Release SunOS (Solaris/Illumos) Official Builds #1718

- TBD -

Ping @athenian200
Depends on MoonchildProductions/UXP#1320

**- TBD -** ---- Ping @athenian200 Depends on MoonchildProductions/UXP#1320

Last I heard this was rejected by Illumos due to some influencer convincing people otherwise?
Has something changed? Are we going to publish them anyway outside of the regular distribution channels for that O.S.?

Last I heard this was rejected by Illumos due to some influencer convincing people otherwise? Has something changed? Are we going to publish them anyway outside of the regular distribution channels for that O.S.?

I say we publish em anyway. There is intrest.

Okay! Suits me fine.

So this mean you will offer binary for OpenIndiana officially? This is exactly what I'm going to ask for. I don't mind having to download the .tar.xz from your website if OI refused to add it to their package repo, though.

BTW, I successfully compiled PaleMoon on OI but it's a painful experience. No, the build is straight forward but it's too slow. It's a VirtualBox which only has 5138MB of ram (my host only has 8G so it's the best I could give to it) and 4 cores. It tooked 4 hours to complete with the generation of the .tar.xz package but I don't feel very confident to use this build daily since because lacked of ram the linking of libxul.so had some warnings. Nevertheless, the resulted palemoon binary worked just fine and is much faster than firefox on my system.

@hoper-a11y

I'm the SunOS (illumos and Solaris) maintainer, and I've been running a self-compiled version on OpenIndiana for the past week or so as my daily driver with no issues since I got 64-bit builds working. I created this in a virtual machine before switching to OI as my primary operating system on real hardware because the included version of Firefox wasn't cutting it for me.

This build is very RAM hungry, even 8GB really wasn't enough for it, and you ideally want more than four cores to compile it. It was less intensive when I first started, but various build/code changes eventually reached a point where I couldn't reasonably compile it in a VM, however the browser runs better than ever despite the long compile times and linker memory usage with libxul.so (I'm painfully aware of that issue, and the warnings there are normal).

There are a few finishing touches I should probably put on the build process, but actually creating binaries is something that slipped my mind because of other issues we're working on. I'm planning to start up SunOS work again soon and hopefully we can have official binaries relatively soon. It definitely motivates me to know that someone out there actually wants to use this.

@hoper-a11y I'm the SunOS (illumos and Solaris) maintainer, and I've been running a self-compiled version on OpenIndiana for the past week or so as my daily driver with no issues since I got 64-bit builds working. I created this in a virtual machine before switching to OI as my primary operating system on real hardware because the included version of Firefox wasn't cutting it for me. This build is very RAM hungry, even 8GB really wasn't enough for it, and you ideally want more than four cores to compile it. It was less intensive when I first started, but various build/code changes eventually reached a point where I couldn't reasonably compile it in a VM, however the browser runs better than ever despite the long compile times and linker memory usage with libxul.so (I'm painfully aware of that issue, and the warnings there are normal). There are a few finishing touches I should probably put on the build process, but actually creating binaries is something that slipped my mind because of other issues we're working on. I'm planning to start up SunOS work again soon and hopefully we can have official binaries relatively soon. It definitely motivates me to know that someone out there actually wants to use this.

Tennative plans are that Pale Moon 29 will be the start of sunos as well as linux gtk3 builds. This does require infra changes on our part and making sure our ducks are in a row.

Until then you can continue to build for your self if you want.

Tennative plans are that Pale Moon 29 will be the start of sunos as well as linux gtk3 builds. This does require infra changes on our part and making sure our ducks are in a row. Until then you can continue to build for your self if you want.

Tennative plans are that Pale Moon 29 will be the start of sunos as well as linux gtk3 builds. This does require infra changes on our part and making sure our ducks are in a row.

Until then you can continue to build for your self if you want.

I will wait for you. I don't have the computer resource to build it easily. It's fast to build on Linux but on OI it seemed to need twice of the resource I currently have to be reliably build. My workstation only has 4 cores and 8G of ram, I can't give the OI VM anything more and OI doesn't support my network card, I can't install it on the physical machine.

> Tennative plans are that Pale Moon 29 will be the start of sunos as well as linux gtk3 builds. This does require infra changes on our part and making sure our ducks are in a row. > > Until then you can continue to build for your self if you want. I will wait for you. I don't have the computer resource to build it easily. It's fast to build on Linux but on OI it seemed to need twice of the resource I currently have to be reliably build. My workstation only has 4 cores and 8G of ram, I can't give the OI VM anything more and OI doesn't support my network card, I can't install it on the physical machine.

