Re: suggested meeting topic
I've been working on a project for the National Archives that uses Open Source software and would be happy to talk about it anytime. It is a bundled (hardware and software) system that tests magnetic tape media (DLT tapes in their specific case) to verify that the tapes do not have so many manufacturing defects as to be risky for long-term data storage. The hardware is all off-the-shelf components and the software is a custom built application that layers above multiple other existing software components. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
How to stop a cascade of mail delivery failure messages
A client seems to have a situation whereby a forged, and invalid, address in their domain was used to send out spam. Some of the recipient addresses are also bogus and therefore cause a mail delivery failure. This bounces back to the invalid user on their mail server and causes another mail delivery failure back to the original recipient's system. I'm getting the impression from my client's description that this bouncing back and forth is not stopping at this point but rather is continuing. So first let me ask, does sendmail actually work this way, with no limit on ping-pong bounces? If it does then there must certainly be a way to stop it, such as silently dropping failed mail. I'm going to guess that a tweak to sendmail.cf will do the trick and am off to research that. I figured I'd check with the group as well just in case I can learn even more about the situation, which is usually the case with this bunch! Thanks! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: ODBC question
I'd second this. Forget the ODBC stuff. It's pathetically slow anyway. I've used FreeTDS very successfully to do just what you want. It performed very well and was pretty straightforward to install and configure. Good luck! Dan David Berube wrote: Hello, Try FreeTDS. http://www.freetds.org/ Take it easy, David J Berube Kenneth E. Lussier wrote: So, I have been tasked with writing some PHP code on a Linux system running Apache, PHP4, etc. However, the database that "they" want the interface for is an MS-SQL database. After doing some research, unixODBC is the way to go. The problem is, the only way that I can find to do this is to use the ODBC to ODBC bridge from EasySoft, which costs $1000. Does anyone know how to connect to an MS-SQL server from a Linux box without spending $1K (or moving the database to MySQL)? TIA, Kenny begin:vcard fn:Dan Coutu n:Coutu;Dan org:Snowy Owl Consulting, LLC adr:;;P.O. Box 349;Derby;VT;05829;USA email;internet:[EMAIL PROTECTED] title:Managing Director tel;cell:802-309-4781 x-mozilla-html:FALSE url:http://www.snowy-owl.com/ version:2.1 end:vcard
Re: free software alternative to Access
Peter Dobratz wrote: Does anyone know of any free software packages that we can use? Basically, we have hikers and teams of hikers that raise money. We want to keep track of how much money each hiker contributed, and keep their names and addresses, so that we can mail them a brochure for next year's hike. Most of the laptops that people own or borrow for the event have some version of windows on them. Some sort of bootable CD with USB flash drives for the datastore might be optimal. --Peter Well, it isn't a replacement for Access because it is a whole lot better :-) but I'd recommend using MySQL. It runs well on Windows, if that's what you have to use, and has an ODBC component to provide easy interfacing to Office applications (assuming that's important to you.) It also has a GUI interface that you can download, for free, from the MySQL website and that GUI also runs on Windows (and Linux, of course.) You can even put the MySQL database on a Linux server and access it via ODBC from Windows if you like. The main problem, of course, is that it doesn't break as often as Access or have the lame restrictions on scalability of Access. But you can probably manage to live with that... Dan begin:vcard fn:Dan Coutu n:Coutu;Dan org:Snowy Owl Consulting, LLC adr:;;P.O. Box 349;Derby;VT;05829;USA email;internet:[EMAIL PROTECTED] title:Managing Director tel;cell:802-309-4781 x-mozilla-html:FALSE url:http://www.snowy-owl.com/ version:2.1 end:vcard
Re: (ext3 + (2.4.18 -> 2.6.11)) = WTF?
