Puppy 5.0 "New Hope"

[KF6SNJ]

ttuuxxx,

I fully support you trying to initiate this project. You know far more about computers that I do. I think you will be able to make this puppy fly. I think you have a good idea of what needs to be done. I would be happy to test the first few alphas to see what I see. Who knows, you may be unto something big.

[MU]

I really would recommend, that localization gets a higher priority.

I installed Puppy 409.
Neither xorgwizard nor the networkwizard includes localization, though I had provided patches in the past.
With xorgwizard, there was a parallel development, so I can understand it.
But networkwizard already was localized in some versions of Puppy, and now is off again, because a fork of the networkwizard is used instead.

It certainly is great to have better wifi support now, but really disappointing to see, that former hard work is completely dropped.

Mark

[disciple]

Once again, don't panic!

1. Don't replace reliable calculators with the bug-ridden (but very nice) galculator. Fix the bugs first
2. Don't add unnecessary libs or large programs like the Gimp. The main Puppy should not be bloated, or he will lose his usefulness to a lot of people, and slowly lose his identity and become just another medium (non-Gnome/KDE) distro. We should instead make much greater use of addon squashfiles. These could be put together by different people, and could be updated more or less frequently than the main Puppy .iso
We could have an openoffice .sfs, a devx .sfs, a graphics .sfs (Gimp, Cinepaint, Pencil, Xnview...?), a FF/Thunderbird .sfs - even a KDE! .sfs and a Gnome .sfs
Since FF/Thunderbird and openoffice are such a big deal for many people, we could even drop Abiword/Gnumeric and/or Seamonkey from the base distro and put them in squashfiles.
This would be much more sustainable than several forks of Puppy (feel free to make bloated forks though ) - even Nathan hasn't been able to really keep Grafpup going.

[technosaurus]

maybe to calm things down a bit

a dingo/T2 fork for more hard core developers (at least for now) using ftp://distro.ibiblio.org/pub/linux/dist ... /t2-dingo/
from Barry's forum post http://www.murga-linux.com/puppy/viewtopic.php?t=32972 and http://www.t2-project.org/packages/
T2 should allow 64 bit and ppc (and more) architectures - start with 4.1 final, devx.sfs and http://dl.t2-project.org/source/t2-7.0-rc2.tar.bz2

and a 3.x slackware version (already in progress) http://puppylinux.org/community/puppy-c ... ick/lassie

and a minimalist 2.x version for older systems

or work on Simplux (puppy + gentoo) while Leachim is working on his Simplicity compiler

Does anyone still want a 2.4 kernel? guess that could be T2 also or 1.x based

OR if you can't play well with others - make your own puplet

[HairyWill]

I really would recommend, that localization gets a higher priority.

I agree. For a large number of users this would be a significant reason to upgrade.

Dougal pointed out to me the other day that when we move to a community developed puppy it would make sense for those involved in creating localised puplets to make their changes directly to the applications before a version is released. At the same time we could include in the repository/installer a mechanism that would allow users to install the mo po files for their language and chosen applications only.

I'm surprised to learn that Dougal hasn't included localisation in the network wizard.

[MU]

I'm surprised to learn that Dougal hasn't included localisation in the network wizard.

Though I usually see many advantages in the "chaotic" structure of the Puppy Linux project, this issue also shows us disadvantages.

Many people work on different versions of the same program, what makes it difficult, to get a version, that includes all the improvements that were made.
Dougal just took one of the different versions as a base for his work, added tweaks, patches, and suddenly there was a parallel development, a fork.
This certainly was not intended by Dougal, it is a general problem of the chaotic development process.

I had the same problem, when I localized xorgwizard.
I used an older version, so I unwillingly created a fork, that never could make it into the official Puppy.

Barry has different priorities than other people have.
E.g. localization always was not in his focus, though he tried to add suggestions from people.
So he drops a version with localization, because another version supports more chipsets.
He has to make a decision, and as his priority as native english speaking man are new technologies, the localization improvements get lost.

So I think one chance of changing the development process to a comunity built solution is, that such issues can be adressed, because the comunity has some different priorities.

A solution in this direction might be the idea, to create a svn repository, where changes to core programs must be committed.
A rule could be, that only programs using that mechanism, will be accepted in the official release.

I think this would encourage people, to keep such changes consistent.
It also would simplify development, because developers would have 1 reliable, "official" base they could use for their patches.

