mklibs -l option (force libs)
Hi Bastian, I've been looking at mklibs revision 49561 where you added the -l option. I found a comment that this had something to do with cdebconf - presumably to support dlopening of /usr/lib/cdebconf/frontend/*.so. But all that's going to happen is they'll be stripped, so I can't figure out the point. Do you remember what the option is needed for? My assumption above might be plain wrong. I ran into the fact that it won't work with a foo.so file, because foo.so does not match so_pattern. -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#484366: rootskel: cttyhack requires the serial console to be /dev/ttyS0
On Wed, Jun 04, 2008 at 03:21:41PM +0200, Frans Pop wrote: On Tuesday 03 June 2008, Daniel Jacobowitz wrote: I tried to install Debian on Versatile/PB using qemu and the armel installer (Lenny beta 1). It is quickly stuck complaining cttyhack: '/dev/ttyS0': No such file or device. What _exact_ qemu command line did you use? How did you connect to the ttyAMA0 serial port? Using minicom or something on the same host system, from a different system or what? Note that D-I works just fine if you just let it run inside a qemu window. I'm not yet convinced that there is either an installer bug or a busybox bug here. If you just run qemu, and SDL was available when it was built, you get a graphical window including qemu's console and a real TTY. So there's no serial console involved at all. My qemu was built without SDL, and I ran it with -nographic. Either of those is sufficient to end up with a serial console install, effectively headless. qemu's standard input and output are the serial port. The qemu command was: qemu-system-arm -m 256 -M versatilepb -kernel qemu-kernel-image \ -initrd initrd.gz -append mem=256M console=ttyAMA0 root=/dev/ram \ -hda qemu-disk-image -nographic -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#484366: rootskel: cttyhack requires the serial console to be /dev/ttyS0
Package: rootskel Version: 1.62 Severity: normal I tried to install Debian on Versatile/PB using qemu and the armel installer (Lenny beta 1). It is quickly stuck complaining cttyhack: '/dev/ttyS0': No such file or device. That's because the serial port on Versatile is named ttyAMA0. I ran qemu with -nographic and put console=ttyAMA0 on the command line. This causes two problems: 1. Initially there is a /dev/ttyS0 but it is device 4/0. ttyAMA0 is 204/64. I could recreate it from the second debug shell or by appending to init-udev-devices. 2. When udev runs, it recreates /dev. udev successfully creates /dev/ttyAMA0 with the proper major/minor. But it naturally does not create /dev/ttyS0. So that doesn't work either and I created the symlink by replacing 'exit 0' with 'ln -s ttyAMA0 /dev/ttyS0' at the end of S02udev. Frans warns that finish-install's 90console script may also have a problem. I haven't gotten there yet. I don't know a clean fix for this problem. It could be detected by checking /proc/devices, most likely - there's ttyAMA but no ttyS. On a related note, I thought cttyhack was supposed to give things enough of a controlling terminal to use job control. But cttyhack /bin/sh still complains that there is no job control... -- System Information: Debian Release: lenny/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 2.6.25-rc9 (SMP w/2 CPU cores; PREEMPT) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/bash -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#469070: mklibs and libthread_db
FWIW, I think adding a libthread_db-specific patch to mklibs (Joseph's part 4) is reasonable. It's part of glibc, for which there are already local hacks in mklibs. And it's a library with a particularly annoying interface, that relies on undefined symbols in the library being defined in the application. Any thoughts on applying the patch? -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: libc dlopening libgcc on arm
On Wed, May 30, 2007 at 11:49:54AM +0200, Bastian Blank wrote: On Wed, May 30, 2007 at 11:31:35AM +0200, Bastian Blank wrote: On Wed, May 30, 2007 at 11:19:38AM +0200, Aurelien Jarno wrote: What I don't understand is why /sbin/init on arm now uses threads? It is /bin/sh, which should be busybox. And it does not use pthread at all. It may be usefull to get a backtrace. Yes. I suspect the relevant wrapped functions are called more frequently on ARM. This is ARM old-ABI, so it uses SJLJ exception handling; that means the _Unwind_* functions are called around functions that might throw, not just when an exception is thrown. Note, that was exceptions, not threads. This code is in libc and libpthread. -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#410065: installation-report: d-i rc1, NSLU2: minor network-console and boot issues
-Size: 28 Maintainer: Debian Install System Team debian-boot@lists.debian.org Version: 0.10 Depends: cdebconf-udeb, apt-setup-udeb, di-utils (= 1.20) Description: Select and install packages Installer-Menu-Item: 70 Package: scsi-core-modules-2.6.17-2-ixp4xx-di Status: install ok installed Priority: standard Section: debian-installer Installed-Size: 184 Maintainer: Debian Install System Team debian-boot@lists.debian.org Version: 1.3 Depends: kernel-image-2.6.17-2-ixp4xx-di Provides: scsi-core-modules Description: Core SCSI subsystem This package contains the core SCSI subsystem for the Linux kernel. Package: tzsetup-udeb Status: install ok unpacked Priority: standard Section: debian-installer Installed-Size: 292 Maintainer: Debian Install System Team debian-boot@lists.debian.org Version: 1:0.12 Depends: cdebconf-udeb, localechooser Description: choose timezone Installer-Menu-Item: 62 Package: usb-storage-modules-2.6.17-2-ixp4xx-di Status: install ok installed Priority: standard Section: debian-installer Installed-Size: 136 Maintainer: Debian Install System Team debian-boot@lists.debian.org Version: 1.3 Depends: scsi-core-modules-2.6.17-2-ixp4xx-di, usb-modules-2.6.17-2-ixp4xx-di Provides: usb-storage-modules Description: USB storage support This package contains the USB storage driver for the Linux kernel. Package: user-setup-udeb Status: install ok unpacked Priority: standard Section: debian-installer Installed-Size: 336 Maintainer: Debian Install System Team debian-boot@lists.debian.org Version: 1.7 Depends: cdebconf-udeb Description: Set up users and passwords Installer-Menu-Item: 63 -- System Information: Debian Release: 4.0 APT prefers testing APT policy: (500, 'testing') Architecture: arm (armv5tel) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.18-4-ixp4xx Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968) installation-report depends on no packages. Versions of packages installation-report recommends: ii pciutils 1:2.2.4~pre4-1 Linux PCI Utilities ii reportbug 3.31 reports bugs in the Debian distrib -- no debconf information -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#397091: debootstrap-udeb: Depends on libgcc_s.so.1 on MIPS
Package: debootstrap-udeb Version: 0.3.3 Severity: important The version of debootstrap-udeb in the archive depends on libgcc_s.so.1. This library is not included in debian-installer; as a result, d-i does not work on MIPS. I'm not sure how this happened; maybe it was built with a transient GCC bug and just needs to be rebuilt? -- System Information: Debian Release: 4.0 APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable') Architecture: amd64 (x86_64) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.19-rc1 Locale: LANG=en_US, LC_CTYPE=en_US (charmap=ISO-8859-1) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#396365: please add a gdb .udeb, for easier debugging inside d-i
On Tue, Oct 31, 2006 at 03:30:23PM +0100, Sven Luther wrote: So, i am not entirely sure how d-i handles the case at hand here, where libraries are part of the actual image, but symbols are removed during reduction, because they are not needed by the .udebs in the image, but they are needed by .udebs loaded later on. My impression is that d-i loads the full libraries later on, or something such. If they did that, you could just wget a full GDB binary. It doesn't need anything else in the package besides the executable. But I don't think d-i does what you describe (since I tried to use this approach for strace recently, and there were missing symbols in libc.so.6). If someone more familiar with d-i than I am can confirm that the udeb will be useful, I can try to get GDB to build one, but it seems like a very strange package to have a udeb. -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#396365: please add a gdb .udeb, for easier debugging inside d-i
On Tue, Oct 31, 2006 at 01:06:12PM +0100, Sven Luther wrote: A version of gdb as .udeb, which could easily be loaded inside the d-i would be very useful for developers. This would come in handy to debug various bugs which are otherwise hard to track, like crashes in the graphical installer, or problems like i am having right now, in parted/libparted, which only are apparent inside the d-i environment. Is this really a good idea? It seems like the wrong way round; you could manually load a full GDB and associated libraries, if you need it for debugging, but having it as part of a d-i image would e.g. change the required bits of libraries. Hoping for feedback. -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#385919: RFA: bogl -- Ben's Own Graphics Library - graphical terminal
Package: wnpp Severity: normal I request an adopter for the bogl package. The package description is: Ben's Own Graphics Library is a small framebuffer library, including basic widgets, support for text in multiple languages, and mouse handling. . This package contains bterm, a UTF-enabled framebuffer terminal. I haven't had a setup where I could test bogl in a long time, nor have I worked with d-i (the only significant user of bogl). Ideal would probably be someone who wants to assist with d-i. I can provide a Subversion repository containing the recent history of bogl to any adopter. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable') Architecture: amd64 (x86_64) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.18-rc4 Locale: LANG=en_US, LC_CTYPE=en_US (charmap=ISO-8859-1) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#359062: debian-installer: bterm is not (yet) accessible for brltty
On Tue, Mar 28, 2006 at 12:57:32AM +0200, Samuel Thibault wrote: In its own memory, which I proposed (in the very first mail of the discussion) to export somehow to brltty, via some /dev/bterm mmap()ed file for instance (very easy to implement in bterm's code: a mere open() then mmap() instead of the current malloc()). Sounds plausible... Unfortunately, I am no longer able to maintain bogl or bterm; I was hoping for someone to adopt it, but he hasn't had time either. Don't suppose someone on the CC list wants to fix this bug and adopt a needy package? -- Daniel Jacobowitz CodeSourcery -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#277092: amd64-libs: Removing package fails and breaks system
On Tue, Oct 19, 2004 at 11:57:42PM +0200, Frans Pop wrote: As it happens I'm a member of the d-i team (actually I was testing d-i when I noticed this problem). Oh good :-) In line with Debian policy, d-i installs _every_ package available for an architecture that a) is in base OR b) has priority 'standard' or higher. As lib64stdc++6 is both 'base' and 'important', it gets installed and pulls in amd64-libs and lib64gcc1 because of it's dependencies. lib64gcc1 would get installed in it's own right as it is priority 'required'. I have no idea how the sections and priorities for these packages have been decided, but to me they seem very high. Especially as I can see no other packages in base or with high priorities that depend on these libs. Moving lib64stdc++6 to libs and giving both lib64stdc++6 and lib64gcc1 priority 'optional' should fix this. You'll have to talk to Matthias about that. I'm trying to upload the amd64-libs fix. I'm having a fistfight with the upload queue, which insists that its checksum is wrong. -- Daniel Jacobowitz -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#237075: Bug not related to Sparc64... Maybe related to Creator?
