One Of The Reasons Why Linux 5.5 Can Be Running Slower

[pgoetz]

Greater than 50% ad block rates.
Less than 1% being premium subscribers.
Less than 1% tipping.

I didn't know about tipping and thought this would be a good way to contribute. Unfortunately a 5-minute perusal of phoronix.com did not provide a clue as to how one might tip. Providing information on how to become a member and how to tip in the top menu bar might have a positive impact on the monetization of the site. Just sayin'.

[Michael]

I didn't know about tipping and thought this would be a good way to contribute. Unfortunately a 5-minute perusal of phoronix.com did not provide a clue as to how one might tip. Providing information on how to become a member and how to tip in the top menu bar might have a positive impact on the monetization of the site. Just sayin'.

Right hand side of all the web pages under "Support Phoronix" should be a "PayPal tip" link.

[JanC]

Problem there is that that link is most likely removed by tracker blockers, because Paypal is one of those scummy privacy-invading companies people try to keep out of their life as much as possible…

[billbo]

No need to wonder when it's on their site.

Actually, that page doesn't answer his question. It says what they do with "individual donations". Not with the (apparently) much larger slush funds that they get from corporations (including Microsoft). See complete list here:

https://www.linuxfoundation.org/membership/members/

Propublica has some summary info on their budget at:

https://projects.propublica.org/nonp...ions/460503801

In 2017, the Linux Foundation spent over $7 million on "executive compensation". Seems like paying Phoronix to do performance regression testing of Linux kernels would be well within their budget. It even sounds like something their corporate membership would seem to want to support (well maybe not Microsoft...).

[pgoetz]

Right hand side of all the web pages under "Support Phoronix" should be a "PayPal tip" link.

OK, just found this at the bottom of the right sidebar (that likely no one ever looks at), so was able to PayPal a donation (the AppArmor discovery is definitely worth $20 to me).

As already mentioned, there's no harm in moving this stuff to the top bar or making it more visible somehow.

[danmcgrew]

Right hand side of all the web pages under "Support Phoronix" should be a "PayPal tip" link.

Two problems:

1) none of the web pages show "Support Phoronix" (probably some sort of fault on my part);

2) Hell will freeze over before I use PayPal for ANYTHING. I'm MORE than certain you've heard this before; I, absolutely, cannot understand what kind of choke-hold PayPal must have on its "service" recipients, to make them not seek alternatives. It got so bad that Clement Lefebvre (Linux Mint) turned into a laughing-stock with all the really ludicrous excuses he made for staying with PayPal as his ONLY method of EVER accepting donations. It's amazing that his position changed as the donations started dropping...

[NotMine999]

It's unbelievable how these awful regressions pass unobserved until Michael finds them...
I think it's clear that the Linux kernel who has millions of users has no quality assurance whatsoever.
I wonder what is the Linux Foundation doing with all those money they get from the companies ???

They should be pouring something like 5000 $ / month to Michael for doing this wonderful job of finding this terrible regressions and raising the alarm.
If it weren't for Michael we would have a much lower quality and less performant kernel.

The LK developer crowd has always impressed as "typical programmers".

They don't write very good documentation ... when they write documentation.

They don't perform useful regression testing on their work ... when they should ... as many rely on volunteers (3rd parties) to test the code to dubious standards.

I guess these genius LK developers can't be bothered to conjure some test rigs for their code, though I have heard a few of them might have 1 or 2 rigs for testing that they timeshare with other tasks. Perhaps these LK developers are waiting for hardware donations or the price of RAM to drop.

I still say the Linux Foundation should send a grant to Michael & Phoronix, or provide a yearly stipend to help offset the costs of his work in exchange for a set amount of testing. Better yet, perhaps the Linux Foundation should fund an effort by Michael to develop some testing platforms for the use of LK developers to judge the impact of their patches & features.

[F.Ultra]

The LK developer crowd has always impressed as "typical programmers".

They don't write very good documentation ... when they write documentation.

They don't perform useful regression testing on their work ... when they should ... as many rely on volunteers (3rd parties) to test the code to dubious standards.

I guess these genius LK developers can't be bothered to conjure some test rigs for their code, though I have heard a few of them might have 1 or 2 rigs for testing that they timeshare with other tasks. Perhaps these LK developers are waiting for hardware donations or the price of RAM to drop.

I still say the Linux Foundation should send a grant to Michael & Phoronix, or provide a yearly stipend to help offset the costs of his work in exchange for a set amount of testing. Better yet, perhaps the Linux Foundation should fund an effort by Michael to develop some testing platforms for the use of LK developers to judge the impact of their patches & features.

There are numerous regression tests being run on the Linux kernel. The CKI project (https://opensource.com/article/19/6/...egration-linux) alone runs over 100.000 kernel test runs each month. And then there is LTP (https://linux-test-project.github.io/) and also various distributions perform their own set of regression testing.

The problem with this one is explained by Michael in the article itself, AppArmor is not used by many outside of Ubuntu. Also 5.5 is not released, it's still in development and no actual distribution is going to use the vanilla version of it anyway.

A second problem is that the vast majority of the regression testing is done for correctness and not performance (not to mention that decrease in performance for one type of load on one type of system could be an increase in performance for a different kind of load on a different kind of system).

[F.Ultra]

EU law doesn't apply on this side of the pond! Internet FTW.

Don't really know why but this is the first time I realised that Michael was American, I've always seen him as either German or Belgian (must be his love for Beer that made me believe that).

Don't really understand why you cheer that you are not subject to pro-privacy legislation?! (And I say that as an EU citizen that browses on Phoronix with ad-blocker disabled).

[xorbe]

LWN does it better by not requiring signing up for a forced recurring subscription. A couple weeks before expiration they show reminders, etc.