Multiple Sound Card Wizard 2.2

01micko posted a fix here:
http://murga-linux.com/puppy/viewtopic.php?t=84298&start=75

The script was originally created by Kirk, and hacked on by various people. Rodin.s internationalized it recently.
I have now added 01micko's fix, and hacked on that a bit -- don't really know if I improved it though!

The PET (2K):
http://distro.ibiblio.org/quirky/pet_packages-noarch/mscw-2.2.pet

Note to 01micko: take that script out of z_base_hacks PET.


Posted on 16 Feb 2013, 20:14


Comments:

Posted on 17 Feb 2013, 8:33 by BarryK
z_base_hacks
I am taking a look at what else is in that package...

/bin/mount looks like it should be removed. It is an old version. Early in 2012 shinobar applied a fix for ntfs-3g, adding the "silent" option (which solves the problem of ROX-Filer reporting failed move/copy):

ntfs-3g $CMDPRMS -o umask=0,no_def_opts,silent

Whereas, 01micko has removed "no_def_opts":

ntfs-3g $CMDPRMS -o umask=0



Posted on 17 Feb 2013, 8:50 by BarryK
z_base_hacks
Hmmm, a lot more scripts in that PET.

It is unwise to have Woof scripts in a PET, that override the official ones, as the official ones keep evolving.

Any fixes should be reported and included in the official scripts.

For example, /etc/rc.d/rc.shutdown. 01micko's version had the last official change in 121104, whereas the official script has had two changes since then, in 121125 and 130128.

...01micko's rc.shutdown has this line after the 'service scripts' are shutdown, whereas I have it before:

umount-FULL -a -t cifs,smbfs,nfs,sshfs

...OK, I have moved it down too. 01micko, please remove rc.shutdown from z_base_hacks.



Posted on 17 Feb 2013, 8:11 by 01micko
re Z
Yes, kept that one in (shutdown, will remove), however, I usually diff them all on a woof upgrade.

You will notice significant differences in /etc/rc.d/rc.sysinit, and rc.update.

This is for 2 reasons.
1. Slackware has moved to kmod to load modules. busybox depmod does not work at all. Also, alsaconf is broken due to this, not a big deal.

2. There is also the "BUG", there is more to post about this but you can see a table I created <--
https://docs.google.com/spreadsheet/ccc?key=0Atag-cdyFbQmdG9ZVkR5VXp1SUxCMzV2T0tQMVJrNmc&usp=sharing#gid=0

You can download that as an xlsx file and it opens in gnumeric.

As I said, more to post about this but I haven't time ATM. I need to go into depth.


Posted on 17 Feb 2013, 8:19 by 01micko
re Z again
That z_base_hacks you have wasn't the final one released in slacko-5.4.0.3 beta. I updated woof before that and it has the correct init script.
There is other stuff too, like the connectwizard that have been removed since inclusion of rerwin's stuff.


Posted on 17 Feb 2013, 8:22 by BarryK
z_base_hack: rc.sysinit
I am now looking at /etc/rc.d/rc.sysinit in z_base_hacks.

There has been one official change since 01micko "forked" the script.

01micko has added this line:

# Iguleder for Slacko based on slackware-14 120806 # jamesbond, link to /tmp
[ ! -h /run ] && ln -sf /tmp /run


It is not just Slackware 14 that is using /run, see here:
http://wiki.debian.org/ReleaseGoals/RunDirectory

So, I have added a top-level /run in Woof, just a symlink to /tmp. This is in the directory hierarchy in rootfs-skeleton in Woof, so the rc.sysinit script does not have to create it.

The only other change that 01micko has made is to comment-out use of the Busybox depmod:

#if [ "`busybox | grep 'depmod'`" != "" ];then
# busybox depmod #use busybox depmod if available.
#else
depmod # 120806 hack for slacko based on 14
#fi


...that's an awkward one. The reason that we prefer the Busybox version is that the full version needs a lot of physical RAM -- it takes ages and even hangs when not enough RAM (I don't recall the amount of RAM that causes the problem, and I seem to recall that a swap partition does not fix it).

Even with plenty of RAM, the busybox version is much faster.

All of my pup builds have coped with the busybox depmod, so I am going to leave it as-is.
Though, if I had to, I could put a test if running Slackware, then use the full depmod.



Posted on 17 Feb 2013, 8:50 by BarryK
z_base_hacks: rc.update
01micko,
Your post overlapped mine, just got in first!

Yes, rc.update has more differences. I will look into it.

I also have run out of time, gotta go and do other things for most of the day.



Posted on 17 Feb 2013, 16:37 by BarryK
Woof scripts
It is unwise to have Woof scripts in a PET, that override the official ones, as the official ones keep evolving.

I intended this comment as a general warning to everybody who uses Woof or who creates PETs.

01micko maintains some "forked" scripts, as he sees some good reasons to do so, and indeed there are some issues that I have not yet addressed such as kmod and a philosophical difference over where some cache files should reside

The main warning to anyone who might want to "fork" a script is that it is a lot of work in the future, keeping it synced with the latest in Woof. Everytime Woof gets updated, you have to check the script for changes and apply diffs to your own script.

There are some PETs out there that do have Woof scripts in them. Hence my warning.
Try as much as possible to get me to incorporate any fixes into the official script, rather than forking it.

I think that some of the early langpacks had some forked scripts, but I hope that is now resolved.



Posted on 17 Feb 2013, 24:02 by vicmz
woof scripts in pets
I do offer some woof updates in a .pet, but separate from langpack_es, mainly internationalized scripts that have been tested in the targent Puppy versions. New features involving deeper testing and follow-up (such as udf support) are not taken, as the goal is to make a static package. I clearly state that such .pet update is not mandatory, that it simply adds language support for certain aplications, and encourage everyone to use the latest Puppy Linux instead whenever possible. Fixes and new versions of packages that have a pinstall are offered separately, too. For anyone using Momanager I tell them to use the latest Puppy build, which is currently Precise 5.5 alpha 1 (by the way, it seems very stable for an alpha).