Lucid Puppy 5.2.7 RC2

[jpeps]

I have tested this again with Lupu 526. I ran editinit, went to the folder puppy-init, changed the permissions on /dev/null and repacked everything.

On the next boot, the permission was correct.

It's not the no-hard-drive patch that breaks the initrd.gz. It's simply unpacking and repacking it that does the damage.

Now the permissions are correct for /initrd/pup_rw/dev/null

[playdayz]

It's not the no-hard-drive patch that breaks the initrd.gz. It's simply unpacking and repacking it that does the damage.

Agreed completely . I confirm that by opening the initrd, editing, saving, but without making any changes, and still had the problem. But I have been using editinit. I will try rcrsn51's exact method.

[jpeps]

remastering the initrd.gz worked for 264 as well (as lucid 525).

[playdayz]

Looks like a Bingo rcrsn51. I was using editinit incorrectly, or only half-correctly. Your explanation made the difference..

So 527 has the usb fix and the pplog fix. Muchas gracias.

It also has a menu entry created for the flashplayer config which is installed when flashplayer is installed.

It will have the latest rc,update.

And it will be smaller than 5.2.5.

And finally, the improved right-click options.

Very little more I hope.

[rcrsn51]

Excellent. I realize now that my false positives occurred because I had already unpacked/repacked the initrd.gz and had inadvertently broken it.

I was curious why this scenario has never been reported before, so I looked at some older Puppies like 431 and Quirky. In those, you can unpack/repack the initrd.gz and /dev/null keeps its 666 permissions. Even though inside the puppy-init folder it's only 644. Maybe it has something to do with recent changes to the layered file system.

[Snail]

So 527 has the usb fix and the pplog fix. Muchas gracias.

527? I thought it would be 526rc2 or final.

I posted about a couple of minor bugs in pfind here: http://www.murga-linux.com/puppy/viewto ... &start=375
It might be worth checking with Zigbert to see if he has any comment.

I cannot open the video here:
http://www.wimp.com/japanesetsunami/

Running from terminal all I see is

Code: Select all

Failed to open VDPAU backend libvdpau_nvidia.so: cannot open shared object file: No such file or directory

Which I presume doesn't tell us much.

EDIT Whaat! on the sixth or seventh try it's suddenly working. I did absolutely nothing to reconfigure or patch SeaMonkey since the last failure???

A bit off-topic... I had to fix my Seamonkey profile. It's in /mnt/home and I inadvertently shared it between 525(SM 2.0.140 and Luci 26X and 525 (SM 2.2), which broke History and Bookmarks, at least . I had 6 "places***.sqlite" files! Deleting them all and starting Seamonkey worked. I got a shock however because all the discussion on the Mozilla sites refers to Firefox and I didn't know that some mail settings are also in the Seamonkey version, at least that seems to be the case, because after restarting Seamonkey I found the mail window was blank. It all came right after a reboot however. Just thought I'd mention it if anyone else has screwed up going back and forth between 525 and 260 during testing.

EDIT Not a complete fix. Seamonkey still looses the icons in the bookmarks. Clicking on the bookmark restores some but not all of them, but it doesn't persist. Very annoying. I'm not a bling fan but these icons do help navigate a big list. I am a compulsive bookmarker and hoarder, so I have a lot to navigate.

[Sage]

Couple of years ago, there was an hiatus over non-detection of just about the most popular PCI plug-in card NIC: RTL8139D. It's back. None of the Realtek NICs in the list will induce it to function - seem to recall, it's all a bit deja vu.
At that time some public spirited guru stepped up to the plate with a 'permanent' solution. This is the first time that it has recurred on any of my systems. Can't remember whether it requires an extra driver in the list or some other manipulation? Whatever, it's still a show-stopper.

Update: RTL8139C auto connected at first ask.

[pemasu]

RTL8139D. I would like to see posting of lspci -nnv content from the comp installed with RTL8139D card.
Searching from the net there are mentions of fake RTL 8139D card which does use other than rtl driver.

[Sage]

fake RTL 8139D

Didn't know that, pm. Looking at it right now and will report infra. Either way, 'D' isn't in the list and my dim memory tells me it was fixed?

OK. lspci shows:
01:04.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10).
Believe the NIC is working. At least it has been on some other distros and its led is alight. Not so sure about that 8139C+ ? Did that become 'D' or was it yet another variant with another driver?!

Another update: whilst I was typing the above (~5mins), it eventually popped up a message: found the NIC as an 8139too. Evidently the auto connect had continued to search. Normally, NICs come up on DHCP immediately as they are detected in the boot script. However, this beast still refused to connect. The log shows messages including timed-out. Tried manual DHCP with BK's original connection utility. After several further negative-looking messages - bingo!

