Slacko 5.3.3 bug reports and feedback

Please post any bugs you have found
Message
Author
simes
Posts: 61
Joined: Sat 29 Sep 2007, 09:35
Location: Cheshire, UK

#141 Post by simes »

Thank you ETP. I've done as you suggest and it helps me a lot.

User avatar
01micko
Posts: 8741
Joined: Sat 11 Oct 2008, 13:39
Location: qld
Contact:

#142 Post by 01micko »

Important announcement in the main post regarding up to date patches.
Puppy Linux Blog - contact me for access

tlchost
Posts: 2057
Joined: Sun 05 Aug 2007, 23:26
Location: Baltimore, Maryland USA
Contact:

Drive Icons

#143 Post by tlchost »

I am running 533 as a bootable USB, I notice that while I can mount a second USB drive (Other than the one I use to boot Slacko), the icon does not appear on the desktop.

Anyone else notice this? Is there a fix? Could someone test this in the beta version?

thanks

Thom

User avatar
rcrsn51
Posts: 13096
Joined: Tue 05 Sep 2006, 13:50
Location: Stratford, Ontario

Re: Drive Icons

#144 Post by rcrsn51 »

tlchost wrote:I am running 533 as a bootable USB, I notice that while I can mount a second USB drive (Other than the one I use to boot Slacko), the icon does not appear on the desktop.Anyone else notice this?
This works correctly for me.
Is there a fix?
Try running the Desktop Drive Icon Manager and doing a Redraw operation.

tlchost
Posts: 2057
Joined: Sun 05 Aug 2007, 23:26
Location: Baltimore, Maryland USA
Contact:

Re: Drive Icons

#145 Post by tlchost »

rcrsn51 wrote:
tlchost wrote:I am running 533 as a bootable USB, I notice that while I can mount a second USB drive (Other than the one I use to boot Slacko), the icon does not appear on the desktop.Anyone else notice this?
This works correctly for me.
Is there a fix?
Try running the Desktop Drive Icon Manager and doing a Redraw operation.
Thanks

That seems to work, although you must restart x to have it take.

Also the option to erase and then redraw and realign tick box does not stay ticked, as well as the tick box for autolaunch handler.

Jasper

#146 Post by Jasper »

Hi 01micko,

I am staying with Slacko 5.3.1 main (it's stable, fast and ideally suited to my needs and preferences) so I very much appreciate the continuing support with patches for versions 530 to 533 reported under the heading "IMPORTANT" in your opening post on page 1.

Conveniently the columns can be sorted by name, date or size; so may I ask if using

http://mirrors.dotsrc.org/slackware/sla ... /packages/

is an alternative way to download the patches/updates?

My regards

PS I'm sure we all wish you well in your upcoming exams.

tlchost
Posts: 2057
Joined: Sun 05 Aug 2007, 23:26
Location: Baltimore, Maryland USA
Contact:

Re: Drive Icons

#147 Post by tlchost »

tlchost wrote:
rcrsn51 wrote:
tlchost wrote:I am running 533 as a bootable USB, I notice that while I can mount a second USB drive (Other than the one I use to boot Slacko), the icon does not appear on the desktop.Anyone else notice this?
This works correctly for me.
Is there a fix?
Try running the Desktop Drive Icon Manager and doing a Redraw operation.
Thanks

That seems to work, although you must restart x to have it take.
It seems that I have to restart x after a reboot to get this to work. Other than placing a command to restart x in the startup folder, which would create an endless loop, wthat command can I add to slcko to restart-x
immediately after boot?

Thanks

Also the option to erase and then redraw and realign tick box does not stay ticked, as well as the tick box for autolaunch handler.[/quote]

npierce
Posts: 858
Joined: Tue 29 Dec 2009, 01:40

#148 Post by npierce »

Subject: Running xorgwizard changes option settings for synaptics touchpad driver.