reassign 237075 kernel-image-2.4.26-sparc64 severity 237075 important merge 237075 245620 thanks On Fri, Mar 19, 2004 at 05:20:15PM +0100, Thomas Poindessous wrote: On Fri, Mar 19, 2004 at 03:18:47PM +0100, Reinhard Tartler wrote: This image still has the problem with unable to open initial console so the option devfs=mount was needed. Not so in Jeff Bailey's build. bterm keeps on segfaulting with this image too. then you indeed found a bug in bterm I think this is the same as 245620. It's a kernel bug, not bterm. -- Daniel Jacobowitz -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: bogl_write_font failed
On Sun, Mar 07, 2004 at 09:47:01PM -0500, Joey Hess wrote: Autobuild of d-i on ia64 is failing like this, do you know what the message means? bdftobogl -b ./tmp/netboot/unifont.bdf tmp/netboot/tree/unifont.bgf.tmp bogl_write_font failed make[7]: *** [tmp/netboot/tree/unifont.bgf] Error 1 It means I had a brain fart. Thanks for the pointer. name_len = strlen (font-name) + 1; name_pad = sizeof (int) - (name_len % sizeof (long)); if (name_pad == sizeof (long)) name_pad = 0; er... that first (int) should be (long), or that quantity may end up negative on ia64. I'll upload a fix now. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: bterm font reloading patch
On Tue, Jan 20, 2004 at 06:49:36AM -0800, Matt Kraai wrote: On Tue, Jan 20, 2004 at 08:55:06PM +0900, Kenshi Muto wrote: At 19 Jan 04 00:23:15 GMT, Matt Kraai wrote: The attached patch should make bterm reload the font when it receives a SIGHUP. Would someone please test it? Thank you Matt!! It works perfectly on my test environment. Great. Thanks for trying it out. I strongly wish this patch will be merged into bogl-bterm udeb. Dan, would you mind if I made an NMU? And I noticed - - bterm-unifont uses awk for HUP signal, but awk doesn't exist on 1st stage. Does this need to rewrite by sed or another tool? I've changed it to use pidof instead in CVS. I'll upload it once the freeze is over. Let me try to take care of it tonight. Importing NMUs into my CVS is a nuisance. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Processed: bterm font problem
[CC me please.] On Sun, Jan 18, 2004 at 11:48:08AM -0800, Debian Bug Tracking System wrote: severity 219333 critical Bug#219333: cdebconf: Letter u grave not displayed Severity set to `critical'. You're trying to do something no terminal emulator in the world will let you do. I don't think that deserves to be a critical bug on the terminal emulator. I still think that bogl-bterm has no business reloading the font. And it seems to me that it would be a good feature for main-menu to be able to move into a new terminal, e.g. to switch from a text to graphical installer after a graphical installer is available. This bug will be as easily and more usefully fixed in main-menu. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#221538: bogl-bterm-udeb: var/lib/dpkg/info/bogl-bterm-udeb.md5sums unwanted
On Sat, Jan 10, 2004 at 10:06:21AM +0100, Petter Reinholdtsen wrote: This bug is the only bug that keep this package from propagating into testing. This is the testing status reported at URL:http://packages.qa.debian.org/b/bogl.html. - 65 days old (needed 10 days) - bogl-bterm (alpha, arm, hppa, i386, ia64, m68k, mips, mipsel, powerpc, s390, sparc) is buggy! (1 0) - Not considered I will upload a NMU version this weekend unless you as the maintainer do it first. Apologies for forgetting about this bug. However, if you are going to make a pretense of warning maintainers, kindly do not do so and then upload an NMU nine minutes later. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
bogl-bterm: font reloading
Can you just restart bterm? Yes, I realize that's got some issues since you're running inside of it. But it's really not set up to re-parse the font. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Processed: reassign 215160 to bogl-bterm
On Sat, Oct 11, 2003 at 06:18:05PM -0500, Debian Bug Tracking System wrote: Processing commands for [EMAIL PROTECTED]: reassign 215160 bogl-bterm Bug#215160: redraw problems with framebuffer and vt switching Bug reassigned from package `cdebconf' to `bogl-bterm'. I don't even have a machine with working framebuffer right now. 2.6 ate my aty128fb. I applied the optimization patches because they looked sane and I was told they worked. Is there anyone who can debug this or do they have to be reverted? -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Always use UTF-8 when running base-config?
On Sun, Sep 28, 2003 at 02:05:51PM +0200, Martin Sjögren wrote: On Sun, Sep 14, 2003 at 05:31:13PM +0100, Edmund GRIMLEY EVANS wrote: I applied Alan Cox's patches to bogl-0.1.12 and found that bogl-bterm-udeb's bterm scrolled much faster. I have forwarded the patches to Daniel Jacobowitz. Anyone else who wants them, let me know. Any progress on this? Dan? Where can I find those patches, Edmund? I'll upload it to unstable today after some local testing. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Processed: reassign
[cc me] Any particular reason? I believe this is fixed in bogl since Mar. 17th, but I want to make sure you don't know something I don't before I close it. On Sun, Mar 31, 2002 at 04:48:08PM -0600, Debian Bug Tracking System wrote: Processing commands for [EMAIL PROTECTED]: reassign 138653 bogl-bterm Bug#138653: boot-floppies: Language chooser '#' on right hand side looks funny Bug#127517: boot-floppies: # instead of a scrollbar Bug reassigned from package `boot-floppies' to `bogl-bterm'. quit Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#134014: bterm causes split screen
On Sun, Mar 17, 2002 at 11:37:59PM +, Philip Blundell wrote: On Sun, 2002-03-17 at 20:00, Chris Tillman wrote: The patch works! Thanks Dan! It also helps the appearance because it's sized properly now. Adam, we need a new bterm to come in for the next build to fix this. If we're going to have a new bterm anyway, I would be inclined to put the patch from #138653 in too. It does make the dialog boxes look quite a lot prettier. Yeah, I intend to. Thanks. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#137734: Missing filesystems for 2.4 kernel on testing cds
On Sun, Mar 10, 2002 at 03:14:49AM -0500, Colin Walters wrote: Package: kernel-image-2.4.16-newpmac Severity: serious [ Dan: basically the gist of this is that the newpmac kernel needs to have NFS and iso9660 built in. Also, (on an unrelated note), we need MESH scsi built in. ] OK. As soon as I can build kernel packages again, I'll upload 2.4.18. This bug was stunningly timely, in that I'd forgotten again that newpmac existed... must have grabbed the wrong source tree. I'm having some trouble with depmod right now. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Floppy boot for powerpc: add video=ofonly?
On Fri, Feb 15, 2002 at 10:22:35PM -0700, Chris Tillman wrote: Many oldworld powerpc computers need the video=ofonly boot argument in order to use the installer. The hfs-boot-floppy image we currently provide does not supply this argument. I used a binary editor (beav) to add video=ofonly into the miBoot System.bin source file in boot-floppies, and verified that on my oldworld machines the modified file enables the floppy to work, where before the monitor couldn't sync up. (I replaced spaces in the file, and didn't change the file size). It's my understanding that this argument would do no harm for any powerpc system, at least if it's just the installer boot we're talking about. If that's right, I'd like to commit the change to cvs so the hfs-boot-floppy image has that argument. Many systems would be enabled for floppy usage that otherwise wouldn't work. Sounds like a good idea. Go ahead; I don't remember any reason not to do this. (that's why I put the spaces in there in the first place!) -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
3.0.19 install issues
I just ran through a clean Woody installation on 3.0.19; here's my observations from it. Hope these are useful. Please CC me on any comments; I no longer read debian-boot. The machine was a Sony Vaio R505JL. It has a roughly firewire-driven expansion bay, but the BIOS successfully masquerades the floppy drive as a real drive, so I was able to use a normal rescue/root set. The CD (DVD) drive wasn't seen, unsurprisingly. I picked the udma100-ext3 flavor; I wanted ext3 for my root FS. I think a fair number of these problems are related to that. In particular: - Mount Previously Initialized Partition failed to mount the FAT32 filesystem. The error was No such file or directory. Not so useful. - Choosing harddisk as an installation medium fails. Error was No such device. No idea. - Choosing mounted defaults to /instmnt but is actually RELATIVE TO /instmnt/. This results in a lot of confusion and some ugly doubled slashes. - Installing from mounted failed to copy the kernel off the rescue disk. The loop device was successfully configured (I checked by mounting it on VT2) but the mount failed again with No such device. del_loop failed repeatably, also - negative return code from LOOP_CLEAR_FD. - bterm glitch: Last line of the screen remained colored. I'll investigate this in a few days. - Some error messages are still written to the screen where dialog is running. I don't know which, since they flash by pretty quickly. My guess is right after the installation failed to read kernel off of rescue.bin, above. - Network install, conveniently, did not have most of these problems. - Perl gave me a lot of locale warnings. These are because LANG is set to English (QWERTY/US). Someone's picked a bad name for a variable, I think. Minor nits: Tasksel still wraps top to bottom in its scroll list, which I find massively unintuitive. Debootstrap alphabetizes downloaded packages by package name rather than by source package / URL; that's just personal preference for organization. Any of these I should file particular bug reports on? -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
More 3.0.19 woes
... and I thought I was done... Base-config somehow got run a second time, after a number of packages had configured. I didn't see any errors, but some (like exim) just never came up. Pretty good otherwise! -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
One last issue
Interestingly, the installer also decided to remove PCMCIA support. I don't know why; the module loaded just fine, and it seems to work. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: powerpc b-f build compiles the kernel
It's not so mysterious. Actually read the part I left below :) On Sun, Jan 06, 2002 at 09:27:05PM -0500, Colin Walters wrote: # Fix .config for ramdisk rooting and kernel ip autoconfiguration. # This image is 1440 KB and is suitable for CD/net booting. sed 's:# CONFIG_CMDLINE_BOOL is not set:CONFIG_CMDLINE_BOOL=y \ CONFIG_CMDLINE=root=/dev/ram ip=off:' \ $builddir/usr/src/kernel-patches/powerpc/config.prep .config.param # Enable kernel ip autoconfiguration sed 's:# CONFIG_IP_PNP is not set:CONFIG_IP_PNP=y \ CONFIG_IP_PNP_ENABLE=y \ CONFIG_IP_PNP_DHCP=y \ CONFIG_IP_PNP_BOOTP=y \ CONFIG_IP_PNP_RARP=y:' \ .config.param .config.ipauto # Enable nfsroot sed 's:CONFIG_NFS_FS=y:CONFIG_NFS_FS=y \ CONFIG_ROOT_NFS=y:' \ .config.ipauto .config # Build CD/net kernel boot image make oldconfig make dep make clean make zImage.initrd cp ./arch/ppc/boot/zImage.initrd $top/bootprepfull.bin # Copy config.prep-.config and fix up for floppy/ramdisk rooting. # This image is =1440 KB and is suitable for floppy booting. sed 's:# CONFIG_CMDLINE_BOOL is not set:CONFIG_CMDLINE_BOOL=y\ CONFIG_CMDLINE=root=/dev/fd0 load_ramdisk=1:' \ $builddir/usr/src/kernel-patches/powerpc/config.prep .config # Build floppy-sized kernel boot image make oldconfig make zImage cp ./arch/ppc/boot/zImage $top/bootprep.bin -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#122738: boot-floppies can't handle devfs
