Puppy's future is still not secure

[davesurrey]

A few weeks ago I realised I was not alone in worrying about the future of Puppy Linux. In a thread started by another about Puppy's diminishing place in Distrowatch's ratings (where today we are even lower) I used the opportunity to say why I thought Puppy was spiralling down. To sum it up I believed it had no clear direction, it's application repository was poorly stocked and finding other apps was not easy, helpful information was scattered around and it was obviously leaderless.

The reaction was fairly predictable. We got "We are Puppy so we are invincible" type verbage and others seemed to equate our chaotic structure with freedom. However I was surprised by the number of folk who privately as well as publically seemed to agree we were in trouble.

Then came a thread from Barry who, IMO, quite correctly stated his concern that there had been no "official" release for a while and by implication there didn't seem to be one in the offing. The reaction was obvious. Folk seemed happy that Barry was back and seemed to support whatever he felt was right. That, to me, seemed to admit that we were leaderless and that folk really had concerns, whether they wanted to admit it or not, about Puppy's future.

There were a few who seemed to believe that the suggestion of a 4.3.2 was a bit of a band-aid that wouldn't tackle the fundamental issues. I admit to being one of those.

Now after all the feedback that Barry requested where the overwhelming majority seemed to be in favour of a 4.3.2 it seems the "solution" is 3 new Puppys to be developed in the next 5 weeks. Interestingly the release would be in sync with the release of Ubuntu 10.04. Distrowatch following perhaps? And that also seems to sound the death-knell of 4.3.2 and 4.4.

On Barry's blog one brave soul wrote today:

And, maybe you precipitate a little with this announcement, because, in respect to the Puppy fans, you should have explained why 4.3.2 and 4.4 will not be released, after all the fussy about?

I fear that Puppy is steadily losing focus on what should be important: stability, usability and good applications that really "do the work".

So, my "personal" opinion is this strategy now is a huge mistake!
But, wish good luck!

I totally agree with him.

Nothing in this plan will resolve the basic problems that Puppy faces, as I have mentioned above. It is yet another "plan" with little chance of success. Barry is clearly a very talented developer but we need leadership and structure as well. And, seemingly pulling the rug under the feet of those that were trying hard to make a new Puppy, does not bode well for a future where others are sought to set the future direction.

I expect utter condemnation for daring to tell the hard truth as I see it but I believe that Puppy has made a wonderful contribution to free software and Linux and dearly would like that success story to continue.

Dave

[8-bit]

I tend to agree with you!
Not to knock anyone's Puppy version, but it seems to me that the user of Puppy has to rely on bug fixes found here for the Puppy he is running and those very fixes and updates get ignored in creation of a new "official release".

[MinHundHettePerro]

As mostly, Dave, a very spot-on observation and (sadly), I believe, an insightful interpretation of the state of Puppy OS.

/MHHP

[bugman]

don't care much about distrowatch, but agree totally with everything else in your post

i went with 1.07 for quite a while before upgrading [4.12?], may go a while before doing so again...

[FlyingRedGoat]

Like all things puppy rises and falls in popularity. Just because it falls, does not mean it's going away. (I hope not, I really like it).

That being said, I'm sad to see 4.4 go away. I was looking forward to it.

One of the reasons cited for puppy's failing popularity was a small selection of software in the repository. Sadly the source did not seem to be aware of the ample sources in the forum. Perhaps a mechanism to select the better and more stable pets from the forum, and add them to the repository would help.

Canis vincit omnia.

K

[davesurrey]

FlyingRedGoat wrote

One of the reasons cited for puppy's failing popularity was a small selection of software in the repository. Sadly the source did not seem to be aware of the ample sources in the forum.

Yes I am aware that there are many apps, as well as pointers to apps, in this forum. The fact they are needed at all seems to indicate how poorly the repositories are stocked.
But my point was that it's very difficult for ordinary users to find these apps and when they do to be sure they are relevant for a particular version of Puppy. I think you agree as you say

Perhaps a mechanism to select the better and more stable pets from the forum, and add them to the repository would help.

which I fully agree with.
Cheers
Dave

[WhoDo]

Barry is clearly a very talented developer but we need leadership and structure as well. And, seemingly pulling the rug under the feet of those that were trying hard to make a new Puppy, does not bode well for a future where others are sought to set the future direction.

As one of the very few to have actually produced an "official" Puppy release, I feel that I am entitled to comment here.

