[Solved] Porteus 5.0 Updater

[Ed_P]

Code: Select all

 Checking patch availability ...
 ######################### 
 Available core updates 

001-core.xzm
002-xorg.xzm
002-xtra.xzm

 Would you like to continue? [y/n]
 
 ...
 
 Modules will be downloaded to /tmp 
Downloading: 001-core.xzm  DONE
[ERROR] md5sum for 001-core.xzm is bad
Removing file ...
Downloading: 002-xorg.xzm  DONE
[ERROR] md5sum for 002-xorg.xzm is bad
Removing file ...
Downloading: 002-xtra.xzm  DONE
[ERROR] md5sum for 002-xtra.xzm is bad
Removing file ...

[ncmprhnsbl]

thanks for reporting
strange.. seems to have lost track of one of it's variables
anyways, have applied a dirty hack fix and reuploaded the live script to the server
should work for you now (with the usual caveat: mirrors will take a day or so to sync)

[Ed_P]

The /updates/core/updates.txt file shows the dates for the 3 updated files as:

Code: Select all

84a04c5b3455a215124033a288b09d3e  ./20221210/001-core.xzm/ 122M
cf36fa5d3da917c8947d535eb0ac30b8  ./20221210/002-xtra.xzm/ 18M
f9dfae7497572aedffdce6b9904eb468  ./20221210/002-xorg.xzm/ 112M

but the dates for the 3 files in /updates/core/20221210/ folder show as:

Code: Select all

001-core.xzm                                       12-Dec-2022 05:43           127332352
002-xorg.xzm                                       12-Dec-2022 05:44           116957184
002-xtra.xzm                                       12-Dec-2022 05:45            18399232

Latest update:

Code: Select all

 All available updates are selected: 

  1+) 001-core.xzm
  2+) 002-xorg.xzm
  3+) 002-xtra.xzm
Deselect an item (again to reselect, ENTER when ready):


The base modules go in your base directory. Unfortunately I am not
able to write to the destination. You can choose an alternative
location to download your base module updates to.

Enter a target directory to download to: 
> /tmp

 Modules will be downloaded to /tmp 
Downloading: 001-core.xzm  DONE
grep: 20220924: No such file or directory
grep: 20221210: No such file or directory
[ERROR] md5sum for 001-core.xzm is bad
Removing file ...
Downloading: 002-xorg.xzm  DONE
grep: 20220924: No such file or directory
grep: 20221210: No such file or directory
[ERROR] md5sum for 002-xorg.xzm is bad
Removing file ...
Downloading: 002-xtra.xzm  DONE
[OK] md5sum for 002-xtra.xzm is good

 Updates complete. 
 Updates are at /tmp 
 Please move them into your base folder. 
 You may now close this window. 

Code: Select all

guest@porteus:~$ ls -oh /tmp/*.xzm
-rw-r--r-- 1 root 18M Dec 12 00:45 /tmp/002-xtra.xzm
guest@porteus:~$ 

[ncmprhnsbl]

hmm, more odd.
i'll have to look into this further..

but the dates for the 3 files in /updates/core/20221210/ folder show as:

this is irrelevant. as long as the date on the folder is newer( and the newest) than the date of the /etc/porteus/*.ver file of your present module is how the correct module is chosen..

EDIT: found it. just a typo.. should be okay now.

[ronriel]

By the way, where can I find the update changelog? Thanks.

[Ed_P]

Success

Code: Select all

Enter a target directory to download to: 
> /tmp

 Modules will be downloaded to /tmp 
Downloading: 001-core.xzm  DONE
[OK] md5sum for 001-core.xzm is good
Downloading: 002-xorg.xzm  DONE
[OK] md5sum for 002-xorg.xzm is good
Downloading: 002-xtra.xzm  DONE
[OK] md5sum for 002-xtra.xzm is good

 Updates complete. 
 Updates are at /tmp 
 Please move them into your base folder. 
 You may now close this window. 

Code: Select all

guest@porteus:~$ ls -oh /tmp/*.xzm
-rw-r--r-- 1 root 122M Dec 12 00:43 /tmp/001-core.xzm
-rw-r--r-- 1 root 112M Dec 12 00:44 /tmp/002-xorg.xzm
-rw-r--r-- 1 root  18M Dec 12 00:45 /tmp/002-xtra.xzm
guest@porteus:~$ 

So is this a 20221210 update or a 20221212 update?

