Lucid Puppy Revitalized as 5.2.8.7 - Nov., 2017 - Final

[TwoPuppies]

I have GQview 2.0.4 installed in Lucid-5.2.8.7.

It has always worked just fine, but the other day I clicked on the Help button in GQview to open the User Manual (something I haven't done for a while), and nothing happened. The User Manual is in HTML format, and usually opens in Seamonkey.

I have Seamonkey-2.48 installed, including the modifications provided here to enable it to run and update correctly in Lucid. A bit of further investigation revealed that the User Manual opens without any problem in Seamonkey-2.46. Trying to open it in version 2.48 results in the following error:

Code: Select all

(gconftool-2:29773): GConf-CRITICAL **: gconf_engine_get_local_for_addresses: assertion `addresses != NULL' failed
XPCOMGlueLoad error for file /usr/lib/seamonkey/libxul.so:
libxcb-shm.so.0: cannot open shared object file: No such file or directory
Couldn't load XPCOM.

This would seem to indicate that the problem lies with the glibc219_runtime modifications that have been applied to get Seamonkey-2.48 working

Any idea how I might get the GQview User Manual up and running in Seamonkey again without needing to revert to Seamonkey-2.46?

Edit: I probably should make it clear that the User Manual displays perfectly well in Seamonkey-2.48 if I go to the actual HTML Index page and click on that. Using the same method, it also works with Midori, Dillo, and presumably any other Browser. It just will not open from within GQview any more.

[Mike Walsh]

@Richard/anyone:-

Been running lupusuper2-5.2.8.7-k3.2.48-20150702 for quite a while now. Very much enjoying it, too. Now, then...

A question:-

Would you (or anyone else, for that matter) happen to have, or know where I can get hold of, the correct devx for this particular build? I've tracked down the kernel sources package in Ally's superb archives.....but I'd like to have both, obviously (and the devx is more often needed for 'straight-forward' compiling.)

TIA!


Mike.

[Mike Walsh]

I have assembled the next version of 5.2.8.7 for 2017. The 2017 release disk images are here:
http://www.mediafire.com/?5il6yef4h3sdj

Hi Rerwin,

Just wanted to give you feedback on this latest release. Mostly all good. Only two issues I can find after using it all weekend and again tonight: redshiftgui loads and runs, but nothing happens (I'll later find the problem with this easily, not a big deal right now) and, second, on this core i3 laptop that I 'frugal' installed lupu too, it cannot find any sound. I know the sound works as many other pups have used this laptop. It's a bit of a problem, but I am not sure what to do to fix it. Thought I'd ask you first before I go rooting around trying my fixes. Also, I will flop this 'frugal' install over to my P4 machine tomorrow and see if the sound (and redshift) problems appear there.

As a side note, I'm sure gonna hate to this lupu end. Why? Simple: it still runs great and I can see your hard work and touches everywhere. It's nice to see a pup creator/ maintainer be like this. You've done great with this, Richard, and deserve credit . Anyhow, here's some pics, the titles explain them, Palemoon installed from the Menu script so fast I couldn't believe it, and the pop-up boxes for everything that you do in Lupu, like I said, your touches & attention to detail are everywhere:

@belham2:-

With regard to your redshift 'problem', it sounds like it might be your hardware that's causing the issue. However, have a look here:-

http://www.murga-linux.com/puppy/viewtopic.php?t=111450

SCT simply allows you to set the colour 'temperature' directly, in the terminal. I've tried this out in a few Pups, and it's as easy as pie to set up. It doesn't have redshift's automatic bluer/redder back-and-forth with the sunrise/sunset and your location, it's true.....but it does work.

Just load your devx, enter the line given in the 'Read.me' into the terminal, hit 'Enter', and it should compile the binary without issue. Then move the binary into /usr/bin, and enter the colour temp value you want directly into the terminal.

Works a treat; it's worth a try.....and it's a long-standing, pre-existing *nix tool that's been around for a long while, apparently.


Mike.

[rerwin]

I have Seamonkey-2.48 installed, including the modifications provided here to enable it to run and update correctly in Lucid.

glibc219_runtime is already part of the beta iso. Please uninstall the pet, to use the embedded version, which might contain a subtle difference.

[rerwin]

A question:-

Would you (or anyone else, for that matter) happen to have, or know where I can get hold of, the correct devx for this particular build?