On Sat, Dec 08, 2001 at 01:33:05AM -0500, Colin Walters wrote: Adam Di Carlo [EMAIL PROTECTED] writes: Unfortunately boot-floppies doens't currently support devfs. Colin found an issue with parsing of /proc/partitions, but I'm afraid that's probably just the tip of the iceburg. Getting devfs support into boot-floppies would set us back at least 2 months. Just to clarify, this issue appears even devfs=nomount, because the /proc/partitions format changes when devfs is compiled into the kernel, even if it isn't mounted. I'm pretty strongly opposed to this idea. Could I persuade you to at least fix the /proc/partitions bug and see what happens? It's not like this bug is a surprise; I remember hearing about it six or more months ago. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#122738: boot-floppies can't handle devfs
On Sat, Dec 08, 2001 at 01:17:33PM -0500, Adam Di Carlo wrote: Daniel Jacobowitz [EMAIL PROTECTED] writes: I'm pretty strongly opposed to this idea. Why? Because devfs is a useful feature, and I get the impression from debian-powerpc that many people are using it on PowerPC. I'm therefore reluctant to remove it from the stock kernel images. Could I persuade you to at least fix the /proc/partitions bug and see what happens? It's not like this bug is a surprise; I remember hearing about it six or more months ago. We can try to fix this bug, sure, but as I said, it's just the top of the iceburg I'm sure. I'm not so confident. Particularly if you either include devfsd or do not mount devfs, it should not change much beyond the occasional /proc message. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: .config for boot-floppy-hfs.img
On Wed, Sep 26, 2001 at 02:39:44PM +0200, [EMAIL PROTECTED] wrote: no I forgot, I am going to try with it. But why the kernel which is in the kernel-image-2.2.19-pmac package has a size of 2.7Mo, and the kernel I generate from its .config has a size of 1Mo? That's not right. Are you using the same kernel type? The 2.7Mb one is 'vmlinux', at the top of the build directory, not a compressed image. And an other question, what is the difference between a vmlinux kernel and a vmlinux.coff kernel? COFF is another bootable format. It's what the floppies need. It'll appear under arch/ppc/ after you build. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: .config for boot-floppy-hfs.img
On Wed, Sep 26, 2001 at 01:20:08AM +0200, [EMAIL PROTECTED] wrote: On 11-Sep-2001 Daniel Jacobowitz wrote: On Tue, Sep 11, 2001 at 03:45:37PM +0200, Damien Morel wrote: Hi, do some of you know where can I find the .config file used to generate the kernel on powerpc's boot-floppy-hfs.img (http://ftp.debian.org/debian/dists/woody/main/disks-powerpc/3.0.13-2001-08-2 5/powermac/images-1.44/boot-floppy-hfs.img to be exact). Thanks in advance. In the kernel-patch-2.2.19-powerpc or kernel-image-2.2.19-pmac packages. I tried with this .config but it doesn't seem to give the same kernel than the bootdisk's one. They have not the same size, and I can't boot with the kernel compiled with this .config on my Performa6400. You did also apply the patch from kernel-patch-2.2.19-powerpc, right? -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: mklibs.py, unresolved symbol
On Sun, Sep 16, 2001 at 05:25:56PM -0400, Adam Di Carlo wrote: Ethan, I was finding other problems with mklibs.py on powerpc, to wit, ld-linux.so.2 was used in a non-portable way. To get it to work on PowerPC, I had to use this patch. Is anyone interesting in fixing these problems? -- ...Adam Di Carlo..[EMAIL PROTECTED]...URL:http://www.onshored.com/ diff -u -u -r1.4 mklibs.py --- mklibs.py 2001/08/11 06:55:18 1.4 +++ mklibs.py 2001/09/16 21:18:49 @@ -352,7 +352,7 @@ base_name = so_pattern.match(library).group(1) # libc needs its soinit.o and sofini.o as well as the pic if base_name == libc: -extra_flags = find_lib(ld-linux.so.2) +extra_flags = find_lib(ld-2.2.3.so) You probably want to check by architecture what to use instead of hardcoding 2.2.3... It's not much work - just: objcopy -j .interp -O binary /bin/ls foo cat foo (being aware that foo will be a null, rather than newline, terminated string). ld-linux.so.2 is standard on some architectures, and ld.so.1 is standard on others. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: .config for boot-floppy-hfs.img
On Tue, Sep 11, 2001 at 03:45:37PM +0200, Damien Morel wrote: Hi, do some of you know where can I find the .config file used to generate the kernel on powerpc's boot-floppy-hfs.img (http://ftp.debian.org/debian/dists/woody/main/disks-powerpc/3.0.13-2001-08-25/powermac/images-1.44/boot-floppy-hfs.img to be exact). Thanks in advance. In the kernel-patch-2.2.19-powerpc or kernel-image-2.2.19-pmac packages. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: build failure on ia32
[wow, can you tell I'm cleaning up a mail backlog?] On Tue, Sep 04, 2001 at 06:48:13PM -0600, Erik Andersen wrote: On Tue Sep 04, 2001 at 04:18:42PM -0800, Ethan Benson wrote: yup but again i still need busybox reduction since none of this reiser stuff is on the powerpc disks, and i still have a problem with space there. Hmm. Then something must be eating up all that space -- powerpc binaries are not that much bigger then x86 binaries... Is library reduction not working on powerpc? No, PowerPC needs extra tools. We have hfsutils and yaboot and quik; I guess these exceed the size of lilo and whatever else x86-only things we need, that's all. hfsutils is quite big. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: build failure on ia32
On Sun, Sep 09, 2001 at 03:45:48PM -0800, Ethan Benson wrote: On Sun, Sep 09, 2001 at 03:13:14PM -0400, Daniel Jacobowitz wrote: Hmm. Then something must be eating up all that space -- powerpc binaries are not that much bigger then x86 binaries... Is library reduction not working on powerpc? No, PowerPC needs extra tools. We have hfsutils and yaboot and quik; I guess these exceed the size of lilo and whatever else x86-only things we need, that's all. hfsutils is quite big. no i ditched hfsutils and yaboot a lng time ago. now ybin/mkofboot/yabootconfig on the root disk are a tiny wrapper that sets LD_LIBRARY_PATH and runs ybin from /target/... /usr/lib/yaboot is a symlink to /target/usr/lib/yaboot that was the only way especially given the discusting bloat of current yaboot binary. Oh, OK, I'm completely wrong again :) So much for that. I wonder why we're so much tighter on space, then. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: boot-floppies 3.0.13
On Thu, Aug 23, 2001 at 10:26:13PM -0400, Adam Di Carlo wrote: Ok, 3.0.12 was kinda a bust because the pcmcia problems were only fixed half-way, and I messed by not tagging for inclusion in the source packages some s390 stuff. Also m68k has 2.2.4, so that is also needed. Due to that, I'm building boot-floppies 3.0.13 now. Below is the changelogs. Quick turn-around, but how can that hurt, eh? Mmkay. Please don't build PowerPC until after dinstall tomorrow, I've uploaded fixed kernels, pcmcia and all. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: boot-floppies_3.0.12_i386.changes INSTALLED
On Tue, Aug 21, 2001 at 04:14:51PM -0400, Adam Di Carlo wrote: Daniel Jacobowitz [EMAIL PROTECTED] writes: Woah. OK, I've been letting this slide a little bit, but are you telling me that we need a new 2.2.19 kernel for the version of boot-floppies now in the archive? What must I do to get you to tell me these things BEFORE committing them, so that we could actually have working kernels? Or is that changelog entry not clear? I'll have to leave this to Ethan to answer. Dan, I do know that PowerPC can't build now due to lack of pcmcia-modules-2.2.19-pmac bug. Do you think we should NMU pcmcia-cs or somehow workaround that? Do you need my help on that front? Fix was installed today, actually. I'm not fixing some things about the build process of the modules for woody; I'll do to 2.2.x in unstable what I've done to 2.4.x, a little further down the line, which will make PCMCIA easier to build. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: boot-floppies_3.0.12_i386.changes INSTALLED
On Tue, Aug 21, 2001 at 02:32:35PM -0800, Ethan Benson wrote: On Tue, Aug 21, 2001 at 11:22:31AM -0700, Daniel Jacobowitz wrote: On Tue, Aug 21, 2001 at 01:43:09PM -0400, Debian Installer wrote: * Ethan Benson - PowerMac uses linux (read i386) keycodes now, use i386 keymaps. This still needs a fixed -pmac kernel, but console-data in woody requires that anyway. Woah. OK, I've been letting this slide a little bit, but are you telling me that we need a new 2.2.19 kernel for the version of boot-floppies now in the archive? What must I do to get you to tell me these things BEFORE committing them, so that we could actually have working kernels? Or is that changelog entry not clear? we either need a kernel with CONFIG_MAC_ADBKEYCODES=n or the user has to pass the following kernel argument: keyboard_sends_linux_keycodes=1 the 2.2.19 in woody is already broken because console-data has made this trasistion and is switching users keymaps and setting the sysctl (which won't last a reboot), it then tells them to use the above append= line or recompile thier kernel. this work has been going on for weeks now has been discussed on the list several times now. it needs to happen one way or the other, b-f is not the first thing to require it. As I said, please do bring this sort of thing to the attention of someone who can do something about it. I'm sorry for not following -boot religiously while work has been sucking my life and I've been busy reconstructing the entire rest of the powerpc port. Should I take this to mean I need to build new kernel packages, then? -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: boot-floppies_3.0.12_i386.changes INSTALLED
On Tue, Aug 21, 2001 at 03:10:15PM -0800, Ethan Benson wrote: On Tue, Aug 21, 2001 at 03:46:28PM -0700, Daniel Jacobowitz wrote: As I said, please do bring this sort of thing to the attention of someone who can do something about it. I'm sorry for not following -boot religiously while work has been sucking my life and I've been busy reconstructing the entire rest of the powerpc port. Should I take this to mean I need to build new kernel packages, then? yes a bug is already filed against kernel-image-2.2.19-pmac Feh. I'm sorry, Ethan, it was held up in a mail spool. I'll take care of it soon. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: pcmcia modules for powerpc not available
On Tue, Aug 07, 2001 at 10:19:43AM -0400, Adam Di Carlo wrote: Daniel Jacobowitz [EMAIL PROTECTED] writes: It's not in unstable or testing. Eek. How'd that happen?? I would guess pcmcic-cs was forced from unstable into testing, causing the older pcmcia-modules dpeneding on the older pcmcia-cs to be booted. I'll try to un-lag it later this week. Excellent. pcmcia-cs does not build. Bug filed; #108823. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Library reduction on Alpha (was: Re: Alpha: aboot rescue disk)
On Wed, Aug 08, 2001 at 06:01:27PM +0200, Goswin Brederlow wrote: the scripts first (I'll check them out of CVS probably tomorrow) to make sure they don't already try that :-) mklibs.py on the other hand includes each symbol only once. Theres a small hack to prefer libc symbols (althout not realy needed since libm seems to be after libc) but otherwise its a first come first use basis. Ok...this script probably does what I'm thinking about then. Is this one working well enough or do you want to shrink it even more? If we can, sure. But the only thing that I see left si to remove versioned symbols and leave only the right version or the minimal set of version required. But does that realy help? Its probably just some stubs that get removed by that. Actually, I'd bet it saves more than you think. However, like I've said, we don't need a profusion of library reducers. We should have at least one packaged independently of the boot floppies. I'm going to try to dig up and package MontaVista's (also written in python). It's quite complete (and strips out unneeded versions, IIRC). Currently the root.bin fits, so unless someone adds more the jobs done, but that just a lazy programmers attitude. :) I'm sure we can find things to add. We're an industrious bunch that way :) -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: libm on root-filesystem?
