silicon, so memory becomes the workhorse.
So, uncovering my crystal ball, I see a tiny 3W VIA 1GHz cpu responding to long banks of embedded EEPROMS running Puppy out of a multiplicity of DRAMs. A power brick on the wall socket should suffice. No fans, no noise. Instant on. The whole lot stuffed inside the lcd monitor or keboard. Oh dear - Alan Sugar worked all that out back in 1980, even before IBM conceived the PC.

#11 Post by MU » Fri 03 Mar 2006, 10:08

There is a "golden rule".
An increase in processor-speed by factor 10 will be "felt" as a computer just twice as fast.
With "normal" applications (office), not with specialized stuff like 3D-rendering.
Reason is, that other factors are more important (speed of videochips, harddisks and so on).
Putting everything into RAM actually is the best way to make things faster.

If you have to calculate:
better slow processor with lots of RAM than vice versa.

Mark

(old)Puppy Linux Discussion Forum

(old)Puppy Linux Discussion Forum

KDE way faster

KDE way faster

KDE, Gnome and Puppy woosh

emergency rant