I'm fine with officially releasing current versions as well if they are stable enough. With our current development for releases on redwood there's little difference between it and the new milestone in terms of OI support, so it can be at least added as an official 3rd party build to the website and linked from there to any binary distribution location desired.

I'm fine with officially releasing current versions as well if they are stable enough. With our current development for releases on redwood there's little difference between it and the new milestone in terms of OI support, so it can be at least added as an official 3rd party build to the website and linked from there to any binary distribution location desired.

I have no problem using a third party build, especially when knowing it's from a member of the team. I only feared having to build it again myself, another 4 hrs? No no I will just use your build 🥰

> I'm fine with officially releasing current versions as well if they are stable enough. With our current development for releases on redwood there's little difference between it and the new milestone in terms of OI support, so it can be at least added as an official 3rd party build to the website and linked from there to any binary distribution location desired. I have no problem using a third party build, especially when knowing it's from a member of the team. I only feared having to build it again myself, another 4 hrs? No no I will just use your build :smiling_face_with_three_hearts:

@athenian200 How are we standing on this? I'm assuming this isn't 29.0 material (as we're in the final stages of that) but is there any news/movement here or is this going to take a while yet?

Well, everything still technically compiles and runs, but it all suffers from a lack of testing and the one banned guy on the forums who used it didn't seem very happy with various things about it.

There are a couple of tweaks I'd like to make, but none of them would fundamentally change what I've already implemented. I'm using the browser daily with no real problems, so I could try building a release but I wouldn't feel right calling this a polished product on par with what we have for Linux.

Well, everything still technically compiles and runs, but it all suffers from a lack of testing and the one banned guy on the forums who used it didn't seem very happy with various things about it. There are a couple of tweaks I'd like to make, but none of them would fundamentally change what I've already implemented. I'm using the browser daily with no real problems, so I could try building a release but I wouldn't feel right calling this a polished product on par with what we have for Linux.

Well maybe start making Beta branding versions and enable updates. The new AUS would be able to accomidate it. Give the builds to me and I will throw em up.

We should get them out there because they need to be subjected to the potentally insane bashing by normal users.

I still feel like we should come up with something that would accomidate not-windows for everyone to use but not mess with the mainsite as that is a much bigger project.

First thing to do is transfer linux specific bits to the dpmo docs and I merely suggest having a download.palemoon.org and just use that as a unified downloads page. Windows would refer back to the mainsite for now and the rest would be basically a single page that can index and serve the not-windows builds. It is much easier to do such a thing when the files are ON that server. linux dot would just redirect links to their new locations. That way de2 isn't bogged down with linux files.

Something that could be done by the time the SunOS builds have been thrashed for a bit.

Well maybe start making Beta branding versions and enable updates. The new AUS would be able to accomidate it. Give the builds to me and I will throw em up. We should get them out there because they need to be subjected to the potentally insane bashing by normal users. I still feel like we should come up with something that would accomidate not-windows for everyone to use but not mess with the mainsite as that is a much bigger project. First thing to do is transfer linux specific bits to the `dpmo` docs and I merely suggest having a `download.palemoon.org` and just use that as a unified downloads page. Windows would refer back to the mainsite for now and the rest would be basically a single page that can index and serve the not-windows builds. It is much easier to do such a thing when the files are ON that server. linux dot would just redirect links to their new locations. That way `de2` isn't bogged down with linux files. Something that could be done by the time the SunOS builds have been thrashed for a bit.

Well something needs to be done if we want to move forward, is my point.
"Just building it and using it yourself" isn't what the plan was so getting some builds out there that aren't primary-tier targets to get the sharks to ~~hug them with their mouths~~ thrash them is what we need to be doing. But that can be set in motion after the milestone release, just think it over how to.

As for a download site... I'm not sure -- people will start linking to that and then just ignore the actual/informational product website and not know what they are getting, most likely.

Well something needs to be done if we want to move forward, is my point. "Just building it and using it yourself" isn't what the plan was so getting some builds out there that aren't primary-tier targets to get the sharks to ~~hug them with their mouths~~ thrash them is what we need to be doing. But that can be set in motion after the milestone release, just think it over how to. As for a download site... I'm not sure -- people will start linking to that and then just ignore the actual/informational product website and not know what they are getting, most likely.

Limited dev time makes this rather unlikely, so putting this on hold. Builds will remain ad hoc for the time being.

Labels Milestones

Release SunOS (Solaris/Illumos) Official Builds #1718