[SLUG] this Friday's meeting - change of location
Hi, Due to construction at Google, SLUG will be at a different building (and room). It's the "Fairfax Media building" just across the park from the normal building. The address is; Cylon Conference Room @ Google Australia Level 2 Fairfax Media Building 1 Darling Island Rd, Pyrmont, NSW 2009 https://www.google.com.au/maps/place/1+Darling+Island+Rd,+Pyrmont+NSW+2009/@-33.864993,151.195801,17z/ See you there! Jiri -- Jiří Baum 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] A call to arms!
Hi, This months SLUG meeting is fast approaching and we don't have any talks lined up. So this is a call to arms . . . If anybody has a subject to talk about please SMS or call me (Robert on 0468 306 239) or Jiri (0413 183 117) or email the slug mailing list (slug@slug.org.au) Even if all you have is an idea of what you want to hear about post it anyway and maybe it will prompt someone to talk about the subject. Cheers Jiri -- Jiří Baum 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Compilation issues with a C++ Internet-of-Things Event-Framework
Hi, David Lyon: > When I compile, I get these errors: > dlyon@dlyon-HP8000SFF:~/IoT/madstuff$ make > CC madstuff.o > madstuff.cpp: In member function ‘void App::setup()’: > madstuff.cpp:33:45: warning: converting from ‘void (App::*)()’ to ‘void > (*)()’ [-Wpmf-conversions] > addTimerEvent(2, (void (*)()) &App::timerevent); A quick Google suggests that converting a pointer to a bound method to a function pointer is a terrible, non-standard thing to do... They're about evenly split between suggesting you don't do that, and suggesting you use -Wno-pmf-conversions to suppress the warning. The thing is, a pointer to a bound method needs to contain not just the function pointer but also information about the "this" pointer and about the virtual table; converting it to a plain pointer throws all that information away. Only gcc will even let you do that (and still call the function afterwards). So, if there's an easy way to avoid it, don't do that. Otherwise, if you're happy with being gcc-specific, you can suppress the warning. Cheers Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] April meeting — unusual date!
Hi all, As announced at the meeting last night, the April meeting will be held a week late, on the 2nd of May. Please don't come on ANZAC Day, because there'll be nobody there and the building will be all locked and everything. This means that there will be two meetings in May, on the 2nd and 30th. The talks at the meetings are still TBC; if there's a topic you'd particularly like to see presented at SLUG — or, indeed, if you'd like to give a talk — please let us know! Cheers Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] connectivity test loop?
Hi, On 12/03/14 09:00, li...@sbt.net.au wrote: > I have a basic script to watching mailbox for new emails with > cygwin/getmail, that works OK, BUT, occasionally, ADSL goes down, > generally, it comes back promptly, though, not always ... > how can I make some basic connectivity test (against the mail server) to > delay getmail until connectivity exists ? Hmm, normally you'd use the ping command, although with cygwin it may be interesting. Apparently there are two different ping commands in a cygwin environment, each with its own issues, so you'd have to try both and see if you can get one of them to do some useful thing (possibly in combination with grep). You could put a "sleep 300" command in the loop, so that it waits 5 minutes before retrying. That may reduce the looping to a level which is no longer a problem. Finally, you should probably put an actual loop in the script rather than calling the script as the last command. Put "while true; do" at the top and "done" at the bottom instead of calling the script again. The way it is now, each time around the loop will run up a new process, which will hang around forever. (Alternately, turn the last line into "exec /home/voytek/sms" so the new copy of the script replaces the old one.) Cheers Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] I can't be the only one.
Hi, David: > Oh.. sorry.. you needed to install that *before* it got stolen... Well, you can install stuff on your phone via the web. No idea whether it can also be remotely launched — it probably can, but whether you can do it, or just Google and the NSA, I've no idea. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] cygwin script file name error
Hi, > I re-created and retyped all 3 lines, actually, pasted one line at a time. > (more than once...) If all else fails, try putting # signs on the end of each line... then the ^M will be comments :-) Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] shell script to parse html and date comp ?
Hi, > Rick, Jiri, > many thanks, I'm trying to get this done; > egrep works good, I was trying to shortcut date comparison thing by > converting to seconds, BUT, hit a snag as my system is set to: > # set | grep LANG > LANG=en_US.UTF-8 check also the LC_* variables: set | grep LC_ > # date --date='20/03/2014' +"%s" > date: invalid date `20/03/2014' > # date --date='03/20/2014' +"%s" > 1395234000 > which I think means assumed format is mm dd yy ? Yes; 1395234000 is 20 Mar 2014. > where should I change it to en_au ? LANG=en_AU.UTF-8; export LANG >> Of course, all of this would be much easier to code in python, perl or >> ruby. > I'll look at that next time, thanks It's probably still worthwhile switching even now — you haven't even started with the ics/vcal generation and so on... Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] shell script to parse html and date comp ?
