(old)Puppy Linux Discussion Forum

amigo

Shouldn't you be that bloke off 'The Fast Show', who kept saying "You don't want to do it like that!"?

Not the best job title in the world, but satisfying in its own way.

'The Fast Show' does not compute here, Will Smith.

My opinion!

Creating a CE edition of puppy linux.

What base to use... I test a lot of puppy's and have for the past several years now. I look for things probably not many testers do, first off
(1) Ease-of-use for new (and experienced) users - And that means not a lot of initial problems that can't easily be fixed if necessary!
(2) Display, Display, Display - Clarity of graphics and pictures, coloration of graphics and pictures and 'types' (.jpg .png. .gif etc.) In this area, there are many that are 'good' and a few 'close' to being perfect .
(3) Programs... what's available from the repo PLUS what other's work from a variety of other repo's without installation problems.
(4) How much RAM is utilized at idle and during operation and the temperature of the CPU and how fast the 'fan' runs at idle and at various program operations.
(5) Response of various programs especially videos in Mplayer... does it 'Buffer' a lot trying to keep up to the display or is it smooth and synchronized with audio and video.

Of course there are other things I test in a distro but these are the first and most important - at least for me.

All my testing is done on 3 computers -
(A) Dell 3000 with intel pentium 4 CPU 2.80GHz graphics.
(B) HP (P3 equiv) with AMD graphics
(C) IBM Thinkpad T22 with intel graphics.

My personal findings over hundreds and hundreds of tests keep pointing to ONE program that I believe would be the best basis for a CE build and that would be 'RACY.'

In fact... what did BK eventually wind up using for his new Quirky-6 build... 'T2' and I wouldn't be afraid to bet if one really looks close at the inner-code you will find a lot of RACY in there. Barry has many years of experience building puppy's I'm sure he selected very carefully what to use for Quirky-6.

Anyway... the latest Racy-5.5 with Kernel 3.0.66 (i686) is a very nice running program and is one of my favorites as it handles everything I mentioned above very well.

There was also a Racy-5.4.91 where BK has the Kernel 3.8.0 built in that could provide a good base for a Puppy-CE. I am just starting to 'play' with this one, probably a few bugs but I would think most are fixable.

Understand... 'Just My Thoughts' but this would be my choice if I were building a Puppy-CE!

>>>---Indian------>

sszindian wrote:
Of course there are other things I test in a distro but these are the first and most important - at least for me.

All my testing is done on 3 computers -

Here's a radical idea...choose the project that most closely matches your hardware in addition to how you use the computer. A project that attempts to run everything always loses efficiency.

jpeps wrote:

sszindian wrote:
Of course there are other things I test in a distro but these are the first and most important - at least for me.

All my testing is done on 3 computers -

Here's a radical idea...choose the project that most closely matches your hardware in addition to how you use the computer. A project that attempts to run everything always loses efficiency.

Hmm, but that means it would be very difficult to produce a "One Size Fits All" OS.

Ghost Dog wrote:Simargl as Package Management Lead Dev

Why does it have to be simargl?
Why not Justin Bieber, or Paris Hilton?
That clown, listed together with real devs, means any community effort will be nipped in the bud.
.

tlchost wrote:

jpeps wrote:

sszindian wrote:
Of course there are other things I test in a distro but these are the first and most important - at least for me.

All my testing is done on 3 computers -

Here's a radical idea...choose the project that most closely matches your hardware in addition to how you use the computer. A project that attempts to run everything always loses efficiency.

Hmm, but that means it would be very difficult to produce a "One Size Fits All" OS.

More like rule it out completely. The job of the developer is far easier when a project is well defined and limited. Applications can be smaller and targeted. That's also in keeping with current trends.

Greetings puppy community edition fans

If a few of the members of the puppy community want to make a community edition enough to actually do something instead of just talking endlessly. It will eventually get done

The people that criticize others but do nothing themselves should be ignored. They are just the meaningless background noise of life.

I have removed the link to the CE repository because i can think of no way to distribute it appropriately and i do not want to post it generally and open it to abuse. if someone has an iso that they wish to share they will have to use their own link.