Michael ODonnell wrote: OK - this is now officially upgraded to a double-WTF. I haven't yet had the time to really sit down and do a bit-bashing debug session but in spare moments I've been trying to gather additional clues and I thought y'all might be amused by this one: even though with 2.6.11 I can't (directly) mount /dev/hde1 anywhere because some chunk of code somewhere along the line keeps deciding that /dev/hde1 is busy or already mounted, I am nevertheless able to mount it (indirectly) thus - mount -o loop /dev/hde1 /mnt/hde1 ...while this is all that's necessary with 2.4.18 - mount -t reiserfs /dev/hde1 /mnt/hde1 Oh, yeah - I forgot to mention that I wiped /dev/hde1 and rebuilt the filesystem as ReiserFS instead of ext3 with the symptoms remaining *entirely* unchanged... (!?!?) Well, I saw this very thing yesterday when I added a new disk to a system. Turns out that I needed to run mke2fs to make a file system on the new paritition. Then all was happy. I'm betting that you're creating the partition but not making a filesystem. Does your mount command complain about finding no superblock? That's a sure clue. I don't know what command you'd use to make a ReiserFS file system, you may need to research that. Dan begin:vcard fn:Dan Coutu n:Coutu;Dan org:Snowy Owl Consulting, LLC adr:;;P.O. Box 349;Derby;VT;05829;USA email;internet:[EMAIL PROTECTED] title:Managing Director tel;cell:802-309-4781 x-mozilla-html:FALSE url:http://www.snowy-owl.com/ version:2.1 end:vcard
Re: Need php help after Core 3 => 4
Steven W. Orr wrote: I have a simple phpinfo.php that still seems to work fine but I have a php app that now does nothing. I don't really know php so I don't know what to do. http://frambors.syslang.net/pscal/index.php When I run it I get a blank screen with no error in the apache logs. If I'm on my server and say php < index.php I get no output. I did log into mysql and accessed the tables fine. Any idea? :-( TIA Debugging PHP can be a royal pain. Check to see if you have an error log setup on your system. There should be a php.ini file someplace on your system, /etc is a common location. Read it to see where the error log is (if it is setup at all). Then go read that log file to see what is wrong. If error logging is not setup then edit the php.ini file to set it up, re-run your php script, and go read the error log. That's the easiest way to find out what's going on. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Need assistance with Cisco PIX configuration
I'm trying to setup a PIX to allow MySQL traffic. My remote connection attempts are failing and I'm trying to determine why. Since my server logs are showing no connection failures I'm trying to verify that the PIX is in fact configured correctly. I can see that I can configure it to do use my Linux server for syslog messages and tried that but it also seemed to not work. If there's anyone out there that can offer assistance it would be greatly appreciated. It's best to contact me directly I'm guessing since it's probably not real likely that everyone here wants to know this stuff. Thanks! Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Experiences with using Active Directory for Linux authentication
Surely there are people in the group that have had experience setting up a Linux system so that it uses an Active Directory server via PAM for login authentication. Since we live in world where Linux has to sometimes take over gradually it is often useful to be able to set this up in environments where otherwise the M$ weenies would rule. How about some tips, hints, war stories, etc. on the topic? Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Funky samba bug
I'm hoping someone may be able to point out to me the obvious detail that I'm clearly missing here. Here's the setup. I have a Red Hat 9 server running Samba 3. It has been working just fine. A client had their hard drive die in their windows laptop. So, get a new hard drive, reinstall windows, try to connect to Samba. For some reason their own login is not working from this newly rebuilt system. Other logins work fine. There is no error message of any sort. They get a login popup window when they try to connect to the samba shares and it reappears after they enter their (verified correct) login credentials. An endless login loop. One thing that seems to be different is the Windows user name is now all in capitals when it used to be all lower case. So I changed the smbusers file to reflect that with a line like this: djc=DJC This maps the lowercase djc Linux login with the uppercase DJC windows login (I hope.) I've used smbclient to connect to the server with that person's login, no problem. They can connect from other machines with that login, no problem. Other people can connect from this rebuilt machine, no problem. Only the owner of the machine can't connect! I think it must be a nano-Bill lurking inside Windows. :-) Seriously, is there something with the system's new SID or some other detail that I may be missing? The samba logs show the successful connections but not the failing connection so they're no help. Thanks, Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Jabber experiences
Has anyone here setup a Jabber server as a company's internal IM solution? It seems from what I can learn to be an ideal solution, particulary for businesses whose primary servers are Linux rather than Windows. Are there pros, cons, or other tidbits that people can share on the matter? Thanks, Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Perl include question
On Mon, 2005-12-05 at 15:25 -0500, Cole Tuininga wrote: > How can I set up my include path (within the perl script) to make sure > that the directory of the executable is in the include path? I can't > hard code it because the location might be different on different > machines (different mount points). > Piece of cake. You need to use the FindBin module that is included in the default Perl distribution. Do this: #!/usr/bin/perl use FindBin qw($Bin); use lib "$Bin"; use myinclude; This assumes that your module is named myinclude.pm and is in the same directory as your perl script. For more details use "perldoc FindBin" on the command line. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Perl include question
On Mon, 2005-12-05 at 15:59 -0500, Thomas Charron wrote: > use lib '.'; > > Thomas > Note that this will use your current directory as the place it looks for your module. The current directory would happen to depend on wherever you happen to have last done a cd to. That may not be what you really want. Chances you really want the module to be in the same directory as the perl script itself, or a directory relative to it. For that FindBin is the way to go. Dan > On 12/5/05, Cole Tuininga <[EMAIL PROTECTED]> wrote: > How can I set up my include path (within the perl script) to > make sure > that the directory of the executable is in the include > path? I can't > hard code it because the location might be different on > different > machines (different mount points). ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Multi-boot, partition label conflict
I'd like to just double check my thinking on a configuration. Here's the setup: An HP Itanium machine comes with RHEL AS 4 factory installed on the internal SCSI disk. It also contains a fiber channel controller card for use in connecting to an existing SAN. I did a new install of RHEL to the SAN (so that it could boot from the SAN rather than the local disk) and upon boot it complains because it is trying to mount / and finding that there are two partitions with that label, one on the SAN and one on the SCSI disk. The ways around this that I see include: 1. Use something like Partition Magic to hide the SCSI partition. 2. Format the SCSI disk (since it won't be used anyway.) 3. Change the /etc/fstab to mount by device name rather than label. 4. Change the label on at least one disk. Are there other options that I'm missing? Has anyone else run into this? Thanks in advance, Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: One more bites the dust
Jon maddog Hall wrote: I saw the announcement of the new Palm Treo 700w today, and thought I might take a look at it. My old phone is getting a bit long in the tooth, so I thought I might go for a new Treo. When I go to the page http://web.palm.com/products/smartphones/treo700w/details.jhtml I saw the agonizing Microsoft logo. Apparently Palm has decided to put their "ease of use" on top of the Windows Mobile platform. It means that every time I would go to use my phone I would have to be reminded of a monopoly gone bad. It is a shame that one of the last bastions of closed source good software felt they had to partner with someone who is basically their biggest competitor. I may never be able to purchase another Palm again. Sadly, maddog Palm has also announced that they have plans to release products that use a Linux-based O/S. That might sit with you better. Now all they have to do is actually release it Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Database question
Paul Lussier wrote:
This question specifically deals with PostrgreSQL and other
SQL-compliant databases. I say this, because the question deals with
foreign keys and constraints, which I'm pretty sure MySQL doesn't deal
with properly, if at all.