1. Perhaps many will now realise just how much effort it takes to produce a polished "official" release and give whoever is leading the effort some long overdue support.

2. Leading the development effort on a release is NOT leading the community on Puppy direction. Barry does that by dint of his status as Creator. No-one else has seemingly earned enough "stripes" to fill those alarmingly large shoes.

3. Efforts to "organise" the community, like efforts to "organise" the development effort, have failed dismally before for many reasons; probably as many as there are regulars here.

I believe Barry is genuinely concerned that Puppy will die on the vine if we don't show some ongoing development effort to the Linux community at large ... Perception is Reality. That is why I welcomed the suggestions of a 4.3.2 release to give technosaurus time to progress with 4.4 (CE or not CE, that is the question Horatio). If that protracted release is not ready, then Barry is right to jump to 5.0 ... the gestation there has been even longer!

I agree with the triple-pronged release, but I think Barry picked the wrong three "prongs" - they should be Upup Lucid, Dpup Lenny and Spup or Tpup (Puppy Puppy). Quirky fits in another niche entirely. Releasing versions of Puppy that were able to access solid repositories at Ubuntu, Debian and Slackware respectively would IMHO give Puppy, and the Woof build system, a far bigger "push" than any single distro release could possibly have.

The single biggest complaint about Puppy has been the lack of a solid application repository. Unless someone is able to offer hosting for all available dotpets, that's unlikely to happen for original Puppy. Caneri tried and ran into insurmountable problems. That's why the Trident approach is the best idea - piggy back off existing repos while retaining the unique nature of Puppy in the underlying structure.

I expect utter condemnation for daring to tell the hard truth as I see it but I believe that Puppy has made a wonderful contribution to free software and Linux and dearly would like that success story to continue.

I expect nothing less these days either. I tried following my sig and "doing" something rather than just "talking" about it and look where that got me. I'm happy with what I delivered but that isn't enough when opinions are as prevalent as arseholes.

Do I think Puppy is dead or dying? Nope. As long as doers like ttuuxxx, Gposil, runtt21 and others are around, there will always be a Puppy of sorts; just not the Puppy we've all come to know and love. One beholden to no other distribution for its "roots" and proudly "different" in how it gets things done.

Dave, your plea for structure and organisation will never be heard while ever Puppy remains a community-driven distro. Look at the real success stories ... Ubuntu, Fedora, OpenSUSE, Mandriva, etc. They survive because they are driven by commercial interests, one way or another. Slackware, Debian and Arch are niche distributions that are unlikely to represent the wishes, hopes and desires of a majority of Linux faithful. Clones may come and go, but the big ones keep going when the clones have gone away. Only Mint and PCLinuxOS are showing any real signs of longevity ... just. Guess what? Both are single developer editions like Puppy is under Barry.

Puppy survives on being different to other distributions and yet remaining familiar to those refugees who are moving away from that other commercial OS and looking for a Linus blanket in the new, Free world. That's why Quirky is interesting but not yet ready for the mainstream IMO.

Until Barry bites the bullet and decides to take the reins again OR anoints his most capable successor, this question will continue to tempt the Chicken Littles to proclaim the Puppy sky is falling. Heaven forbid it should become a self-fulfilling prophecy! If it takes a little parochial declaration from time to time - "We are Puppy. Resistance is futile" - then what harm will that do?

[Bert]

I too have to agree with Dave.

(Dave, is it a coincidence that we met in the Slitaz forums, where several other puppies seem to be active..? )

But I don't think anyone is to blame for the current situation. Puppy has always been experimental, innovating. However, time has moved on. What I found exciting 5 years ago, has become more mainstream nowadays.

The unexpected success of net-books has led all major distros to review their strategy and develop versions for low-power machines. What chances does Puppy 5.0 have against eg Mint LXDE or Lubuntu, both fast booting and fast running completely usable distros with very low ram and processor requirements. Those are just two of the big names. ( There's also Crunchbang, Slitaz, Jolicloud, etc..as contenders)

What exactly will Puppy 5.0 add to Debian or Ubuntu? Surely not stability, repos or localization.

This discussion reminds me of the day tombh left Puppy. His blog explaining why, can still be found here. I now think he was right.

But I still love Puppy and its community! It's just that I now see it for what it is: a laboratory for new ideas, a breeding place for Linuxes -to-come. Not something suitable for my real computing needs.

