Average forum topic

[dimkr]

This attempt to find the smallest list of sufficient and necessary conditions to declare a distro an 'official' Puppy release or force preinstalled Samba into this list is futile. Some things in life don't need to be defined, and Puppy is a do-ocracy: it's the thing built by the people who build it. Official or not, there won't be any Puppy Linux if this forum becomes a pile of useless posts and dies.

[Grey]

And at this time in the commentary booth of the stadium:
We are conducting our today's report from the Forum Championship on... combat SambA sports Sambo.
So, @dimkr, with the help of social engineering methods studied at work, managed to lure @Clarity into a separate topic on the edge of the forum stadium. Excellent technique, the first four posts consisting of one word, dimkr is at the peak of his athletic form.

But Clarity has a secret weapon and it's natural stubbornness. A good attempt to throw the opponent through the thigh on the wrestling mat with the help of several powerful posts with a lot of arguments.

Oh, the end of the first half. I'll see you in the second half when the teams rest. Where is the buffet and pies with tea?!

[Clarity]

@dimkr I dont buy that argument.

My post supercedes your current argument presented, as my posts lean more to "What IS a Official PUP?" Are you now personally re-defining what a PUP is? Or are you diverting to something else?

I do understand, clearly, that WoofCE is a generator, BTW.

And DONT take this personal. It is NOT my intent!!!

[williwaw]

... and do other things I don't want to.

This is the developer FREEDOM I have spoken to often in my past. I strongly agree with this premise.

EVERY forum distro developer who presents distros to this forum, does what they feel others will find usefulness from their work........

Since I am NOT a generator of code, my view from this discussion, that people like me are 'unworthy' ........

Why not do a remaster instead?

[wiak]

Why not do a remaster instead?

Yes, that would be my overall recommendation, aside from the documentation side of things, too. Remasters are also a long term aspect of the work done by forum members and a useful one. In fact, I feel this forum should perhaps encourage and better support remastering. Many distros here include some remastering functionality, but it would be good to see a more unified small set of tools that includes final iso production. Even better might be one forum remaster utility to rule them all that included functions that catered for all current major distros on here. Some use different iso-making under-the-hood utilities and of course different initrd files (including alternatives with overlayfs and others with aufs for layering) and sfs composition. Nevertheless, the general idea of creating a remaster is pretty similar for all. Maybe one version of same overall remaster tool, but in 'flavours' to match each major distro here. All the KL distros, for example, no matter the upstream repo and package manager can be remastered using same approach.

Personally I prefer to produce and offer what might be called a 'core' distro that does function quite well out-of-the-box but is not what most would call 'feature-complete'. That is certainly different from traditional Puppy that went to great lengths to include all generally wanted functionality albeit in slim and sometimes less than fully functional versions of components and apps. But I prefer a somewhat complete (and thus immediately usable) core distro, that relies on package management of some sort (including sfs addons and other portable forms) to add optional additional features. But what is then missing is indeed simple, but complete remastering utility app or apps. The idea should be that once a user has created their own topped-up variant of the distro, the remaster utility app should simply need run to complete the whole iso rebuild, which makes the result then easy also to publish for others to use.

In the KL-world we do have a relatively simple plugin builder approach, which even relatively novice coders could use to rebuild or make a root filesystem to their own specification and needs. However, that isn't quite so easy as simply adding new packages/configs and then remastering. In either case a remaster utility is still required to create the actual iso; most of us who produce such isos do use relatively simple one line commands to do so, but some distro components (such as mkisofs/xorriso and so on) do need to be present in the underlying host builder system. Okay, I'm not offering to build the remaster utility I'm describing - maybe one day, but maybe not since I have other priority additions I already don't have time to do. But would be nice to see the forum more formally supporting remaster creation, rather than the painful idea that only certain types of build have worth or official value, which does cut out non-coders from active valued contribution. It doesn't need a person who can code really to remaster a distro into a better or alternative shape and form. Of course, the more remasters there are, the better we need to find a way of documenting what has been created, and that includes trying to keep all distros relevant and up-to-date or usable for as long as possible - it is very frustrating to download and try a distro only to find it is well past its sell by date!

[williwaw]

In either case a remaster utility is still required to create the actual iso

would an iso be required if the remaster were to utilize this download and run solution?

[Clarity]

No. ... per se. I assume you refer to using QEMU or Ventoy.

These 2 will directly boot either an ISO file or an IMG file. While SG2D and ISObooter are geared for booting ISO files directly. Thus, one merely downloads and boots the file(s) to desktop with no user intervention required to have the expected normal frugal operations we have with forum distros. Those products are time savers for users.

