X.Org Server No Longer Allowing Byte-Swapped Clients By Default

[phoronix]

Phoronix: X.Org Server No Longer Allowing Byte-Swapped Clients By Default

Following the recent discussions around Fedora planning to disable byte swapped clients support for the X.Org Server in order to close another "large attack surface" with the aging X11 server codebase, the upstream X.Org Server has now dropped this support by default...

https://www.phoronix.com/news/X.Org-No-Byte-Swapped-Clients

[kpedersen]

I wonder if the modernized and combined Xenocara will follow suite. This project started out not as a full fork but merely a way to undo the messyness introduced by xorg a while back when it went modular. However it does tend to choose different defaults in the build configuration.

[skeevy420]

If you do find a need for say a bing endian X11 client to connect to a little endian X.Org Server

I know MS has been making some inroads round these parts...

Michael

[Danny3]

No problem!
KDE Plasma works great on Wayland for 2-3 years already.

[cb88]

No problem!
KDE Plasma works great on Wayland for 2-3 years already.

Ironically KDE Plasma is a desktop... not an end user application, of which there are mountains of legacy x11 applications.

[bug77]

I like how Michael loves to underscore X11's age with phrasing like "the aging ... codebase". Ironically, code that lives as long as X11 does is some of the most successful code on the planet. Sure, it's not modern anymore, sure, few people know how to maintain it today. But who here wrote code that lived for half of X11's lifetime? And if you did, how much of all the code you wrote fits into that category?

That said, LE/BE and the various ways to end a line of text still look like huge problems the industry has inflicted on itself for no good reason.

[andyprough]

Which moronic hackers are "using IBM s390x or PowerPC 64-bit for remote X11 use of graphical applications" for their hacking? This doesn't make sense as a "security" concern, I can't imagine anyone doing their black-hat hacking this way. Maybe I'm just naive?

Or is this just more of the "security theater" that accompanies any discussion of Xorg? More press releases from the wayland crowd - "LOOK at all the GHASTLY stuff we are SAVING you from!!! Now give us more money!"??? I could understand it from that perspective, which is probably incredibly common in tech.

[kpedersen]

No problem!
KDE Plasma works great on Wayland for 2-3 years already.

Do the Plasma or existing Wayland developers even know what a "Byte-Swapped Client" is?

[oiaohm]

That said, LE/BE

There is a historic good reason for Little endian and Big endian.
If you are on a serial port and are receiving stream of bytes differences following come into play:
Input validation is normally greater/less than operations. Big endian can solve sooner because the byte will be turning up in the right operations.
Math operations like addition, subtraction, divide multiplication little endian has the bytes turn up in the right order to start the operations from.

Patent over MMU memory regions being defined endian has expired so we could have application transparent endian conversion if common CPU vendors decide to add it to the MMU..

[oiaohm]

Do the Plasma or existing Wayland developers even know what a "Byte-Swapped Client" is?

Yes the core Wayland protocol developers do due to being X11 developers prior. They made Wayland protocol local only to avoid this problem.