Fatdog64-700/701 [April 22 2015] [CLOSED]

[L18L]

setfont -m 8859-2 lat2-16

Great! I knew there must be a simpler & cleaner way to achieve this.
So, we can mark this issue as solved.

solved for Polish.
What do you think about this (copied from init of initrd in precise5.7.1):

Code: Select all

 #120216 L18L suggests load these, instead of what is below...
 case $PLANG in
  en*) echo ;;
  ar*|iw*) #L18L no Greek
   setfont /lib/consolefonts/LatArCyrHeb-16.psfu.gz -C /dev/tty1
   FONTMAP='LatArCyrHeb-16.psfu'
  ;;
  *) #L18L All European languages; new default ?!
   zcat /lib/consolefonts/LatGrkCyr-8x16.psfu.gz | loadfont
   FONTMAP='LatGrkCyr-8x16.psfu'
  ;;
 esac

Might be even simpler.

Greetings

[SFR]

Hey L18L

The font itself is ok, too, although still needs ISO-8859-2 codepage for some langs (pl, croat, slovene...):

Code: Select all

setfont -m 8859-2 LatGrkCyr-8x16

This codepage also covers de/de-latin1, btw.

Hmm, I'm wondering is there some mechanism that would automatically pick a corresponding codepage, so we wouldn't have to define it for every language group...

Greetings!

[L18L]

Hmm, I'm wondering is there some mechanism that would automatically pick a corresponding codepage, so we wouldn't have to define it for every language group...

just surfed the web...utf-8 codepage number

http://webhelp.esri.com/arcpad/8.0/refe ... _local.htm

The UTF-8 code page equivalent is 65001.

Specifying the code page for UTF-8 provides a number of advantages:

Characters from different code pages can be stored and displayed using a single code page for UTF-8. For example, with UTF-8, it is possible to display both Traditional Chinese and Simplified Chinese characters at the same time.
Some languages are only supported via Unicode and so can only be displayed when using shapefiles if the code page is set to UTF-8.
Any Unicode character can be displayed on a device that supports Unicode, providing that the system font required for the characters is present on the device.

For example, many PCs in Hong Kong are configured for Traditional Chinese Encoding (code page 950), while the shapefiles are deployed on English Windows Mobile devices that do not support the 950 code page. Specifying the code page as 65001 (UTF-8) and copying the required system font to the device enables the English Windows Mobile devices to display characters in Traditional Chinese.

Glad to know you are caring about these issues

[smokey01]

would that help finding dups?

Do you mean duplicates, dupes? If so that would be yes.

[Aung]

Thanks for the new Fatdog.
In every one of the 700 series eztables has had this problem. I have never been able to achieve a tick in the 'Running' box even after clicking 'Start'.
On reboot eztables dont 'start' at all even though the box remains ticked.
However upon starting manually eztables does appear to be working with a sea of green for Solicited TCP Packets and Unsolicited Packets:
at shieldup.
Is ping reply failing a problem for being on the net, in 600 series cant recall that failing.
Is it possible to have something in the tray to show firewall status seeing that eztables is tucked deep into control panel?

Ping Reply: RECEIVED (FAILED) — Your system REPLIED to our Ping (ICMP Echo) requests
Solicited TCP Packets: PASSED — No TCP packets were received from your system
Unsolicited Packets: PASSED — No Internet packets of any sort were received from your system

[JustGreg]

Up to now, I had no problems with Fatdog64 700 final. I tried installing my Thunderbird package from the repository. It installed correctly. But, when I try to run it it I get this error:

Code: Select all

dash: 1: /usr/lib64/thunderbird/thunderbird: not found

At first, I thought I may have done something wrong. I did check my notes and the package did work with the beta versions. I try changing somethings in my script that runs thunderbird as spot. There was no luck on with the changes. I added some echo commands to the run-as-spot script. The error occurs when the if command (line 8 ) is started. I have tried adding /usr/lib64/thunderbird to the PATH, but, no luck. Any thoughts or suggestions will be greatly appreciated. Thanks in advance for any help on this one.

[SFR]

@JustGreg: do you have 32bit-slacko-3.sfs loaded?
http://www.murga-linux.com/puppy/viewto ... 001#819001

Greetings!

[JustGreg]

SFR, you were correct. I did not have 32bit-slacko-3.sfs loaded. But it is now. Thanks for the help. I had to remove it for some work with the Arduino 1.6.0 IDE. I will have to check again with the Arduino 1.6.0 IDE to ensure it works. I am starting to hate the 32 bit/64 bit libraries problems. Thanks again for the help.