On Sat, Aug 04, 2001 at 04:12:41AM +0200, Falk Hueffner wrote: Adam Di Carlo [EMAIL PROTECTED] writes: Adam Di Carlo [EMAIL PROTECTED] writes: Thimo Neubauer [EMAIL PROTECTED] writes: On Wed, Jul 25, 2001 at 09:42:28PM -0400, Adam Di Carlo wrote: Thimo Neubauer [EMAIL PROTECTED] writes: Ouch. So its that libslang itself should be reduced, and based on that, the usage in libnewt should be reduced? Almost :) First libnewt needs to be reduced, then libslang and after that libm, so that only the really needed math-symbols get in :( I'll take a swipe at it. Ok, this is done in CVS. Can some people test this? Hm, I've not tested it, but in any case mklibs.sh seems to be getting more and more ugly. I've considered rewriting it in Python, using an iterative method where I simply add symbols until everything is satisfied. Shell is just not the language for such stuff. Does that look like a good idea? Would add another build dependency, but we have a gazillion already ;) Rather than writing Yet Another library optimizer, which I see you and Goswin have done, I think we should investigate some of the existing ones. I know Lineo has one (is it free?), and MontaVista has one too (which is, I believe, supposed to be free but not released yet - I'll see if I can make headway on this on Monday). -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: problem with dhcp-client
On Sun, Jul 15, 2001 at 11:39:01AM -0600, Erik Andersen wrote: On Sun Jul 15, 2001 at 05:44:33PM +0100, Philip Blundell wrote: David Kimdon [EMAIL PROTECTED] wrote: So in fact the new version might be too big for boot-floppies. And the old version doesn't work on powerpc... so? powerpc should switch to use pump, everyone else stick with dhcp-client-2.0, IMHO. Well, how hard can it be to make dhcp-client 2.0 work on powerpc? Has anybody investigated what the actual problem is? How hard can it be to make pump work? I'm tempted to grab a copy of RFC 2131 and RFC 2132 and fix pump so it works... What exaxtly is the problem with pump that caused us to switch? -Erik Please, do that. We had no good reason aside from a few it doesn't work for me reports, as I recall. Using an obsolete version of dhcp-client deliberately is just bad news. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#104798: Build failure on hppa (at least)
On Fri, Jul 13, 2001 at 11:08:42PM -0600, LaMont Jones wrote: Package: boot-floppies Version: 3.0.7 Priority: serious devel/boot-floppies_3.0.7 by bdale-hppa [extra:uncompiled] Reasons for failing: [Category: none] || mkdir -p /archive/debian/download mkdir: cannot create directory `/archive': Permission denied [ -d /archive/debian/download ] || mkdir -p /archive/debian/download mkdir: cannot create directory `/archive': Permission denied Can we stop boot-floppies from being picked up by autobuilders somehow? PLEASE? -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#104798: Build failure on hppa (at least)
On Sat, Jul 14, 2001 at 09:57:39AM +0100, Philip Blundell wrote: Can we stop boot-floppies from being picked up by autobuilders somehow? PLEASE? I don't think there's anything that debian-boot can do to stop autobuilders from having a go. The people who run the build daemons need to add boot-floppies to @no_auto_build in buildd.conf. Or of course we could fix boot-floppies to make it autobuildable. I don't think this would be all that hard, but it probably isn't worth the effort. I don't think we could. It has some rather specialized build dependencies, and last I checked still needed real root to create images. The build daemons mostly use fakeroot now. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Install Report, powerpc oldworld 2.3.6 dtd 6/21 06:11 (woody)
On Sun, Jun 24, 2001 at 10:35:04PM -0800, Ethan Benson wrote: On Sun, Jun 24, 2001 at 03:21:20PM -0700, Chris Tillman wrote: These are some notes as I attempted to install the woody b-f. After selecting Partition a Hard Disk, an alert advised about there being no need for a boot partition and ... and the root partition must be on the first disk. I have had Debian installed on this machine before, on the second external SCSI drive (address 6), so I think this message might be in error? I'm not sure what the 'first disk' would mean except only sda, not sdb or sdc. its for quik, many versions of OldWorld BrokenFirmware(TM) are incapable of booting from anything but the master IDE disk, im not sure if that matters on scsi or not, maybe we can kill the check for scsi systems. Indeed, it's not at all true for SCSI. We should try not to complain about booting off of any scsi disk. -- Daniel Jacobowitz Carnegie Mellon University MontaVista Software Debian GNU/Linux Developer -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: testing with powermac/oldworld (probs in miboot, debootstrap, kernel)
On Fri, Jun 08, 2001 at 10:03:17AM -0500, Stephen R Marenka wrote: On Thu, Jun 07, 2001 at 11:29:41PM -0800, Ethan Benson wrote: On Thu, Jun 07, 2001 at 09:20:41PM -0500, Stephen R Marenka wrote: powerpc-utils should still be fixed, that diversion needs to go away. File a bug, I guess (oops, master's down). That doesn't prevent other packages from having conffiles that want replacement does it? i did right before master died, and its back now btw. i think dan maintains powerpc-utils now but its still showing up as belonging to espy I suppose that means Dan hasn't uploaded a new version yet? I also filed a bug with a patch because powerpc-utils doesn't build on woody. If you forward me these bugs I'll try to deal with them this weekend; the BTS is still not spooling. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#99652: dialog boxes confusing - active choice not clear
On Sat, Jun 02, 2001 at 12:28:50PM -0700, David Whedon wrote: Another alternative might be to put some sort of indicator under the active choice, like '^', '*', or '---^---', etc. I don't know how to easily implement this with libnewt (the windowing library boot-floppies uses). An easy option which may or may not be any better is to not use the 'compact' version of buttons for yes/no boxes. An example of this can be found here: http://people.debian.org/~dwhedon/yesno_proposed.png for reference this is what the box looks like now: http://people.debian.org/~dwhedon/yesno_current.png Comments? I think it's a good idea. Do we have any serious space issues with this? I don't remember any. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: powermac boot-floppy patches
On Wed, May 30, 2001 at 03:57:00PM -0800, Ethan Benson wrote: one other side issue, do we plan to support IBM CHRP hardware? (RS/6000 and such) these machines also use yaboot, but install it a bit differently. yabootconfig and ybin already support that. the difference dbootstrap is interested in is they use msdos disklabels and a type `0x41 PPC PReP Boot' bootstrap partition, yaboot is directly dded to it (mkofboot does that). however the 64 bit versions of this hardware needs a yaboot compiled for them specifically, the yaboot package currently does not do that, and yabootconfig does not yet know about using different versions of yaboot. (i would really prefer to get yaboot fixed to work on both machines.. the only difference is a tiny fragment of ppc asm to force the thing into 32 bit mode) I don't know about the 64 bit line yet, but we certainly intend to support the 32 bit CHRP/PReP machines. I'll have one for testing in a few days, with luck. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Systems with no floppy
On Fri, Jun 01, 2001 at 02:34:53PM -0400, Adam Di Carlo wrote: Ethan Benson [EMAIL PROTECTED] writes: On Wed, May 30, 2001 at 03:39:37PM +0200, J.A. Bezemer wrote: I suspect that booting off a Zip disk is more like booting off a floppy than booting off a CD, which means that this probably is a debian-boot issue; forwarding it there. using the same method as the bootable CDROMs will make a bootable zip disk, however it must be done from GNU/Linux and the zip disk must never be mounted in MacOS or else MacOS will render it unbootable. one can always boot the bootloader (yaboot) manually as described many times in the debian-powerpc archives. Maybe Chris Tiller can add this to the Potato documentation? Would this ZIP disk be unbootable on OldWorld PowerMac, just as the CD is? I don't suppose there's a way to just take the HFSimage and burn it to Zip? I'm pretty sure you'll have the same problems booting a ZIP as booting a CD. The HFS floppy image only works because floppies aren't expected to provide their own drivers. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to boot-floppies by dan
On Wed, May 30, 2001 at 12:47:19AM +0200, Marcin Owsiany wrote: On Sun, May 27, 2001 at 06:54:29PM -0700, Daniel Jacobowitz wrote: Probably the best way is to leave it in EXTRACT, but not SMALL_BASE_LIST, and conditionally copy it. Does that make sense? [...] Someone can fix this, or I will on Tuesday. Committing now. Hopefully I didn't break too much. Yes, we now need libbogl-dev from sid rather than woody (tomorrow's sid, even) - but only if LANGUAGE_CHOOSER is on Does that mean that they contain 'reduce-font'? Yes, the libbogl-dev in sid today contains reduce-font. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: powerpc-utils and DEBIAN_FRONTEND
On Thu, May 24, 2001 at 02:23:58PM -0500, Stephen R Marenka wrote: While working on woody boot-floppies on my oldworld powermac, I noticed that powerpc-utils seems to ignore the DEBIAN_FRONTEND=Noninteractive. It thus fails to configure. This is a required package. How do I figure out what's at fault and fix it or where should I file a bug? Huh? Read the powerpc-utils postinst. All I call is update-rc.d and #DEBHELPER#... no debconf at all. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: woody bf-utf8 (Re: slang, boot-floppies, and wide character support)
On Sat, May 26, 2001 at 02:09:35PM +0100, Edmund GRIMLEY EVANS wrote: Marcin Owsiany [EMAIL PROTECTED]: Thanks! Now we have to check newt-utf8 by Enrique, then check the problem in bogl-bterm which impede bterm working without libutf8. So do we actually stick to libutf8 or try to make it work with glibc? We make it work with glibc. Done. CVS log for bogl-term.c: With glibc-2.2.3, mbrtowc(wc, s, 0, ps) seems to return 0 instead of (size_t)(-2). I think this is a bug in glibc-2.2.3. Make sure we work either way. As I said on this list several times, I would appreciate it if someone would make boot floppies use libbogl0 and bogl-bterm instead. Every patch that goes in to bogl in this archive runs a high risk of being lost. I'm not set up to build boot floppies right now, but since no one else has done this, I guess I'll try to do it tomorrow. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: libutf8, bogl, bgf fonts and tools vs boot-floppies