Mark

[cb88]

a slackware build would be quick and dirty

a T2 build would pay off in the long run in development hours

I have nothing against ttuuxxx helping (note helping not leading... this project has never had that kind of leadership nor has it ever needed it) but i guess we all just need to chill out a lot

in short what i mean is that say 2.14R it had significant development go into it but was mostly handled by two people there was no real leadership other than what ever they thought was good.... that shows that puppy development can go on there is nothing stoping us from having mutiple bases as far as that goes say gentoo T2 and slackware it just isn't THAT hard who knows maybe ttuuxxx will even do a debian base (please not ubuntu ) since lost of people happen to like it a lot

barry seems to have intended to make it easier to build the base of puppy note the T2topet scripts and the such he didn't do this so we would could just build a base/core puppy IMO he did it so we can all really get our hands in and get them dirty at the core if we want and build our own puppy from scratch puppylinux is a learning project just as much as an easy to use project

note i strongly say learning and ttuuxxx you should listen up there I understand about learning I've done a lot since i first came here and plan to do more as I'm sure you do as well but I don't offer services that i don't have right now to give I'm sure that you can eventually figure out just about everything in building a custom puppy but you don't know all of that yet so what is so wrong with being contented as a fellow puppy developer and not the coordinator of an unwilling band of semi rouge developers who are gonna pretty much do thier own thing anyway regardless?

with repect to the last part you can see that in just about every CE project we have done ... we have a release manager yes but every body else does just about whatever they feel like not what the project leader feels

contrary to what you say that we need a coordinator i say that we need none since people get the most done when they are doing what they like to do and not we bossed around which is how things have been ran so far with barry

with barry gone do we need a core developer?

personally i think that we won't need one for quite some time puppy 2.14 is pretty up todate and we can probably coast on it for at least a year or more even then updates to gtk etc... have been done before nothing new

so essentially i don't think there is this sense of immediacy at all that some people are feeling

[alcy]

so essentially i don't think there is this sense of immediacy at all that some people are feeling

Exactly.

In fact, how about those wild ideas that you had in mind to change Puppy your way, but somehow never did. Like, (I am not sure on this), Bruce B started a thread sometime back regarding Puppy to be installed inside the RAM or something like that. Maybe people should spend time experimenting whatever way they want, share with the community, let the opinions build and then probably agree upon something.

There's just no need to rush !

[ttuuxxx]

a slackware build would be quick and dirty


personally i think that we won't need one for quite some time puppy 2.14 is pretty up todate and we can probably coast on it for at least a year or more even then updates to gtk etc... have been done before nothing new

so essentially i don't think there is this sense of immediacy at all that some people are feeling

2.14r Lots of people had 1 issue with that, The default menu structure was changed and wouldn't take a new window manager, Like icewm wouldn't work on it. The menu order was out of its default locations. If this was reverted back a lot of users would be happy.

Hey I'm not leading anything on this release also, only offering ideas and packages and maybe a few graphics, you can use them, delete etc.
Debian, Slackware is all good also I'm not hell bent on ubuntu, actually I hate how they break up single packages into 2-6 packages, not including the needed libs.
As for a big brother specs, just a few version numbers, Just enough to be binary compatible would be great. Glibc, qt4 version, kernel etc
ttuuxxx

[bugman]

Does anyone still want a 2.4 kernel? guess that could be T2 also or 1.x based

YES

[see my endless whining about scanner not working with new kernels]

[cb88]

@ttuuxxx one issue? come on there are more than that in regular releases

[HairyWill]

My understanding of the menu system in 2.14R is that it is compliant with freedesktop.org standards whereas the normal puppy menu system is not completely.

Adhering to standards can have performance/development and maintenence costs but is often crucial for interoperability.

edited for spelling

[tronkel]

Ever wondered why Barry has decided to retire at this specific point in time? Has he thought of something that we haven't? Why is he thinking so much in terms of a tiny Puplet (Unipup)?

Could it just be that he thinks that the days of PC's, laptops and even mini-laptops are numbered anyway?

Could be that he is really thinking about stuff like a Puplet that would offer something similar to Google's latest Android platform?


So instead of putting all this time and energy into "conventional" Puppy versions, should we really be not be thinking about a completely new Puppy platform that would provide an OS/software stack for one or more reference mobile devices.

These latest generation devices are not going to displace the more conventional platforms immediately, but I'll bet that in say 2-3 years a large percentage of internet communications that are currently done using PCs and laptops will be done on mobile hand-held devices such as Blackberry, Google G1 and iPhone style units. Would be good if Puppy was there too.