Hope this is helpful.

[williwaw]

I assume you refer to using QEMU or Ventoy.

I do not know enough about either to say.

I suppose your bootloader of choice could be packaged in an .img file along with your networking favorites? Maybe something media center oriented?
Something that could utilize Easydd for simple deployment and minimal user intervention?..........

..........With your own project, you would be bringing your suggestions into the community rather than suggesting forum members look outside the community for solutions

[Clarity]

Hi @williwaw

You ask:

I do not know enough about either to say.

I suppose your bootloader of choice could be packaged in an .img file along with your networking favorites? Maybe something media center oriented?
Something that could utilize Easydd for simple deployment and minimal user intervention?...

I think this may address all 3 of your questions.

The item mentioned; namely QEMU, SG2D, ISObooter, and Ventoy are merely tools to make life simple for members. They are positioned to reduce the effort of the members who download a modern forum distro to get to desktop with little to no effort. In the case of developers they would use such to quickly check their development's behavior using, say, QEMU.

These are only designed to boot modern ISO-IMG files for normal frugal operations. So you can expect that using FATDOG, DebianDOG family, KL family, and WoofCE family, using the instructions found on the forum, have a USB that is created once which will serve to contain all your future ISO files and allow you to select them when you boot the USB.

All of this intends to make life simple for users with one the USBs or QEMU subsystem. These do NOT place demands on forum developers for their production of ISO files or IMG files. Thus they do NOT impact those who product remasters or direct development for their distros. In other words, SG2D, ISObooter, or Ventoy are NOT installed within a distro. These are merely external bootable USBs that contain ISO files downloaded by the user.

Thus, the intent is to increase our productivity by reducing efforts to get immediately to desktop while at the same time reducing much of the problems seen in the past in trying to manipulate contents and re-code files in order to boot.

Choose any one and determine if it helps you to accomplish what has been documented thus far. I and other members have been using these approaches since 2019, If it helps, run with it. You can expect help should you run into any problems as members do use these 4 mentioned ISO boot tools.

Hope this helps as I try to contain all your questions in this single post.

[williwaw]

Hi @williwaw

You ask:

I do not know enough about either to say.

I suppose your bootloader of choice could be packaged in an .img file along with your networking favorites? Maybe something media center oriented?
Something that could utilize Easydd for simple deployment and minimal user intervention?...

I think this may address all 3 of your questions.

The item mentioned; namely QEMU, SG2D, ISObooter, and Ventoy are merely tools to make life simple for members. They are positioned to reduce the effort of the members who download a modern forum distro to get to desktop with little to no effort. In the case of developers they would use such to quickly check their development's behavior using, say, QEMU.

These are only designed to boot modern ISO-IMG files for normal frugal operations. So you can expect that using FATDOG, DebianDOG family, KL family, and WoofCE family, using the instructions found on the forum, have a USB that is created once which will serve to contain all your future ISO files and allow you to select them when you boot the USB.

All of this intends to make life simple for users with one the USBs or QEMU subsystem. These do NOT place demands on forum developers for their production of ISO files or IMG files. Thus they do NOT impact those who product remasters or direct development for their distros. In other words, SG2D, ISObooter, or Ventoy are NOT installed within a distro. These are merely external bootable USBs that contain ISO files downloaded by the user.

Thus, the intent is to increase our productivity by reducing efforts to get immediately to desktop while at the same time reducing much of the problems seen in the past in trying to manipulate contents and re-code files in order to boot.

Choose any one and determine if it helps you to accomplish what has been documented thus far. I and other members have been using these approaches since 2019, If it helps, run with it. You can expect help should you run into any problems as members do use these 4 mentioned ISO boot tools.

Hope this helps as I try to contain all your questions in this single post.

I actually didnt have any questions, and will edit my post above to retain the more salient point you wish to ignore.

[dimkr]

I do understand, clearly, that WoofCE is a generator, BTW.

woof-CE is not a generator. It's a tool you feed with configuration files, scripts, packages and other things to build a distro. Take any pair of 'official' releases and you''l see that the difference in woof-CE itself is relatively small compared to other differences. Support for Samba, QEMU and other things you mention is not an integral part of woof-CE: Samba comes from a Samba package and QEMU comes from a QEMU package. No, woof-CE is not a mind-reading generator with a magic 'build an official Puppy' checkbox that adds Samba out of thin air.

Your understanding of what woof-CE is and how woof-CE works is simply wrong. Maybe try to read https://github.com/puppylinux-woof-CE/w ... /README.md?

[Clarity]

... Support for Samba, QEMU and other things you mention is not ...

