Puppy 4.1.1 release candidate

[panzerpuppy]

Barry,have you seen the last post by happypuppy in the '(SP1) Gparted fix' thread? I'm experiencing the same problem in Puppy 4.1.
I wonder if this bug is already fixed for 4.1.1(RC).

Even more bugs to report

There's a strange bug in the Puppy Boot Configuration wizard:

In the wizard where you manually select additional modules to be loaded at bootup,if you add more than one module to the box on the right, none of those modules will get loaded after a reboot.

Instead of adding each module in a separate line,the wizard places all modules in one line (delimited by a space) as they were additional parameters for the first module.
Same for the box where you are asked to specify a parameter for a module.They're all mixed up.

A small but very annoying little bugger.

panzerpuppy

[timcriger]

I've also had the package manager hangs when trying to download... well, when i was attempting to download libstdc++.so, the only attempt i've made.
The orange dialog came up, but was never able to connect to ibiblio or nluug or any other. It seemed to be trying to connect to an odd IP address: 0.0. or something.
Can anyone else duplicate this on their computer?
Thanks,
Tim

[dogone]

I've also had the package manager hangs when trying to download... well, when i was attempting to download libstdc++.so, the only attempt i've made.
The orange dialog came up, but was never able to connect to ibiblio or nluug or any other. It seemed to be trying to connect to an odd IP address: 0.0. or something.
Can anyone else duplicate this on their computer?
Thanks,
Tim

Precisely my experience, but consistent and many times over. I don't recall the IP in the orange dialog precisely, but it contains mostly zeros and is most certainly invalid. I would suggest that you exercise PM a bit to convince yourself one way of the other. If your luck is anything like mine, you won't need to worry about disk space .
This issue would be a show stopper if real, so we need to pursue it. Is anyone else having such problems - or not?

[BarryK]

Barry,have you seen the last post by happypuppy in the '(SP1) Gparted fix' thread? I'm experiencing the same problem in Puppy 4.1.
I wonder if this bug is already fixed for 4.1.1(RC).

Even more bugs to report

There's a strange bug in the Puppy Boot Configuration wizard:

In the wizard where you manually select additional modules to be loaded at bootup,if you add more than one module to the box on the right, none of those modules will get loaded after a reboot.

Instead of adding each module in a separate line,the wizard places all modules in one line (delimited by a space) as they were additional parameters for the first module.
Same for the box where you are asked to specify a parameter for a module.They're all mixed up.

A small but very annoying little bugger.

panzerpuppy

This thread is for reporting 4.1.1rc bugs. That bug was (supposedly) fixed and announced on my blog just prior to release of 4.1.1rc.

[BarryK]

I've also had the package manager hangs when trying to download... well, when i was attempting to download libstdc++.so, the only attempt i've made.
The orange dialog came up, but was never able to connect to ibiblio or nluug or any other. It seemed to be trying to connect to an odd IP address: 0.0. or something.
Can anyone else duplicate this on their computer?
Thanks,
Tim

Precisely my experience, but consistent and many times over. I don't recall the IP in the orange dialog precisely, but it contains mostly zeros and is most certainly invalid. I would suggest that you exercise PM a bit to convince yourself one way of the other. If your luck is anything like mine, you won't need to worry about disk space .
This issue would be a show stopper if real, so we need to pursue it. Is anyone else having such problems - or not?

Well, the problem is that I cannot reproduce your problem, so I cannot fix it. You are really going to have to try and get into the code and try to find out what i wrong.

The script is /usr/sbin/petget, and 'wget' is used to download .pet packages. So, there may be some problem with wget. Maybe you could locate the line where wget is run, then try to execute that on its own in a terminal.

Looking in the code, I see that 'ping' is used to verify that a site exists, for example:

# ping -c 1 -q ibiblio.org ;echo $?

if that fails, that is, returns non-zero value, the code tries this:

# wget -t 2 -T 20 --waitretry=20 --spider -S ibiblio.org -o /dev/stdout | grep '200 OK'

...it's looking for that cring "200 OK"

The actual line that does the download is:

wget $PKGSRC/pet_packages-${PUPREPO}/${APKGNAME}.pet

Doing some example substitutions:

# wget http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet

So, try those in a terminal, see what happens.
(remember, highlight the text, then use middle-mouse-button to paste into the terminal window).

