Google Chrome/Chromium Begins Landing POWER PPC64LE Patches

[ynari]

I don't disagree, I was a fan at the time, but it's now basically on life support. However, it's a wider point that when a large project stops supporting something, it may not be practical for another group to pick that work up. This is working in reverse here, where a small group are convincing a much larger project to support them, as their longer term prospects are very interesting.

[starshipeleven]

* The "cast" feature (for Chromecast)

This is a desirable feature connected to one of their hardware devices, AND does mine information as Chromecast does need to download codec or stuff from Google servers depending on what it is actually doing.

* The PDF viewer

It's less shit than Adobe, and for most documents it's perfectly fine. I would still say it is a desirable feature.
It's an unsafe piece of shit, granted, but this is secondary.

* WebUSB

Indirect revenue through the fact that if you want to use some hardware's features you MUST use Chrome and therefore you get the usual data mining, plus there is a good chance that normal people would actually convert to Chrome, even more data mining.

* WebGL

Potential indirect (ads/tracking/app payment share) revenue through future web applications that take the place of what now is a locally-installed application.

If we're talking indirect methods of generating revenue, features on Linux like video acceleration very much subscribes to that.

You mean Windows video acceleration perhaps? The issue here is marketshare. Linux desktop has jackshit marketshare, and its users are much more likely to have adblockers and tracking blocks in place. That's not really interesting or desirable market to even care about. They maintain a token presence, but it's not much more than that.

Maybe so, but if other developers have already done the work and people using these builds are not experiencing problems, why wouldn't Google apply them in mainstream builds? The work was already done for them, for free. So what do they have to lose?

The same any opensource developer has to lose by accepting random contributions. That's code that can and will break sometimes, and once it is in the project it becomes their own responsibility to make sure it works. If they accept it and then let it rot, it's worse than not accepting the contribution at all.

Also, all the code added will add complexity to the component it is added to, and this means mainteneance of parts they actually care about will become harder.

I'm not saying they can't afford it or that it would be an issue for them. Google can very well afford to hire a dozen developers dedicated to Chromium on Linux Desktop development, and do an amazing job and be loved by the children without even noticing the costs of this additional staff on its bottom line.

The point is, does it make economical sense to do so? Eeeeh, no I don't think so.

[starshipeleven]

and Warpzilla will be end of lifed in favour of a browser with a Chromium engine.

I'm not sure why this is even bad for anyone. I mean who cares about OS/2 to begin with, let Mozilla do something new and potentially interesting and leave old crap alone.

[ynari]

When Mozilla stopped OS/2 builds, a team forked it into Warpzilla and so it continues.

Point of fact. When Mozilla stopped OS/2 builds, a third party company started up Warpzilla builds and was contracted by Arca Noae to continue the work.

Mozilla has now got to the point where it can't be maintained under OS/2 (New builds need Rust, and large memory support. OS/2 is resolutely 32 bit), and Warpzilla will be end of lifed in favour of a browser with a Chromium engine.

There simply aren't enough OS/2 users to support the infrastructure some people would prefer.

[schmidtbag]

What features you think people didn't ask for and are not generating ad revenue?

Just to set things straight, I mean features nobody asked Google to put into Chrome by default; the features themselves could still have value or be desirable (I personally like some of them) but my point is they didn't need to be integrated:
* The "cast" feature (for Chromecast)
* The PDF viewer
* WebUSB
* WebGL
There are plenty of other "unnecessarily integrated" features too, but they're not bloated or insecure enough for me to warrant bringing them up. And of course, there are features that Chrome comes with that people didn't ask for (or explicitly don't want), but, those are revenue-generating (like the data collection).

Because there is more than just end users (also developers targeting browsers with web apps, Android "web-technology" apps, or chrome extensions), and many other indirect ways of generating revenue (through data collection in new services they can offer, through getting a cut in paid apps/extensions or ads served in-app and so on, through getting a lot of free input to test their AI algorithms that are going to be monetized in a not-so distant future).

If we're talking indirect methods of generating revenue, features on Linux like video acceleration very much subscribes to that. Keep in mind, I know there are features Linux users desire that I don't necessarily think is worth Google's time, like proper Wayland support or PPC64 support.

Also, even potential future revenue if a new feature catches on is still better than 100% sure jackshit in secula seculorum, which is what they get from (third party) linux builds of chromium.

Maybe so, but if other developers have already done the work and people using these builds are not experiencing problems, why wouldn't Google apply them in mainstream builds? The work was already done for them, for free. So what do they have to lose?

[starshipeleven]

Mind explaining how exactly is Google paying bills for features people didn't ask for and don't generate ad revenue?

What features you think people didn't ask for and are not generating ad revenue?

Because there is more than just end users (also developers targeting browsers with web apps, Android "web-technology" apps, or chrome extensions), and many other indirect ways of generating revenue (through data collection in new services they can offer, through getting a cut in paid apps/extensions or ads served in-app and so on, through getting a lot of free input to test their AI algorithms that are going to be monetized in a not-so distant future).

Also, even potential future revenue if a new feature catches on is still better than 100% sure jackshit in secula seculorum, which is what they get from (third party) linux builds of chromium.

[schmidtbag]

The point here is that Google is paying the bills with the additional features, not people using Chromium on Linux.

Mind explaining how exactly is Google paying bills for features people didn't ask for and don't generate ad revenue?

[Charlie68]

You may want to read my last sentence again.

But this is the beauty of open source software, if it does not Google, others can do it. Google for its part has explained the reason for this choice, which is related to too many problems and fragmentation of Gnu / Linux. It is no coincidence that openSUSE had to even patch Mesa to make it work properly.

[-MacNuke-]

In openSUSE, VAAPI hardware acceleration in Chromium is enabled by default and works well on my PCs.

You may want to read my last sentence again.

[curaga]

There are newer options than old Macs if you want to play with Power. I linked yesterday a refurb POWER8 server (now sold out unfortunately), such options appear once in a while. Some Xbox 360s can run Linux, and all WiiUs are hackable. Raptor's Blackbird is new and POWER9, but the mobo+cpu combo is over 1k.