MySQL 4 and newer does handle foreign keys. MySQL 3 and older didn't.
I have the following table, which most other tables reference:
hosts=# \d machines
Table "public.machines"
Column | Type | Modifiers
--+--+-
id| integer | not null default nextval('ids'::text)
itemtag | text| not null
model | integer | not null
location | integer | not null
monitor_temp | boolean | not null
serial_number | character varying(64) |
Indexes:
"machines_pkey" PRIMARY KEY, btree (id)
"machines_itemtag_key" UNIQUE, btree (itemtag)
Foreign-key constraints:
"machines_model_fkey" FOREIGN KEY (model) REFERENCES machine_models(id)
"machines_location_fkey" FOREIGN KEY ("location") REFERENCES locations(id)
And this table:
hosts=# \d classes
Table "public.classes"
Column | Type | Modifiers
--++-
id| integer| not null
class | text | not null
Indexes:
"classes_pkey" PRIMARY KEY, btree (id)
Out of curiosity why don't you define the id column here the same way
you did in the machines table with a default value. I *always* define
primary keys like that, it saves so many headaches down the road.
I want to create a table which has the following:
hosts=# \d class_members
Table "public.class_members"
Column | Type | Modifiers
--++-
id| integer| not null
member| text | not null
Indexes:
"classes_members_id_pkey" PRIMARY KEY, btree (id)
However, I want to restrict the member column by restricting the data
in it to also exist *either* in machines.id *OR* in classes.id. The
reason for this is that a class member can either be a machine or
another class (think netgroups here). Does anyone know how to do
this, or if it's even possible?
Let me ask this, why can't a machine be a special type of class? You
could add a colum to the class table that indicates the class type and a
class type of machine could then indicate a cross-reference into the
machine table.
If you do this then the class_members foreign key constraint becomes easy.
You could then take this a bit further to allow a class to contain
sub-classes, which might be useful too because you could then create
groups of classes that could be manipulated all at once. If you go this
route then it might become a good idea to break down the class table
into two pieces, class and class_detail where class_detail might contain
(but doesn't have to contain) a class id value, machine id value, or
some other id value. If you structure the class_detail table such that
individual column types are foreign keys to specific tables then you
have full constraints checking enabled.
Another possible approach might be to chuck the class_member table
entirely and instead have multiple tables such as machine_class,
class_class, etc. that map one table to the other (machine to class,
class to class, and so forth.)
Anyway, food for thought. Hope it helps.
Dan
I suppose one solution is to just not have nested classes and
explicitly list each machine that's a member of any given class as
such, but, well, that's not overly elegant :)
Thanks for any insight.
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Database question
Paul Lussier wrote: Dan, Is this what you meant: class_types: id | integer | nextval name | text| not null primary key: id classes: id | integer | nextval name | text| not null type | integer | not null primary key: id foreign key: type references class_types(id) members: id| integer | not null class | integer | not null type | integer | not null primary key: name foreign key: class references class(id) foreign key: type references class_types(id) Then, based on the value of members.type, I could figure out which table to look up the member(id) in. If it were of type 'class', then the id would map into that table, if of type 'machines', the id would map into the machines table, etc./// Thanks! Yes, this would work. Because the members table is provided a many (members) to one (class) relationship you can build data structures of arbitrary depth. Further, by using the class type indicator you can later expand the logic to include things that you have not yet considered in the same way that you're including the machines table here. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: PHP Templates and/or Web Frameworks
Ted, I've created lots of PHP code and love the SMARTY templating system. I've used a number of template systems in both Perl and PHP and this is the best one I've found anywhere. Look for it at smarty.php.org Note that this is different from a framework, it simply provides a mechanism for templates. I've not yet found a good general purpose framework for use with many things. Instead I tend to build on a framework that's focused on the desired job at hand. So for example Mambo is a good framework for a content management based site. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux Dist for use with 64 Bit AMD System
Mark Rousseau wrote: Hi All, Does any one have recomendations for a distro best suited for a 64 bit AMD system? I'm looking to build a server for home with raid drives which will act as a file, mail, and a development server. Any recomendations for hardware vendors for budget hardware( under $1500 ). thanks, Mark I'm using SuSE on my AMD 64 bit (Athlon) laptop. The boxed set provides both 32 and 64 bit installation choices. It runs great, no problems at all. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux Dist for use with 64 Bit AMD System