Hi, Voytek: > I would like to fetch date/time from html file, and use date comparison > and make an ics/vcal file eventually Hmm, sounds like you might want to use a different programming language, like python, perl or something, rather than just a shell script. It can be done in shell script, but it'll be advanced shell scripting, while it would be fairly basic python or perl or whatever. > the date comes as so: > Start Date Time: rowspan="1">20/03/2014 1400 Thursday > is 'grep -o' the way to go ? what regex do I need where I put ? You might want to use "grep" first to get the correct lines out, so you're not doing multiple things in one command, then either grep -o to get the date out, or the cut command. Something like: grep 'Start Date Time' | cut -d '>' -f5 | cut -d '<' -f1 > what do I need to do with date to be able to compare it to a date range? You have two options here: * convert it to an integer, with the date command (date -d $x +%s), then work with them as integer number of seconds, or * touch a temporary file, again with the -d option, then check which file is older. Cheers Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Issues with Ubuntu 12.04 LTR
Hi, Maxim Zakharov: > For Ubuntu 12.04 it makes sense to install Lubuntu on an old laptop, which > is lightweight version of Ubuntu using LXDE Desktop > See > http://complete-concrete-concise.com/ubuntu-2/ubuntu-12-04/ubuntu-12-04-how-to-install-the-lubuntu-desktop And of course the advantage of this is that you can do it with your existing installation. Install it, and then you can choose between them every time you log in. All it takes is a bit of disk space, and most of us have plenty of that these days. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Issues with Ubuntu 12.04 LTR
Hi, Michael Chesterton: > There's no downgrade option, you can restore from backups, but I would > recommend fixing the problem rather than going backwards. As the song goes, "boldly going forward 'cause we can't find reverse!" Sorry... Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Printing Dir of NAS
Hi, > I access it via Firefox ( http://192.168.1.2) or Gigolo ( > smb://192.168.1.2 - to mount directories and access them with a > filemanager) > I want to be able to print a directory listing to a file - on a HDD > attached to my PC I use "ls" but dont have Terminal access to my NAS. Couple of ways: * once you've mounted the directory, you can use the normal "ls" command * the "smbclient" program will let you run various commands Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Brief Intro to Programming
Hi,
As promised at Friday's meeting, I'm sending a copy of the code from the
second part of the talk.
I'm sending it in two forms:
* SLUG-2014-01.ipynb is the IPython Notebook file, which is the
interactive exploratory programming environment I was using during my
presentation.
* SLUG-2014-01.py is an ordinary Python program, converted from the
above. This is the more common form of Python programs.
I've also uploaded my slides here:
http://www.baum.com.au/~jiri/slug/what-is-programming.pdf
Cheers
Jiri
--
Jiří Baum
Sabik Software Solutions Pty Ltd
0413 183 117
http://www.baum.com.au/sabik
#!/usr/bin/python
# -*- coding: utf-8 -*-
# 3.0
#
print 'Hello World'
print 'How are you?'