I have never requested QEMU to be a part of an OOTB WoofCE blend. (at least, not to my knowledge).

I hope everyone understand this.

I have, from time to time, issued inquiry of SAMBA as I have seen it as a feature since "official" @01micko's Slacko, @kirk's Fatdog, TaZoC's Lighthouse64, and many more forum distros over the years since 2019. This was one of the many features of forum distros that give forum distros matching OOTB skills in comparison with all major OSes. It was one primary reason, as much as I can tell, for its inclusion AND being embraced. It, OOTB, participates with all the others on home networds with all home units and PC which come standard with smb protocol features. Further with its inclusion occurring, little to NO user installation issues have happened/happens since OOTB inclusion. The OOTB inclusions that developers have done for 'official' releases have been 'spot-on'.

Hope this clarification is observed.

P.S. I do comment and show examples when testing. I show behaviors in boot, distro operations, and shutdown-reboot processing. You may not like my showing such, but I consider it important for knowing results found.

[wizard]

@geo_c

This is a late answer to your query about what it was like to setup Samba before it was included OOTB.
In short, it was not intutitive, it was frustrating and a PITA to get it working. I've run a Puppy Samba fileserver since Puppy Lucid 5.25. I believe Puppies since Slacko 6.3 have all included 2-3 networking tools that make using Samba shares easier. They are:
-Samba Simple Management (sets up shares on the Puppy computer)
-YASSM (mounts Samba shares on Puppy or any other OS with smb shares)
-Pnethood (also mounts Samba shares on Puppy or any other OS with smb shares)

For the record, setting up a fileserver has been one of the most productive, time saving projects I've ever done. It allows me to access key and important files from any device anywhere in my home. In recent years I have also been running a "media server" that contains all my audio, photos and video files. This has been a huge hit with all other members of the household. These both run on old single core laptops that make small, quiet, inexpensive home "NAS". PC's, phones, and tablets runnng MS Windows, Android and ChromeOS can all seamlessly access, manage and stream their content.

Smb is the common protocol that does these things better than other choices such as FTP or using a web server setup.

If request by @Clarity helped get the above tools included, then I thank him for that.

wizard

[dimkr]

Further with its inclusion occurring, little to NO user installation issues have happened/happens since OOTB inclusion.

If all this endless noise eventually boils down to some Samba installation failure you had in the past, will you reduce the number of Samba-related posts if it's fixed? (This is not an offer)

[rockedge]

If all this endless noise eventually boils down to some Samba installation failure you had in the past, will you reduce the number of Samba-related posts if it's fixed? (This is not an offer)

Around where I live that would be "This is not an offer you can refuse"...if you know what I mean.

@dimkr I was thinking about your buzz kill surrounding making distro's manually and well tough luck. I for one will continue doing it because woof-CE is a real hassle to steer when it's not doing the default recipe's.

There would be no Lucid 5.2.8 no Tahr-6.05 no UPUP 3.9.9.2 no Xenial or Bionic64 OR ANY of peebee's creations. It's my reckoning that Puppy Linux would have fizzled out 15 years ago without manual polishing.

[dimkr]

Around where I live that would be "This is not an offer you can refuse"...if you know what I mean.

What I meant is exactly what I wrote - I'm not going to do this work because it's not important enough, and it's probably a problem with this particular ancient Samba package on that particular ancient Puppy.

@dimkr I was thinking about your buzz kill surrounding making distro's manually and well tough luck. I for one will continue doing it because woof-CE is a real hassle to steer when it's not doing the default recipe's.

(I don't think it's hard to customize woof-CE's build output)

I believe Puppy will die if the woof-CE project dies, because sharing tarballs in a forum is an awful way to develop something together. I'm urging others to use development tools like git, use collaboration tools like GitHub and make their existing builds reproducible (or, closer to that) because the warning signs are already here. And I'm adding another warning sign - as you can see in the commit log, I'm 100% focused on Wayland and not as involved as I used to be.

[rockedge]

@dimkr I am beginning to work with and experiment on woof-CE and hope to commit some progress soon.

Once I can figure out how to transfer the progress made manually on a distro type to woof-CE and I can produce a working system with those additions much more effort can go into it. And if I can show how it was done by examples, we'll be in a better place

[Clarity]

Reading between the lines, I view, that submitting a patch on WoofCE GIT to allow an OOTB build including SAMBA, FOR EXAMPLE, may not be accepted due to the 'smb' displeasure expressed by this thread's author.

Someone in WoofCE, 'with authority', would have to approve

This thread is surfacing something disturbing.

[Grey]

may not be accepted due to the 'smb' displeasure expressed by this thread's author.