And does it include the updater-test-22-10-31.xzm changes?

[donald]

Hmm...
From what I understand the script search for the fastest mirror to download the files.
..and the md5sum check? -- IMO the md5sum should be better checked against the file
from the porteus server.
Why:
Because if the md5sum file from the mirror is used the script will always report that
everything is OK even if the download comes from a mirror with manipulated files.
(and correspondending md5sums)

As it is now the script only tell you that the download itself was OK, not that you received
original (not manipulated) files.

[Ed_P]

And we are live with:

Code: Select all

~#  uname -a
Linux porteus.example.net 5.18.8-porteus #1 SMP PREEMPT_DYNAMIC Sat Jul 2 10:05:31 MSK 2022 x86_64 11th Gen Intel(R) Core(TM) i5-1135G7 @ 2.40GHz GenuineIntel GNU/Linux

~#  cat /etc/porteus/*
001-core.xzm:20221211
002-xorg.xzm:20221211
002-xtra.xzm:20221211
003-cinnamon.xzm:20220925
initrd.xz:20220928

~#  ls -l /mnt/live/memory/images
000-kernel.xzm : Jul 2 03:32
001-core.xzm : Dec 11 03:47
002-xorg.xzm : Dec 11 03:50
002-xtra.xzm : Dec 11 01:13
003-cinnamon.xzm : Sep 25 01:21

So is this a 20221210 update or a 20221212 update?

Apparently it's a compromise, it's a 20221211 update.

IMO the md5sum should be better checked against the file
from the porteus server.

An interesting point donald.

[ncmprhnsbl]

By the way, where can I find the update changelog? Thanks

Porteus-v5.0 updates is the closest to that.

From what I understand the script search for the fastest mirror to download the files.

although there seems to be a function for that present, i'm pretty sure it isn't implemented (as is a bunch of other stuff in it eg. patches).
so it uses whatever the SERVER variable is (set in /etc/porteus.conf) which defaults to our server if fastest-mirror hasn't been used.
even so, what you say about mirrors, md5sums, and manipulated files is valid (insomuch as it's not beyond the realms of possibility) and i'll see what i can do about it.
just saying: i didn't write this script, brokenman introduced it way back in v4.0 and seems like something worth using.

[babam]

I hope for 5.0.1

[burdi01]

I hope for 5.0.1

As my use cases are running the ISOs in VirtualBox, QEMU and dd'ed to a USB stick I would be grateful for 5.0.1 ISOs.

[Rava]

even so, what you say about mirrors, md5sums, and manipulated files is valid (insomuch as it's not beyond the realms of possibility) and i'll see what i can do about it.

I think using the correct and relevant md5sum .txt from the original server instead of the one from some mirror would do the trick, yes?
Loading that one small text file from the original porteus server should not put any undue load on our own server and give all users more security about manipulated files and mismatches in md5sums.

If you would point me to the script that does the download and md5checking I could look into it myself and thus take a bit of work load out of your hands. (Probably better via PM)

[donald]

I think using the correct and relevant md5sum .txt from the original server instead of the one from some mirror would do the trick, yes?

Yep, that's what I suggested.

I hope for 5.0.1

As my use cases are running the ISOs in VirtualBox, QEMU and dd'ed to a USB stick I would be grateful for 5.0.1 ISOs.

You can easily update the isos with a tool like iso-master.

[Ed_P]

You can easily update the isos with a tool like iso-master.

Code: Select all

guest@porteus:~$ getmod -m iso-master
Reading Package Lists...Done
No such package: iso-master
guest@porteus:~$ 

[Rava]

Code: Select all

guest@porteus:~$ getmod -m iso-master
Reading Package Lists...Done
No such package: iso-master
guest@porteus:~$ 

Code: Select all

root@porteus:/Lsfind# getmod -s isomaster
isomaster-1.3.16-x86_64-1gv  (an open-source, GUI CD image editor for Linux and BSD)

It needs gtk+2 2.24.33-x86_64-2 as well… but maybe other programs you use also need gtk+2 2.24.33-x86_64-2 so maybe you want one module including the gtk+2 stuff when you only activate isomaster, and one module with only isomaster when you already activated a module that has the needed gtk+2 dependency because of a different program. Just my 2 cents.