The kernel source and devx SFS files are in "Supplemental SFS Files" in the third message of this thread. The devx file is for all releases of all sulus:
http://www.mediafire.com/file/pbqakqs1p ... _002-4.sfs

[Mike Walsh]

A question:-

Would you (or anyone else, for that matter) happen to have, or know where I can get hold of, the correct devx for this particular build?

The kernel source and devx SFS files are in "Supplemental SFS Files" in the third message of this thread. The devx file is for all releases of all sulus:
http://www.mediafire.com/file/pbqakqs1p ... _002-4.sfs

Thanks for that, Richard. Much appreciated.

Cheers!


Mike.

[TwoPuppies]

glibc219_runtime is already part of the beta iso. Please uninstall the pet, to use the embedded version, which might contain a subtle difference.

I was actually using the previous (December 2016) version, not the current 2017 Beta version. So, following your advice, I started from scratch by booting the 2017 Beta version. I then installed GQview 2.0.4 and Seamonkey-2.48, and found that the GQview User Manual issue was solved. It seems that you were right regarding a difference in the embedded version of glibc219_runtime. Thanks for your help.

The above is actually the first time that I have tried out the new 2017 Beta version. I guess you would value some feedback:

There is a display issue. I have an Acer Aspire 5741 with a display resolution of 1366x768 (see below for more details).

The December 2014 release of Lucid booted straight to the desktop with the correct display resolution and without any issues: x-server: versa

The December 2016 release did likewise: x-server: fbdev. The intel driver also works here if it is chosen manually. Both the 2016 and 2014 versions also work OK if I download and install Xorg_High-1.1-Lucid.pet.

By contrast, the current 2017 Beta release boots to a blank screen, and then hangs. The only way to proceed is to turn off the computer and then start again. By booting to the command prompt instead, and then playing about with Xorgwizard, I can make it to the desktop, but only by using the versa driver with a low display resolution of 1024x768. I can't obtain a higher resolution using any method that I have thought of.

Any ideas?

Display details:

Code: Select all

  2.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller (rev 02)  
  oem: Intel(R)Ironlake Mobile Graphics Chipset Accelerated VGA BIOS
  product: Intel(R)Ironlake Mobile Graphics Controller Hardware Version 0.0

X Server: Xorg   Driver: intel
X.Org version: 1.7.6
  dimensions:    1366x768 pixels (361x203 millimeters)
  depth of root window:    24 planes

By the way Richard, I just wan to say how much I appreciate all the time and effort that you (and presumably a few others, I guess) have been putting into keeping Lucid going. Don't ever feel disheartened when things get hard. You are doing a great job. Lucid is easily the best version of Puppy Linux that was ever created. That's why so many of us continue to use it. Keep up the good work.

[rerwin]

By contrast, the current 2017 Beta release boots to a blank screen, and then hangs. The only way to proceed is to turn off the computer and then start again. By booting to the command prompt instead, and then playing about with Xorgwizard, I can make it to the desktop, but only by using the versa driver with a low display resolution of 1024x768. I can't obtain a higher resolution using any method that I have thought of.

Thanks very much for your feedback and encouragement. Curiously, the xorg drivers for your card are unchanged between the 2016 lupu and the beta. Since your graphics is Intel, please try booting afresh using the boot argument:
i915.modeset=0

If it still hangs, also use pararmeter pfix=nox and then xorgwizard to choose the intel xorg driver (as you have been doing).

To give me more to work with, please post the result of the command:
lspci -nn | grep 'VGA'

Also, run pupscan and post the entry for the VGA PCI interface, to confirm which kernel driver/module is used.
Richard

[sheldonisaac]

.. the xorg drivers for your card are unchanged between the 2016 lupu and the beta. Since your graphics is Intel, please try booting afresh using the boot argument:
i915.modeset=0

If it still hangs, also use pararmeter pfix=nox and then xorgwizard to choose the intel xorg driver (as you have been doing).

To give me more to work with, please post the result of the command:
lspci -nn | grep 'VGA'

Also, run pupscan and post the entry for the VGA PCI interface, to confirm which kernel driver/module is used.

Thanks, Richard; I'll try those on this computer

---------
UPDATE Oct 8
I did all those tests, and got the same results as TwoPuppies, at
http://www.murga-linux.com/puppy/viewto ... 243#970243. I guess our video hardware is similar.

Is it something to do with drivers? firmware? (I don't understand exactly what they are),

[rerwin]

I have uploaded Lucid Super 2 as version 20171007 and made updates to the change list on the previous page here:
http://www.murga-linux.com/puppy/viewto ... 772#968772

I found that my GeForce 6150SE graphic card was not handled correctly in sulu2 even though it works in lupu. It sometimes used the newer nv xorg driver but if probed for, would use the older version, which worked.