Persuade someone to send a patch. The initiator of the topic is not such a villain

[Clarity]

This thread is surfacing something disturbing.

I like to hope you are right, but looking over this thread, there is no request of such. Instead ...

There is a constant affirmation that " ... no interest"

Another example attacks its age while this was developed in concert with Linus's Linux in the 1993-4. Is Linux consider undesirable because of its age?

Then there is the attack of security. SMB is just one of the many protocols that have been around and continue for years. Others protocols like ARP, PING, DNS, DHCP, HTTP, SSL, TLS, SMTP, SMS, etc have each, too, have vulnerabilities from time to time but the world's developers quickly and constantly address them. Thus another example to single out and distort reality.

What does that behavior say to you?

BUT, in spite of what has been written shows, I hope @Grey your are correct. You may not be a villain for something that important as somoeone "suggesting in WoofCE", but I hope you are portraying accurately.

[dimkr]

@Clarity Have you ever tried to install Samba on a Puppy that didn't have it preinstalled?

[mikewalsh]

Reading between the lines, I view, that submitting a patch on WoofCE GIT to allow an OOTB build including SAMBA, FOR EXAMPLE, may not be accepted due to the 'smb' displeasure expressed by this thread's author.

Someone in WoofCE, 'with authority', would have to approve

This thread is surfacing something disturbing.

@Clarity , in case you're not aware, Woof-CE no longer HAS a whole 'team' of people busily working on it. When you talk to - or about - @dimkr - you ARE talking to/about the person "with authority".

The community is not where it was 10 or 15 years ago, when we had scores of enthusiastic individuals who all eagerly interacted and tried to help each other out in the process of advancing Puppy's "standing" for all. I guess there's probably only a handful still extant now...

Mike.

[mikewalsh]

@Clarity Have you ever tried to install Samba on a Puppy that didn't have it preinstalled?

@dimkr :-

I don't know about Clarity, but I know I did in my very early days with Puppy. To be brutally frank, it was a f***ing nightmare....

Never, EVER again. Especially given there's far simpler and easier methods available for sharing data these days.

Mike.

[Clarity]

Hello @dimkr ... YES!

But, this thread is about a whole slew of things you've raised.

Maybe its best to just let this thread go dormant. As you have shared your position succintly. You seem to be upset should anywhere on the forum you see entries about any of the products you mentioned when you opened this thread.

So far, the majority of the forum distros are already built with a what one might consider standard subsystems, OOTB.

Hope your seeing "any" comments, in the future, of those products won't upset you as much as it has.

[wiak]

@dimkr I am beginning to work with and experiment on woof-CE and hope to commit some progress soon.

For several years, one individual, I forget his/her handle beavered away on woof-CE, but indicated in some of their comments that they realised it was a lonely and thankless task. There were a couple of others that occasionally added some work, but then they all (inevitably in such circumstances, I'd say) faded away. Dimkr, for some reason came back from his Iguleder past and became effectively the follow-on one-man band. Perhaps, there is some feeling of control or glory, but again, actually, woof-CE is just seen as an impossible to understand fortress by most people. Despite only four scripts needing to be run, underneath these involve a whole host of other woof-CE-containing scripts and config files and skeleton bits and pieces. In this day and age, is it so surprising that even the many coders capable of contributing are not willing to expend the time and energy getting to grips with that multi-part system? Hence not so surprising only one or two ever nowadays even think of trying since, why??? The likes of Peebee (and he is a vanishing breed too it seems) don't really seem to involve themselves much at all in actual woof-CE patch submissions and so on, but rather utilise what they understand on how to build a Puppy via a recipe using what woof-CE then creates for them. So end result is that hardly anyone at all is involved in that do-ocracy, and most of what we see as Puppy 'activity' is simply seen on the forum as people doing Puppy 'remasters'.

I continue to believe that the best for Puppy future would have been for FatDog to have been recognised as the successor to old traditional Puppy, if only because in look and feel it reminds you strongly of Puppy, but underneath it is a different and very nicely put together and maintained beast. Well, I understand that FatDog is FatDog and would never be called 'Puppy' (and yes, it is a different distro entirely really so stands on its own), but I also understand that FatDog continues to have a solid development team (I think... though releases seem far apart...) and because of some of its innovative facilities it has a good number of supporters. I don't use it, but I certainly respect its design.

Perhaps rockedge will be the next lonely woof-CE developer - seems like a bit of a fortress up there - a lonely castle. Is it even possible to lower the draw-bridge to make entry easier and more attractive? Hmmm... I can't say I'm convinced it is worth the amount of energy and little reward involved; status from being 'in charge'... well, that would seem more like 'imagination' to me.