[pa_mcclamrock]

The PETget package manager on Puppy 4.1.1rc sometimes displays only the packages from Puppy 4; when I click the radiobuttons for Puppy 2 or 3 packages, there's nothing. Other times, the Puppy 2 and 3 packages show up correctly. I'm pretty sure the problem is not with my Internet connection (DSL modem + router + Ethernet), which works fine.

David McClamrock

[dogone]

I've also had the package manager hangs when trying to download... well, when i was attempting to download libstdc++.so, the only attempt i've made.
The orange dialog came up, but was never able to connect to ibiblio or nluug or any other. It seemed to be trying to connect to an odd IP address: 0.0. or something.
Can anyone else duplicate this on their computer?
Thanks,
Tim

Precisely my experience, but consistent and many times over. I don't recall the IP in the orange dialog precisely, but it contains mostly zeros and is most certainly invalid. I would suggest that you exercise PM a bit to convince yourself one way of the other. If your luck is anything like mine, you won't need to worry about disk space .
This issue would be a show stopper if real, so we need to pursue it. Is anyone else having such problems - or not?

Well, the problem is that I cannot reproduce your problem, so I cannot fix it. You are really going to have to try and get into the code and try to find out what i wrong.

The script is /usr/sbin/petget, and 'wget' is used to download .pet packages. So, there may be some problem with wget. Maybe you could locate the line where wget is run, then try to execute that on its own in a terminal.

Looking in the code, I see that 'ping' is used to verify that a site exists, for example:

# ping -c 1 -q ibiblio.org ;echo $?

if that fails, that is, returns non-zero value, the code tries this:

# wget -t 2 -T 20 --waitretry=20 --spider -S ibiblio.org -o /dev/stdout | grep '200 OK'

...it's looking for that cring "200 OK"

The actual line that does the download is:

wget $PKGSRC/pet_packages-${PUPREPO}/${APKGNAME}.pet

Doing some example substitutions:

# wget http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet

So, try those in a terminal, see what happens.
(remember, highlight the text, then use middle-mouse-button to paste into the terminal window).

Thanks Barry. Here's what I see in response to those command strings. Note that resolved address and the fact that the second wget command hangs until I break it:

# time ping -c 1 -q ibiblio.org ;echo $?
PING ibiblio.org (152.46.7.80): 56 data bytes

--- ibiblio.org ping statistics ---
1 packets transmitted, 1 packets received, 0% packet loss
round-trip min/avg/max = 219.791/219.791/219.791 ms

real 0m0.260s
user 0m0.000s
sys 0m0.004s

# time wget -t 2 -T 20 --waitretry=20 --spider -S ibiblio.org -o /dev/stdout | grep '200 OK'
HTTP/1.0 200 OK
200 OK

real 0m0.489s
user 0m0.004s
sys 0m0.004s

# time wget http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet
--19:57:24-- http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet
=> `bluefish-1.0.7.pet'
Resolving distro.ibiblio.org... 1.0.0.0
Connecting to distro.ibiblio.org|1.0.0.0|:80... ^C

real 1m41.282s
user 0m0.000s
sys 0m0.000s

# time wget http://distro.ibiblio.org/pub/linux/dis ... pt-1.1.pet
--20:11:00-- http://distro.ibiblio.org/pub/linux/dis ... pt-1.1.pet
=> `bcrypt-1.1.pet'
Resolving distro.ibiblio.org... 1.0.0.0
Connecting to distro.ibiblio.org|1.0.0.0|:80... ^C

real 1m21.677s
user 0m0.000s
sys 0m0.004s

# time wget http://distro.ibiblio.org/pub/linux/dis ... 0.96.1.pet
--20:13:34-- http://distro.ibiblio.org/pub/linux/dis ... 0.96.1.pet
=> `dia-0.96.1.pet'
Resolving distro.ibiblio.org... 1.0.0.0
Connecting to distro.ibiblio.org|1.0.0.0|:80... ^C

real 2m22.896s
user 0m0.004s
sys 0m0.000s

[Lobster]