I now look forward to the new year filled with many enjoyable hours working on linux and puppy

Merry Christmas to all and to all a good night

Wanderer

wanderer wrote:Greetings puppy community edition fans

If a few of the members of the puppy community want to make a community edition enough to actually do something instead of just talking endlessly. It will eventually get done

The people that criticize others but do nothing themselves should be ignored. They are just the meaningless background noise of life.

I have removed the link to the CE repository because i can think of no way to distribute it appropriately and i do not want to post it generally and open it to abuse. if someone has an iso that they wish to share they will have to use their own link.

I now look forward to the new year filled with many enjoyable hours working on linux and puppy

Merry Christmas to all and to all a good night

Wanderer

First Wanderer I want to say thank you for trying to start the initiative of a CE Distro but to be honest what you were proposing is a Pupplet not really suitable for a mainline Community Distro and to be honest a System like the Community model is only as good as those involved with it is on the whole a successful system on what many great distros have been built on.

I don't think any criticism of the community system is warranted and really falls or succeeds in the hands of that community. I still believe the frankly the silly Benevolent Dictator Model is not what Open Source/Foss is about.

Unfortunately this Thread got diluted into a wish fest and some insults - Just to long in the debate dept. You need to have debate to get ideas thrashed out but also know when to say ok we must decide to go in this direction and go with it.

Anyway Wanderer Merry Xmas and to all those in this Community and who knows what might come out next year for example Dejan and an updated DPUP from him and the ongoing 5.286 project well done Rerwin and co - Level headed stuff. Anyway good luck Puppy Community for the new year and simargl with your distro - just don't bely your talents with wasteful silly trolling.

Best wishes all


Darren

darry1966

For some odd reason, I had no notifications of postings by users, grrr that gets me, anyways The next CE I think should be be using the same kernel, glibc, and xorg as debian, or ubuntu or Slackware, But forget all the massive backend things, keep it gtk2 and we will compile what is needed after we have jwm.
then we could move on.
ttuuxxx

Correct me if I'm wrong but puppy "Community Edition" has always been, as far as I know just a latest official stable puppy polished and fixed by community.

If there was stable puppy 1.08, it was 1.09CE
For puppy 2.14, there was 2.15CE
For puppy 4.1x there was 4.2.1 CE

So earlier when someone like you started thread for CE edition, puppy community said just "let's do it!" and there was no question in base, latest puppy was served by Barry and left for community to play with.

I think the reason we don't see new CEs is because a lot of polishing nowadays is done directly to woof and now that there's woof-CE I think development will only go for better. (I remember when doing puppy 4.2.1 lots of fixes done for that version weren't incorporated back to official puppy because Barry did not monitor changes and in meantime worked on something else that became next puppy, but now we have woof-CE)

The other reason is there isn't one mainline puppy anymore but with variety of choices for binary compatibility we now have dpup, spup etc... and there are developers that took role of maintaining them and part of community that follows that specific project.

So wanderer why such haste for "CE" ?
Are you afraid puppy is going to disappear? Because it's not.

As far as future development of puppy generally goes I think we should do:
Developers should continue work on the projects they normally do and send back fixes to woof-CE
Users can test and suggest new fixes

I don't think puppy needs a total rewrite, puppy as we know it is very good just needs more polishing.
As far as binary compatibility goes woof was created because there was always issue with lack of puppy packages and updating/recompiling them for next release. Puppy is not supported and managed by some big company. There are many good developers here who like contributing to puppy but there's noone to tell them "Maintain this, this and this package for xxx future puppy versions" but that's the beauty of puppy, there are still many awesome puppy releases.

Puppy is beautifull
Thanks to Barry and everyone who ever contributed to it and keep the good work guys.

Dejan

Here is a what I want.

Most of the Puppy commands >> puppy pfix=ram ...etc

Why are we still having to type it.

Particularly the obscure ones for video cards.

Why not numbers.

We can hit F2.

Why not an F3 >>> then a quick description.. >> then hit a number.

I think Shinobar had a really early JP release with much more options.

About time we came out of the caves.

Was it you ttuuxxx who did the early work on the first boot logo??

To me this is an obvious deal breaker for newbies.

And quite honestly I hate that command crap on booting up.

Cheers to all.....Chris.

cthisbear wrote:And quite honestly I hate that command crap on booting up.

That command crap is awesome and important. You can't get any info off of a bloated splash screen program that appeare(s/d) to (be/) freez(en/on).

