The MkLinux Journal
 The MkLinux Journal
Think Differently. Think MkLinux
Home | About | News | FAQ-O-Matic | Getting Started | Using MkLinux | Advanced Admin

Current MkLinux Project Status, a.k.a. How's it cooking?

This article is an attempt to describe the status of the MkLinux project as a development team. Eventually, other articles will discuss the possible organization of MkLinux as a legal organization for purposes of donations, fund raising, equipment loaning, etc. This is the first in what promises to be a series of articles on the MkLinux web site. For more information, check out http://www.mklinux.org

Performa Support:

Issues related to compiler and toolchain may be causing problems. Either that or fixes for G3 upgraded x100's may be to blame for splash screen freezes. After this is solved, much interrupt handling support is done, with some details remaining. I _think_ the hardware addresses for the onboard IDE are correct, not sure about 53C94 SCSI. Video addresses should be correct. I'm not sure about ADB (Cuda). I think serial info is correct, but again, not sure, and this could cause the splash screen problem, too.

Overall status: medium rare.

PDM floppy support:

Can now get as far as reading some sort of track headers or something (not sure about the terminology). No actual data appears to be being read.

Overall status: medium well.

NuBus card support:

NuBus video cards are, in some cases, supported, even though there was previously no framework for handling any other NuBus cards. A framework is being constructed, and many of the machine-dependent parts are coded in rough form. For example, the code to detect that the machine has NuBus support hardware (BART controller) appears to detect the chip and read its version code correctly. It is unclear whether the code will work correctly at detecting the absence of the chip.

There is still no high-level NuBus support code. This includes, but is not limited to, code for detecting a device in a NuBus slot and copying information about that card into an in-kernel table (although the table exists in a rough form, the field list is by no means inclusive or final).

Until that code is written, it is not reasonable to actually begin construction of drivers based on the NuBus framework.

Overall status: medium rare.

PlanB video input support:

PCI machines should have video in a window, no capture, as I understand it. I have not tested it, and am not developing that in any way. PDM machines with AV cards will need some additional attention (obviously).

Overall Status: medium.

iMac support:

BMac+ ethernet is probably finished. I've patched the pmap code to do the right thing, I think. Hopefully, everything should work except USB. However, in the several weeks since I posted a request for testers, I have not received any replies, I don't think, so the actual level of support is as yet unknown.

Overall Status: medium well on the outside, but may be bloody in the center.

B&W G3 support:

Only speculation other than the iMac support fixes. As of today, there's a developer who is known to have one, so maybe this will finally go somewhere.

Overall Status: raw, just placed in a preheated oven.

HPPA support:

From all accounts, stronger than the Intel port by a large margin. Still not up to par with the PowerMac port.

Overall Status: Apparently medium to medium-well, but left out to spoil on OSF's table. Almost past time to put it in the CVS freezer.

Intel support:

Supported in the basic configurations only and not kept up to date. This port needs a real shot in the arm. Hopefully, an expanded web presence and the availability of a CVS source tree will provide much-needed support for this port. We need to be vigilant when porting drivers to consider making them compatible with the Intel port if possible. I'm not sure how feasible this is, or what it will involve, though.

Overall status: half-baked.

Other Status Notes

The development pages need a few tweaks, especially in light of hopefully expanding the site to include the Intel and HPPA ports.

Every area of MkLinux development needs volunteers. It's just a question of which areas need volunteers most. :-) The area that needs work that would be easiest to do for someone without a lot of driver experience would be porting ethernet drivers. We already have a mostly working tulip driver, but there are other cards (3Com comes to mind) that aren't supported yet, and won't be unless someone takes time to port a driver in. Ethernet drivers are fairly similar between NetBSD and MkLinux. For porting information, look at if_tulip.c and compare it to its parent, if_de.c under NetBSD.

The timing between updates is generally dependent on the number of changes since the last update. If there's something major like floppy support, I'll often have an official kernel/server pair out in a couple of weeks after some testing. If the only thing happening is a few people updating Linux server versions from one pre-patch to another, it could be a while between releases.

Right now, I'm about to move to California (leaving Sunday, arrive Friday to a large amount of email, I suspect), so I won't release a kernel/server pair until I'm at least half-way situated with some semi-regular net access just because I don't want to deal with a bunch of bug reports while I'm hauling heavy machinery.

That's a brief summary of the MkLinux project status as of 05/27/1999 to the best of my knowledge.

David Gatwood
dgatwood@mklinux.org


webmaster


Home | About | News | FAQ | Getting Started | Using MkLinux | Advanced Admin

Contact the web team at mklinux-www@mklinux.org.

Copyright (©) 2001 MkLinux Developers Association. Permission to freely distribute this document is hereby granted, provided that it is distributed at no charge other than normal connect-time charges. All rights other than those specifically granted above are reserved.