I am sure I am doing something wrong

but I booted up with 'pfix=ram' just to check
this bug is only related to Geany
leafpad or other source is OK

Highlighting text and middle clicking (to copy)
in terminal has no effect

I know there was some problem with Geany, Barry mentioned
and as far as I remember this specific problem was not in 4.1

[Dougal]

Please add the following line to xorg.conf, after the MaxTapMove-option line:

Code: Select all

	Option "MaxDoubleTapTime" "220"

Try that with both pads, so we can know whether it impacts the ALPS pad that does not already need it.

Since you have both types of touchpad, with both connected and active, would you post the content of /proc/bus/input/devices? I am looking for a way to differentiate the two, in case that becomes necessary.

The xorgwizard differentiates between alps and synaptics and inserts different blocks into xorg.conf, so you can obviously use the same grep command it uses...

A while ago I thought of writing a gtkdialog gui for configuring all of the touchpad options, but I'll need to google and find out what they all are and what are their ranges... also, gtkdialog doesn't support "scrollboxes" (or whatever they're called, the ones where you press the up/down arrows to increment/decrement the number), so it's bound to end up pretty ugly.

[puppyjim]

Barry could you get in touch with me - internal mail is fine - its important

[rerwin]

The xorgwizard differentiates between alps and synaptics and inserts different blocks into xorg.conf, so you can obviously use the same grep command it uses...

Thanks, Dougal, for pointing that out. I overlooked the synaptics part of the xorgwizard code. So, we don't need to ensure MaxDoubleTapTime is safe with whichever pad doesn't need it.

I am now concerned that dogone did not see the touchpad setup in xorg.conf. If we need to add a delay before testing /proc/bus/input/devices, I need feedback from him so it could be added to 4.1.1. I also need confirmation that it is the Synaptics (internal) pad that does not currently support double-tap-and-drop, by his posting of that file for each of his pad cases, and his test of the MaxDoubleTapTime option for whichever type needs it.

Time is running out! (if it's not too late already)
Richard

Different subject: Barry, thanks for fixing remaster2. Please consider the subsequent post regarding a similar drive-icon issue with the installer, if it is an issue. We might as well clean up all such impacts from the new drive icons.
Richard

[dogone]

The xorgwizard differentiates between alps and synaptics and inserts different blocks into xorg.conf, so you can obviously use the same grep command it uses...

Thanks, Dougal, for pointing that out. I overlooked the synaptics part of the xorgwizard code. So, we don't need to ensure MaxDoubleTapTime is safe with whichever pad doesn't need it.

I am now concerned that dogone did not see the touchpad setup in xorg.conf. If we need to add a delay before testing /proc/bus/input/devices, I need feedback from him so it could be added to 4.1.1. I also need confirmation that it is the Synaptics (internal) pad that does not currently support double-tap-and-drop, by his posting of that file for each of his pad cases, and his test of the MaxDoubleTapTime option for whichever type needs it.

Time is running out! (if it's not too late already)
Richard

Different subject: Barry, thanks for fixing remaster2. Please consider the subsequent post regarding a similar drive-icon issue with the installer, if it is an issue. We might as well clean up all such impacts from the new drive icons.
Richard

Thanks Richard. The attached contains all.
I should remind you that after an early hiatus, double-tapping works with both pads. The problem remaining is the lack of double-tab and drag with the internal Synaptic pad.
One other point. After booting with the Synaptic enabled and Alps connected, both initially work (but for the above). The Alps however, stops working properly as soon as the internal pad is used. The system is obviously then confused. This is understandable and not a real problem - but FYI.
I've not pursued your "MaxDoubleTap" code tests as double tapping is no longer the issue.