To get it to work with the newer version, I implemented "modeset0" rules for that card family but only for the sulu2 kernel (k3.2.48). If anyone else has problems with the nVidia cards, please report them here, including the output from the command:
lspci -nn | grep 'VGA'

But also try the "nouveau.modeset=0" kernel parameter to see if that makes it work. If it does, I can add that to the list of cards requiring KMS mode 0, to make the kernel parameter unnecessary.
Richard

[TwoPuppies]

please try booting afresh using the boot argument: i915.modeset=0

Result: Same as before. Boots to a blank screen and then hangs.

If it still hangs, also use pararmeter pfix=nox and then xorgwizard to choose the intel xorg driver (as you have been doing).

Result: Same as above.
I can still only reach the desktop using the Versa driver at 1024x768.

To give me more to work with, please post the result of the command:
lspci -nn | grep 'VGA'

Code: Select all

00:02.0 VGA compatible controller [0300]: Intel Corporation Core Processor Integrated Graphics Controller [8086:0046] (rev 02)

Also, run pupscan and post the entry for the VGA PCI interface, to confirm which kernel driver/module is used.

Code: Select all

DESCRIPTION: VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller 
VENDOR: 8086  DEVICE: 0046

I have also just downloaded and tested lupusuper2-5.2.8.7-k3.2.48-20171007-beta, and that is working perfectly, so I'm not really getting all that bent out of shape about lupu-5.2.8.7-20170923-beta. Even without the display issue, I would probably end up using lupusuper2 anyway, so don't feel that you need to invest too much time and effort in this.

[rerwin]

TwoPuppies,
Thanks for the good news of success with sulu2!

DESCRIPTION: VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller
VENDOR: 8086 DEVICE: 0046

Was there no line showing the kernel module for it?

Anyway, this card seems to be common. So we can maybe help those who need to stay on lupu by working this problem. Could you still get me a pdiag file after starting a fresh ('pfix=ram,nox' or pupsave 0) lupu bootup to the black screen, entering xorgwizard, choosing i915, testing it and starting xwin? If that fails, without rebooting, use xorgwizard to use xvesa (however you have been doing that -- the xvesa button or 'choose' selection -- tell me which).

I need that file so I can look in the logs and other places for clues. TIA
Richard

[sheldonisaac]

TwoPuppies,
Thanks for the good news of success with sulu2!

DESCRIPTION: VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller
VENDOR: 8086 DEVICE: 0046

Was there no line showing the kernel module for it?
I need that file so I can look in the logs and other places for clues. TIA

Richard, I'll do those tests also, hopefully today. And will try the new SuLu2, and it'll probably work.
I'm in tahr just now, and did pupscan; here's part of the results:

DESCRIPTION: Intel Corporation Core Processor Integrated Graphics Controller
VENDOR: 8086 DEVICE: 0046 KERNEL MODULE: i915

[Walter Dnes]

@rerwin can you apply additional CFLAGS/CXXFLAGS when you do the Puppy build? I'm thinking specifically of "-fno-unwind-tables -fno-asynchronous-unwind-tables". This noticably reduces the size of object files and binaries. See messages http://lists.busybox.net/pipermail/busy ... 78326.html and http://lists.busybox.net/pipermail/busy ... 78331.html for details. The "busybox" maintainers have incentive to produce smaller binaries, because busybox is often used in memory and space-constrained environments.

Older machines targetted by Puppy linux could use this help too. Executables/libraries will load faster, due to being smaller. And once loaded, they chew up less memory... win-win all around. This does not sacrifice speed like the "-Os" option. The only downside is that it might interfere with advanced debugging. BTW, I've been using this option on my Gentoo linux machines at home for years without problems. Gentoo builds almost everything from source. If this option caused problems, I would've found out "the hard way" a long time ago.

[keniv]

@rerwin

I have just installed your updated beta version of lucid from the beginning of this thread. So far I've always been able to run the updated versions of lucid on what is a very limited old laptop (800MHz 320MB). I know there has been some discussion as to which browser can be used in such a limited machine. I cannot run Seamonkey 2.48 supported, using watchdog's technique with glibc 2.19 as I get what seems to be random distortion of drop down boxes and unreadable text previously mentioned. I know glibc 2.19 is included in the beta version of lucid. I have tried with both the glibc 2.19.pet uninstalled and installed but get the same result. I am posting using the Seamonkey 2.46.pet which you provided but is a bit slow. I have used QtWeb in the past which has it's problems but is fast and OK for my use of a browser. However, it won't run on this updated version of lucid. I also like netsurf 3.6 which is still being supported. There is a pet for version 3.0 but this crashes all the time. There is also a deb for this. Is it possible to get one or other of these "lite" smaller browsers to run on the new updated lucid?