This is a bug in xorgwizard that affects Slacko 5.3.3 (and any other Puppy that uses an Xorg X server version 1.9). This won't affect future Slackos or other Puppies that use an Xorg X server version 1.10.0 or newer, or a version older than 1.9.


o To reproduce problem:

1. Assuming that your PC uses the Xorg X server and the synaptics driver, boot Slacko 5.3.3 without a save file and try your touchpad, noticing its speed and acceleration.

2. Menu -> Shutdown -> Exit to prompt

3. Run xorgwizard and make your usual choices.

4. Run xwin and try your touchpad. Notice that the speed and/or acceleration will have changed.


o Background:

In June and August of 2010, the developers at X.org made significant changes to the way that the synaptics touchpad driver handles acceleration. These changes were first released in version 1.3.0 of the synaptics driver on August 31, 2010 (See synaptics-1.3-branch), in the same month that Xorg X server 1.9.0 was released. Because of these changes, the values for the "AccelFactor", "MinSpeed", and "MaxSpeed" options previously specified in xorg.conf for use with earlier versions of the synaptics driver were no longer appropriate. Using those values with the new driver would make the pointer movement noticeably sluggish.

Puppy's xorgwizard-automatic script handles these changes by adjusting the "Synaptics Mouse" section of xorg.conf that it creates according to the version of the Xorg X server being used.

Slacko's /etc/DISTRO_SPECS file defines DISTRO_XORG_AUTO='yes', and the Xorg X server in Slacko 5.3.3 is version 1.9.5. So, when xwin sees that the xorg.conf file needs to be created, and that an Xorg X server version 1.7.0 or newer is to be used, it runs xorgwizard-automatic instead of xorgwizard. So far, this is good.

When xorgwizard-automatic detects a supported touchpad, and sees that an Xorg X server version 1.9.0 or newer is to be used, it creates an "InputDevice" section in xorg.conf for the "Synaptics Mouse", which doesn't load any options for the driver. So the driver uses its default settings. This is also good.


o The problem:

But, if the user decides to run xorgwizard (not xorgwizard-automatic) to choose a different screen resolution or whatever, xorgwizard behaves differently. This is because, when creating the "Synaptics Mouse" section, xorgwizard compares the Xorg X server version to 1.10.0, not 1.9.0 as xorgwizard-automatic does.

When xorgwizard detects a supported touchpad, and sees that an Xorg X server version older than 1.10.0 is used, it creates an "InputDevice" section in xorg.conf for the "Synaptics Mouse", which loads various options for the driver. Some of these settings are very different from the driver's default settings, since they were originally used with older versions of the synaptics driver. The user will likely notice that the pointer moves much more slowly than previously. This is not good.

The fix is simply to change this line in xorgwizard:

Code: Select all

 if vercmp $XORGVER lt 1.10.0;then #110701
to:

Code: Select all

 if vercmp $XORGVER lt 1.9.0; then #110701, 121014
(Having an XOrg X server 1.9.0 or greater is not an absolute guarantee that the synaptics driver is 1.3.0 or greater, or vice versa, but it is highly probable. Of course, there was a brief window in late August of 2010 when someone could have grabbed the latest versions and got the X server 1.9.0 and the old synaptics 1.2.0 driver, but I'm not aware of any Puppies with that unlikely combination.)

Note that if the user ran flsynclient before ever running xorgwizard, didn't adjust the acceleration, but did click Save, the good values will have been saved and should be restored when rebooted. In this case, one might reasonably expect that even if the user later runs xorgwizard, the user would experience a sluggish pointer only briefly until the load-touchpad-settings script ran and restored the good settings.

Unfortunately, that is not the case. Although the good settings are restored, the actual speed and/or acceleration of the touchpad are not the same as they were when first booted. Without digging deeply into the source code, my best guess is that one or more internal values are now calculated when the driver first starts, based on the speed and acceleration values at that time, and those internal values don't get recalculated when MinSpeed, MaxSpeed, and AccelFactor values later get reset. That's just a guess at what's happening, but whatever is happening, the bottom line is that if you change the values that are used when the driver is first loaded, values loaded later will cause different behavior than the same values did previously.

