woof-CE needs you

[darkcity]

Had a try at it today, Slackware 14.2 failed with loads of packages not found, Debian only list Wheezy, Should also list Jessie and Stretch, since Stretch is now Alpha 7 State. Barry used try and release the Next puppy version based a major distro a week before they actually released it. Well we are truly behind the ball on updating this. Plus a lot of other distro's are no longer listed, like Wary, T2, Quirky . No retro kernels like 2.6.32.59 Wary's last kernel. Does anyone have a working link to Bary's patched 2.6.32.59 kernel, I've found a few links but they are all dead.
Not trying to criticized this project, I'm just a bit frustrated.
ttuuxxx

Which links are dead? Ally has archived a lot of stuff at archive.org

Kernels are here:
https://archive.org/details/Puppy_Linux_Kernels

Quirky/Wary stuff here:
https://archive.org/details/Russoodless ... quirkywary

Files from meownplanet have been copied:
https://archive.org/search.php?query=su ... dle%27s%22

Puppy Linux category:
https://archive.org/search.php?query=su ... uppy+Linux

Note: .Pets are listed under 'ADVANCED AUDIO CODING'

[foxpup]

@jlst and @ttuuxxx

Much of it I do not understand, but this conversation is really exciting!

I have an old machine (Palomino/geforce256) and I hop from tahr to wary and back. Wary supports my gcard well (nvidiadriver, legacy71xx, compiles and xorg is right) and is really swift. But lately it needs gtk3 for latest Firefox and newer glibc/gcc for latest libreoffice. I don't need them, but my wife does for her schoolwork. Tahr has the right libs, but I have to run it on vesa (still better than nouveau!), so it is less snappy.
The best of both worlds, old and new, would really hit it indeed!

I think a computer is like a pair of shoes:
old shoes walk better than new ones.
Once one gets used to a machine and gets to know it, one wants to hold on to it.

btw, technosaurus thinks about pupBSD
http://www.murga-linux.com/puppy/viewtopic.php?t=107542

[ally]

http://archive.org/download/Puppy_Linux ... s-1-w5.pet

http://archive.org/download/Puppy_Linux ... ufs-w5.pet

http://archive.org/details/Puppy_Linux_Wary5_Repository

http://archive.org/details/Puppy_Linux_Racy

http://archive.org/details/puppylinux?s ... []=ttuuxxx

[jlst]

.

[jlst]

.

[jlst]

.

[Robin2]

I have been using TahrPup exclusively for about 18 months having switched from Xubuntu. I particularly like the fact that it can be installed by copying 4 files and editing syslinux.cfg and also the fact that all the changes go in a save file so the whole thing is easy to backup. I also like the idea that TahrPup could use all the programs available to Ubuntu so that I would be very unlikely to need something that i could not get.

One of the things that particularly irritated me about Ubuntu was all of the updates that appeared daily or weekly. It was never clear if an update might break some existing thing and there was no easy way to undo it. Whereas with TahrPup I can just go back to a previous savefile. On the other hand if I did not update as I went along it would be a real problem upgrading to the next version when it came along.

Originally I had assumed TahrPup had the additional advantage that I could transfer all my settings and all the programs I had installed to a different laptop just by copying the save file - but that proved not to work if the two laptops have different hardware (such as video drivers). Also (as with Xubuntu) if I upgrade to a newer version of Puppy I probably also need to upgrade all the programs - rather than copy over the old savefile.

In the last couple of months I got an Acer cloudbook and then a HP15 laptop. I had great trouble getting Puppy to work on the Acer to the point where I gave up. The HP15 nearly works fine with TahrPup, except that the Puppy driver for the internal Broadcom wifi does not work properly and I have to use a USB wifi dongle.

I thought the emergence of XenialPup might solve my problems but it is clearly still a work in progress.

In contrast Xubuntu 16.04 seems to work perfectly on the Acer and Linux Mint Mate works perfectly on the HP15 - including using the internal Broadcom wifi.

This got me thinking that maybe keeping up with the new versions of Ubuntu may be too much for the small team of people developing Puppy and that got me wondering if it may be time to review the rationale for PuppyLinux?