This is an old Gigabyte with 950Celeron and 384mem. and on-board video (wouldn't be using it from choice, but this a testing regime). The NIC seems to drop out sometimes, but at present is UP but not connecting to any URLs through Dillo. Suspect there might be HW issues with both this chassis and the driver in use since an 8139C had no issues on the same board.

[Took the precaution of sticking 8139D into slot 2 since there might've been conflict with video in slot 1. Notwithstanding, 'C' NIC worked perfectly with all the same HW limitations in slot 1].

[Snail]

Pemasu, I have just installed your Right-click-4.9.6.4-luci pet. Thanks for a huge amount of work. I noticed a few broken symlinks that need patching. They are in :

........./SendTo/.application-vnd.sun.xml.calc(All)
........./SendTo/.application_x-desktop
........./sendTo/.application-x-wav
........./sendTo/.image.tiff
........./SendTo/.inode_directory

The links point to files in /mnt/sda6/IcepupnewStuff/Right-click etc.

I hope I've got all the little critters. A clever script to find them is beyond me so this is only a manual check.

[Sage]

The Dutch repo is missing. This is the fastest for Europeans, and probably US after ~6pm GMT.
See BK's blog of about 2mnths ago. Didn't this site drop out by mistake when there were some changes at ?ibiblio? and others; has since been re-instated.

[Bert]

Hi Sage,

Yes, there's a problem.
The Dutch mirror is still there though:

http://ftp.nluug.nl/ibiblio/distributio ... pu-526.iso

So downloading with a browser or wget is possible. And you can keep the pets for later re-use

OOPS wrong address..this one is for the .pets:

http://ftp.nluug.nl/ibiblio/distributio ... ges-lucid/

[nooby]

Place holder for report on success of booting the new RC. I booted with nosmp just in case. All went well. dld Opera for a change and flash just worked on our local tv station and on youtube. looked at some pics on Australian Dingo and African Cat. Looked good. Have not tested if the Duch repo is up. I agree usually it is a fast server that just works. Not sure where Opera was taken from.

[Sage]

Ah yes, I remember now. The Dutch site was omitted during some registration event. Someone pointed it out to BK, he wrote to nluug and they agreed to permit access again. Just a few oversdights now rectified. It should go onto our main list as all the other derivatives have been adjusted (I think?!). The entire epistle is recorded on the BK blog, anyway.

[pemasu]

Snail. Thank you of your piercing eyes. I swear I checked all the symlinks that they point to the /usr/local/apps. But....when you have 5-6 folders open, there seems always happen mistakes.
Now I went through the symlinks and I swear now again that they point to the /usr/local/apps. I fixed your found wrong symlinks.
I hope again that the package is now right.

Don570 and I have been doing the same work. I have added mime types, right clicks and and so on....
The difference is that I have freezed to 4.9.6 version and only tried to improve it. Now I hope all the links are absolute links to /usr/local/apps.

I have done my package especially for next Lucid, not done overall improvements like don570. He adds all the time new stuff to the right clicks like zathura. I appreciate don570 work high. I have noticed now how tedious it is to get all symlinks right, when there are hundreds of them.

[pemasu]

Sage.

Code: Select all

01:04.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10). 

That one is Realtek chip eth card. It should work with 8139too driver fine.

[ASRI éducation]

Bizarre, I can not use the local FR with gnome mplayer 1.0.3.
I tested several .mo, but none works.
However, with 1.0.2, I had no problem locating.

[Sage]

That one is Realtek chip eth card. It should work with 8139too driver fine.

Wish I had your confidence in the correct ID, pm! The writing on the chip itself is usually definitive. Sometimes the internal ID microcode is just copied from a prior batch of earlier designs within a organisation - after all, manufacturers don't expect anyone to read it! Fake chips often miss off the logo or have tatty legends. My card has all the correct text and logo - it says RTL8139D, so I would expect it to be detected as such and use an 8139D driver.

[aarf]

got this little script that mounts partitions that i need at startup

Code: Select all

#!/bin/ash
mkdir /mnt/sdb2 2>/dev/null
mount /dev/sdb2 /mnt/sdb2 2>/dev/null

works in puppee. comes from startmount (which i cant find in 526.)
anyhow when i right click it (right-click pet installed) and run in terminal i get

Code: Select all

/tmp/runinterminal-9896: line 3: /mnt/home/user/mount-drives.sh: Permission denied
Script completed hit RETURN to close window.

note the /dev/null has been permission modified as per pplog.

[playdayz]

Right-click-4.9.6-5-luci.pet

Got it, thanks to pemasu and snail.