Let me clarify. Although the values saved by flsynclient (that previously worked well before the user ran xorgwizard) will result in different behavior after the user runs xorgwizard, it may be possible for the user to then use flsynclient again and achieve acceptable (but not ideal) behavior by changing the MinSpeed and Max Speed values. And, if saved, those new values will continue to work on future restarts of X, unless the synaptics options in xorg.cong are changed again (as would happen if the user ran xorgwizard-automatic again).

Unfortunately, any attempt to adjust the acceleration with flsynclient will probably result in less acceleration than desired. This is because flsynclient was designed for the older versions of the synaptics driver, which used smaller values for "AccelFactor". The limit in flsynclient is 0.01, but a typical value for use with the new driver (when using a Synaptics touchpad) is 0.04. (If the value is already above flsynclient's 0.01 limit, flsynclient will save that value OK, but any attempt to change the value will result in a value of 0.01 or less.) If a user runs into this problem, it can be fixed with synclient, like so:

Code: Select all

synclient AccelFactor=0.04
(If one has time for such things, one can observe some of the confusing behavior of the new synaptics driver by setting the MinSpeed option in xorg.conf to any value other than one. (To see this, the /root/.flSynclient file, if any, will need to be temporarily removed or hidden to prevent the values from being overwritten by load-touchpad-settings after restarting X.) After X is restarted, synclient -l will list the current values. Confusingly, the values that were given in xorg.conf for MinSpeed and MaxSpeed will have been changed by the driver. They have been multiplied by the reciprocal of MinSpeed. (Looking at the source code, I see that MinSpeed isn't actually multiplied by its reciprocal, but is simply set to one, which is of course the same result.) Is this any way to design a user-friendly interface? :))


o Not a problem, but while we are in the neighborhood:

As a side note, the following line that appears in three places in both xorgwizard and xorgwizard-automatic is now unnecessary:

Code: Select all

	Option "Device" "/dev/psaux"
I believe it was used with Linux 2.4 kernels. The line that follows it sets the "auto-dev" protocol, which automatically sets a /dev/input/event* device when used with 2.6 and newer kernels.

So recent Xorg X servers (including Xorg 1.9.5 in Slacko) mostly ignore the Option "Device" "/dev/psaux" line. But for some reason the Xorg 1.11.0 X server is confused by this line, if it finds it in xorg.conf, and will fail. But this is NOT a problem for any Puppy, since neither xorgwizard nor xorgwizard-automatic put that line in xorg.conf when an Xorg X server 1.10.0 or newer is to be used. So no change to Puppy needs to be made for that.

User avatar
01micko
Posts: 8741
Joined: Sat 11 Oct 2008, 13:39
Location: qld
Contact:

#149 Post by 01micko »

Bravo npierce!

Nice post and a link in the main post has been added directed to your post.

Thanks.

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

I have developed a simple updates manager to complement the posted patches repo.

See main post
Puppy Linux Blog - contact me for access

ICPUG
Posts: 1308
Joined: Mon 25 Jul 2005, 00:09
Location: UK

#150 Post by ICPUG »

Great work npierce.

I have noticed the behaviour and used flsynclient to sort myself out. Now I know why it happens. You must have spent a lot of time researching and checking code.

Thanks for that.

npierce
Posts: 858
Joined: Tue 29 Dec 2009, 01:40

#151 Post by npierce »

01micko and ICPUG,

You're welcome. Thanks for the kind words.

The new updates manager is a handy addition. Thanks 01micko.
ICPUG wrote:You must have spent a lot of time researching and checking code.
Yes. Sometimes I start looking into what I think is a simple issue, and before I know it I've run into other issues, and hours or days have gone by. :)

nooby
Posts: 10369
Joined: Sun 29 Jun 2008, 19:05
Location: SwedenEurope

#152 Post by nooby »

I'm testing Thin slacko just now.
I reused a Grandr pet I found on Smokey
and that one worked well.

