Is TCL's philosophy of minimalism going the wrong way!?

[xor]

Is TCL's philosophy of minimalism going the wrong way!?

TCL's philosophy of minimalism should not be adopting the latest kernel, but the most stable, oldest-based, minimum-consumption kernel!

low-level legacy users; clearly looking for early TCL releases, my suggestion is to support both the latest and oldest versions at the same time!

Would our TLC community consider a democratic vote on this?

[Rich]

Hi xor

... Would our TLC community consider a democratic vote on this?

Absolutely. All forum members are encouraged to exercise
their right to vote in the attached poll.

Good luck xor, I hope you win.

[gadget42]

Should xor be required to support the oldest versions of Tinycore?

go big or go home, how about this:

"Should xor be required to maintain a complete and up-to-date TinyCoreLinux mirror?
(including any/all versions that are removed for space constraints. see https://forum.tinycorelinux.net/index.php/board,49.0.html for additional information)

[gadget42]

speaking of space...

@curaga, will there be a post regarding version 4 shortly?
(last year's post was dated October 20th)

[CentralWare]

@xor: Is it going the wrong "way" -- not really.

Is that way antiquated?
Maybe a little.

Bare in mind, Tiny Core Linux / DSL was spawned back in the day when the general public DID NOT have computers in their pockets, when a USED 80486 computer was still astronomically priced compared to today's machines, back when RAM was still a commodity due to natural disasters overseas (and "under" seas) but most importantly, the general PEOPLE were still clueless for the most part and WinXX was their tool of choice to break in the new age.  DSL/TCL was visualized because people wanted to take USED machines and give them NEW life.

Now, in the past couple of decades things have progressed, become cheaper (price AND quality), more portable, etc.  Tiny Core Linux is still leaning toward the 486/586 days mentally, though.

Is a CLEAN, LEAN, RIPPED operating system a good thing?  Of course it is.
Does it have to be byte-counted?  Probably not as much as it is currently, but to do anything different would open debates that can't be resolved as someone out there is going to ask "how much is too much?"

Hard drive storage, Flash storage and RAM are a fraction of the cost they were when DSL (2005) and TCL (2009) were launched so there's room to wiggle unless you're trying to resurrect a 486dx and turn it into a kiosk or something...  a feat to have fun doing, no doubt, but for day to day use...  not the wisest path to take if you have other means.

Does it have to have CURRENT kernel releases?  Most definitely.

You have to understand, the KERNEL isn't just basic functionality and foundation...  it's HARDWARE SUPPORT.
The newer the hardware, the newer the kernel.  Period.
If you have older hardware, you can get away with an older kernel.

Should Tiny Core offer multiple kernels?
Technically, it already does.  Pick a TCL version from 5.x to the recent 15.x release - you have plenty to choose from!

Now, if we were to keep EXTENSIONS in each of those versions up to date...  THAT would be a feat!
A very unlikely one, though, as some extensions depend on kernel content so the extension cannot be any "newer" than the kernel without it potentially breaking.
It's a blissful thought, though!

You have a valid question, but it's not really an opinion based topic in need of a vote.

Tiny Core doesn't have to be "as tiny" as it is - but I do like the methodology behind it.
Tiny Core doesn't HAVE to be as "new" (kernel) as it is...  but if it weren't, it wouldn't support things like USB3.x, the latest graphics cards, the latest CPUs, etc. so it would become an operating system that ONLY supports OLD hardware...

[curaga]

Thanks for the reminder gadget42.

[gadget42]

@CentralWare: Very Well Said! Ditto!

[mocore]

Thanks for the reminder gadget42.

https://forum.tinycorelinux.net/index.php/topic,27321.msg175797.html#msg175797

if the logic driving the removal of old versions is reduce space  could / should tcz/src (build scripts) be retained ?...
*more* on that tangent :

reinstating  ./src from the removed/archived repository's ?

& ftr the downloads.html still links to 1/3 .x

[xor]

Thx

Hi xor

... Would our TLC community consider a democratic vote on this?

Absolutely. All forum members are encouraged to exercise
their right to vote in the attached poll.

Good luck xor, I hope you win.

[xor]

It may be a very idealistic approach,
but my personal opinion is that the latest hardware can be supported by preserving the kernel architecture of the 2.6 series.

[CentralWare]

if the logic driving the removal of old versions is reduce space  could / should tcz/src (build scripts) be retained ?...

@Curaga: @mocore brings up a good point regarding sources and build scripts
For the time being, at least until we have a solid source repository/archive, please LEAVE source directories when disposing of antique versions.
x86/x64: ./release/src and ./tcz/src
arm/etc: ./releases/RPi/src and ./tcz/src

It may be a very idealistic approach,
but my personal opinion is that the latest hardware can be supported by preserving the kernel architecture of the 2.6 series.

@xor: You do realize that 2.6 was given a warrior's funeral around eight years ago?
Maybe I am not seeing something you're trying to get across...  please enlighten me?
I have kernel.org on the screen to my right showing that the oldest long-term kernel still being supported with security/bug patches is v4.19
Please be aware, 4.19 is old enough that it probably doesn't know what USB3.2 ports are or how to communicate with them, so loading that kernel on a newer motherboard is probably not a good idea unless you have no use for the newer hardware chips.