Code: Select all

root@porteus:/Porteus_modules# TMP=/Porteus_modules getmod -M isomaster
Reading Package Lists...Done
The following NEW packages will be installed:
  gtk+2 isomaster 
0 upgraded, 0 reinstalled, 2 newly installed, 0 to remove, 0 not upgraded.
Need to get 6.7MB of archives.
Do you want to continue? [y/N] y
1/2 Get http://slackware.uk/salix/x86_64/slackware-15.0/ gtk+2 2.24.33-x86_64-2 [6.6MB]...Done
2/2 Get http://slackware.uk/salix/x86_64/15.0/ isomaster 1.3.16-x86_64-1gv [179.0kB]...Done

Done
Verifying package isomaster-1.3.16-x86_64-1gv.txz.
Installing package isomaster-1.3.16-x86_64-1gv.txz:
PACKAGE DESCRIPTION:
# isomaster (an open-source, GUI CD image editor for Linux and BSD)
# 
# isomaster is an open-source, GUI CD image editor for Linux and BSD.
# Basically you can use this program to extract files from an ISO,
# add files to an ISO, and create bootable ISOs
# all in a graphical user interface
# 
Package isomaster-1.3.16-x86_64-1gv.txz installed.
Verifying package gtk+2-2.24.33-x86_64-2.txz.
Installing package gtk+2-2.24.33-x86_64-2.txz:
PACKAGE DESCRIPTION:
# gtk+2 (multi-platform GUI toolkit)
#
# This is GTK+, a multi-platform toolkit for creating graphical user
# interfaces. Offering a complete set of widgets, GTK+ is suitable for
# projects ranging from small one-off projects to complete application
# suites.
#
Executing install script for gtk+2-2.24.33-x86_64-2.txz.
Package gtk+2-2.24.33-x86_64-2.txz installed.
Parallel mksquashfs: Using 8 processors
Creating 4.0 filesystem on /Porteus_modules/isomaster-1.3.16-x86_64-1gv.xzm, block size 262144.
[===============================================================\] 442/442 100%

Exportable Squashfs 4.0 filesystem, xz compressed, data block size 262144
	compressed data, compressed metadata, compressed fragments,
	compressed xattrs, compressed ids
	duplicates are removed
Filesystem size 3193.39 Kbytes (3.12 Mbytes)
	17.67% of uncompressed filesystem size (18077.38 Kbytes)
Inode table size 3584 bytes (3.50 Kbytes)
	23.79% of uncompressed inode table size (15064 bytes)
Directory table size 4560 bytes (4.45 Kbytes)
	44.07% of uncompressed directory table size (10347 bytes)
Number of duplicate files found 2
Number of inodes 458
Number of files 388
Number of fragments 15
Number of symbolic links 10
Number of device nodes 0
Number of fifo nodes 0
Number of socket nodes 0
Number of directories 60
Number of ids (unique uids + gids) 1
Number of uids 1
	root (0)
Number of gids 1
	root (0)

====> '/Porteus_modules/isomaster-1.3.16-x86_64-1gv.xzm' was created successfully.

resulting in this:

Code: Select all

root@porteus:/Porteus_modules# l -o isomaster-1.3.16-x86_64-1gv.xzm 
-rw-r--r-- 1 guest 3272704 2022-12-18 01:47 isomaster-1.3.16-x86_64-1gv.xzm

This version contains its gtk+2 dependency, even when getmod usually not tells the user that by the module name.

This would be the version without the GTK-2 stuff

Code: Select all

root@porteus:/Porteus_modules# ls -o isomaster-1.3.16-x86_64-1gv.xzm 
-rw-r--r-- 1 root 188416 2022-12-18 01:54 isomaster-1.3.16-x86_64-1gv.xzm

while I renamed the originally created module:

Code: Select all

root@porteus:/Porteus_modules# ls -o isomaster-1.3.16-x86_64-1gv+dependencies_gtk+2-2.24.33.xzm 
-rw-r--r-- 1 guest 3272704 2022-12-18 01:47 isomaster-1.3.16-x86_64-1gv+dependencies_gtk+2-2.24.33.xzm

Added in 4 minutes 49 seconds:
This is after activating the 3.12 MB module I renamed into isomaster-1.3.16-x86_64-1gv+dependencies_gtk+2-2.24.33.xzm :

Thanks donald for the tip.