[jamesbond]

JustGreg, this version should be fully 64-bit and doesn't need 32-bit stuff. But I haven't tested it myself. Apparently the linux 64-bit build is not advertised in the official download page https://ftp.mozilla.org/pub/mozilla.org ... ux-x86_64/

[olinm]

VLC 2.2.0 has just been released

http://www.videolan.org/vlc/releases/2.2.0.html

http://www.videolan.org/press/vlc-2.2.0.html

Major new release of VLC with a new hardware acceleration mechanism, GPU 0-copy decoding, faster and implementations for Linux, Android, and Raspberry Pi.

------------------^

I seem to be the only one that thinks this is awesome news.

Linux guys seem to be a different breed to Window users , they can't install new versions of anything quick enough.

[jamesbond]

@SFR: "Hmm, I'm wondering is there some mechanism that would automatically pick a corresponding codepage, so we wouldn't have to define it for every language group..."

There is, and Barry did the hard work long time ago. If you zcat /usr/share/keymaps/*/*/pl.map.gz, you will see that near the top, there is "charset 8859-2". And if you look at the languages in the lat2 group (in rc.sysinit) - ie cz*|hu*|pl*|ro*|sk*|croat*|slovene*), then you will see that *all* of them are under this charset too. I wasn't the one who originally wrote the separation between lat1 and lat2 - Barry did. So he must have known something about it too.

Unfortunately, the keymaps for initrd is binary only and you can't fish this information from there.

@Aung - eztables is one of the few "run-at-boot-but-not-continue-running-afterward" services. If you have enabled it to run at boot, rest assured that it will continue running even though it doesn't seem so (there is another service like that, ntpd-boot that syncs your time from Internet time servers). "Alsa" service is another one of those.

To check whether it's running, run this in terminal: "iptables -L -n". If eztables is not running, you will see nothing. If you see plenty of rules, eztables is running.

As for the ping - by default ping reply is allowed. Please feel free to modify /etc/eztables.conf if you want to disable ping reply.

@olinm: kirk did look at vlc 2.2.0 and attempted to build it, but it requires ffmpeg update. Updating ffmpeg in itself is not a big deal, but *tons* of other programs depend on ffmpeg and many of them will break and require fixing/updating to run with a newer ffmpeg. So this will take time. It will happen when it happens.

[olinm]

@olinm: kirk did look at vlc 2.2.0 and attempted to build it, but it requires ffmpeg update. Updating ffmpeg in itself is not a big deal, but *tons* of other programs depend on ffmpeg and many of them will break and require fixing/updating to run with a newer ffmpeg. So this will take time. It will happen when it happens.

They'll never be a quick update script in the control panel for that then.

You would have thought they would have made it less integrated for mobile versions.

[L18L]

@SFR: "Hmm, I'm wondering is there some mechanism that would automatically pick a corresponding codepage, so we wouldn't have to define it for every language group..."

There is, and Barry did the hard work long time ago. If you zcat /usr/share/keymaps/*/*/pl.map.gz, you will see that near the top, there is "charset 8859-2". And if you look at the languages in the lat2 group (in rc.sysinit) - ie cz*|hu*|pl*|ro*|sk*|croat*|slovene*), then you will see that *all* of them are under this charset too. I wasn't the one who originally wrote the separation between lat1 and lat2 - Barry did. So he must have known something about it too.

charset 8859-1 and charset 8859-2 are just 2 language groups.
charset UTF-8 is for some more languages.

__________________________________________
https://forums.opensuse.org/showthread. ... -I-m-using

[kirk]

@olinm: kirk did look at vlc 2.2.0 and attempted to build it, but it requires ffmpeg update. Updating ffmpeg in itself is not a big deal, but *tons* of other programs depend on ffmpeg and many of them will break and require fixing/updating to run with a newer ffmpeg. So this will take time. It will happen when it happens.


They'll never be a quick update script in the control panel for that then. Sad

No, they don't provide binaries, if they did they would have to stactic link ffmeg. Which actually is probably a good idea with ffpeg. Anyway, I did report the problem with ffmpeg and it's now fixed in the VLC git trunk, so now it builds with our ffmpeg. However the audio buffer bug has not been fixed yet. https://trac.videolan.org/vlc/ticket/10422 and the patch I made for 2.1.x will no longer apply. So it's not usable at this point unless you install something like pulseaudio that does larger fixed buffers. Running some abstraction on top of alsa is not something I'm interested in.