[rerwin]

dogone,
Thanks for the data. All but one file indicate that the touchpads are not detected. The exception is when both are active, the "devices" file shows "Alps". So none of the controls that support touchpads is built into xorg.conf. The behavior you see is what the pads do as default. The issue now becomes how to get them recognized, at all.

Time to call in the "big guns". It does not appear we can fix this for 4.1.1, but I hope Barry and Dougal can help out here. Is there a better way to detect the pads? If dogone took any time at all to capture the /proc/bus/input/device files, it does not seem that a delay before testing would change anything. However, the "Alps" item in "devices" for both pads active might be a clue; but it was apparently not included at the time xorgconfig read it, because xorg.conf would have shown the Alps, too.

Maybe the problem is with the new underlying (udev?) structure, that just doesn't set up /proc/bus/input/device completely. But dogone did say (didn't he?) that he also tried the retro 4.1.1-rc. Right now that area is beyond my reach. Does anyone at all see the "tapping control" checkbox in the input/mouse wizard?
Richard

[dogone]

dogone,
Thanks for the data. All but one file indicate that the touchpads are not detected. The exception is when both are active, the "devices" file shows "Alps". So none of the controls that support touchpads is built into xorg.conf. The behavior you see is what the pads do as default. The issue now becomes how to get them recognized, at all.

Time to call in the "big guns". It does not appear we can fix this for 4.1.1, but I hope Barry and Dougal can help out here. Is there a better way to detect the pads? If dogone took any time at all to capture the /proc/bus/input/device files, it does not seem that a delay before testing would change anything. However, the "Alps" item in "devices" for both pads active might be a clue; but it was apparently not included at the time xorgconfig read it, because xorg.conf would have shown the Alps, too.

Maybe the problem is with the new underlying (udev?) structure, that just doesn't set up /proc/bus/input/device completely. But dogone did say (didn't he?) that he also tried the retro 4.1.1-rc. Right now that area is beyond my reach. Does anyone at all see the "tapping control" checkbox in the input/mouse wizard?
Richard

Boy, thanks for the effort, Richard. This is a bit of a niggler.

I captured those files immediately after X settled following a reboot (in each case). I copied xorg.conf and "cat" the proc file. No problem in either case. And for the record, I have not looked for this problem under 4.1.1. retro. I can easily do so and provide the same files again.

Barry's just rising as I type this so I'll wait until this evening GMT-7 and check for his take on your last. I'm happy to do whatever it takes.

This problem is not a show stopper, but double-tap and drag is just a natural for selecting/scrolling via a pad. Let's leave it to Barry to consider priorities. Barring a fix, all touch pad users have to do is drop a digit on the left button and drag.

As you suggest, this may have deeper implications re device detection. Me thinks Barry would have to be part of that investigation that so, again, over the Msr. K. re priorities.

[tazoc]

I am sure I am doing something wrong

but I booted up with 'pfix=ram' just to check
this bug is only related to Geany
leafpad or other source is OK

Highlighting text and middle clicking (to copy)
in terminal has no effect

Hi Lobster,
I noticed the same thing in Puppy 4.1.1retro-rc. After looking through Barry's blog I see he's aware of this:
http://www.puppylinux.com/blog/?viewDetailed=00429
there is a workaround there:

Posted on 18 Oct 2008, 4:28 by GeoW
Neat Geany Workaround
Try this:

Open your file in Geany and highlight some text

Open a terminal and confirm that you cannot copy

Now, with the text still highlighted, in Geany

Search>Find>Close

Now you can drop the text into the terminal!

I don't have a fix.
-TazOC

[Lobster]

Search>Find>Close

many thanks for geany bug info

[tazoc]

I grabbed rxvt from Puppy 3.01 and paste from Geany works--no workaround needed! It's rxvt v2.6.4 and it's smaller (92k) than the one in Puppy 4.1.1rc.
-TazOC

[BarryK]

Precisely my experience, but consistent and many times over. I don't recall the IP in the orange dialog precisely, but it contains mostly zeros and is most certainly invalid. I would suggest that you exercise PM a bit to convince yourself one way of the other. If your luck is anything like mine, you won't need to worry about disk space .
This issue would be a show stopper if real, so we need to pursue it. Is anyone else having such problems - or not?