[davesurrey]

Hi Bert,
Yes I do inhabit the SliTaz forum along with the TinyCore, Crunchbang, Arch and some other forums, although Puppy is still my main interest.

I think those distros are good examples of how a one-man led (or a very small group led) distro can have structure and a well stocked repository. Those that say it isn't possible should get out more.

For the sake of clarity, by structure I mean something like a single website which acts as a stopping off point to a forum, a repository and some sort of info centre/wiki. Nothing more radical than that. Non of those above are perfect but Puppy's is the poorest IMO.

Plus a general direction where the distro is heading. Then any developer in the community can focus their efforts on bug chasing, adding apps to the repo, developing new scripts/apps, making improvements to the base even or whatever is needed in a focused way. If they don't like the general direction they can leave and go somewhere else of course. But if they do then they can make a significant contribution knowing their talents are not spread so thinly.

It's just that I now see it for what it is: a laboratory for new ideas, a breeding place for Linuxes -to-come.

This may well be what Puppy has/will become. In which case so be it... but there's one further point even of this is the case.

These distros, with the exception of Crunchbang, are not based on another distro's foundation. The core developers, like Barry, have produced the core and part of the success or otherwise of the distro is based on them keeping abreast of kernel and other developments in tjhe Linux world and changing the core appropriately. As I have said before all the puplets so far have been based on Barry's work. I can't see what will come after that. But staying with the same core is not an option for the future.

Dave

[racepres]

This discussion reminds me of the day tombh left Puppy. His blog explaining why, can still be found here.

While I agree wholeheartedly with what Nelson Kondo says on that very page!!
RP
Take it from a guy who runs an outfit that is Not the Biggest, nor Most Popular. We are the Funnest!!! Period! If our Family is not the Biggest... OH well!

[jemimah]

I think the drama has not played out yet and you are reacting slightly too early.

Barry has contacted me and suggested that Puppeee could be an official Puppy release. I am quite excited about this prospect.

Apparently this privilege might come with access to the ibiblio repository. Which means I might be able to add new official packages, or create an additional repository. I don't know the details yet so I may be speaking too early.

I definitely haven't earned enough stripes to fill the "alarmingly big shoes," but I still have an interest in working on Puppy 4 and also a huge interest in working on, coordinating, or providing a "direction" (if that is truly necessary) for Puppy for Netbooks if anyone besides me would like to work on it.

[nubc]

Puppy's main weakness is that it isn't easily extensible. Nobody wants to whine and beg for help, just to install an app. Needs a centralized, organized, reliable repository.

[Béèm]

Nice post Dave with a lot of truth in it.
Already some 1 or maybe 2 years ago, I made several post stating that puppy lacks a product manager.
I was told then, that managing by anarchy was the good way to go.

I am not happy to see some scenario's I was afraid of then tend to be coming to reality, but on the other hand, I think Barry will take things in hand.

He seems to be very sensitive on the distrowatch ranking which will drive him to take some kind of action.

[RetroTechGuy]

There were a few who seemed to believe that the suggestion of a 4.3.2 was a bit of a band-aid that wouldn't tackle the fundamental issues. I admit to being one of those.

I guess that I thought 4.3.2 was a test to produce a version of 4.3.1 via a different source route -- as was indicated here http://bkhome.org/blog/?viewDetailed=01391:

I must reiterate that I did not create this with the intention that it will be the next official 4.3.2 release. I'm mostly just wanting to know that Woof builds a sane Puppy 4.x.

It makes sense to me, that when migrating code forward, that first you make sure that your new code produces a valid/correct output (i.e. nominally, if not identically, the same as the original method).

From there, your newly created tools permit you to move into a new version, with reasonable confidence that you didn't break something in the process.

(in years past, I was coding a stat. mech. program, which ended up performing operations on an order 9 or 10 tensor -- I could reduce this code to perform a calculation on a known lower rank tensor, to confirm it was working correctly -- so the new code went through an extensive testing phase, to reproduce results from the simpler code)

[RetroTechGuy]

Apparently this privilege might come with access to the ibiblio repository. Which means I might be able to add new official packages, or create an additional repository. I don't know the details yet so I may be speaking too early.

BTW, while poking around in the ibiblio Linux collection, I noticed a folder named ttuuxxx... Does ttuuxxx already have a distro point (it is currently empty)?

ftp://distro.ibiblio.org/pub/linux/dist ... ns/ttuuxxx

[davesurrey]

RetroTechGuy
Your quote from Barry's blog is dated 8 Feb when his idea was just as you described ie an experimental unofficial pup.