#
a = 1
print a
b = 2
print b
print a+b
b = 3*b
print b
#
farmyard = ['dog', 'hen', 'horse', 'sheep', 'cow']
print farmyard
#
print sorted(farmyard)
#
def find_my(needle, haystack):
""" Narrate a search for needle in haystack. """
for animal in haystack:
if animal==needle:
print 'This is my %s.' % (animal)
return
else:
print 'This is not my %s! It is a %s!' %(needle,animal)
print 'There is no %s!' % (needle)
#
find_my('cow', farmyard)
#
find_my('shipwreck', ['coral', 'clownfish', 'anchor', 'mermaid', 'nemo'])
#
find_my('ostrich', farmyard)
#
{
"metadata": {
"name": "SLUG-2014-01"
},
"nbformat": 3,
"nbformat_minor": 0,
"worksheets": [
{
"cells": [
{
"cell_type": "code",
"collapsed": false,
"input": [
"print 'Hello World'\n",
"print 'How are you?'"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"Hello World\n",
"How are you?\n"
]
}
],
"prompt_number": 1
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"a = 1\n",
"print a\n",
"b = 2\n",
"print b\n",
"print a+b\n",
"b = 3*b\n",
"print b"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"1\n",
"2\n",
"3\n",
"6\n"
]
}
],
"prompt_number": 2
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"farmyard = ['dog', 'hen', 'horse', 'sheep', 'cow']\n",
"print farmyard"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"['dog', 'hen', 'horse', 'sheep', 'cow']\n"
]
}
],
"prompt_number": 3
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"print sorted(farmyard)"
],
"language": "python",
"metadata": {},
"outputs": [
{
"output_type": "stream",
"stream": "stdout",
"text": [
"['cow', 'dog', 'hen', 'horse', 'sheep']\n"
]
}
],
"prompt_number": 4
},
{
"cell_type": "code",
"collapsed": false,
"input": [
"def find_my(needle, haystack):\n",
"\"\"\" Narrate a search for needle in haystack. \"\"\"\n",
"for animal in haystack:\n",
"if animal==needle:\n",
"print 'This is my %s.' % (animal)\n",
"return\n",
"else:\n",
"print 'This is not my %s! It is a %s!' %(needle,animal)\n",
"pr
Re: [SLUG] Working with PKCS7
Hi, Amos: > Perhaps: > openssl pkcs7 -text -noout -print_certs Well, that prints out the certificates (with -inform DER). I was more after verifying the signature and extracting the actual content of the message, though... I can see the message in there, it's an XML file with another XML file hex-encoded within it... I just haven't found a way to get it out. > http://www.madboa.com/geek/openssl/ Hmm, that does sound more promising... although the smime subcommand is refusing to recognise my file as an S/MIME message. Which is probably not surprising, given that the pkcs7 subcommand is accepting it as a DER. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Working with PKCS7
Hello, Anyone have any hints for how to work with PKCS7 (.p7s) files? So far I've managed to find that openssl can create them, and convert them between two different encodings, but nothing that would actually check the signatures and/or extract the text... Any suggestions? Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] script help with grep or regex ?
Hi, > the page is reasonably constant (until any fixes?) Yeah, that's always a danger... > all the values I want are in a single 6000 char line, how do I break the > 6000 char line into individual vaules, 'grep any_value file' gives me the > whole 6000 chars ? Several ways, probably the easiest is to use the --only-matching option on grep. That way, it will only return the parts it matched. For instance, you might write: grep -o '[0-9]*' file Another way would be to use "sed" to add a newline at some reasonable points, for instance at the beginning of each tag. Finally, if you use Perl, you can just search for /([0-9]*)<\/foo>/ which will give you the parts you want directly in $1. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] script help with grep or regex ?
Hi, > thanks. I think you might have anwsered my next question already: > what I'm doing is like: wget url > html; lynx html > text Ah, skip the lynx step. Just work with the html directly. All the tools (sed, awk, grep) can work directly with html. > even it's somewhat outside of my abilities, I guess I could try Perl? > (if I can find some sample code) To some extent it depends on how variable the original page is. Once you skip the lynx step, you might even find that it's trivial. For instance, the values you want might all be in a tag with a single, fixed "class" or "id"; you can just grep for that, then strip off the HTML with sed. This is especially likely if they're all from the same website, generated from a single template. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Client Management Systems and data base
Hi, Johannes Nielsen: > I am in the process of setting up a Data Base for my Marketing clients and > and I am wondering what people on the list use and what anyone could > recommend for a small to medium sized business where I will be managing > client data and I would as well like to back it up to a cloud server. For a traditional database, you can either use PostgreSQL or one of the MySQL/MariaDB family. I tend to use PostgreSQL, but I get the impression it's not a big difference; either of them will do a good job. > I am also looking for a client management system where I can set up events > such as automatic reminders, and Birthday greetings and the like and I am > as well looking to back that up to a secure cloud server. Hmm, there are several CRM systems for Linux (cross-platform), but I haven't had a chance to evaluate and compare them recently. I think last time I set one up it was OpenERP, there's also SugarCRM and a couple of others. It may be worthwhile to get one which works with the database you picked in point 1 (or vice versa), to simplify administration. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] bzr multi-move