On Sun, May 27, 2001 at 10:28:10PM +0100, Edmund GRIMLEY EVANS wrote: Daniel Jacobowitz [EMAIL PROTECTED]: We need to have the stuff in subject included in bf bould process somehow. The question is: how? AFAIK bterm has its own package (bogl-bterm), as does bdf fonts' source. (Now we either need to put libutf8 in its package or make bterm work without it - I'd say we should do the former, since it's probably easier and so require less time). No, the latter. Since I did it several months ago; see bogl-bterm's changelog. Oops. Wait. I forgot to upload that; I'm sorry. I'll do it soon. I may have committed a fix for the same bug today. Please check. Indeed you did, and it's a similar fix. This is what I originally did: @@ -206,18 +206,11 @@ -for (; (k = mbrtowc (wc, s, n, term-ps)) = 0; s += k, n -= k) +for (; n 0 (k = mbrtowc (wc, s, n, term-ps)) = 0; s += k, n -= k) Your patch seems rather more complete; should I use that instead? Sorry for snapping at you earlier, I missed this message. On Mon, May 28, 2001 at 01:00:09AM +0200, Marcin Owsiany wrote: But you commited to bf CVS, right? So should we use the packaged or CVS version? I'm confused. Packaged. I'll fix it. BTW: now that we use glibc's UTF8 support, what locale should we use? I've found en_IN.UTF-8 (in my /etc/locale.gen) to work, but I don't even know what that means. Indian, I think... -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: woody bf-utf8 (Re: slang, boot-floppies, and wide character support)
On Sun, May 27, 2001 at 05:22:52PM -0700, Matt Kraai wrote: On Sun, May 27, 2001 at 04:32:00PM -0700, Daniel Jacobowitz wrote: As I said on this list several times, I would appreciate it if someone would make boot floppies use libbogl0 and bogl-bterm instead. Every patch that goes in to bogl in this archive runs a high risk of being lost. Why don't we remove it from boot-floppies CVS then? That way, no one will be tempted to do the wrong thing. Is there some reason it has to be here? Because it won't build out of the box with the packages :) I'm working on this right this moment. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to boot-floppies by dan
On Sun, May 27, 2001 at 06:50:36PM -0700, [EMAIL PROTECTED] wrote: Repository: boot-floppies who:dan time: Sun May 27 18:50:36 PDT 2001 Log Message: remove bogl Files: changed:rootdisk.sh todo Please don't hate me. I did this in perhaps not the brightest way - bterm (33K) is now always included. That makes no sense, but I didn't want to figure out how to conditionally extract a package, since I need to run. Probably the best way is to leave it in EXTRACT, but not SMALL_BASE_LIST, and conditionally copy it. Does that make sense? Someone can fix this, or I will on Tuesday. Hopefully I didn't break too much. Yes, we now need libbogl-dev from sid rather than woody (tomorrow's sid, even) - but only if LANGUAGE_CHOOSER is on, otherwise any libbogl-dev/bogl-bterm at all will do. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: removing bogl from bf CVS module
On Sat, Apr 28, 2001 at 01:33:14AM +0400, Peter Novodvorsky wrote: Hello! Daniel made a package with bogl frame-buffer library and bterm, so we don't need to keep them in boot-floppies CVS. But still I think that we should have bogl and stuff, but in another module, Dan? Well, I could move my local CVS tree to cvs.d.o, certainly, if people want me to. Do you want to tackle making woody b-f build with my packages? Mostly boxes.c got gutted; that wants to stay in b-f and link to -lbogl. I'd rather not give others commit access, though... -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#94458: depmod after drivers installation fails because directory is incorrect
On Wed, Apr 18, 2001 at 07:28:31PM -0700, Chris Tillman wrote: Package: boot-floppies Version: 2.2.23 After Install Operating System Kernel and Drivers completed, an error message informed me that the drivers installation had failed. Looking in the error log, I noted the error depmod: No modules were found in /target/lib/modules/2.2.19 that could be configured. I went to Execute a Shell and navigated to /target/lib/modules and asked for a ls, with the results 2.2.18pre21 Apparently a directory was created but with the incorrect revision number to match the installation package. Are you absolutely sure you have a bootdisk and rootdisk from the same build? The 2.2.23 I uploaded had a /target/lib/modules/2.2.19: block/ fs/ ipv4/ ipv6/ misc/ net/ pcmcia/ scsi/ usb/ I'm betting you used an out-of-date mirror for a network install. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: new potato boot-floppies
On Sun, Apr 15, 2001 at 01:51:29AM -0700, Andrew Sharp wrote: For people who care about powerpc stuff While the changelog below mentions powerpc, these are of course only ia32 floppies. Having no patience, I have taken the liberty to Of course, I've built them eight times in the past two days... You -could- just have asked on the list. I didn't upload them because I was in the middle of debugging... build a boot floppy for oldworld apple powermacs that solves the dreaded no response from keyboard when prompting to insert root floppy problem. It seems to work for doing installs on the one this. machine that I was able to test it on, an Apple 7200/75. It may spew out a few error messages at one or two places about missing symbols, can't load modules. This should be ignored, unless someone discovers a piece of hardware that isn't supported during the install phase, in which case I'll fix it. It should be fine, though. This works for the 2.2r2 Debian release for powerpc It SHOULD NOT BE USED. You're avoiding a problem by using an old kernel where the problem didn't occur; I've fixed the problem. The new floppies will probably be on mirrors by tomorrow. You'll hit a lot of random issues by the fact that you're not using the kernel that the floppies were built with. On Sun, Apr 15, 2001 at 03:49:50PM +0200, Martin Schulze wrote: Ethan Benson wrote: his image is just the old 2.2.16 boot floppy since the potato r2 boot floppies are broken. drow is trying to fix that in the r3 boot floppies. ARGS. Is he "trying to fix this" or did he "actually fix this"? Drow, it would be a good time to appear now... :) Fixed. I just booted the same floppies I put in Incoming half an hour ago and got a good way into an install. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: new potato boot-floppies
On Sun, Apr 15, 2001 at 12:24:25PM -0700, Andrew Sharp wrote: Perhaps you all should have read my post more fully. It is NOT an old kernel, it is 2.2.18pre21 and I tested an install. It works fine for 2.2.r2 Debian as I said. It SHOULD be used unless you are posting yours for 2.2r2. They've already been installed into the archive for 2.2r3, in fact. If it is 2.2.18pre21, I'm interested in what you did to make it work. Change the wait to a delay? Or did you actually duplicate what I've been working on for the past two days and fix the initialization bug in the keyboard layer? You should have read the whole post. It clearly says otherwise. As Ethan pointed out, no, it didn't. You nowhere said what you were doing, and everything I've seen posted before this was using 2.2.16. I'd appreciate it if you could explain what you did to fix the problem. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: new potato boot-floppies
On Sun, Apr 15, 2001 at 08:29:26PM -0700, Andrew Sharp wrote: boot-floppies build system works. ~:^) Contrary to popular myth, 2.2r2 can't be obsolete until r3 is released. It hasn't been released, has it? Believe it or not, I'm trying to help out here. It has been finalized. I believe it should be out this time tomorrow. "Note that this link used to point to a 2.2.17 based floppy I built a while ago that really wasn't all that useful." My impression was that the 2.2.17 floppy had the input problems, and had been downgraded back to a 2.2.16 based one. Oops. Sorry. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: patch to fix debootstrap invocation on local archive
On Fri, Apr 13, 2001 at 04:07:35PM +1000, Anthony Towns wrote: * progress -- how much has downloaded (x%), what's happening (I: installing essential packages), etc It would be really nice to have some %'s, like apt-get. Is it reasonable for boot-floppies/dbootstrap to do the display of that (with slang and such), or does debootstrap need to do it? Or does debootstrap need to have some sort of module or something to do boot-floppies-like-slang itself? I'd like to see dbootstrap have a way to do it. When we get to debian-installer and frontends, we should have a defined way for frontends to display progress bars... I could also make the summary/informative output look like: I: Getting libc6 2.2.2-1... P: 30% Downloading base system if you were able to draw percentage bars somewhere on the screen, say. How about I make a debootstrap that directs dpkg's output to stderr, prefaces all the informative stuff to stdout with an I:, and ends with error code 1 after printing E: something, all iff the first arg is "--boot-floppies"? That can be extended later to have P: lines for percentages or whatever too, if desired. P: would be nice... -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Boot floppies 2.2.23
Is there anything else holding up 2.2.23? Can I tag it, or do you want to? -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.2r3 and boot-floppies
On Tue, Apr 10, 2001 at 03:07:01PM -0800, Ethan Benson wrote: On Tue, Apr 10, 2001 at 04:32:27PM -0400, Adam Di Carlo wrote: Martin Schulze [EMAIL PROTECTED] writes: is *anything* missing for 2.2r3 wrt. boot-floppies? Not sure I understand the question -- I guess not as far as I know, but I'm not the one who is finding security issues in base packages or kernels. I don't think all ports have rebuilt yet. PowerPC seems to have serious kernel troubles. its a misconfiguration, sysrq is turned on and this happens to break the arrow keys on some keyboards. sysrq needs to stay off on powerpc for the timebeing, none of the powerpc kernel hackers use it so its almost invariably broken. Under what conditions does this happen? -I-, in my occasional role as a kernel hacker, use it constantly. I can build new kernels without it if necessary, though. the other issue is the boot-floppies are broken since the root disk prompt won't accept enter (or anything) to continue. Working on it. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: powerpc boot-floppies