However if you followed his blog daily you'd have seen one on 19 March where he has cleary changed his mind and 4.3.2 is a potential official pup

So, I know that I am supposed to be retired from the frontline of releasing official puppies, but considering all of the above, should I come back in and bring out 4.3.2?

Beem:
Thanks for your support.

I am not happy to see some scenario's I was afraid of then tend to be coming to reality,

Yes let's hope things improve as, like you, it would give me no pleasure seeing such professies come true.

jemimah:
That's good news and well deserved due to all your efforts, even if Puppeee is rather a specialised distro for eee's and other netbooks.

I think the drama has not played out yet and you are reacting slightly too early

You may be right and I hope you are but I prefer reacting too early than too late and given some of the apathy around here IMO I think the early option is best.

Cheers
Dave

[FlyingRedGoat]

As one of the very few to have actually produced an "official" Puppy release, I feel that I am entitled to comment here.

1. Perhaps many will now realise just how much effort it takes to produce a polished "official" release and give whoever is leading the effort some long overdue support.

Requires technical expertise to work "magic" on linux, social skills to stop the developers from turning on each other, and lots of free time to handle the constant request from everybody. I have seen how much effort it takes from watching you. I think you did great. Sadly there will always be backseat drivers in the development world. They call out any errors (whether they exist or not). They did not make the choice, so they feel safe from criticism.

You are right. We need to find a way to support our leaders. Even Barry mentions the jibes on his log.

Until Barry bites the bullet and decides to take the reins again OR anoints his most capable successor, this question will continue to tempt the Chicken Littles to proclaim the Puppy sky is falling. Heaven forbid it should become a self-fulfilling prophecy! If it takes a little parochial declaration from time to time - "We are Puppy. Resistance is futile" - then what harm will that do?

This Puppy ultimate leader position is a big job for just one person. Especially if he / she has to hold down a job. Maybe a small close knit group could lead? Or maybe we can figure out how to raise money so the leader wouldn't have to hold down a job?

K

[linuxsansdisquedur]

puppy (unique nomoneymaker & totally amateur distro which run on old hardware faster as all other ones do on last pc ever) is down (#10!) on distrowatch (just after all commercial linux distro) means it gonna go to death ?
no way ! even if it's libertarain way of live (don't need :newer pc, blings apps, terabyte something, unique interated os ) is not uptodate way of thinking.
No enough apps with puppy ? just seach the forum (or ask for it in the forum) !
perhaps make sort of updated PSI ( http://www.murga-linux.com/puppy/viewtopic.php?t=10960 ) for any puplet/official to find it more simply...
linuxsansdisquedur write

yeah running on puppy is NOT running any other commercial os as 'dows does...
it s a bit ascetic way of life, keeping light, keeping old'ware user,
it's better elegant stuff! even if it means not to be able to running right now any new aps uptodate as everybody as dows user need to do....
no matter it's more human relationship with puppy community
no dictatorship of the money and uniform thinking of commercial web...
resistance ! we can find an other way to life in...
i'm still running pizzapup (with real pupy 2.14 inside) I can do every thing i want (and my wife too...) even if i can't run flash 10 website-needing-the-last-one : i find others with underneeding , even if my gaim don't run msn protocol right now: gotta run meebo...no repo with all apps can run with my os gotta try something or lose my mind or forget it ... anything simple for dows user is a workabout...
so what, gotta install windows 7 ? no no no gona try and try to run light and fun os puppy!

[jemimah]

This Puppy ultimate leader position is a big job for just one person. Especially if he / she has to hold down a job. Maybe a small close knit group could lead?

Agreed. The trick is not to create a "good old boys club" (even if it's coed, it's easy to make it appear more exclusive than it is). One of the best things about Puppy is the open community.

I think a central community "contrib" repo would be really useful (especially if you could get to it from the package manager), and testing the packages wouldn't be that time consuming. There's probably a fairly long list of people that could be entrusted with testing and approving packages. It seems even with Puppy 5 much of the software needs to be repackaged in order to work correctly with minimal dependencies. However, something like this would need official sanction from Barry to take off.

[davesurrey]

I think a central community "contrib" repo would be really useful

I think that's an excellent idea jemimah.
Dave