Hi, Robert: > b) bzr mv --auto should dtrt all at once. So it does! Except it can't handle more than about 10k files at a time (only 8G RAM on the box), but that's OK, I can split it into 10-20 batches. More worrying is why some of the files end up modified after "mv" and "bzr mv --auto", but that's probably a separate problem. Possibly a much more serious separate problem. I'll give it a go on a different box and see if it works better. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] bzr multi-move
Hi, Zenaan: > Ah ok, I remember now, git figures out renames implicitly, so in git ... As it turns out, you can ask bzr to do implicit renames, and this appears to work. It can only do ~1 at a time, but that's OK, easy enough to break up into chunks. "bzr mv --auto" Thanks for the help! Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] bzr multi-move
Hi, Jiří: >> I need to rename 100,448 files in bzr, but running "bzr mv" 100,448 >> times is taking too long. Anybody know if there's some sort of >> multi-move? Zenaan: > I only know git sorry. If there is any similarity, then some > assumptions needs to be provided by you, eg how are the files > structured (one directory, or multiple?), what depth of directories?, Various directories in a tree structure. > and how are you doing a single rename? Well, I have a list of old-new filename pairs (and a script to produce them, if they need to be in a different format). The first idea was just to make a shell script that runs "bzr mv" for each pair, but that takes too long. I think the problem is that bzr does a lock and sanity check each time it starts up. Hence the request for a multi-move... > If export-to-git (eg apt-cache show git-bzr bzr-git) and back to bzr > is quick enough (probably not) that might be a way. Probably not, and it would probably also change all the commit IDs so all the other copies of the repo would have to be fixed up. > Perhaps figure a quick way in bash (with find or some such) and > perhaps bzr can then auto-delete the "old" files and auto-add the > "new" (renamed) files (ie, with a single bzr command).. Yeah, I'd prefer a rename rather than re-adding all the files. With 10 files adding up to 2G, it's worth doing it right. I may try "bzr mv --auto", though; Depending on what algorithm it uses, it may either be perfect, or completely useless. > PS: you might try stackoverflow dot com or whatever it's called > (google site search may be useful etc). Hmm, I tried general search but not site-specific; I'll give it a go. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] bzr multi-move
Hi, I need to rename 100,448 files in bzr, but running "bzr mv" 100,448 times is taking too long. Anybody know if there's some sort of multi-move? Some other trick for moving the files all at once rather than one at a time? Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Beginner Workshop and Installfest
Hi, > I wish I went... would've been handy, just developed a problem, > when I log in over putty, I get: > login as: voytek > voy...@xx.cc.vv.bb's password: > Last login: Fri Nov 22 15:28:50 2013 from 11.22.33.44 > -bash: /etc/profile: Permission denied > # ls -al /etc/profile > -rw-r- 1 root root 1796 Nov 22 16:08 /etc/profile > tried chmoding above, didn't seem to work You probably know this, but just in case - you need to use sudo, since it's a system config file: sudo chmod a+r /etc/profile Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] apt-database
Hi James Linder wrote:, > I have a customer who has indulged in extreme stupid stuff. ... > So I want to scrap and rebuild the package database. ... > Is there any sane (telling him to manually edit the status file is > not sane!) way to recover? It may not be sane, but it's probably better than trying to "scrap and rebuild" the package database... I'm not even sure that's a thing, much less a sane thing. You don't even need to edit much in there, just go to the x11vnc package and take x11vnc-data out of the Depends / Recommends line (and possibly vice versa as well). Once you break the loop, dpkg and even apt will start working again. You can then use them to --reinstall the packages that are broken. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Bogus PPPoE length ?
Hi, Michael: ... > while true; do > d=`date +%s` > tcpdump blah options -c 1 -w tout-$d > if dmesg | grep "Bogus PPPoE length"; then ... Ah, of course. Should have thought of that. Thanks. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Bogus PPPoE length ?
Hi, Jiří: >> Can anyone tell me what it means when I get the error message "Bogus >> PPPoE length field (1502)" a few hundred times a day, usually in >> bunches? Always the same number, 1502, never any other. Jeremy: > Have you tried negotiating a differently sized MTU to see if it changes > the length field displayed in the error message? No, I haven't... Part of the problem is that I don't know what triggers it, which makes it hard to debug. It appears from time to time in the log, usually in a bunch, but not associated with any activity that I've been able to notice. Also, I'm on the wrong end of the link; if I mess something up, I'm fixing it over the phone... > Also, it is possible that TPG support RFC 4638, [0] and your PPPoE > implementation doesn't understand that. But then it should fall back to 1492, especially since that's what I have in my config? Also, if it negotiated a bigger MTU, it would show in ifconfig output? > On a side note, while we're on the topic of MTUs -- changing the MTU > rarely has any effect. I recommend that you fix the MSS value on your > router, so your LAN clients don't hit the MTU limit. > > Something like this out to do the trick: > > # iptables -A FORWARD -p tcp --tcp-flags SYN,RST SYN \ > -j TCPMSS --clamp-mss-to-pmtu OK... Although I've set the MTU on the LAN side to 1412, so that should already cover it? Anyway, I'll give that a try tomorrow. > I'd be interested to see a full packet capture of the PPPoE negotiation > while it generates that error. Hmm, not sure how to capture that, given that I don't know how to trigger the error. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Bogus PPPoE length ?