So all this arguing about how the current Puppys should be built and by whom, may in time, turn out to be a waste of time and energy. The world of technology just moves on regardless.

[jason1308]

I am reading this as a relative newbie to Linux and puppy and only switched to puppy from Mandriva because of all the reviews about it being good on older machinesand at the time I was looking for something to put on a older laptop.

But now I am worried, because although i am quite content with running it on my desktop machine, I wonder whether I should switch back to something more mainstream, that although more bloated than puppy, has a direction, and a larger selection of programs from a central repository, and adds missing dependancies automatically.

All you lot seem to do is argue over petty little items like calculators etc or trying to dis-credit each other.

All I can see is a load of different forks and puplets being developed, making the choice for any newbie even harder which package to install.

It would be better to build on what you have, develop a central software repository which also checks automatically for dependencies. Add better hardware detection, enable the number lock to come on at boot, auto mount all the drives, when it switches off, then make sure it actually does, and don't follow Ubuntu (Its Crap, bloated abd buggy).

If you all worked together you could have some fantastic product. This is proven by all the good advice and PET builds in the Forum.

[cb88]

you see i can't get a T2 build up because the only place i have linux access is at home on dialup...

someone could get a debian based system to work with unleashed or even gnomeslack

also pizzasgood is trying to get svn figured out so we can get somewhere with that although im sure he has time constrains similar to my own

I guess some progress is better than none ... while i wouldn't consider an ubuntu base progress. also we haven't consider an Archlinux base it has a ports system but it isn't as mature as T2 but has an excellent package manager already

also slitaz somehow fits most of the puppy software in 30mb how does that work?

[Trobin]

But now I am worried, because although i am quite content with running it on my desktop machine, I wonder whether I should switch back to something more mainstream, that although more bloated than puppy, has a direction, and a larger selection of programs from a central repository, and adds missing dependancies automatically.

I don't know if you need to worry just yet. Puppy may seem to be coming apart at the seams, and may eventually just do that, but I'm sure most would help with any problems you might have regardless of which puplet ypou have.

All you lot seem to do is argue over petty little items like calculators etc or trying to dis-credit each other.

I suppose this infighting is a natural evolutionary step between Barry's retirement and a possible future for a mainstream Puppy.

[cb88]

the calculator battle is a battle settled long ago by barry himself since it is a FACT that they are soo small that even if you remove them there is hardly any difference in the ISO size

i think galculator is a little bigger and it is better to have 3 good tiny calculators than one slightly bigger one

fact us galculator itself being one of the larger calculators is only a 64kb pet... there are better places to be triming puppy than the calculator selection

[raffy]

Disagreement is not really disaster. What is important is that ideas are kept coming.

I guess if someone figures how to start SVN at sourceforge, we will be in new business. Hey, Chase (cb88), have you added MU to the puppylinux site at sourceforge?

Trimming the number of essential applications for Puppy sits well with keeping the load low while maintaining an uptodate repository. Perhaps we can focus now on how to setup the SVN at sourceforge? (The SVN is being updated this Sept 30, incidentally.)

[cb88]

a brain dead fish could compile a calculator ... so stuff like that doesn't really increase the load but stuff like ffmpeg or other utulities are what adds to the load not much you can do about that

personally i would ditch geaney since i tend to use mp or nano and lately have gotten the hang of emacs ...die vi die j/k.... but I'm sure there are a lot of people that like geaney or need a program that can display larger fonts.

I really don't see anything we can ditch... also im my experience mplayer is faster than xine on slackware that goes against what you would think but it just is.... so is the media player up for debate?

VLC ... the best ...requires more libs
Mplayer very good .... resource usage questionable?
xine slow ...crashes a lot or locks up ...xine ui works better but there are still issues

what were barry's reasons for sticking with xine? interface consistency?

on sourceforge we have the following users: http://sourceforge.net/project/admin/us ... _id=235995

[alienjeff]

also we haven't consider an Archlinux base it has a ports system but it isn't as mature as T2 but has an excellent package manager already

Arch Linux's pacman is outstanding! As for maturity, stop fretting. Arch is over six years old and was inspired by Crux Linux which is over eight years old. T2 started around 2002 as a fork of Rock Linux. So they all kind of grew up together.

also slitaz ...

Speaking of Slitaz, though a very young distro, its movers and shakers have done an excellent job with documentation. Check it out.

edited 29Sept08@1747GMT - Arch/Crux relation clarified - jwb