Firefox has Namoroka 3.6.28 instead of FF 16.0.01
that my Lupu have now.

Does that make my browsing insecure maybe?
Any suggestion? Should I try to get FF instead
or what?
I use Google Search on Puppy Forum
not an ideal solution though

npierce
Posts: 858
Joined: Tue 29 Dec 2009, 01:40

#153 Post by npierce »

At http://www.murga-linux.com/puppy/viewto ... 385#647385,
Baldronicus wrote:The problem:
A right-click on a clear area of the desktop (under the default configuration) brings up the menu, as expected. After this is cleared (by a left-click), subsequent right-clicks on the desktop icons (including the drive icons) result in the desktop menu appearing, as if they were right-clicks on a clear area of the desktop (or as if the icons were not there, or simply not identified).
A subsequent left-click (or in some/many cases two spaced left-clicks) on one of the icons results in it working as expected. Following this, right-clicking on the icons also works as expected (as if the "problem trigger" had been reset).
Another right-click on a clear area of the desktop recreates the problem, and subsequent left-clicking on an icon clears it again.
ETP wrote:Bug confirmed. (Also exists in 5.3.3.2.) It would seem that we have all been living with it!
Minor as you say, but very confusing for anyone new to Slacko.
Jasper wrote:It is the same with Slacko 5.3.1 . . .
That is the ROX-Filer "ungrab bug" (a.k.a. "multiple root menus problem"). This has been patched. See Barry's blog entry: ROX-Filer multiple-jwm-menus fix

Note that the .pet offered in that blog entry is no longer the latest ROX-Filer .pet. For the latest, see Barry's blog entry: ROX-Filer squashed log-window

But those .pets may not be best for use with Slacko, since 01micko reports that "For some reason rox needs to be compiled in Slacko or else Fido refuses to work." (See his comment under Barry's blog entry: ROX-Filer focus problem fixed) So it is probably best to compile on Slacko from the patched source, which can be found in the file rox-filer-20111022-patched3.tar.gz at http://bkhome.org/sources/alphabetical/r/.

EDIT: Also note that installing a .pet would have side-effects. For instance, your PuppyPin file would be overwritten with the PuppyPin used for Wary and Racy. So, again, it's best to compile your own from the patched source, or wait for someone to make a Slacko version of the .pet. (Even then, if you have customized your PuppyPin, you would want to back it up before installing the new .pet.)

npierce
Posts: 858
Joined: Tue 29 Dec 2009, 01:40

PPM in Slacko: Some repo check boxes don't work.

#154 Post by npierce »

The following lines in pkg_chooser.sh prevent the use of quirky and wary5 repositories in slacko, even though the database files for both are included in the distro, and so check boxes are available in the configuration dialog -- they just don't do anything:

Code: Select all

# #100903 if wary5, want to list quirky repo, as has same code base...
# [ "$BASEREPO" = "Packages-puppy-quirky-official" ] && [ "$DISTRO_COMPAT_VERSION" != "wary5" ] && continue #100126 already handled above. 100903
# [ "$BASEREPO" = "Packages-puppy-wary5-official" ] && continue #100903 already handled above.
# [ "$BASEREPO" = "Packages-puppy-lucid-official" ] && continue #100911 already handled above.
To reproduce:

1. In Slacko 5.3.3, run PPM and choose Configure package manager (or, in the Ziggy UI, the button with the screwdriver and spanner).

2. Click either of the check boxes for puppy-quirky-official or puppy-wary5-official. (If you already had 5 boxes checked, you will need to uncheck one so as not to exceed the limit of 5.)