On Wed, Apr 11, 2001 at 10:56:06AM -0400, Adam Di Carlo wrote: Dan, let me know when you have all the Potato source in good shape. At that point we can burn the new version, I can do a source upload and i386 binary version. I already have the hdg/hdh bug fix in CVS which I think makes it worthy of potato update. Sorry about dropping the ball on this. I've got it almost prepared; I'm testing now. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#48778: powermac bf issues (was Re: Bug#48778: Feature has been implemented)
On Wed, Apr 04, 2001 at 10:21:42PM -0800, Ethan Benson wrote: On Thu, Apr 05, 2001 at 01:15:23AM -0400, Adam Di Carlo wrote: Ethan Benson [EMAIL PROTECTED] writes: it is possible NOW to have a fully working `make linux bootable from the hard disk' step on newworld powermacs. right now that step is 100% broken on this varient. Is anyone going to try to fix that for Woody at least? i certainly hope so, dan was talking about it at one point, but he is so busy... if anyone else wants to try it out i would be happy to explain how mkofboot works and how dbootstrap needs to run it in order to get a bootable system. I hope so, also. I've not got time for woody boot floppies right now, but after 2.2r3 goes out the door I'll take a look at it, hopefully. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [PATCH] nearly successful powerpc build
I assume you're talking woody here? I've been working on potato boot floppies for the past week steadily and not had time to look at the woody set. On Sat, Apr 14, 2001 at 01:12:55AM -0400, [EMAIL PROTECTED] wrote: I got a build that ran all the way up to release.sh before failing to install some documentation. Here are the relevant patches. I'm not too sure about the second one, because I don't understand what '$system' is supposed to be for (it's currently always passed in as null from the higher-level scripts, and the value the lower-level scripts want is in the subarch, i took a wild guess which to fix). powerpc-prep-apus.diff - Added powerpc/rules, to deal with prep and apus oddities. - Modernized bootprep.sh. - config: Changed default powerpc kernel version to 2.2.17. Why on earth would you do that??? It should have been 2.2.18presomething in the archive; the 2.2.18pre was there for a reason. It should also just be changed to 2.2.19 now. They're in incoming. powerpc-hacks.diff - rescue.sh: Switch on $subarch, not $system for powerpc. - rootdisk.sh: Added 'charset=latin1' (HACK--where should this come from?) -- James Deikun, Techie(tm), CSI Multimedia The opinions expressed c. # Patch kernel source - FIXME: allow for other patches cd kernel-source-$kver -zcat $builddir/usr/src/kernel-patches/powerpc/*.diff.gz \ - | patch -l -s -p1 +# zcat $builddir/usr/src/kernel-patches/powerpc/*.diff.gz \ +# | patch -l -s -p1 Deleting lines of code is usually a bad idea. In this case, a very bad idea. This comes from kernel-patch-ver-powerpc; I assume it was part of your attempt to go to 2.2.17? ifeq "$(architecture)" "powerpc" -kver := 2.2.18 -pcmcia_kver := 2.2.18-pmac +kver := 2.2.17 +pcmcia_kver := 2.2.17-pmac apuskver := 2.2.10 endif Like I said. Why on earth do you want to do a downgrade? ifeq "$(architecture)" "sparc" @@ -57,7 +57,7 @@ # Top-level directory where we will look for external stuff required # to build the boot-floppies. It has to be an absolute PATH or else # the basedisks.sh script will fail. -ftp_archive := /archive/debian +ftp_archive := /space/james/archive/debian Careful of including this with submitted patches! If you feel like adding a config.local, more power to you... please... -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.2r3 and boot-floppies
On Fri, Apr 13, 2001 at 08:50:03PM -0800, Ethan Benson wrote: On Fri, Apr 13, 2001 at 02:04:53PM -0400, Daniel Jacobowitz wrote: Under what conditions does this happen? AFAIK (based on ramblings from benh on #mklinux) it seems to affect the recent iMac keyboards. for some reason the kernel decides that the arrow key is sysrq so stops functioning as an arrow key. you might ask him about it maybe he fixed it. For now, I'm turning it off. I don't care enough to fix it. I think I fixed the oldworld floppy issue, though; we'll know in the morning. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [PATCH] nearly successful powerpc build
On Sat, Apr 14, 2001 at 01:36:52AM -0400, [EMAIL PROTECTED] wrote: Hm. It seems that you are right. I noticed the 2.2.18 per se was missing, ran 'grep-available' to check for other packages, and only 2.2.17 came up, I don't know why. Going through the lists apt downloaded by hand revealed ample evidence of 2.2.18pre21 and 2.2.19pre17, either of which would have been suitable. It should also just be changed to 2.2.19 now. They're in incoming. It happens :) But try to look a little closer in the future, please. No one should use anything before 2.2.19 any more, for security reasons. # Patch kernel source - FIXME: allow for other patches cd kernel-source-$kver -zcat $builddir/usr/src/kernel-patches/powerpc/*.diff.gz \ - | patch -l -s -p1 +# zcat $builddir/usr/src/kernel-patches/powerpc/*.diff.gz \ +# | patch -l -s -p1 Deleting lines of code is usually a bad idea. In this case, a very bad idea. This comes from kernel-patch-ver-powerpc; I assume it was part of your attempt to go to 2.2.17? The patch seemed to either have already been applied in the source package, or was getting applied twice in the build. I planned to take a closer look at it tomorrow. In the meantime it was building with that change and freaking out without it. This is another thing I have to ask you to look closer at. Please don't commit or post stopgaps like this; they get forgotten about, and then when I put an important fix in the kernel patch package it'll take me weeks to figure out why boot floppies are still crashing. D'OH! This should teach me to submit patches after midnight. Sheesh. At least I remembered to delete it out of the other patch ... ... I think I'll be adding a config.local now. Thanks! -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: PPC boot-floppy building
On Thu, Apr 12, 2001 at 03:36:58PM -0500, [EMAIL PROTECTED] wrote: Maybe we can get ppc cleaned up! For the time being I just commented out mac-de-latin1-nodeadkeys.kmap.gz in keymaps.sh. I didn't see any good solution for what looked like a missing maintainer. (It doesn't look like he's doing anything upstream either.) The proper fix would be to finish with our conversion to PC keycodes; 2.2.19 should have everything necessary, as should our 2.2.18 packages. Then we can lose those keymaps entirely. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Booting to floppy
On Tue, Apr 10, 2001 at 12:40:07AM -0400, Adam Di Carlo wrote: Michael Schmitz [EMAIL PROTECTED] writes: The 'rescue floppy' doesn't even work as this on a whole other bunch of architectures I bet. Can you boot from this floppy on sparc? Sure. Only on architecture, PowerPC, is in bad enough shape to require one image for actual booting (when possible at all) and other one to supply the kernel to dbootstrap. Wouldn't it be easier to just get the kernel directly from one of the 'linux' files lying around, in the case that the user is installing by a method other than floppies? This might be an easy, general fix we can do for all arches. Definitely! But the only times we've tried to address this are when I need to get potato b-f's out, and I'm not about to try that in potato. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: boot-floppies progress
On Wed, Mar 21, 2001 at 09:42:35PM -0500, Adam Di Carlo wrote: I apologize for my recent lack of involvement in boot-floppies maintenance. I have been allowing myself the rather guilty pleasure of working through the new policy of Debian SGML/XML stuff. Which is a whole lot more fun for me than boot-floppies hacking. Anyhow, I'm right now going to tag/release 2.2.21 for Potato 2.2r3. I need to back out David's changes bumping i386 kernel to 2.2.19pre17 because pcmcia modules are not yet available for the compact and idepci flavors. So I'm shipping bf 2.2.21 i386 using 2.2.18pre21. I know that's old but that's the last complete set we have for i386, and there's not much I can do about it. When a new complete set is ready, it's a pretty simple thing to make a new boot-floppies. I would really rather someone fix this, even at the cost of delaying boot floppies further. 2.2.19pre17 is a significant security upgrade. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team "I am croutons!" -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Booting to floppy
On Mon, Feb 26, 2001 at 08:21:18PM -0800, Andrew Sharp wrote: Ethan Benson wrote: On Mon, Feb 26, 2001 at 02:48:40PM -0800, Andrew Sharp wrote: Um, yeah, that must be what I meant. Actually, what I meant is this: the docs constantly talk about booting the rescue floppy for this that and the other. But the "rescue" floppy for the powerpc port is an ext2 file system, and doesn't boot at all on old world macs, and I just assumed that it must be for new world macs. If not, then the rescue floppy image is a complete hoax for powerpc. the rescue floppy has only one purpose on powerpc, and that is to provide the kernel image that is installed in the `install OS and kernel modules' step. the bootable rescue floppy for oldworld macs is the hfs-boot.img (or whatever its called). I've never needed rescue.bin for that. Granted I've only done two installs. ~:^) But there are two images called driver-1.bin and driver-2.bin which one might guess have drivers on them. Never used those either. You've probably done either network or CD installs. They both use rescue.bin; they don't require it be on a floppy, that's all. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Booting to floppy
On Mon, Feb 26, 2001 at 02:28:55PM -0800, Andrew Sharp wrote: Welcome to the powerpc boot floppy fiasco. The rescue floppy will only work on New World macs, and yours is an old world. The That's just not true. The rescue floppy is not meant to be booted off of on this architecture at all. New World macs don't have a floppy drive. You mean the boot CD only works on New World. -- Daniel Jacobowitz Debian GNU/Linux Developer Monta Vista Software Debian Security Team -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Request for addition: USB mass storage support
On Tue, Feb 20, 2001 at 03:57:33PM +1100, Glenn McGrath wrote: So you probably need a kernel that has usb support compiled in to read the root disk, which probably means you need a 2.4 kernel, would that be correct ? But even in 2.4 USB storage support is... um... a bit dubious. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: what is the whole purpose of kernel-image-di?