• It seems to me that it is very useful to be able to launch Puppy from a usb stick for recovery purposes - but even that may be under threat from UEFI and secure-boot etc.
  
  Undoubtedly it is also very useful to have a version of Linux that runs well on older hardware - but that does not require anything more recent than TahrPup.
  
  But for any PC that needs Ubuntu 16.04 there is no real value in making Linux smaller.
  
  I do appreciate that it is fun to find out if the Puppy treatment an be applied to Ubuntu 16.04 - but I confess I don't see any practical use for it beyond that.

I have learned a great deal about Linux while using TahrPup that I would defintely not have learned if I had continued with Xubuntu. the irony is that that knowledge also makes it easier to use "regular" Linux.

...R

[jlst]

.

[learnhow2code]

with NO disrespect intended to robin intended, the main thing missing from puppy isnt developers, but users that understand the road to quality is someone trying the distro out and reporting problems (so they can be addressed.) strictly for dev morale, the best time to report these things is *before* youve given up (a couple tries, a couple reports, a couple more users trying things would go quite a ways.)

in the sig i use these days, theres a hint as to partly why i think such reporting is more difficult or less instinctive than it used to be-- a (partial) disconnect between devs and users.

this leads to user frustration and dev frustration, which furthers the disconnect... but we have enough developers (at least almost) and we DONT have enough communication between users and devs. (imo.)

this, not forum "politics" is imo the #1 issue for puppy right now. it cant be put on devs or on users, but everyone should take a good look at it. above all, its fixable.

and the fix is more communication between devs and users. but as i started out saying-- that still requires users to try out puppy, and report issues, so they can be fixed. it doesnt work the other way-- unreported things dont get fixed so they can be reported, eh? i know there IS a shortage going on there. for several reasons. im not trying to be a tease-- i will make the suggestion, and im happy to elaborate. but really, this is almost certainly what the matter is. and its utterly fixable.

[jlst]

.

[darkcity]

Developers will develop what interests not what users want. Also, developers will have limited time to spend volunteering on a project. So I don't think the lack of communication is the key issue. If the users' wants roughly coincide with developers' interests then there can be some mutual benefit.

This is how it works unless a project is setup as a foundation with clear aims. A foundation can also help to cover administration and some develop cost, with ongoing or one time fundraising.

For example, Software in the Public Interest helps Debian. The Document Foundation supports LibreOffice.

https://en.wikipedia.org/wiki/Software_ ... c_Interest
https://en.wikipedia.org/wiki/The_Document_Foundation

Another model is companies developing software on the back of getting paid for other services. Like Canonical with Ubuntu, and Red Hat with Red Hat.

[learnhow2code]

Developers will develop what interests not what users want.

this is very true in more cases than not-- i was actually just explaining to someone in irc that woof-ce is not only the source of the official pups, its something ANYONE can use to make a pup their own way.

Also, developers will have limited time to spend volunteering on a project. So I don't think the lack of communication is the key issue.

you dont think a decreased level of reporting bugs is responsible for a decreased level of bug fixes? because thats all i was saying.

If the users' wants roughly coincide with developers' interests then there can be some mutual benefit.

im pretty sure there is some overlap-- not in all cases (you cant judge puppy by my own hybrid that isnt even made using woof) but i would at least like to know what the experience is like for users.

This is how it works unless a project is setup as a foundation with clear aims. A foundation can also help to cover administration and some develop cost, with ongoing or one time fundraising.

im just going based on how it was run by barry for many years. the foundation never happened, and barry cared until he gave up.

then several people stepped up and took over development. are you in fact suggesting that those developers dont care what the user experience is? because i wont put those words in your mouth-- i can refute them without accusing you of saying them.

also, its a valid opinion if you do think its the case-- but i still doubt it.

For example, Software in the Public Interest helps Debian. The Document Foundation supports LibreOffice.

as someone who has raised funds for spi, i dont think they are helping anymore. the debian mailing lists make it pretty clear theyre LESS interested in what users have to say about systemd than the devs are.

in this case i think a foundation could have even the opposite effect you say-- BUT I THINK IT DEPENDS, case-by-case basis, fwiw.