Well, the problem is that I cannot reproduce your problem, so I cannot fix it. You are really going to have to try and get into the code and try to find out what i wrong.

The script is /usr/sbin/petget, and 'wget' is used to download .pet packages. So, there may be some problem with wget. Maybe you could locate the line where wget is run, then try to execute that on its own in a terminal.

Looking in the code, I see that 'ping' is used to verify that a site exists, for example:

# ping -c 1 -q ibiblio.org ;echo $?

if that fails, that is, returns non-zero value, the code tries this:

# wget -t 2 -T 20 --waitretry=20 --spider -S ibiblio.org -o /dev/stdout | grep '200 OK'

...it's looking for that cring "200 OK"

The actual line that does the download is:

wget $PKGSRC/pet_packages-${PUPREPO}/${APKGNAME}.pet

Doing some example substitutions:

# wget http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet

So, try those in a terminal, see what happens.
(remember, highlight the text, then use middle-mouse-button to paste into the terminal window).

Thanks Barry. Here's what I see in response to those command strings. Note that resolved address and the fact that the second wget command hangs until I break it:

# time ping -c 1 -q ibiblio.org ;echo $?
PING ibiblio.org (152.46.7.80): 56 data bytes

--- ibiblio.org ping statistics ---
1 packets transmitted, 1 packets received, 0% packet loss
round-trip min/avg/max = 219.791/219.791/219.791 ms

real 0m0.260s
user 0m0.000s
sys 0m0.004s

# time wget -t 2 -T 20 --waitretry=20 --spider -S ibiblio.org -o /dev/stdout | grep '200 OK'
HTTP/1.0 200 OK
200 OK

real 0m0.489s
user 0m0.004s
sys 0m0.004s

# time wget http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet
--19:57:24-- http://distro.ibiblio.org/pub/linux/dis ... -1.0.7.pet
=> `bluefish-1.0.7.pet'
Resolving distro.ibiblio.org... 1.0.0.0
Connecting to distro.ibiblio.org|1.0.0.0|:80... ^C

real 1m41.282s
user 0m0.000s
sys 0m0.000s

# time wget http://distro.ibiblio.org/pub/linux/dis ... pt-1.1.pet
--20:11:00-- http://distro.ibiblio.org/pub/linux/dis ... pt-1.1.pet
=> `bcrypt-1.1.pet'
Resolving distro.ibiblio.org... 1.0.0.0
Connecting to distro.ibiblio.org|1.0.0.0|:80... ^C

real 1m21.677s
user 0m0.000s
sys 0m0.004s

# time wget http://distro.ibiblio.org/pub/linux/dis ... 0.96.1.pet
--20:13:34-- http://distro.ibiblio.org/pub/linux/dis ... 0.96.1.pet
=> `dia-0.96.1.pet'
Resolving distro.ibiblio.org... 1.0.0.0
Connecting to distro.ibiblio.org|1.0.0.0|:80... ^C

real 2m22.896s
user 0m0.004s
sys 0m0.000s

Well, I haven't got a clue why wget is failing for you. Do you have a proxy server?

I don't know anything about proxy servers, never had to deal with one. Don't know if that would account for the problem ...but the first wget did work. Anyone else got an idea why distro.ibiblio.org is resolving to 1.0.0.0?

[BarryK]

I grabbed rxvt from Puppy 3.01 and paste from Geany works--no workaround needed! It's rxvt v2.6.4 and it's smaller (92k) than the one in Puppy 4.1.1rc.
-TazOC

That's interesting! Would you mind gzipping it and post here so others can give it a quick test to confirm the fix? Well, why can't I do that...

Attached, gunzip it and place at /usr/bin. Try it guys, let me know if it seems to do all the things that rxvt is supposed to do.

[Lobster]

let me know if it seems to do all the things that rxvt is supposed to do.

Well it works with Geany - so that is a start
ls is in colour - that is probably busybox though

rclock is not there (not sure if it is used)
http://en.wikipedia.org/wiki/Rxvt