Rich, that would be SuSE 10. I've not tried to use the modem on my HP Pavilion zv5000 so I can't say whether or not it works. I believe my system uses the same modem as yours. Dan Richard A Sharpe wrote: Dan What version of SUSE do you use? I tried but can not get my internal modem to work with SuSE. The internal modem is in a hp pavilion notebook running an AMD 64 Athlon CPU the modem is an Agere Systems AC97 Modem. Thanks Rich Richard A Sharpe 8 Meadowview Lane Merrimack, NH 03054 "Treat everyone with politeness, even those who are rude to you, not because they are kind, but because you are." -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dan Coutu Sent: Monday, February 06, 2006 5:05 PM To: Mark Rousseau Cc: gnhlug-discuss@mail.gnhlug.org Subject: Re: Linux Dist for use with 64 Bit AMD System Mark Rousseau wrote: Hi All, Does any one have recomendations for a distro best suited for a 64 bit AMD system? I'm looking to build a server for home with raid drives which will act as a file, mail, and a development server. Any recomendations for hardware vendors for budget hardware( under $1500 ). thanks, Mark I'm using SuSE on my AMD 64 bit (Athlon) laptop. The boxed set provides both 32 and 64 bit installation choices. It runs great, no problems at all. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Crontab entries
Paul Lussier wrote: Hi all, Is there a requirement to place parens around a set of commands delimeted by a ';' for crontab entries? At some point I got in the habit of creating entries like: 30 02 * * * (foo;bar;baz) But nowhere can I find documentation even mentioning the use of parens. Of course, I haven't even seen mention of having cron exec'ing multiple commands either... Using the parens definitely works, I'm just trying to figure where I picked up that habit... Thanks, You picked up that old habit because the shell runs the set of commands within parentheses as a subprocess. This insures that the group of commands within the parens are run within the same context. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Unkillable processes?
Okay, here's a strange one. On a Red Hat 9 system I've encountered a situation where there are two processes that I cannot kill when using kill -9 (or any other value, for that matter.) What's the deal with that? The only kind of process that I've ever run across that I could not kill was a zombie and neither one of these is a zombie. Just to add more confusion to the mix, or maybe a useful clue, the system load average is about 4 but top shows 97% system idle time. Strange. Any ideas? Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Unkillable processes?
Ben Scott wrote: On 2/17/06, Dan Coutu <[EMAIL PROTECTED]> wrote: On a Red Hat 9 system I've encountered a situation where there are two processes that I cannot kill when using kill -9 (or any other value, for that matter.) Do a "ps aux" and note their status. It's "D", right? That means they're in "uninterruptable sleep" -- waiting for system call to finish something that cannot be interrupted. The "D" stood for "driver" or "disk" originally. Bad hardware or buggy device drivers are the most common cause of a process stuck in this state. The only thing you can do is wait or reboot the system. If the syscalls ever complete, the kernel will immediately process the kill signals you sent, so those processes are dead, they just don't know it yet. :) -- Ben Hmm, the I/O wait seems likely. We've been having trouble with an IOMega REV 10 disk autoloader ever since we bought the thing. Even swapped it out for a new one but still get flaky behavior. Maybe it's time to send the thing back... Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: change file names
Bill McGonigle wrote: On Feb 21, 2006, at 10:51, Jon maddog Hall wrote: I use "apropos" a lot, but it is just not the same as the printed man pages. You're on to something there. It got me thinking, "so what's the good modern-day equivalent that doesn't involve dead trees"? At first I thought, well there must be an RSS feed for "Man Page of the Day". There's one fella who has a sub-section of his blog dedicated to that which he updates once every several months. For a long time I've liked the xman program in order to browse man pages. I typically invoke it with the -notopbox switch to minimize screen clutter. It shows you all of the available man pages, sorted alphabetically, in a given section. Handy. Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
SCSI hardware question
I'm a bit fuzzy on the specifics of different types of SCSI and in particular the compatibility and termination requirements when mixing devices. I believe that I'm seeing intermittent problems due to hardware being configured wrong. I had someone setup a new system and it isn't behaving right so I'm trying to troubleshoot it. Here's the setup: RedHat 9 is the O/S Using an Adaptec 29160 SCSI controller. This provides a variety of SCSI connection possibilities. The external connector is a 68 pin LVD. There are three internal connectors that provide the following: 68 pin LVD, 68 pin Ultra Wide, 50 pin Ultra SCSI. Connected to the internal LVD connection is a Tandberg SLR60 which is specified as LVD/SE Wide Ultra2 SCSI by the manufacturer. Connected the external LVD connection is an Iomega REV 1000 autoloader which is specified as LVD Ultra3 SCSI. Terminators are in place on the tape drive and REV drive. Are these devices, the tape, REC, and controller, compatible with each other as detailed above? I'm not at all clear on the impact of the tape drive being single ended (SE), for example. The differences between Ultra2 and Ultra3 are also unclear. Thanks for any insight that can be provided! Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
What the heck is a dbus?
On a new installation of Centos 4.2 (a RHEL clone from source) I'm
seeing the following error in the messages file:
Mar 3 15:55:06 hanka dbus: Can't send to audit system: USER_AVC
pid=2521 uid=81 loginuid=-1 message=avc: denied { send_msg } for
scontext=root:system_r:unconfined_t tcontext=user_u:system_r:initrc_t
tclass=dbus
I've never heard of a dbus. So first of all, what's a dbus, and
secondly, what does this error mean?