Regards,

Ken.

[rerwin]

I cannot run Seamonkey 2.48 supported, using watchdog's technique with glibc 2.19 as I get what seems to be random distortion of drop down boxes and unreadable text previously mentioned. I know glibc 2.19 is included in the beta version of lucid. I have tried with both the glibc 2.19.pet uninstalled and installed but get the same result. I am posting using the Seamonkey 2.46.pet which you provided but is a bit slow. I have used QtWeb in the past which has it's problems but is fast and OK for my use of a browser. However, it won't run on this updated version of lucid. I also like netsurf 3.6 which is still being supported. There is a pet for version 3.0 but this crashes all the time. There is also a deb for this. Is it possible to get one or other of these "lite" smaller browsers to run on the new updated lucid?

I ran into the same problem with SM 2.48 and have no solution for it. I reported that here (2 messages):
http://www.murga-linux.com/puppy/viewto ... 510#963510

That's why I have reverted to 2.46 in my main pupsaves and why I added support for palemoon, which does not (yet) require the glibc219 addition. Please try palemoon via: menu > Internet > palemoon installer. It seems to need about 100MB of pupsave space. I also upgraded sylpheed in case it can replace SM mail.

Taking a cue from your posting here:
http://www.murga-linux.com/puppy/viewto ... 461#952461
I experimented with the netsurf 3.6 Wheezy package. So far, it does not complain about needing glibc 2.17. It indicates the need for libssl and libcrypt 1.0.0 and apparently detects that the link to 0.9.8 won't do. I expect to try to find them, soon, to see if that helps. I am avoiding all of the chromium-based browsers because that just looks like a "can of worms" beyond my capability to troubleshoot.
Richard

[Robert123]

Just a quick note regarding upgrading Seamonkey to latest version from previous ones. There is a warning regarding upgrading to 2.48 from previous versions so a fresh install of Seamonkey 2.48 is required.

https://www.seamonkey-project.org/relea ... onkey2.48/

[keniv]

Helo rerwin,
Thanks for your quick reply. You must have a better memory than I have as I did not remember making the post you quoted from but it does illustrate the difficulty in getting a usable browser

Please try palemoon via: menu > Internet > palemoon installer. It seems to need about 100MB of pupsave space.

I was able to download and install palemoon using the installer but it will not run from the menu. I tried running it from a terminal but the output was illegal action. I also tried 3.6 Wheezy package and as you said it wants libssl and libcrypt 1.0.0.
I have also tried to install netsurf 3.6 from debs without success. I am not sure if I've got all the dependences. Thanks for the help in trying to get a lite browser to work with this latest update of lucid.

Regards,

Ken.

[rerwin]

Robert123,
That page is displayed whether or not you upgrade or install the full package. I found when upgrading from 2.46, an "update" was downloaded but after restarting SM the update is reported as "failed". But then a full package gets downloaded and installed. On restart, the big red message is still shown.

It didn't always do that, but 2.48 seemed to work except for sometimes showing drop-down lists or pages slanted.
Richard

[rerwin]

keniv,
I found a reason for the libssl problem with netsurf, so am continuing experimentation. So give me a little more time to work it.

I am surprised that palemoon fails for you. Please copy the last few lines of /tmp/xerrs.log (into a reply), which should have any error messages about it.

UPDATE: I resolved the libssl and libcrypto issues but am stumped by an undefined symbol message:

Code: Select all

netsurf: symbol lookup error: netsurf: undefined symbol: g_static_resource_init

It appears to tolerate these messages:

Code: Select all

netsurf: /lib/libssl.so.1.0.0: no version information available (required by netsurf)
netsurf: /lib/libcrypto.so.1.0.0: no version information available (required by netsurf)
netsurf: /usr/lib/liblber-2.4.so.2: no version information available (required by /usr/lib/libcurl.so.4)
netsurf: /usr/lib/libldap_r-2.4.so.2: no version information available (required by /usr/lib/libcurl.so.4)

I use watchdog's glibc219 wrapper technique, but that does not help with the undefined symbol. I'll have to "sleep on it".