Re: Outstanding pre-production issues and next dev. cycle...

Alan Cox (alan@cymru.net)
Wed, 28 Feb 1996 10:19:02 +0000 (GMT)


> It looks like we're getting very close to a production release, given
> the apparent stability of 1.3.68. So, an obvious question comes to
> mind: What priority bugs remain in 1.3.68? Does anyone have a
> complete list of what needs to be done before we call it 1.4/2.0?

Sorry we are NOWHERE near it. 1.3.69 is still very wobbly in places.

> 1) Performance:
> Can we get more performance out of networking?

Undoubtedly. We should be able to fill a 100Mbit/second lan by rewriting the
tcp_input/tcp_output queue handlers and adding proper header prediction.

> 2) Modularity
> Are there any drivers that could be made to work as modules but
> aren't?

IPX, AX.25, NetROM, Appletalk

> Are there any deficencies in the current module scheme?

Lack of documented interfaces.

> 3) Standards compliance
> Where does linux deviate from POSIX? Can/should these areas be
> rectified?

Not much. If you want to buy us a all a pile of Posix compliance tests (ie
you are _very_ rich).

> Others?
POSIX.12 once its a real stable spec for sockets.

> 4) Other
> BSD ffs?
This is a real now project. Getting SparcLinux mounting Sun FFS would be
very nice. As to BSD ffs see the ufs patches on tsx-11 as well. Not up to
date, read only but a start.
> linux able to run BSD binaries?
It can already run some FreeBSD binaries. If you are interested in that have
a look at the BSD emulation bits in the iBCS2 code and fill them out. There
is also much to borrow from the SparcLinux project.
> anything else?
Ermm Ability to mount V7 and CP/M file systems, run PDP/11 V7 binaries, run
on an 8086 ;)

Alan