Thanks for any insights that you can share!
Dan
___
gnhlug-discuss mailing list
gnhlug-discuss@mail.gnhlug.org
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Linux on an AS/400?
I have access to a fairly cheap AS/400 and am wondering if Linux will in fact run on it. Preliminary research indicates that it might but I thought I'd ask here in case anyone actually has experience with it. Do I need a version of Linux, like RHEL, that is blessed by IBM? Dan ___ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
USB 'disk drives'
Well, I got one of those cute little USB 'thumb drive' things that is supposed to look like a removable disk on Windows and is supposed to also work with Linux. So I plugged it in and began the adventure of trying to figure out how to access it. I see that using usbview shows me the thing and recognizes that it is a storage device with the correct amount of free space. Further it appears that it is seen as device /dev/sda. Okay. So the intuitive command 'mount /dev/sda /mnt/usbdrive' (after first creating /mnt/usbdrive of course) wants to have a filesystem type specified. That's where I'm getting stumped. It seems to use a FAT filesystem but uses a 12-bit FAT instead of 16. None of the filesystem types I've tried will work. I'd love to setup the thing to automount but can't do that if I can't mount it at all! Has anyone already wrestled this alligator and figured out how to use it? If not I'll dive into research mode and learn more about using USB things with Linux. Thanks! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: USB 'disk drives'
geoff allsup wrote: Hi Dan The drive should be something like /dev/sda1 (not sda); my USB pen drive is set up as 'auto' in my fstab, and mounts as 'vfat' out of the box. Hope this helps, geoff That was it! I could have sworn that I previously had no /dev/sda1 device, maybe I wasn't paying enough attention. Anyway, once I used /dev/sda1 instead of /dev/sda it mounted like a charm without me having to specify the filesystem type. Thanks! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: What do you use for documents with equations?
Bill Freeman wrote: I've got a little article with a bunch of math to write, and want to use an open source GUI tool. I've used lyx, and it's OK, but I use it so rarely that I need to re-learn it every time. Does abiword or open office do a good job on equations? Are there other tools that I should look at? Ideally it will mix not only text and equations, but also plotted (geometry text style) figures. (I guess that you can embed an image, at least, in any of them.) Thanks, Bill I've not used it for real but KFormula lets you write equations and complex math expressions. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: awk assistance
Michael O'Donnell wrote: find / -type f | while read f; do basename $f; done This does not meet the requirement of providing UNIQUE instances of filenames though. Easily fixed by piping it through sort and then uniq ala: find / -type f | while read f; do basename $f; done\ | sort | uniq The nice thing about this approach is that it uses NO code of any sort. If using awk is the 'old' way then this approach must be the 'ancient' way since it actually uses the original UNIX design thinking of pipes and filters that often gets lost nowadays. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Custom printing from web form data?
I have an interesting challenge to meet. I can think of one way to solve it but was wondering if others knew of already existing tools or techniques that might make it easier. What I'm trying to accomplish is to print certificates (to be snail-mailed) based on data collected from a form on a web site. I think it would be most efficient to print from a cron job using the data that had been previously collected and stored in a useful place, like a MySQL database. The tricky part is that the certificates should look fairly classy, which means they can't be just plain ASCII text. They should include graphic elements such as your typical certificate border with the fancy artwork. Of course multiple colors are required too. My thinking was to use a wee bit o' perl along with Latex and pre-made graphic elements to crank out a PostScript file for each certificate, then queue it up to the color printer. Any other ideas out there? -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Custom printing from web form data?
[EMAIL PROTECTED] wrote: On Mon, 25 Nov 2002, at 8:57am, [EMAIL PROTECTED] wrote: My thinking was to use a wee bit o' perl along with Latex and pre-made graphic elements to crank out a PostScript file for each certificate, then queue it up to the color printer. It would likely be cheaper/faster to purchase pre-printed certificate stock (paper) and just print the name, etc., on a black-and-white printer. Yes, it would. But then all the certificates would have to look the same. I guess another requirement in my head that I didn't realize was that I need to be able to generate a wide variety of certificates with a single program. So they can't all look the same. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Networking help
[EMAIL PROTECTED] wrote: In a message dated: Mon, 25 Nov 2002 16:10:41 EST "Ken D'Ambrosio" said: I shouldn't be an ARP issue -- if it were, then the other machine sending pings wouldn't work. Namely: Keep in mind. The pinging machine, systemA *cannot* ping systemC, but *can* ping systemB. B and C are on the same subnet, A is not. By 'cannot ping' I mean, I type 'ping systemC' and it just sits there. However, by ssh'ing to systemB, and from there to systemC, I run 'tcpdump -i eth1 icmp' and I can see that systemC *is* in fact receiving the "icmp echo request" packets. systemC just isn't replying to them! So we need to examine the possibility that System C doesn't know how to reach System A even though A does know how to reach C. I'd check netmasks on all the systems involved. If I am remembering right you're going from a Class A network (10.whatever) to a Class C network (192.168.whatever) here. Getting netmasks right is critically important in this kind of environment. I do, however, have to wonder if you're routing correctly. Can you ping the remote subnet's router address? I very much wonder if it isn't a router configuration problem. I note that the 192.168.*.* addresses are reserved for "private" networks that are often NAT'd behind a router that does NAT. It can be either really easy to setup this or, if setup wrong, really a pain. I'd first examine very closely the configuration of the router that bridges the two networks. If for some reason you find that there are TWO routers, or more likely a computer with two NICs bridging the two networks, then routing can get really funky as the rogue route sometimes handles traffic and sometimes doesn't. I've seen stuff like that cause me sleepless nights... :-( -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Internationalization with perl?