3. Click OK.

4. Exit from the PPM.

5. Run PPM again and note that the newly selected repo is not listed under Repositories.


Perhaps this is intentional to discourage users from installing stuff in Slacko from these older repositories. If that is the case, their database files should probably be removed to prevent them from showing up in the configuration dialog. (Although I am guessing this isn't intentional since the even older Puppy-2, -3, -4, and -5 repositories are available, and don't have this issue.)

User avatar
Lobster
Official Crustacean
Posts: 15522
Joined: Wed 04 May 2005, 06:06
Location: Paradox Realm
Contact:

#155 Post by Lobster »

Openshot with 3D rendering of animated titles from Blender
http://www.murga-linux.com/puppy/viewto ... 947#661947
Puppy Raspup 8.2Final 8)
Puppy Links Page http://www.smokey01.com/bruceb/puppy.html :D

Maian
Posts: 2
Joined: Wed 07 Nov 2012, 02:10

Ziggy_patch-2.pet fixed Puppy Package Manager!!!

#156 Post by Maian »

Thank you!!! I am a newbie to linux and thought I had broken my puppy or it's tail beyond all repair. 01micko's post fixed that. I installed the ziggy_patch_ppm-2.pet and that fixed the broken Puppy Package Manager. Tail wagging! :) Opps sorry-- I think I posted this in the wrong place.

sonicpuppy
Posts: 2
Joined: Fri 09 Nov 2012, 21:29

#157 Post by sonicpuppy »

Hi there !

My first post, so here goes the mandatory intro: hello to all around here doing a nice job of bringing the super-complicated linux system into puppy level ;-).

And now to the point. Perhaps not a bug as such (although I did stumble upon a few along my 3-day long puppy linux career) but can anyone tell me how to make the download manager in ff/seamonkey open downloaded files according to shell settings (or whatever) and open containing folders (that's in the right-click menu) ?

Using slacky 5.3.3 windows installer (from this forum) clean install, seamonkey as included and namaruka post-installed from a .pet.

EDIT: allright, have managed to get "open containing folder" working by selecting /usr/local/bin/rox on the Open With dialog (and setting as default); should be pre-configured within the puppy, methinks

npierce
Posts: 858
Joined: Tue 29 Dec 2009, 01:40

panel-buttons script corrupts .jwmrc-tray

#158 Post by npierce »

The /usr/local/jwmconfig2/panel-buttons script supplied with Slacko 5.3.3 adds one extra comment and deletes one line in /root/.jwmrc-tray every time a button is added to the tray.

The comment added is always a copy of this existing comment:

Code: Select all

		<!-- Additional TrayButton attribute: label -->
The line removed is the first non-blank line following the last button added.

If the original /root/.jwmrc-tray supplied with Slacko 5.3.3 is used as a starting point, adding buttons causes the following behavior:

The first time a button is added, the line deleted is just a comment, so the tray is created as expected. But when a second button is added, this line is deleted:

Code: Select all

                <Pager/>
So the pager buttons (for switching desktops) disappear.

Adding a third button just deletes another comment, but adding a fourth button deletes this line:

Code: Select all

                <TaskList maxwidth="25"/>
So the task list disappears.

Adding a fifth button deletes this line:

Code: Select all

                <Dock/>
And the dock is gone.

Adding a sixth button just deletes another comment, but adding a seventh deletes only the first line of a comment, which confuses JWM, and everything else vanishes -- in this case, the swallowed program (xload) and the clock.

Adding an eighth and ninth button deletes the last two lines of that same comment, so JWM is no longer confused, and the swallowed program and the clock reappear.

OK, I think I need not continue adding buttons. You get the idea.


The reason that lines are being deleted is that the number of lines remaining in .jwmrc-tray after the last line with a TrayButton definition is sometimes calculated incorrectly. And that happens because the total number of lines in .jwmrc-tray is sometimes calculated wrongly. The total is found using wc -l which counts newline characters, not lines. If the last line in the file has no newline character, the value returned by wc -l will be one shy of the actual number of lines.