YOU DO NOT NEED ANY OF THAT CRAP.
You can boot straight to X+jwm in less than 10 lines of code in about 0.2s and let the other "garbage" run in a terminal.

basic steps
0 export variables
1 autologin
2 mount sys proc dev and dev/pts
3 start Xvesa or xfbdev (or Xorg but it takes longer and is larger)
4 start wm
5 use the wm's startup command to run the rest in an rxvt -e ... in jwm this is <StartupCommand>
6 While that is running pop up a configuration gui for all the stuff requiring user input including the possibility to switch to Xorg, change language, keyboard etc...

using a splash is a waste of time ... just set a default bg image for jwm in the /etc/system.jwmrc (or /usr/etc/system.jwmrc if a less informed packager built your jwm)

Continue discussing the color of the bike shed if you want, I have already demolished it and paved over it.

technosaurus, do you have alternative rc.sysinit / xwin /xinitrc that you use? Have you already posted it somewhere?
See now we're talking, let's get some code suggested for improvement and maybe it will be accepted in woof-CE if it still provides same functionality but with more efficient startup times etc...
Is there a woof-CE thread or are developers supposed to send commits directly to git?
Why not have both (git for developers and thread for regular users and suggestions?)

But wait, I want to use /usr mounted on NFS and run without X as I am building a server. Can your init system do that? And, if I run as you suggest, what happens if the X server won't start on my video hardware? And what about if I want a sane multi-user system where every users needs to login?

amigo, I said "if it provides same functionality" meaning if it stays in puppy fashion of doing things

I said I think puppy doesn't need rewrite but no reason not to optimize/improve code in scripts if in the end it has same effect.
If I understand correctly technosaurus said that most stuff that user sees on startup could be run in background while starting X faster?

EDIT: On the multiuser thing: pizzasgood proved that it can be done but Barry has gone in the wrong direction with implementation of spot and not giving users choice to add more users with own home directories.
I think most of puppy users want to run as root, it's more of the linux community outside of puppy that always criticize puppy because of it but I think it's a feature that could be added to puppy (again, it's mostly puppy scripts that need to be changed, not totally rewritten to use different directories instead hardcoded /root/, you can already create new users in puppy but you can't run Xorg and other puppy scripts won't work well)

dejan555 wrote: On the multiuser thing: pizzasgood proved that it can be done ..... I think it's a feature that could be added to puppy (again, it's mostly puppy scripts that need to be changed, not totally rewritten to use different directories instead hardcoded /root/,

I think it would be very handy for a CE puppy to have some sort of "sandbox" (is that the same thing as spot and fido??) where a user could run a browser.

Even if this was not a truly multiuser functionality it would be nice to be able to let the kids tinker online knowing that they couldn't damage system files (and that the browser couldn't damage them either...)

Could this be achieved via an sfs? A sandbox_browser sfs??

greengeek, dpup 484/485 has a "SafeBrowse" script that runs browser as user "nobody" in /tmp and delete files after close, it doesn't work for me anymore with newer seamonkey versions as seamonkey complains about missing profile, but maybe can be fixed or made to use with other browser? Here's the code
There's also this:
http://murga-linux.com/puppy/viewtopic.php?t=45884
(haven't used or tested)

Running as spot/fido is running as restricted user, so it is more secure too.

Hi again,

Sorry if it seems I'm going over the same material, but it occurred to me, unless I've got it wrong again, that wanderer and I aren't very far apart. Our visions are, in fact, compatible.
First an apology to goingnuts. I represented, without citing a webpage reference, that he wasn't sure pUPnGO was a good idea for a Community Edition. As I understand wanderer's vision, it is to use pUPnGO as the base, onto which different flavors of SFSes could be loaded. [The pUPnGO base, itself, would be a fully working OS capable of running on old, resource-limited computers. The flavors could include one or more capable of using the most recent applications and taking advantage of the resources of recent computers]. This is not very different from my previous suggestion. http://murga-linux.com/puppy/viewtopic. ... 752#742752. Back in August, bark_bark_bark suggested a somewhat different approach to modularity. See the discussion beginning here: http://murga-linux.com/puppy/viewtopic. ... 197#722197, as to which goingnuts responded: “You are sure pupngo is a healthy starting point? (I'm not).