Have any of you illustrious folks had experience with doing Internationalization (I18N) within perl code? I know that by default perl now processes strings as unicode but that's just the tip of the iceberg. Sniffing the locale from the environment (or the web browser settings in the case of CGI and mod_perl code) is also straightforward. What is less clear is what mechanisms (CPAN modules etc.) are available for things like message catalogs. Any pointers would be most welcome! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Apache mod_perl Problem ? Help!
Ben Boulanger wrote: > On Mon, 2002-12-23 at 09:18, Vince McHugh wrote: > >> How do we re-compile the www.necs.biz apache to include mod_perl >> > > You don't really need mod_perl - it's just an accelerator. Check the first > line of the CGI to ensure that it reads something like: > #!/usr/bin/perl > > where /usr/bin/perl is the location of your perl binary. > > Secondly, make sure this CGI is in the appropriate spot on the webserver (/cgi-bin/) - there's a binary (suexec, I believe) that can whack cgi's > in other places - if anyone knows why this is, I'd love to hear it. > > Ben > > Truth? Not quite. It is very possible to create a perl script that will run properly under mod_perl but not as a CGI. Yes, mod_perl can be thought of as an accelerator but that is not all it does. It also provides a number of other useful features that you just can't get in a CGI context, such as persistent database connections. The suggestion about using the RPMs sounds like your best bet. See where that gets you. In wandering through the httpd.conf on my Red Hat system I see that by default it will process scripts in /var/www/perl using mod_perl. (This might be dependent on the above mentioned RPM stuff.) Besides that, you can run CGIs in any directory you choose by making the appropriate annotations in your httpd.conf file. It involves adding ExecCGI to a Directory directive. Read the Fine Manual page for specific details. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: CUPS under RH 7.3/8.0
Mark Komarinski wrote: 1) Get apt-get for RPM. Gives the best combination of using existing RPM files while getting the network/dependency/updating features that APT uses. You can get a copy at freshrpms.net. Get the one for your version, then apt-get update and apt-get upgrade. Let it finish. 2) apt-get install cups cups-devel gimp-print-cups cups-drivers There may be other dependencies that apt-get will install. Hmm, so I tried the above two steps on my 7.3 system and get: # apt-get install cups cups-devel gimp-print-cups cups-drivers Reading Package Lists... Done Collecting File Provides... Error! E: could not open RPM package list file /var/state/apt/lists/apt.freshrpms.net_redhat_7.3_en_i386_base_pkglist.os: (no error) E: Problem opening /var/state/apt/lists/apt.freshrpms.net_redhat_7.3_en_i386_base_pkglist.os E: The package lists or status file could not be parsed or opened. It would appear that I get this error because the file that it can't open doesn't exist! Any idea why? Or how to create it? Nothing in the apt related man pages provides a command or switch that corrects the problem. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: CUPS under RH 7.3/8.0
Mark Komarinski wrote: Did the apt-get update work? -Mark Duh, I flew right past your mention of that in the text. I've got it switched to using CUPS now. BTW, there is no redhat-switch-printer utility. What I found was: /sbin/service lpd stop /sbin/service cups start /usr/sbin/alternatives --config print The last command seems to do what I'm guessing that redhat-switch-printer was supposed to do, permanently switch my printing system to CUPS. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Apache: unrestricting content-length?
[EMAIL PROTECTED] wrote: Hi all, I'm trying to get Apache to accept files for upload to a server. Everything worked fine while testing on my local system, but as soon as I moved the cgi script to the production server I began getting errors about: Request content-length of 224690 is larger than the configured limit of 75000 Interstingly, I can not find where this limit of 75000 is set. There's nothing in my httpd.conf file that I could find. I'm using Apache 1.3. Anyone have any ideas? The default limit is set at compile time. You can raise the limit to as much as 2147483647 bytes (which is 2GB). Do this by adding this to your httpd.conf file: LimitRequestBody 2147483647 Of course you can use whatever other number you like... -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: CD-Rs?