On Fri, Feb 16, 2001 at 10:56:24AM -0700, Randolph Chung wrote: libdetect. Someone KILL libdetect. It has fledgling PPC support, but it's (A) hacked together awfully (B) a little lacking in correctness (C) nowhere near compiling. I got it to build once, with two hours work, but not function. hmm.. drow, I'd like to talk to you about this more... am looking into doing autodetection stuff for woody b-f using libdetect, but that might not be such a good idea if it's very i386 specific. In fact, I just saw that other message in which you mentioned the idea and was about to answer... libdetect COULD be made more cross-platform friendly, by giving it some sane rewriting. It's a mess. cdebconf. Randolph, please do something about the warning and error I showed you from the preinst (?)! been on a road a lot this last few weeks. will fix on tuesday when i get back. Thanks. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: dhclient and kernel-image-di
On Wed, Feb 07, 2001 at 11:32:34AM -0800, Joey Hess wrote: David Whedon wrote: dhclient needs CONFIG_PACKET and CONFIG_FILTER (or at least the docs say it does, I haven't gotten it to work on the floppy, so ther emay be something else lurking). CONFIG_PACKET can be a module, should it be in it's own udeb? Guess so unless you can think of something it makes sense to bundle it with. CONFIG_FILTER can't be a module, or at least 'make menuconfig' doesn't want to let me modulize it, so it just makes the kernel bigger? Fraid so. I'm building kernel-image-di with CONFIG_FILTER on. It will include a packet-socket.udeb, unless someone comes up with a better name. Perhaps prefix these with something indicating that they are kernel modules? Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: installer serious oversight or on purpose?
On Mon, Feb 05, 2001 at 05:36:21PM -0800, Joey Hess wrote: I'm forwarding this to to appropriate list. This problem must be particular to powerpc, accessing the cd certianly works on i386, since it uses the same kernel after install as it used to install from cd. Is something weird going on with the ppc kernel? CONFIG_ISO9660_FS=y From the kernel we install onto the CD... I can't imagine that it's getting a different one from somewhere... corey wrote: I just picked up a cd distro of Debian Linux PPC while at LWE. I found a serious issue that keeps me from getting the thing installed, and I'd like like to know if there's a reason this exists in this state Step 1. I boot from the cd (which of course, has iso9660 built into the kernel). Go through the whole process of setting up system, etc. Choose my modules, and such. I reboot, using kernel placed on hdd during initial install. Do you mean during this particular install? What version of kernel is this? Does dmesg show anything interesting? It -should- be 2.2.18pre21, and it should be the same as what you get from the kernel-image-2.2.18pre21-pmac package. My guess: How are you booting? From a previous linux install attempt, perhaps? You must not be getting the kernel that the CD gave you. Make sure to run mkofboot or ybin before rebooting (see the installation instructions, this is underdocumented). Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: what is the whole purpose of kernel-image-di?
On Tue, Feb 06, 2001 at 09:09:28PM -0800, Joey Hess wrote: Ben Collins wrote: I can understand it being the first milestone, but it seems to be the only focus, and nothing is being considered as to how it affects other ports. That's not entirely true; I know of a group (who I cannot name or go into any detail on since they told me about this privatly) who is using the the basic debian-installer design and code as the base for something (very cool) involving installation on powerpc. The pa-risc porters were also looking at using d-i some months back, but it wasn't far enough along at that point. blink Now I'm curious, curse you! Oh by the way Dan has been involved earlier in porting most of the d-i modules to the powerpc, or rather, compiling them on the powerpc. They seemed to work ok, but this was a while back, before the system did anything. And I'm going to go back to this shortly, now that we've got an x86 bootable floppy. PowerPC is not big on bootable floppies (or floppies at all) but I can probably get a nice ramdisk going, and maybe a netboot... This is one of the nice things about the debian-installer using standard debian sources that generate deb files: autobuilders automatically treat d-i modules just like regular packages. Anyone want to go check the build logs for problems to see if some of the missing items failed to build or have just not need attempted yet? libdetect. Someone KILL libdetect. It has fledgling PPC support, but it's (A) hacked together awfully (B) a little lacking in correctness (C) nowhere near compiling. I got it to build once, with two hours work, but not function. cdebconf. Randolph, please do something about the warning and error I showed you from the preinst (?)! b. Send me an appropriate kernel config and related information, and build kernel-image-di once I integrate it. 2.4.1 kernel.org does not build on PPC. Sigh. c. Some things such as network card involve hardware autodetection. We have concentrated on i386 since it has the most variety (and crap) hardware. udebs are generally available which let the user specify it manually instead of using hardware autodetection. Other than those and some minor crossovers, other architectures are on their own, and someone will have to work on it. Other architectures will want a radically simplified detection scheme. PPC (mostly) has no legacy ISA or such; everything should show up in the kernel's PCI probes. That should make things much nicer. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: initrd build problems
On Wed, Jan 31, 2001 at 07:20:22PM -0800, Joey Hess wrote: Joey Hess wrote: When I try to make an initrd, it always fails like this: Writing inode tables: done Writing superblocks and filesystem accounting information: done mount -t ext2 -o loop ./tmp/debian-installer ./mnt/debian-installer cp -a debian-installer/* ./mnt/debian-installer/ cp: writing `./mnt/debian-installer/usr/lib/libdebconf.so.0.1.0': No space left on device cp: cannot create directory `./mnt/debian-installer/usr/lib/cdebconf': No space left on device cp: cannot create directory `./mnt/debian-installer/usr/lib/debian-installer': No space left on device cp: cannot create directory `./mnt/debian-installer/var': No space left on device make: *** [initrd] Error 1 I wonder why it's working for you.. If I add a fuzz factor of 122 to your initrd size calculation, it seems to work: dd if=/dev/zero of=$(TMP_FILE) bs=1k count=`expr $$(du -s $(DEST) | cut -f 1) + 122` On reflection, I suspect this is ext2fs overhead. The fuzz factor at home is 127, not 122, probably because I added some more udebs in.. Almost certainly. There's both a fixed filesystem overhead and a size-dependent one, I'd imagine; just making a ramdisk the size of the contents really shouldn't work. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: No Keyboard when booting the powerPC bootdisk
On Mon, Jan 29, 2001 at 10:27:18AM +0100, John McLoud wrote: Hi! Im having a "little" problem, when booting from the dick, found in /install/powermac/boot-floppy-hfs.img on the official Debian-CD. I dont have any access to my keyboards (one is connected to the ADB, the other to the USB port of my PowerMacintosh 4400). Some time ago, I downloaded an earlier version of the bootdisk, based on kernel 2.2.17 which workes fine with both ADB and USB keyboard, but I cant use this disk for setting up the modules of Debian 2.2r2. See the archives of debian-powerpc. I'm trying to find a fix for this. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bogl forked (was Re: [patch] bogl memset_var for 32 bpp screen)
On Thu, Jan 25, 2001 at 08:54:38PM -0500, Adam Di Carlo wrote: Daniel Jacobowitz [EMAIL PROTECTED] writes: On Tue, Jan 23, 2001 at 08:27:51PM -0500, Adam Di Carlo wrote: Daniel Jacobowitz [EMAIL PROTECTED] writes: Sorry for not being clear - yes, I know that bterm has the same bug. I'd appreciate it if someone on this list would look into using the bogl and bterm in that package, so that we can scrap the one in debian-boot CVS; having both makes me nervous about lost work. I guess we'll plan on using straight bogl from woody for the woody boot floppies. I hate forking. I'm not sure why we forked it... Maybe Marcin (or was it Marcel?) can explain it? We didn't fork it for boot floppies; it was originally there. I added new bogl packages to woody for debian-installer, though, a month ago. At that point it was effectively forked - mostly my fault. Oh. Ok. So -- am I correct in concluding that boot-floppies for woody should rely on using bogl from woody (the .deb, b-f can't use udebs) and not the internal stuff? If at all possible, yes! It will take a little doing, because I removed the boxes code; that just needs to link to the provided libbogl. Then I can go through and sweep up any pending patches in the b-f bogl repository, and delete the whole directory. I don't have time to do this right now, alas :( Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bogl forked (was Re: [patch] bogl memset_var for 32 bpp screen)
On Tue, Jan 23, 2001 at 08:27:51PM -0500, Adam Di Carlo wrote: Daniel Jacobowitz [EMAIL PROTECTED] writes: Sorry for not being clear - yes, I know that bterm has the same bug. I'd appreciate it if someone on this list would look into using the bogl and bterm in that package, so that we can scrap the one in debian-boot CVS; having both makes me nervous about lost work. I guess we'll plan on using straight bogl from woody for the woody boot floppies. I hate forking. I'm not sure why we forked it... Maybe Marcin (or was it Marcel?) can explain it? We didn't fork it for boot floppies; it was originally there. I added new bogl packages to woody for debian-installer, though, a month ago. At that point it was effectively forked - mostly my fault. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [patch] bogl memset_var for 32 bpp screen
On Wed, Jan 17, 2001 at 10:10:23PM +0800, ha shao wrote: A minor bug. It only effect 32 bpp framebuff which I was using when worked with bterm. Index: bogl-cfb.c === RCS file: /cvs/debian-boot/boot-floppies/utilities/bogl/bogl-cfb.c,v retrieving revision 1.1 diff -u -r1.1 bogl-cfb.c --- bogl-cfb.c1999/04/17 00:52:47 1.1 +++ bogl-cfb.c2001/01/17 14:03:05 @@ -63,7 +63,8 @@ /* Size of an "unsigned int" in pixels: == sizeof(unsigned int) / b / 8 */ const int intsiz = sizeof(unsigned int) * 8 / b; - const int mask = (1 b) - 1; + /* long long is needed for b == 32 */ + const int mask = ((long long) 1 b) - 1; unsigned int fill; /* MUST be signed! */ ssize_t xlen; I'd have thought it would work anyway (since 1 b is usually 0 for b = 32, although this is not well-defined...). Odd. Could you please look in to using the bogl-bterm package in the archive instead of the included bogl in CVS? It should work fine, although you'll have to fiddle a little to do the font reduction. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Thoughts on the installer (early)
On Tue, Jan 09, 2001 at 03:30:23PM +0100, Luca De_Vitis wrote: I mean: you (I hope to join and help you ASAP) should try to make the smallest, feturefull engine you can, and provide as many front-end and back-end as possible and let other to choose what to build together for the installation system as default. Not objecting at all. I'd love someone to write the gtk frontends. I admit I'd rather have a less hokey widget set, but the other advantage of BOGL is a compact utf8 and i18n library; does GTK have something similar? I think this is not the point and taking the discussion on "this is better and this is not" whill not help. I hope I did not annoyed anyone. That's not what I'm saying. As far as I'm concerned, without utf8 and i18n gtkfb would not be an option at all - we need to support this. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Thoughts on the installer (early)