[SFR]

snapmergepuppy

Problem with processing filenames that contain single quotes:

Code: Select all

# snapmergepuppy 
12331 (process ID) old priority 0, new priority -20
deleting newly deleted files
deleting old whiteouts
rsync-ing
xargs: unmatched single quote; by default quotes are special to xargs unless you use the -0 option

In effect, no files (after this error occurs) are being removed from pup_rw.
Second problem with xargs is that it won't process filenames with whitespaces, produced by closedfiles func.
Btw, I tried the suggested '-0' option, but got "argument list too long" - I suppose it would also involve modifying closedfiles func to set \0 as a delimiter..?

For now I just added this sed statement, so, at least, single quotes won't break the execution:

Code: Select all

		closedfiles |
		grep -v -E ".wh..wh.orph|.wh..wh.plnk|.wh..wh.aufs|.wh..wh..opq|$TMPFS/var/run" | 
		tee /tmp/filesmoved.txt | sed -e "s/'/\\\'/g" |
		xargs rm -rf

Greetings!

[Aung]

@jimyb
I'm still mot sure, theres not 'plenty' there, and whats there is double dutch to me, am I covered to do internet banking from a 700 live CD, aint going anywhere near any config file, (cant read double dutch)

To check whether it's running, run this in terminal: "iptables -L -n". If eztables is not running, you will see nothing. If you see plenty of rules, eztables is running.

# iptables -L -n
Chain INPUT (policy ACCEPT)# iptables -L -n
Chain INPUT (policy ACCEPT)
target prot opt source destination

Chain FORWARD (policy DROP)
target prot opt source destination

Chain OUTPUT (policy ACCEPT)
target prot opt source destination
#

[olinm]

No, they don't provide binaries, if they did they would have to stactic link ffmeg. Which actually is probably a good idea with ffpeg. Anyway, I did report the problem with ffmpeg and it's now fixed in the VLC git trunk, so now it builds with our ffmpeg. However the audio buffer bug has not been fixed yet. https://trac.videolan.org/vlc/ticket/10422 and the patch I made for 2.1.x will no longer apply. So it's not usable at this point unless you install something like pulseaudio that does larger fixed buffers. Running some abstraction on top of alsa is not something I'm interested in.

Opened 14 months ago
Severity: major
Work status: Not started

WOW

Thanks for the explanation.

[gcmartin]

@Kirk, I think I remember you were responsible for the latest upgrades to Hardinfo. @Barry makes a request here for help.

[jamesbond]

@SFR: thanks, fixed. The changes:

Code: Select all

# jamesbond 2011, 2015 - print list of files in TMPFS which is closed
BEGIN {
	# list of files to ignore when processing
	ignore_list = ".wh..wh.orph|.wh..wh.plnk|.wh..wh.aufs|.wh..wh..opq|$TMPFS/var/run";	
	
	# load our "list of open files" before processing - assume they all are in $TMPFS
	while ("find /proc -type l -wholename \"/proc/*/fd/*\" 2>/dev/null | xargs readlink 2>/dev/null | grep -vE '^socket:|^pipe:|^anon_inode:' | sort -u " | getline) {
		openfiles["$TMPFS" \$0]=1
	}
	
	# now compare this with the list of files in $TMPFS
	CMDLINE = "find \"$TMPFS\" -not -type d"
	while (CMDLINE | getline) {
		if (\$0 ~ ignore_list) continue;
		if (openfiles[\$0] != 1) printf("%s\0",\$0);
	}
}
EOF

and

Code: Select all

closedfiles | xargs -0 rm -rf

@SFR: I've found the reason why porteus can do "setfont" without specifying the console charset mapping. This was ignited by L18L's call to use utf8 all-throughout, so I went to check what's really inside a keymap and how I can convert it to utf8 (or why it isn't done already in 2015). Not so much documentation there; apparently not many people work with consoles Anyway, I found out how to convert the keymaps to utf8 just comment out (or remove) the "charset" line in the keymaps. When this is done; doing a setfont doesn't need to specify the locale-mapping, and doing a "loadkeys -u -b" will produce the correct utf8 keymaps for use in initrd, which I have done as well. I have also heeded L18L's call to use LatGrkCyr for all latin keymaps (I leave out LatArCyrHeb because I can't find the keymaps that goes with that).

@Aung: no, your eztables doesn't have to work. The output should be much longer than that (about a screenful or more). Try this from console: "service eztables start" and do the iptables -L -n again, see if you've got any change. Did you try this without a savefile (or using a new savefile)? I can't reproduce your problem here, when I enabled eztables I've got it working.

@gcmartin - posted the link for hardinfo repo in Quirky's thread.

[olinm]

en_GB

or

en_GB.UTF-8

Which one should be used as i'm confused by this keymap talk?