Michael O'Donnell wrote: I usually look in the Sunday circulars. Somebody is usually selling packs of 100 for about $7 or $8 after rebate. They are typically either generic CompUSA discs, or Imations. I have never had a problem with them myself. Just FYI, quality does sometimes appear to be an issue with some drives. At work we have a stack of el-cheapo blanks that burn just fine but are then unreadable in about half the drives we try them in. After I went out and bought those Maxell and Memorex disks at Wal-Mart they worked perfectly in all the same units where the el-cheapo disks failed. I'm not saying that Maxell or Memorex are necessarily better quality than any el-cheapo disks you might purchase (and indeed, some el-cheapos may simply be unlabelled/surplus Maxell or Memorex disks, or better) just that quality sometimes does matter. ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss I have seen the el-cheapo disks delaminate quite a lot of times. I simply will not waste my money on them anymore. I watch for sales on the brand-name disks, like TDK, Maxell, Memorex, etc. and get those instead. I've never had a problem with one of them delaminating (or forming bubbles inside the plastic of the disk.) -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Recognizing touch tones
Bob Bell wrote: Someone has approached me about writing some software that would basically provide a phone menu. It would recognize touch tones (phone button presses), gather information, provide voice prompts, and take action based upon the information entered. Any idea what it would take to accomplish this? I'm pretty sure I could record WAVs/MP3s/etc for the voice prompts and such, but the biggest question for me is around the use of the phone. Specifically: 1) How do I respond to incoming calls? 2) How do I get an audio file to play into the telephone? 3) How do I recognize touch tones as input? Bob, I don't know about the software side of things but do know that you can buy off the shelf hardware packages to do what is called Computer Integrated Telephony (CIT) and they handle the touch tone recognition and tie in to the computer for doing the voice response parts. Perhaps with a published API creating the software would be workable. I'd starty by tracking down the CIT hardware and existing package solutions first. Then you'll have an idea of what is out there before you dive in a write a bunch of code that other people have already written a few times over. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: CGOs and SSL?
[EMAIL PROTECTED] wrote:
Has anyone else ever had problems running CGI scripts under an SSL
enabled Apache? I'm trying to use the usemod-wiki, which is small,
self-contained wiki written entirely in perl. Everything works just
fine if I'm running it on port 80 without SSL. As soon as I turn on
SSL and run over port 443, it stops work (i.e., I can no longer edit
pages).
I'm guessing you already checked your web server configuration for the SSL
instance of the virtual host. If not, do.
I'm guessing that something to do with the SSL is interfering with
the passing of CGI params, however, I can't seem to pin it down.
Does anyone have any suggestions for debugging CGIs? (I turned on -w,
didn't help much).
Something I find to be very useful is to add (if it's not already there)
'require CGI::Carp' to the perl code. Then you can add carp('yourmessage')
calls wherever you like in order to see what is really going on. The
messages will show up in the web server's error log (typically
/var/log/httpd/error_log on many Linux systems.)
Additionally, if you wrap them in a statement like:
carp("Received name with value = " . $Name) if $Debug;
Then you can enable/disable them with one easy edit by assigning $Debug the
value of either 0 or 1.
--
Dan Coutu
Managing Director
Snowy Owl Internet Consulting, LLC
http://www.snowy-owl.com/
Mobile: 603-759-3885
Fax: 603-673-6676
___
gnhlug-discuss mailing list
[EMAIL PROTECTED]
http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Bit depths in X
Mark Komarinski wrote: VNC is nice in that I could have it start up a DTE different from the main display. But that was a bad thing, as now the users had to fight with GNOME/KDE and FVWM. The bigger problem is I had to show them to start VNC, connect to the proper display, and then (important!) shut down VNC when they were done with it. These are postdocs who stay logged in for weeks at a time. I had visions of 150 VNC sessions running at once. Are you saying that you have 150 different people needing to use VNC? If you have a manageable number then you can do this: At system startup time create a VNC server for each individual user. So Joe gets display :0, Ann gets display :1, Mary gets display :2, etc. Each one can be setup to require the user to login in order to access their display, this helps to enforce people using only their own assigned server. Then if they wish to stay logged in for weeks at a time it's no big deal. Another advantage to this method is that each user can have their session customized to to their own preferences (screen resolution, applications that run at login time, window manager, etc.) The command line for starting the VNC server is amazingly flexible. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Bit depths in X
Mark Komarinski wrote: On Thu, Mar 27, 2003 at 09:01:25AM -0500, Dan Coutu wrote: Mark Komarinski wrote: VNC is nice in that I could have it start up a DTE different from the main display. But that was a bad thing, as now the users had to fight with GNOME/KDE and FVWM. The bigger problem is I had to show them to start VNC, connect to the proper display, and then (important!) shut down VNC when they were done with it. These are postdocs who stay logged in for weeks at a time. I had visions of 150 VNC sessions running at once. Are you saying that you have 150 different people needing to use VNC? No, but I could see them following my instructions to start VNC and have 150 instances for one user because they forget to stop it. -Mark Okay, maybe I wasn't clear. The idea is that that they *do not stop* the server. They just disconnect now and reconnect later. The VNC client lets you specify which display server to connect to on the remote machine. So each user gets one assigned server and that is the only one they can use. You can prevent getting lots of servers running by not allowing users to have sufficient privileges to run the server on their own. Simple. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ Mobile: 603-759-3885 Fax: 603-673-6676 ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
[Fwd: June 2 talk - privacy and public policy]