This forum is a different matter altogether, but that is because it has the eventual good fortune, in my opinion, to be a place of much greater scope than any single distro involved in its many discussion threads. Trying to force it into a Puppy-only venue seemed to me like a good way to eventually close down the forum itself - at least it is still somewhat active despite maybe never likely to reach the activity levels of its now distant past. Others will no doubt disagree with that opinion, but doesn't really matter who is correct or wrong - time will decide all such matters eventually.

[Grey]

time will decide all such matters eventually.

Not a bad speech I was reminded of a recent speech by the head of the modern Communist Party of Russia. "There are capitalists around us who seized the country, but there are still people who remember how 70 years ago... We're still... Don't bury us... A bright future..."
Time will really tell

[rockedge]

@wiak I don't think I will be able to ever really get deep enough into woof-CE to make it work as well as I can get FirstRib to make a distro right now. With easy to create and modify PLUG files I whipped out a fully functioning useful KLV-Spectr. The original build took 15 minutes, with 10 more adding the kernel components and adjusting the /boot directory parts.

woof-CE is complicated and hard to steer into anything different than the default recipes.
FirstRib is easy to use easy to understand for me so I don't see me really being a woof-CE guru because the time it will take to get a grip on it enough to make it fun.

I really don't see a problem with making a woof-CE generated foundation distro using the default recipies as is and then collaborating on taking that base and building it up and polishing it. Call it a remaster...I don't label them one way or another, jsut give'em a cool name and if the thing is useful and fun to use...share it. This I can do now and seems most effective. Keeping good quality distro's is the name of this game...however it's done...is a positive in my book.

[geo_c]

Since this is one of those topics where people are just talking freely more or less, I feel like thinking out loud.

So take it all as just that.

The recent posts have got me thinking I should personally step up and try and use woof. I know @mikewalsh recently did so and got a working puppy out of it.

I should be qualified. I've compiled a good ten packages of software in the past year. I even figured out that if you type ./ in front of a script or application name in the directory you're running from, it will run the script, even if that directory is not in the system path.

That was a revelation.

I am only a musician of course, but we musicians tend to be good learners. Trial and error, practice makes perfect, etc...

But do we really want guys like me trying to build OS's and possibly sharing them with a wider community?

People with coding knowledge, experience, and training would be such a better option. There was recently a poster, a few months back, who said he just graduated school, and wanted to build a pup and asked a lot of questions about it. Is he still here? I should do a search, but I don't remember his handle or any of the post titles.

Marketing, what about that? There seems to be a lot going on around this forum that younger computer grads should be excited about. Recent efforts to make desktops more appealing by people like @pp4mnklinux come to mind, being that there's initiative to 'Get it out there,' seems like something at least a few computer nerds would be interested in.

I think puppies, Kennel-linuxes, dogs, remasters are the best thing since the invention of ice cream. They all deserve to live and thrive, and not wither up for lack of interest.

All of the people developing on the forum, @dimkr, @peebee, @rockedge, @amethyst, @fredx181, @wiak and all the people I'm forgetting to mention, really do have an impressive amount of output.

Focus seems to be the missing ingredient. For sure everyone is highly focused on their projects, but community wide the lens is cloudy.

There's actually so much going on that the war is very wide indeed, as was evidenced by the efforts to get a pup up on distrowatch.

What would it take to attract a younger, wider interest in building pups, especially using woof?

I don't think anything like that could happen without a community ethos that allows for submitting the interests of individual projects to a larger goal. Being that puppy is one of those "herding cats" cultures, seems unlikely for an ethos like that to take hold.

Still I can dream. It's one of the skillsets of musicians.

Maybe as community of shared interests, we could craft and agree on at least of selection of important questions. Things like:

1) Should puppy be the relegated to the domain of 'runs on old hardware?'
2) What's the feature set that makes a forum OS out-of-the-box user friendly?
3) Are update friendly OS's and rolling releases the future? Or is the feature set of puppy the fact that it's not updateable, and a whole new build is necessary?
4) What produces the most marketable, feature complete, and polished OS most quickly and efficiently?
5) What's the best way to develop as team? More hands make lighter work.

Stuff like that.

Not that anyone would even agree that my list of questions is relevant.

But I'm just talking. Why not?

[amethyst]

@dimkr, @peebee, @rockedge, @amethyst, @fredx181, @wiak

I develop the blues every now and then, that's all.

[fredx181]

@dimkr, @peebee, @rockedge, @amethyst, @fredx181, @wiak

I develop the blues every now and then, that's all.

Where would we be without the blues ?!!