[Manjaro32] Getting rid of 32bit Manjaro based on Archlinux

[philm]

Hi Community,

from today on we have removed Manjaro for 32-bit as you may know it. Couple of days we already had removed links to our 32-bit ISOs for given reasons. So lets talk about how large Manjaro was in that architecture, what our plans are for 32-bit architecture and if you have to move on to another Linux distribution cos of that.

https://manjaro.org/wp-content/uploads/2017/11/manjaro-32.png

To go back in history we started as a really small team of five. Roland, Guillaume, Wlad, Allesandro and me. We went dark before we came back online. So it always takes time to get the good stuff rolling.

As I archived v17.0.6-x32 it resulted in a size of 37 GB:

1.8M	./v17.0.6-x32/core/i686
1.8M	./v17.0.6-x32/core
21M	./v17.0.6-x32/extra/i686
21M	./v17.0.6-x32/extra
48M	./v17.0.6-x32/community/i686
48M	./v17.0.6-x32/community
70M	./v17.0.6-x32
34G	./sync
2.2G	./overlay
37G	.

This includes all any-packages and i686-packages of Manjaro based on Archlinux. So this area is gone for good now. If you want to have those packages saved on one of your servers, please send me an email to talk about details.

So how do we go on from there? Well, our team has grown, we started a new partnership and even started our new laptop line with the spitfire and will connect to other partners in future. So given by that the brand Manjaro is now known to many of you.

We did polls about dropping 32-bit for good. The community and our development team both agreed to axe it for good. However, some of you have still some reasons to hang on to that architecture.

Myself not being a throw away and all is good-guy I thought about it a lot. Archlinux32 is a small project, as they only started couple of month ago. Manjaro has established a bigger Mirror-Infrastructure and will have even a better cloud based mirror-system soon. So I thought, why not continue the boxit-way as we do with Manjaro not also with the new Manjaro32 project. So I secured the github page and @jonathon registered the URL for it.

Jonathon himself had following idea to keep the 32-bit version alive. That gave me an idea to even improve it. Adding some code lines to boxit gave us the opportunity not to only syncing from Archlinux but also on the same infrastructure from Archlinux32. Since Jonathon already added our overlay packages to his [manjaro32] repo, I did it the classic way as we already do with Manjaro in 64-bit.

Together it results now in following footprint:

2.0M	./x32-testing/core/i686
2.1M	./x32-testing/core
25M	./x32-testing/extra/i686
25M	./x32-testing/extra
57M	./x32-testing/community/i686
57M	./x32-testing/community
84M	./x32-testing
2.1M	./testing/multilib/x86_64
2.1M	./testing/multilib
2.0M	./testing/core/x86_64
2.0M	./testing/core
23M	./testing/extra/x86_64
23M	./testing/extra
52M	./testing/community/x86_64
52M	./testing/community
78M	./testing
2.1M	./stable/multilib/x86_64
2.1M	./stable/multilib
2.0M	./stable/core/x86_64
2.0M	./stable/core
23M	./stable/extra/x86_64
23M	./stable/extra
52M	./stable/community/x86_64
52M	./stable/community
78M	./stable
2.1M	./x32-unstable/core/i686
2.1M	./x32-unstable/core
25M	./x32-unstable/extra/i686
25M	./x32-unstable/extra
58M	./x32-unstable/community/i686
58M	./x32-unstable/community
85M	./x32-unstable
2.1M	./unstable/multilib/x86_64
2.1M	./unstable/multilib
2.0M	./unstable/core/x86_64
2.0M	./unstable/core
23M	./unstable/extra/x86_64
23M	./unstable/extra
52M	./unstable/community/x86_64
52M	./unstable/community
78M	./unstable
2.2G	./pool/overlay-32
39G	./pool/sync
38G	./pool/sync-32
2.9G	./pool/overlay
82G	./pool
82G	.

So the Manjaro-64 is 41,9 GB and Manjaro-32 is now 40,2 GB in size. Any-Packages are normally meant to be used in both architectures, however, since we have now two different projects we have to keep them now separate.

Manjaro-32 may not get as frequent updates as Manjaro-64 will get, but it will be still maintained. Some might have now noticed that there is no x32-stable repo yet. This we did on purpose. If we compare both the old Manjaro v17.0.6-x32 with the new Manjaro-x32 you will see that not all packages are at the same level. A list of differences you may find here.

So it is good to read up on this thread to know all about the new community project. And as always, you're more than welcome to join it.

Tell us what you all think about it and if our additional effort to keep Manjaro Linux in 32-bit is wise to do so.

• Yes, I'm happy that you guys still support 32-bit
• I don't mind. Good luck with Manjaro-32, though
• Seems to be a waste of time on your end, though

[mirh]

As usual hats off for considering less lucky users, for as much reasonably feasible.

p.s. I think you misspelled alessandro

[Th3Z0ne]

As I am in the (lucky) situation of not owning a 32bit macine anymore, I don't mind.
But I think its a great move for those people that might still have a old PC and can't get a new one

[philm]

@mirh

I think you misspelled alessandro

Na, his name is Allesandro Calo. See here.