The original .jwmrc-tray file supplied with Slacko 5.3.3 has no newline character in the last line.

While supplying a .jwmrc-tray file that ends with a newline character would make this problem go away at least initially, it wouldn't be the most robust solution. A user might manually edit the file and leave off the final newline -- something that is not ruled-out by the JWM documentation, and something that JWM itself is perfectly happy with. It would be better if the panel-buttons script could tolerate a configuration file that didn't end with a newline character.

An earlier version of the panel-buttons script (in jwmconfig2-20111110.pet in the puppy-common-official repository) included an October 2011 solution from Karl Godt, which appended a newline character to the file if it didn't already end with one. That works.

An alternative solution would be to use sed -n $= to count the lines, instead of wc -l. Using sed -n $= counts lines, not newline characters.

So line 62:

Code: Select all

total=`wc -l /root/.jwmrc-tray | sed -e 's/ *//' | sed -e 's/ .*//g'` ### get total number of lines
could be replaced by

Code: Select all

total=`sed -n $= /root/.jwmrc-tray` ### get total number of lines

As for the extra comment that gets added every time a button is added, that doesn't cause any problems (other than making the file grow unnecessarily), so could be ignored. But it could be fixed by filtering the comment out of CURRENTBTNSCODE by replacing line 24:

Code: Select all

CURRENTBTNCODE=$(grep "TrayButton" $HOME/.jwmrc-tray|grep -v "root:3")
with:

Code: Select all

CURRENTBTNCODE=$(grep '^[[:blank:]]*<TrayButton' $HOME/.jwmrc-tray|grep -v "root:3")
Line 25 is fine:

Code: Select all

CURRENTBTNS=$(grep "TrayButton" $HOME/.jwmrc-tray|grep -v "root:3"|\
But it could be given the same treatment as line 24, by changing it to:

Code: Select all

CURRENTBTNS=$(grep '^[[:blank:]]*<TrayButton' $HOME/.jwmrc-tray|grep -v "root:3"|\
That way, if the user happened to add a single-line comment containing the word "TrayButton", it would not become part of CURRENTBTNS, even if it didn't also contain the word "Additional". This would also eliminate the need for line 26.


By the way, although I ran into this problem in Slacko 5.3.3, it looks like the same panel-buttons script is in jwmconfig2-121105.pet, which I think is used by Slacko 5.4.
Attachments
panel-buttons.diff.gz
suggested changes for script
(702 Bytes) Downloaded 549 times
panel-buttons.gz
script revised with suggested changes
(1.84 KiB) Downloaded 562 times

User avatar
don570
Posts: 5528
Joined: Wed 10 Mar 2010, 19:58
Location: Ontario

#159 Post by don570 »

pupmd5sum has been upgraded to version 1.0

It can save the md5sum to a text file just like gtkhash


http://murga-linux.com/puppy/viewtopic. ... 673#672673

_________________________________________________________

User avatar
01micko
Posts: 8741
Joined: Sat 11 Oct 2008, 13:39
Location: qld
Contact:

#160 Post by 01micko »

This is the latest gtkdialog from Thunor.

It is the same as version 0.8.3 even though it is 0.8.4, it comes straight out of svn which carries the next version number.

This is required for the latest Pburn and the upcoming Pmusic form zigbert (although an ffmpeg update will be required too for that)

Enjoy!

NOTE: the VTE package is mandatory! INSTALL IT FIRST!!! thanks to 8-bit for that tip :wink:
Attachments
gtkdialog-0.8.4-release-slacko-5.3_legacy.pet
the main package INSTALL VTE first!!!!!
(92.26 KiB) Downloaded 717 times
gtkdialog_DOC-0.8.4-release-slacko-5.3_legacy.pet
Documentation and examples
(172.56 KiB) Downloaded 436 times
Puppy Linux Blog - contact me for access

Post Reply