Another model is companies developing software on the back of getting paid for other services. Like Canonical with Ubuntu, and Red Hat with Red Hat.

dont get me started... actually i have no intention of talking about that in this thread (beyond this mention of it, that is.)

[jlst]

.

[learnhow2code]

you may wonder why Red Hat is the main contributor to the linux kernel, etc. why ubuntu is big in the linux world and has eclipsed debian.

"measuring programming progress by lines of code is like measuring aircraft building progress by weight." - bill gates

[darkcity]

@learnhow2code

Thanks for the information about Debian and Software in the Public Interest.

Yes, most developers care that the user experience is good, and there software has less bugs.

Less bugs reported -> to less bugs fixed is true. Is there any evidence of less bugs reported?

@jlst

at least your drug helps others

[learnhow2code]

Yes, most developers care that the user experience is good, and there software has less bugs.

ok, so i think we agree on that part.

Less bugs reported -> to less bugs fixed is true. Is there any evidence of less bugs reported?

thats a really great question-- especially since i said i think its the "#1 issue" with puppy these days, eh?

my very frank answer is that im NOT sure its an issue. (in which case, puppy does not have *any* serious problems i can point to-- at least not more serious than that not-serious one...) the reason i think its an issue is because i hear from time to time from users that it is. and (this is said in the greatest sincerity) they could actually be mistaken. though i think its more likely they have a point, vis a vis the top of this post, where i quote you. how relevant all that is, really and truly is up for debate. when cornered by an excellent question such as yours i in fact said "im not sure." and thats the truth!

[jlst]

.

[Robin2]

with NO disrespect intended to robin intended, the main thing missing from puppy isnt developers, but users that understand the road to quality is someone trying the distro out and reporting problems (so they can be addressed.)

I do understand that but to be honest my level of knowledge is too far below the Puppy experts that I can't keep up with the discussions. And I guess, because of my low level of expertise, I don't have any great interest in being a Linux developer - whereas I am interested in Arduino programming.

And I think I was trying to say that I don't really agree with the direction of Puppy development. I don't see the need to transform Ubuntu Xenial into a Puppy. If that work was stopped some experts would have time for other stuff.

Maybe it also follows from that that WoofCE is no longer needed since a very serviceable TahrPup already exists?

...R

[learnhow2code]

I do understand that but to be honest my level of knowledge is too far below the Puppy experts that I can't keep up with the discussions.

fair point-- imo you should be able to report something and then answer some questions, but its possible the questions will be too difficult or cumbersome (even for me) at which point i can only sympathize-- youve wasted your time in that case (hypothetically.)

And I guess, because of my low level of expertise, I don't have any great interest in being a Linux developer - whereas I am interested in Arduino programming.

rest assured i wasnt implying that you should necessarily become a dev to fix it yourself.

I think I was trying to say that I don't really agree with the direction of Puppy development. I don't see the need to transform Ubuntu Xenial into a Puppy. If that work was stopped some experts would have time for other stuff.

would it matter if puppy had more or less always worked that way?

puppy is by barry, though it has always grabbed kernels or packages from other sources like vector, slack, ubuntu, etc.

Maybe it also follows from that that WoofCE is no longer needed since a very serviceable TahrPup already exists?

im glad you think thats possible since my own puppy-related efforts work on a similar assumption. nonetheless tahr is built using woofce, and would probably not be better (certainly not more puppy-like) without it.

woofce is a fork of what barry used (what barry authored) to create puppy-- some of the efforts are around making it easier to maintain, which isnt in itself a terrible idea at all.

its possible that quite a few of the problems that people perceive with development are due to not factoring in the role that woofce plays-- woofce is the thing that makes puppy. it would be interesting to know if barry ever produced a new pup without tweaks to his woof scripts.

devs have to either tweak the scripts or tweak what comes out of them-- those are the two steps in making any official pup.

[666philb]

Maybe it also follows from that that WoofCE is no longer needed since a very serviceable TahrPup already exists?

...R

without woofce there wouldn't have been a tahrpup