[cimarronline]

Much appreciated! I've received a number of 32-bit machines the past few years, as people have moved on. Linux keeps them alive, and I was impressed by how well Manjaro ran on them even though they are often RAM-limited. I've been able to give people a loaner when they're between laptops, or as a gift when people can't afford a new one, or for kids. Thanks so much for making it easy to keep them going with Manjaro.

Will the older kernels also be maintained for 32-bit computers? Sometimes hardware support gets dropped in the newer kernels. Kernel choice is one of the great advantages of Manjaro. But if it's too much work, I certainly understand.

[philm]

@cimarronline: I've to see on how it pays of. Manjaro-32 is at its project start. It will be a bumpy ride for sure. It depends what the Manjaro-32 Developers team decide. I assume they will reduce the effort to only two kernels. One LTS and the current one. We have to see. Which kernel do you use on your machines?

[cimarronline]

I know at least one laptop needs a pre-4 kernel (for touchpad support), so I've been using 3.16. It's supported until 2020. So I'd recommend that, if you want to keep one older kernel available.

But I understand there may not be enough demand to make it worth the effort. You've already gone beyond my expectations for supporting older hardware.

[d7rk]

Well if it's for 32-bits machines it does make sense to keep an older kernel right?

[eugen-b]

I don't know if it is in fact neccessary. For example antiX is using quite a new kernel and it is targeted on old machines. I don't know exactly how it is patched though.

[ryanmusante]

What is that laptop pictured in the beginning?

[gustav]

I can't throw some numbers but around me there are still A LOT of 32-bits computers, especially for not-techie people who don't change their PC every 10 years (I know, how incredible it is!).

And I've still got my Dell mini 10: Works like a charm!

So very good initiative!

[philm]

@ryanmusante

What is that laptop pictured in the beginning?

Toshiba Libretto 50CT

[tuerda]

I don't use a 32 bit machine and will never use Manjaro-32 myself, but I was pleasantly surprised by this news. I think this is another great example of the Manjaro team's commitment to supporting their user base. You guys are great!

[txhammer68]

i like the idea of keeping Manjaro 32bit around, I rebuild older PC's for ppl all the time and always install Linux, I usually use Mint Cinnamon x86 edition b4 i discovered Manjaro.
Most recently i had the idea of building an rescue USB Linux disc like Partition Magic, since its not free anymore, and Manjaro XFCE or LXQT 32 bit would be my goto DE of choice to build this rescue distro... so yea retaining 32bit manajro is a great idea, 32 bit is not dead yet, maybe 5 more years...

[jimdunn]

I am grateful for the "push"... I finally got rid of my last 32 bit machine... it's a nice feeling.

[eugen-b]

Why so?

[anon71843593]

I'm really quite torn on this issue between 'I don't mind' and 'It's a waste of your time'. While I applaud the efforts and sentiment I also question how beneficial it is to carry on with it too.

you are much better spending time with more OEMs and getting manjaro shipped as a pre-installed OS like you have with the spitfire. Additionally, if you could do so with desktop systems and more modest i3 or upcoming AMD Ryzen based APU notebooks for example in addition to the premium stationX spitfire model, you will of course increase the user base further. not everyone has the funds to spend on a notebook starting at £1k.

Reason being, how much longer will the affected legacy hardware actually be economically maintainable? spare parts for systems from even a decade ago are becoming much harder to source and therefore more expensive. So, factoring this scenario into the equation, how long do you really want to carry on supporting 32-bit architecture as a community project before it gets to the point you only have double or single digit users for that architecture?

64-bit hardware is already flooding the refurbished pc market and being shipped to the third world where redundant 32-bit hardware used to be.

[eugen-b]

You are very right in general. But there are some counterarguments.

• People are more likely to ditch Windows on old machines. - It is desirable to leave them this opportunity. Other Linux or BSD or similar distros can be a good option, but for some "stupid" reasons many people chose Manjaro. They might be not entirely wrong.
• Spare parts for 32bit netbooks are still abundant. I recently bought a battery for 15€, before that it was the charger for 7€. You can 2GB SODIMM DDR2 667 RAM for 10-15€. The netbooks themselves don't get much cheaper than 50€ for several years, there seem to be some demand. Ok, I wouldn't recommend anybody to buy a 32bit machine anymore, not even for these prices.
• The maintainers of archlinux32 might become Arch maintainers when they will have automatised or abandoned their project.

[linux-aarhus]

I don't like the 'use-and-throw-away' mentality which is a dominant mentality - at least in the more wealthy parts of the world population.

I think that the guys - with @jonathon as a driving force and the arch32 guys - a small team is doing great by keeping the computer vaste down. Lets use those 32-bit computers till they die of old age.

My old Thinkpad z60 will live a little longer

[jhagger84]

I would to see a Manjaro -32
Updates and New ISO