On 06/11/13 17:19, Menno Schaaf wrote: > Sounds like it might be the MTU? It should be set to 1492 for PPPoE. > <http://en.wikipedia.org/wiki/Maximum_transmission_unit#cite_note-9> Yes, the MTU is set to 1492. ~> /sbin/ifconfig ppp0 ppp0 Link encap:Point-to-Point Protocol inet addr:60.241.25.199 P-t-P:202.7.179.58 Mask:255.255.255.255 UP POINTOPOINT RUNNING NOARP MULTICAST MTU:1492 Metric:1 RX packets:354083 errors:0 dropped:0 overruns:0 frame:0 TX packets:287596 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:3 RX bytes:368792090 (351.7 MiB) TX bytes:83216762 (79.3 MiB) On the underlying interface, MTU is set to 1500. ~> /sbin/ifconfig eth1 eth1 Link encap:Ethernet HWaddr 00:05:1c:20:df:da inet6 addr: fe80::205:1cff:fe20:dfda/64 Scope:Link UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1 RX packets:15503248 errors:6 dropped:0 overruns:0 frame:0 TX packets:15711458 errors:0 dropped:0 overruns:1 carrier:0 collisions:0 txqueuelen:1000 RX bytes:2928508149 (2.7 GiB) TX bytes:2025599612 (1.8 GiB) Interrupt:17 Base address:0xd800 Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Bogus PPPoE length ?
Hi, Can anyone tell me what it means when I get the error message "Bogus PPPoE length field (1502)" a few hundred times a day, usually in bunches? Always the same number, 1502, never any other. Debian squeeze (oldstable), TPG, pppoe 3.8-3, ppp 2.4.5-4 The connection works, but I'm wondering if maybe it might work better, or if it indicates some problem. Am I doing something wrong? My google-fu fails me... Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] For my next trick ...
Hi, Maxim: > md5sum also has the binary mode, So it does. That's... amazing. Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] For my next trick ...
Hi, Grant: > Just make sure you are also transfering in binary mode etc Well, the md5sum matched, so that shouldn't be the problem... Perhaps the unzip versions are different? I would have thought it's a fairly stable format, though... Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Beginner workshops / installfests
Hi, At Friday's meeting, the idea was floated that SLUG should perhaps run a beginner workshop or installfest or something of that sort, perhaps as a once-off, perhaps regularly (monthly?). We are wondering whether there would be general interest in such a concept and (if so) what precise shape that it should take. Any comments? Obviously, if there is interest in principle, we would also need to put together a bunch of practical details. * Venue and location - at a minimum, we'd need a room with chairs and tables. Internet is optional but would be good. * Date or dates - if it's monthly, what day of the month should it be? If it's once-off, what would be the best date? * Sponsorship - if the venue is not free or if we want to do something extra that will cost money, we'll have to cover it somehow. Closer to the event we'd have to put together install CDs and DVDs, volunteers, extension cords and power-boards, but these shouldn't be too difficult. Any suggestions? Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
[SLUG] Future talks
Hi, The SLUG committee is continually looking for talks for our meetings. Our talks cover anything from programming and software, to desktop, mobile, gaming, userspace, community, government, space and education. They may be introductory or deeply technical. As long as they are related to free software and open source, SLUG is the place for them! SLUG is what the speakers and the community make it. If you have a talk you want to present, or know someone who has something to say, or even just suggestions for talk topics, please e-mail! Jiri -- Jiří Baum Sabik Software Solutions Pty Ltd 0413 183 117 http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] python in linux mint
Hi, Ben Donohue: > If I type python at the shell prompt I get a python >>> > However I'd like something that they can type the program in and another > window opens and displays their program running... or something like that. The "ipython notebook" can be nice. It's nicer to use than the prompt (you can go back, change the code and run it again) and lets you save what you're doing. However, it doesn't do the "displays their program running" and there can be a lot of hidden state (what variable has what value is not visible). Depending on the age of the children and their strengths and interests, something like Scratch may be a more gentle introduction. Jiri -- Jiří Baum http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
Re: [SLUG] Bi-directional rsync?
On 14/05/13 09:16, DaZZa wrote: > The current command being used is > > rsync -v -rlt As Michael wrote, you probably want the -u option, otherwise it's even chance whether a new file will be copied or overwritten. Jiri -- Jiří Baum http://www.baum.com.au/sabik -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