From the sound of it, you're asking to resurrect kernel v2.6.32 (which we were using back in Tiny Core Linux v2.x! Kernel 2.6.29 specifically.)
In YOUR eyes, what would be gained by making the kernel from ~2010 available today?
(I can send you a copy of Tiny Core 2.x, but I'm not sure the general public has much use for the same.)

[CNK]

I'm posting from a PC running old DSL with kernel 2.4. A fair bit of software does still build for such old kernels/glibc, though things like current Firefox versions are way out.

My main use of Tiny Core is running software like Firefox which needs an up-to-date environment. For some other systems I run old TC versions as long as I can get away with because it's not worth the time of upgrading and there's little advantage. I don't care about recent hardware support, the cutting edge has been beyond my needs for decades now.

Ideally I want a distro like Tiny Core based on the work of the CIP Core project that aims to keep Linux kernels and a base set of Linux software maintained for ten years. Ten years is a sensible upgrade cycle in my opinion. I'm happy with hardware much older than that, especially running older, and therefore less bloated, software. If that existed, and projects like Firefox saw the light and made an effort to keep working with those old kernels, I wouldn't need Tiny Core and its modern kernels anymore. It's very unlikely that projects like Firefox would do that though (Debian have even struggled to build new Firefox for their stable releases in the past), so really I want Tiny Core to keep doing what it's doing letting me run a bare minimal system for flashy software, while I just use old unmaintained Linux for other tasks where flashy modern software (and good security) isn't needed. It's not ideal for me, but beggers can't be choosers.

[Stefann]

My 2cents,
I have been running DSL on a board with 500MHz ultra low power Eden processor since 2008. This had 2.4 Linux kernel. This summer I upgraded to Linux 6.6 (which forced me to tiny core because dsl is no longer maintained) for simple reason “it was no longer practical to stay on 2.4”.

With that I mean:
- sure everything keeps running if you don’t want to upgrade apps.
- but extending became close to impossible.
- I for instance was not able to use https calls.

The board runs my home automation system that I have programmed from scratch in C. In itself “it works” so I could have chosen to “not extend further”, but new devices come to the market. New protocols get introduced. So….. I had to choose to either go for a more modern board like raspberry pi OR get Linux kernel updated. I was happy that tiny core allowed me to upgrade.

With all that said…
- I think it is very helpful to keep earlier versions of tiny core downloadable.
- very old hardware can than at least brought to life.
- I also think it’s helpful to keep the associated tce extension library downloadable, and especially the programming tools (gcc). That allows such board to do something useful instead of just showing a desktop and will allow to program simple appplications perfectly suited for outdated hardware.
- but trying to keep the newest applications runnable on that old Linux I would consider a bridge too far. It likely runs into many “new kernel must haves”.

In other words:
- there is nothing wrong with the ability to still download a 15 year old ecosystem to use on hardware that once did cost 500 to 1000euro or dollar and can now be achieved for under 10.
- there can be lot of fun in operating that hardware on that 15 year old ecosystem. I remember I had lots of fun 10 years ago. No problem to relive that.
- but trying to bring that hardware into a modern ecosystem I consider a bridge too far.

While of course: if someone want to try: fun! But just not as a “general strategy”.

And as said…
I say that as a guy who had spent about a year to think about best way to “re vitalize” his low power board with linux 2.4. Finally concluding that it either had to go to 6.6 or abandoned.

If your interested, I posted some of that transition here: https://forum.tinycorelinux.net/index.php/topic,27188.0.html

[gadget42]

at this point in the thread progression perhaps a good question to have answered definitively would be is there a way for someone to pay-for-and-receive-via-mail-post-etc a hard-copy of releases that have been removed?

naturally it is to be expected that whomever might offer/provide this service should be compensated for all time/expenses incurred.

to that end, if someone wanted just the TinyCoreLinux 1.x release in its entirety, how much data is that actually? and if compression is used, checksums would probably be a good idea for both the originals and the compressed tarball?

realistically, this is/might/would-be a very rare occurrence but i can see it having value moving forward into the future as more releases are taken offline.

let the fun begin...lol.

[CentralWare]

@gadget42: Tiny Core 2.x ISO files and TCZ extension library would fit cleanly on a DVD (does not necessarily include tcz/src, the archive directory, updates, etc.)
The TCE extensions are another gig and change
The SRC directories are unreliable in size as many are uncompressed
All versions older than 5.x have their own "way of doing things" and IMO should not be supported for this debate.

That said, it may be more prudent for DVD or USB pen drive sales as opposed to hosting the older content live 24/7

@Stefann: Please note which version(s) of TCL you want/need from the archive (v4.x and older)
@CNK: Please do the same
@xor: Please do the same
Gadget42: Ditto

naturally it is to be expected that whomever might offer/provide this service should be compensated for all time/expenses incurred.

Anything "hard copy" by post would incur a cost; how much depends on media, time and shipping costs - I MAY have a postage-free method in mind for the time being.

I'm working on a couple ideas right now as to how we can accommodate archive members; bare with me!