Firefox 49.0 requires SSE2 instructions

[middings]

Mozilla Firefox 49.0 apparently ended FF support for x86 versions that lack SSE2.
SSE2 was introduced in the Pentium 4 (2001). AMD processors followed in 2003.

See:
https://bugzilla.mozilla.org/show_bug.cgi?id=1271759#c0
https://bugzilla.mozilla.org/show_bug.cgi?id=1271794#c9

Rationale for requiring SSE2:
https://chuttenblog.wordpress.com/2016/ ... fox-users/

[SomeGuy]

That is about as poor rational as they used when they removed drop down menus. Just because usage numbers are small doesn't mean a feature is unimportant.

If they designed cars, they would completely remove the hazard lights because "no one uses them". (On my car the switch for that is a big honking button right in the middle of the dash where I can hit it easily if needed)

They also don't consider that people with non SSE2 CPUs might also be smart enough to disable their phone-home stuff. Or that those needing non SSE2 functionality don't happen to do much browsing on those systems, but still really really need the browsers to be there.

At the rate they are going, Mozilla should just pull the plug, go home, and stop supporting anybody at all.

[Black_Fox]

smart enough to disable their phone-home stuff

In my own personal experience, we pat ourselves on our backs for being smart and disabling this phone-home stuff... then we complain about new changes made - based exactly on this data we opted out of...

And frankly, if those people use 10+ years old CPUs and refuse to upgrade, what's the issue with them having to refuse to upgrade Firefox beyond v48?

[dizt3mp3r]

ReactOS would be happy to have the usage numbers that .5% of users that firefox are abandoning, represent. It is a huge number of users.

You should never alienate any part of your userbase, an alternative in functionality should be provided even if it is only an older version of FF that will keep running,

[Black_Fox]

That's why I ask, what's wrong with staying on Firefox 48?

[dizt3mp3r]

The usual.

[Z98]

I found the rationale pretty reasonable, there comes a time when you have to decide whether the effort needed to support older hardware is worth it. It is impossible to satisfy everyone, and trying will just result in a compromise that no one likes.

[dizt3mp3r]

Yes, I suppose that is the case. Especially if it slows Firefox and the other browsers have already taken the same route.

[MadWolf]

That's why I ask, what's wrong with staying on Firefox 48?

security holes are not patched

That is about as poor rational as they used when they removed drop down menus. Just because usage numbers are small doesn't mean a feature is unimportant.

If they designed cars, they would completely remove the hazard lights because "no one uses them". (On my car the switch for that is a big honking button right in the middle of the dash where I can hit it easily if needed)

They also don't consider that people with non SSE2 CPUs might also be smart enough to disable their phone-home stuff. Or that those needing non SSE2 functionality don't happen to do much browsing on those systems, but still really really need the browsers to be there.

At the rate they are going, Mozilla should just pull the plug, go home, and stop supporting anybody at all.

i abandoned firefox when they change to the australis interface and i started to use Pale Moon and Pale Moon needs a 7th or later generation CPU with SSE2 support.

[erkinalp]

Just compile it yourself. You can thusly target any hardware.

[Alahndro]

I've switched to Pale Moon over a year ago. As the standard version is combiled for SSE2, they also provide special builds for SSE and IA-32-only.

[Black_Fox]

That's why I ask, what's wrong with staying on Firefox 48?

security holes are not patched

I would say that's a good point unless this was part of the article

Yes, they almost all are on Windows XP.

The users don't care about the security holes anyway if they stay on XP... unless there are system updates for some 5.1 kernel Windows SKU still being released that can be pulled in by using a registry tweak.