Librepup 6.0.2.2

[musher0]

Which package have you used? Try to install lm-sensors from PPM. If this isn't the trisquel package, I'll investigate.

No, not the one from the Trisquel repo. The one in rcrsn51's CPU Temp pet
archive, version 1.9. (The URL for that thread is in my previous post.)

[musher0]

Hello everyone.

I don't know the cause of this (see attached pic). Is it the devx that isn't
loaded properly (boot manager says it's loaded) or the included Python
packages that are not complete?

I wanted to test a python program in LibrePup. I chose the up and coming
file manager Sunflower -- I couldn't run it. I tried the Debian
package (first one in that list) and the "generic" package (last one).

I tested Sunflower before on various Puppies [upupRaring-3.9.9.1 and
PPrecise-5.4.3, for ex.]; sometimes it runs and sometimes not.

BFN.

musher0

[Iguleder]

Try to install python-gtk2.

[01micko]

Re ca-certificates:

I suppose you tried something like this in the pinstall.sh? (untested!)

Code: Select all

#!/bin/sh
CWD=`pwd`
if [ "$CWD" != "/" ]; then #woof
	chroot rootfs-complete update-ca-certificates --fresh 1> /dev/null 2> /dev/null
else
	update-ca-certificates --fresh 1> /dev/null 2> /dev/null
fi

-----------------------------------


I think I see why my Qupzilla buid fails with HTML5 audio/video...

Trisquel's qtwebkit4 is built against gstreamer-1.0 whereas slackware's is built against gstreamer-0.10 (my qupzilla/HTML5 works in slacko). On some googling it turns out that it is a bug in webkit that isn't going to get fixed. I tried with every dep related to gstreamer-1.0 install (and rebuilt Qupzilla) and it still didn't work . I still think it's the best choice for default browser.

[Iguleder]

Haven't tried that yet, but yes - that's the general idea.

[01micko]

Anyway - my Dell Optiplex box is much happier with Librepup - intel graphics. I think I'll forget about the radeon box since I've killed 3 pupsaves trying to get acceleration going in the most 'libre' way possible. . I just can't get drm happy and the firmware blob loading. I wish these darn OEMs would at least open source their FW.

Any one know of a decent graphic chip vendor (modern) that has free FW? Intel graphics kind of sucks.

[Iguleder]

- off-topic -

I decided to postpone the first bug fix release until I'm done with my optimizations in woof-CE. I'm currently rewriting debdb2pupdb in C - even now, my version runs for about 100ms, while the original, Bacon-based one, takes three seconds. This should speed up the slowest part of 0setup, in both woof-CE and PPM.

[Billtoo]

- off-topic -

I decided to postpone the first bug fix release until I'm done with my optimizations in woof-CE. I'm currently rewriting debdb2pupdb in C - even now, my version runs for about 100ms, while the original, Bacon-based one, takes three seconds. This should speed up the slowest part of 0setup, in both woof-CE and PPM.

Drat...was looking forward to the bug fix release but will wait patiently for your next release.

Thanks

[01micko]

I don't like the new kernel (the latest one posted in this thread) on my dell/intel. No acceleration, crap resolution so I reverted to the original version. I couldn't be bothered stuffing around.

This was a good test of my kernel script. The backups worked fine. Actually, I could probably extend it to roll back to a backup if the new kernel sucks.

[musher0]

Try to install python-gtk2.

That fixed it. Many thanks.

[musher0]

- off-topic -

I decided to postpone the first bug fix release until I'm done with my optimizations in woof-CE. I'm currently rewriting debdb2pupdb in C - even now, my version runs for about 100ms, while the original, Bacon-based one, takes three seconds. This should speed up the slowest part of 0setup, in both woof-CE and PPM.

It's not off-topic. Take the time you need.
We'll all benefit in the end, with a higher quality pup.

[musher0]

Hello all.

In line with pemasu's post about evince, I ran the two script lines
at the bottom of his post and got evince to work properly.

However, the console reported that the 48x48 evince.png icon (please see
attached) was missing at /usr/share/icons/hicolor/48x48/apps.

BFN.

musher0

[mavrothal]

I'm currently rewriting debdb2pupdb in C - even now, my version runs for about 100ms, while the original, Bacon-based one, takes three seconds. This should speed up the slowest part of 0setup, in both woof-CE and PPM.

There is an old thread on this with an amazing conversion speed routine by Wosh that might be useful.
Had induced several changes by BK back in 2009 (1, 2, 3, 4)

[01micko]

The change kernels script is now out there for all to use and abuse.

[musher0]

Which package have you used? Try to install lm-sensors from PPM. If this isn't the trisquel package, I'll investigate.

Hi Iguleder.

Still no luck. As you suggested, I installed the Trisquel lm-sensors package
from PPM. This time, some sensors were detected. However, the sensors-
detect script can't append the list in file /etc/modules because
/etc/modules is a folder in LibrePup.

BFN.

musher0

[musher0]

I was trying out Trisquel 7.0 64 bit version and the default browser is Abrowser.
I installed it in PPM in Librepup and it works well, firefox under a different name I guess.

(...)

Hi, Billtoo.

Thanks for sharing your discovery.

Compared to icecat, abrowser loads in only 4 seconds from click on icon
to full display. I noticed a lot of python libs supporting it during the PPM
install; maybe that's why. In any case, I think I might use it as my default
browser on every Puppy.

With its html5 extension, abrowser plays songs on youtube without a
glitch. (Please see attached.) No need for the old Flash plugin.

Many more extensions than icecat, too.

Again thanks. BFN.

musher0

[Iguleder]

Had induced several changes by BK back in 2009

Right now, I'm ahead of the competition

The change kernels script is now out there for all to use and abuse.

Good stuff! Maybe you should add it to rootfs-skeleton.

[01micko]

The change kernels script is now out there for all to use and abuse.

Good stuff! Maybe you should add it to rootfs-skeleton.

That's the eventual plan.

[musher0]

@01micko:

I feel a bit responsible for the time you lost reconfiguring and testing the
new kernel. Looking back, it should have dawned on me why the ATI
option in xorgwizard broke Xorg: because the ATI|Radeon firmware is
closed code, it has no place in a totally open-source|libre distro.

Anyway, "live and learn", as the saying goes.

Suggestion: that any ATI option be removed in LibrePup's xorgwizard so
people with ATI|Radeon cards won't be tempted to try it -- and break X.

@Iguleder:
Is there a way that the "satellite" download consoles close automatically
once the PPM download and installation are finished? At first I didn't dare
close them by hand for fear of up

[01micko]

@01micko:I feel a bit responsible for the time you lost reconfiguring and testing the
new kernel.

Don't

Looking back, it should have dawned on me why the ATI
option in xorgwizard broke Xorg: because the ATI|Radeon firmware is
closed code, it has no place in a totally open-source|libre distro.

Anyway, "live and learn", as the saying goes.

Suggestion: that any ATI option be removed in LibrePup's xorgwizard so
people with ATI|Radeon cards won't be tempted to try it -- and break X.

I disagree. What has to happen, (bearing in mind the $$$ vote) for free software to succeed is that users have to try this and complain as loudly as possible.

--------------


My wife was listening to Taylor Swift tonight. Something like her first and second album. I remarked that the second sounds no different to the first. She objected.

Fine. The point is this. If we are dictated to then we accept that standard. I can play three chords on guitar and can sing terribly too. If the standard we accept is three chords with terrible singing then we don't have much hope.