On Mon, Jan 08, 2001 at 12:52:51AM -0500, Chris Ruffin wrote: From what I gather today after playing around with gtk-fb today is that you would have to have the appropriate versions of several libraries available- the gtk gdk libs, compiled for frame buffer support, glib, the pango libs, and the freetype libs. That's a little overhead, yes, but I do think that since gtk-fb's motivation is for embedded devices that it would be adequate for our purposes. From first glance at looking at the new debian installer, it looks like size is becoming less of an issue, since the appropriate modules (udebs) can be retreived while the installer is running. At first glance, it seems feasible to me. The way I see it, the benefit of using this approach over something like bogl (sorry, I've haven't seen this yet), is that it provides a complete set of widgets that we can work with to produce a very intuitive, very clickable, very friendly interface that looks really good. I don't think it would be wise to scrap the bogl efforts and move towards something else, but what I'm asking is if it is feasible to move the UI that bogl implements over to use gtk-ui, as opposed to the lightweight, from scratch interface that is there already? I don't think so, at this time. The motivation for the BOGL stuff I'm writing is size; the complete bogl library strips down to 90K right now on x86, and will probably not go above 120K after I'm done adding widgets. GTK is nice, but it's not going on a floppy. This will. I admit I'd rather have a less hokey widget set, but the other advantage of BOGL is a compact utf8 and i18n library; does GTK have something similar? Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: comments on cdebconf
On Sat, Jan 06, 2001 at 10:43:45AM -0700, Randolph Chung wrote: Looks like lots of interesting things happened to d-i over Christmas. Very good job everyone! I tried out Dan's bogl interface. Looks cool once we have the "select" handler we can do dpkg-reconfigure debconf with cdebconf :-) I knew I was missing a big one. I'll try to work on that later, and any other missing ones; afterwards I can think about the redesign I want to give it. one thing i saw was that if the bogl interface segfaults (for example on an unknown question type) the screen doesn't get cleaned up properly. is there a way to handle this more gracefully? (perhaps a simple SIGSEGV handler will work) Erk. Thanks for the note. I'll think about that. It's a little nervous, since bogl_done() is a bit complex; I'll give it a shot, though. I'm also considering making it run on its own VC. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
graphical frontend committed (preliminary)
I've got a rough sketch of the framebuffer based frontend in CVS now - it requires libbogl-dev to build. Get the one in incoming if you do it before dinstall on the 27th. Must be run as root, obviously, since it hammers on the framebuffer. Oh, and currently you'll need helvR10.bdf from the bogl source package in your current directory when you run it. I'll fix that in a future release, once I have a better idea what font to use as standard. Currently, it can handle note, string, multiselect, and boolean questions - enough to get a feel for it. The interface on booleans is wrong - I want to use two radio buttons, but that means adding a widget to bogl. I'll do that in a few days. I don't really like the multiselects, either, but I'll think about that too. I'm strongly considering a more extensible widget architecture for bogl, at the risk of a little bloat. Comments on the code are welcome, and especially comments on the interface. Does anyone know a good way to draw disabled-looking text? And do people think that rounder-looking buttons would be worth the effort? Any better preferences for fonts? I made one change to the mouse handling code - if GPM is running, or if the unified mouse driver is present and has at least one mouse, we skip the mouse detection and use what we have. Does this sound ok? I know that what this means for my laptop is that the trackpad won't work; I could easily try for PS/2 also before feeling the need to put up a progress bar, since probing the serial ports is the only really time consuming bit. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: BOGL based frontend for cdebconf
On Wed, Dec 20, 2000 at 05:22:45PM +0100, Tollef Fog Heen wrote: * Daniel Jacobowitz | I've begun designing a BOGL based frontend for cdebconf. With the risk of sounding stupid: what is BOGL? neither apt-cache, dict or apropos knows about it. I'll fix that shortly :) BOGL is Ben's Own Graphics Library, a toolkit Ben Pfaff wrote for graphical installation on the old boot-floppies; the main customer right now is bterm, which uses it for i18n console display. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
BOGL based frontend for cdebconf
I've begun designing a BOGL based frontend for cdebconf. I've not actually got anything useful yet, but I plan to take the project home with me over break; I'll see how it goes. In order to do this, though, I need to separate BOGL out from boot-floppies. Ben, would you mind if I packaged it for the main archive? Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [VA-Debian] Comments from a first-time Debian install.....
On Sun, Dec 03, 2000 at 01:49:26AM -0500, Adam Di Carlo wrote: Well, for instance, even with booting with the 'compact' set, which has a lot of ethernet cards included, on my system at home, I need the 3c509 driver installed. So, if I want to install the rest of the system over the network, I need to load the module here. I guess I should try to clarify in the UI that this step is for loading modules which are needed for the installation process. Well, that's not really true, is it? Don't we write out /target/etc/modules here? I don't think that we touch it any other time. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Boot floppies 2.2.19
On Wed, Nov 22, 2000 at 07:29:52PM -0500, Adam Di Carlo wrote: I just discovered a but in the vanillla kernel/pcmcia set for i386. Right now the pcmcia modules for the standard kernel are broken. Figures, that's one of the two I didn't test. I'm hoping Randolph can rebuild the pcmcia-modules for 2.2.19 boot-floppies i386. Daniel, thanks a lot for being so responsive and active, although I was disappointed to see that none of the PowerPC bugs are being closed in 2.2.19 changes, including RC 69161 ? Is that a changelog oversight? Or are the bugs still extant? Is anyone working on fixing PowerPC boot-floppies bugs? I did fix a few of them. The majority of others I am willing to let slide in favor of debian-installer; they'd require major rework. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [patch] dbootstrap documentation
On Tue, Nov 21, 2000 at 06:40:21AM -0500, Adam Di Carlo wrote: trying this again, attached is a patch for dbootstrap's documentation regarding installation of the bootloader on NewWorld PowerMacs. there is a couple things that need to also be added to the root floppy for this to work/be accurate: /dev/nvram /sbin/nvsetenv (from powerpc-utils_1.1.3-2_powerpc.deb) /usr/sbin/ofpath (from yaboot_0.9-0.29_powerpc.deb) also /boot/yaboot (or /usr/lib/yaboot/yaboot) must remain on the root image. (2.2.16 boot floppies have it in /boot) please CC any replies as i am not subscribed to debian-boot atm. I really need a PowerPC person to deal with this. I am not a porter nor do I know much about Linux/PPC. The documentation updates are tied to changes in the powerpc version, so I can't apply those either. Dan, can you deal with this? Can and will (well, mostly did). Is there any reason for me not to just tag a 2.2.19 and upload for powerpc? All my changes are architecture-specific. In fact, if no one objects in the next few hours, I'll do that. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: One or two floppies (was: first weekly debian-installer status report)
On Thu, Nov 02, 2000 at 10:02:50AM -0500, Vaidhyanathan Mayilrangam wrote: Just some prior info: The base kernel takes 385K and all the net drivers take 575K. It essentialy means it is a two floppy install (unless there is someway a user could choose only the drivers he/she needs). It is a catch-22 here.. We need to know the NIC card to choose the driver, but we need the module before that.. 4. Can we fit in a code to detect NIC and activate it in about 400K.. If so, we can have one floppy with kernel and the net drivers and 400K code which will activate the NIC and start downloading everything ?? Probably... Similarly, we can probably manage a disk with enough drivers to access a CD and run a retriever. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: first weekly debian-installer status report
On Wed, Nov 01, 2000 at 08:11:52PM -0800, Joey Hess wrote: - disk partitioner [UNCLAIMED] Not started. Some kind of a UI to let the user partition their disks and decide what the partitions are used for. - disk formatter [UNCLAIMED] Not started. Formats disks, turns on swap, etc. I'd kind of like to work on these, via parted. I'm not sure I'll have a chance, though. - library reduction [UNCLAIMED] Not started. There is code in the boot-floppies that may be usable, but this is one of the hardest parts. I can work on this if necessary, too. How you can help: * Aj and Randolph need someone, especially someone who knows curses well, to help with cdebconf. Why curses? Aren't we sticking with slang's pic library? Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: debian-installer main menu almost done
On Sat, Oct 28, 2000 at 07:12:10PM -0700, Joey Hess wrote: Having looked at where d-i is upto, and how far it's got to go, I'm inclined to think that aiming for d-i to be the installer for woody (for all architectures, for all environments) isn't completely unreasonable. We'll probably have to make a definite decision which one to go with for woody by the end of November, though, to make sure we've got time to hack on b-f if we need to. You're more optimistic than me, but I guess we'll see where we are and how many people are contributing code by the end of November. Well, I agree that it's a little optimistic, but let's see what we can do... I will try to get PowerPC support up and running in the near future. Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Why .udeb won't work.
On Thu, Oct 26, 2000 at 02:47:24PM -0700, Joey Hess wrote: Erik Andersen wrote: why not cd app-1.0 dpkg-buildpackage cd .. mv app-1.0.foo.deb app-1.0.foo.udeb * signed changes files that contain ".deb" * reproducability by autobuilders Do it during debian/rules? Edit the files list in place? Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: busybox in main
On Fri, Oct 27, 2000 at 08:42:00AM +1000, Anthony Towns wrote: Maybe reserving disks-* for boot-floppies and creating a new install-* for debian-installer would be a good thing? If not, just s/install-/disks-/g. Calling them udebs might be helpful to avoid naming conflicts in the pool. Probably _something_ is necessary to avoid naming conflicts in the pool at any rate, if not .udeb, then...? I'm of the opinion that udeb and install-arch are both very good ideas, FWIW. The release notes should go in a regular debian package, as far as I'm concerned. There is no point in using a udeb unless we want to install the release notes onto the installer for some reason. Wouldn't it be useful to be able to read the release notes before/while you install? I'd guess that would have to stay byhand then... Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: busybox in main
On Wed, Oct 25, 2000 at 03:28:42PM +1100, Glenn McGrath wrote: Whats the point in following a policy that just gets in the road. I really dont understand why our (the installer team) work has to get pushed away into some dark hidden corner. Because the whole point of this argument is that they should not be installed on a normal system, and so they shouldn't be where they might accidentally be? My personal opinion is that Joey is trying to make a compromise by having a seperate area for installer packages, but as far as i know the installer team has no idea whatsoever if or when there will be a seperate area to put installer specific modules. Did you read Joey's original message? He said that he had settled the details with ftpadmin! Dan /\ /\ | Daniel Jacobowitz|__|SCS Class of 2002 | | Debian GNU/Linux Developer__Carnegie Mellon University | | [EMAIL PROTECTED] | | [EMAIL PROTECTED] | \/ \/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
