wishlist for TCL16

[GNUser]

Hi curaga. If possible, here are 3 features I think would be nice to have in TCL16:

* remove the useless /linuxrc symlink
* implement /usr merge
* turn on support for wifi usb adapters with rt53XX and rt55XX chipsets, which have been working great with the in-kernel rt2800usb driver for years

For third item, the kernel config options are CONFIG_RT2800USB_RT53XX and CONFIG_RT2800USB_RT55XX

[curaga]

I don't really deal with the first two much anymore, that'd be Juanito's alley. For wifi modules, are they still marked experimental? Generally we follow that mark.

[Juanito]

I’d prefer to continue using /usr for everything in the base and /usr/local for extensions.

[GNUser]

Hi curaga. I'm not sure if still marked "experimental" today. But I and others have tested these chipsets and they among the few that are stable enough on linux to recommend without reservations. The driver is already provided by wireless-KERNEL.tcz. Only users trying to use these chipsets would be affected if these config options were enabled.

Hi Juanito. Absolutely. That's not what /usr merge is about. The merge is about:

* eliminating /bin (which becomes a symlink to /usr/bin)
* eliminating /lib (which becomes a symlink to /usr/lib)
* eliminating /sbin (which becomes a symlink to /usr/sbin)

/usr/local and everything under it would be unaffected.

/usr merge simplifies the file system, leads to fewer broken absolute paths (for example, an application can look for either /bin/foo or /usr/bin/foo and both would work) and would increase compatibility between applications and TCL in the long run.

All that being said, TCL is already the best distro as-is. This would be a minor upgrade.

[GNUser]

Juanito, if /usr merge feels like too radical a change, then please disregard. One of TCL's strengths is that it is conservative and avoids big changes.

[curaga]

Okay, if you've tested those drivers, they can be enabled.

[GNUser]

Okay, if you've tested those drivers...

I have tested them and they work great.

...they can be enabled.

Many thanks! It will remove a long-standing rock in my shoe (recompiling this kernel module with each TCL release).

[gadget42]

with respect to the /usr merge i have reviewed the following:

https://forum.tinycorelinux.net/index.php/topic,26879.0.html
https://wiki.debian.org/UsrMerge
https://www.debian.org/releases/bookworm/amd64/release-notes/ch-information.en.html#a-merged-usr-is-now-required
https://www.freedesktop.org/wiki/Software/systemd/TheCaseForTheUsrMerge/
https://fedoraproject.org/wiki/Features/UsrMove

Understanding the bin, sbin, usr/bin , usr/sbin split - Rob Landley, 2010
http://lists.busybox.net/pipermail/busybox/2010-December/074114.html

and i also support Bruno's wishlist (as long as Juanito approves, naturally)

[GNUser]

    [Edit]: Continued from here https://forum.tinycorelinux.net/index.php/topic,27363.msg176504.html#msg176504  Rich

Hi Rich. What curaga was saying is that he doesn't want to move anything out of /usr/sbin into /usr/bin. I agree with him. That's called sbin merge and is not what polikuo and I are proposing.

usr merge is about doing away with the pointless splitting of some things into /usr/bin and other things into /bin, for example. This serves no purpose other than to break things (e.g., an application looks for /usr/bin/foo but cannot find it because on that distro it happens to be /bin/foo). usr merge forever fixes this kind of breakage while also simplifying the file system. It's win-win.

[GNUser]

Hi Rich. Please feel free to move these posts from this thread to the "wishlist for TCL16" thread here:
https://forum.tinycorelinux.net/index.php/topic,27276.0.html

After all, /usr merge has nothing to do with provides.sh (other than both being on TCL16 wishlist)

[Rich]

Hi GNUser

What curaga was saying is that he doesn't want to move anything out of /usr/sbin into /usr/bin. ...

Sorry, I missed that part.