I know this is short notice but I just learned about it. I think this may be of interest to many members. Those of you were in UEG will remember Wendy, who is sponsoring this lecture. Directions: From Manchester and north. Get onto Rt. 101 West, follow it through Bedford into Amherst. Watch for an exit sign that reads "Rt. 122 Amherst" and take that exit. Turn right at the end of the exit ramp. You'll pass the Police station on your left, continue to a traffic light. Turn right at the light and go to the stop sign two blocks down. Turn left at the stop sign and the library will be on your left about a half block down. From central Mass. Head for Rt. 13 North and follow it to Milford, NH. Then get onto Rt. 101 East. Take the next exit (Rt. 101A East). Turn right at the end of the exit ramp and get into the left lane. At the next traffic light turn left onto Rt. 122 North. Follow 122 to the next traffic light (not a blinking light, a real traffic light.) Go straight through the light and turn left at the next stop sign. Library is on your left as noted above. From Keene, NH. Take 101 East to Milford. Then follow directions above for Central Mass beginning with the Rt. 101A East exit. From Nashua, NH and points East. Take 101A West to the intersection of Rt. 122. Turn right onto 122 North and follow directions as noted above in the Central Mass. block starting with 'Follow 122 to'. I think that covers everyone except private pilots... :-) Original Message Subject: June 2 talk - privacy and public policy Date: Mon, 02 Jun 2003 20:59:37 -0400 From: Wendy Rannenberg <[EMAIL PROTECTED]> To: undisclosed-recipients:; Feel free to forward this to anyone who might be interested. If anyone has questions please send me email. Barbara Simons. professor of Technology Policy at Stanford University, will give a presentation, Privacy and Public Policy at the Amherst Town Library on June 3rd from 6:30 to 8:30PM. This will be an interactive discussion and attendees are encouraged to come with questions. The talk is free and open to the public. As we all know privacy is a concern from the mis-use of medical and insurance information to the stealing of another's identity. Yet there is much more to this issue than meets the eye. The new and evolving Defense Advanced Research Projects Agency (DARPA) Total Information Awareness (TIA) program is to be accomplished by the development of new technologies, including systems that would mine large quantities of data in order to identify potential terrorists. The U.S. Public Policy Committee of ACM (USACM) issued a letter that raises concerns about security, privacy, economic, and personal risks of TIA. This seminar will provide an overview of personal privacy issues and how they are affected by not only the TIA program but other changes in federal laws. Join us to learn about your privacy rights, policies, laws, and what you can do to protect those rights and your personal information. Barbara Simons, past President of the Association for Computing Machinery (ACM), founded ACM's US Public Policy Committee (USACM), which she currently co-chairs. Recently, Dr. Simons has been teaching technology policy at Stanford University. She is a Fellow of ACM and the American Association for the Advancement of Science and she has received numerous awards including the Outstanding Contribution Award from ACM, and the Pioneer Award from EFF. She was selected by c|net as one of its 26 Internet "Visionaries" and by Open Computing as one of the "Top 100 Women in Computing." Dr. Simons has testified before both the U.S. and the California legislatures and at government-sponsored hearings. She co-founded the Reentry Program for Women and Minorities in Computer Science at U.C. Berkeley. Simons served on the Presidents Export Councils Subcommittee on Encryption and on the Information Technology-Sector of the President's Council on the Year 2000 Conversion. She is on the Board of Directors of the U.C. Berkeley Engineering Fund, Public Knowledge, the Math/Science Network, and the Electronic Privacy Information Center, as well as the Advisory Boards of the Oxford Internet Institute and Zeroknowledge, and the Public Interest Registrys .ORG Advisory Council. She has testified before both the U.S. and the California legislatures and at government sponsored hearings. She was runner-up in the first election for the North America seat on the ICANN Board. She co-founded the Reentry Program for Women and Minorities in Computer Science at U.C. Berkeley. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: [Fwd: June 2 talk - privacy and public policy]
Gee, it does say the 2nd doesn't it? But given that I spoke with Wendy in person, after 6:30 *yesterday*, and she told me it was *today* I'm going to jump to the rash assumption that the date was a typo. I can't believe she'd miss the very talk that she was sponsoring... [EMAIL PROTECTED] wrote: On Tue, 3 Jun 2003, at 9:50am, [EMAIL PROTECTED] wrote: I know this is short notice but I just learned about it. [...] Subject: June 2 talk - privacy and public policy U, short notice is nothing new for this group, but if I read that correctly, this talk was *yesterday*. Correct? Unfortunately, my TARDIS is in the shop, otherwise I'd attend. ;-) -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
A useful perl command line skeleton
As a follow-on to my talk about Perl last month I thought I'd put
together a handy code framework that is very useful for creating command
line or cron based utilities.
The skeleton doesn't do much by itself but I have found myself making
command line utilities over and over based on this same structure. It
provides easy command line parsing as well as built-in online help and
self-documentation by using pod. Oh, and it also provides a convenient
version variable that is CVS friendly so that you don't have to update
versions by hand each time you check your code into CVS.
You can get details for each of the packages used by using the perldoc
command like this:
perldoc Getopt::Long
perldoc Pod::Usage
and so forth.
I hope this helps those of you who are trying to make sense of perl.
--
Dan Coutu
Managing Director
Snowy Owl Internet Consulting, LLC
http://www.snowy-owl.com/
#!/usr/bin/perl -w
#
# Author:
# Created:
#
# Describe the purpose of the code here.
#
use strict;
use Getopt::Long;
use Pod::Usage;
# Provide a version variable that works nicely with CVS
our $VERSION = (qw( $Revision: 1.0 $ ))[1];
# Command variables
my $Help;
my $Verbose = 0;
# Parse any command line switches.
GetOptions("help|h|?" => \$Help, # A flag (boolean) value
"verbose|v=i" => \$Verbose # An integer value
) || pod2usage(2);
pod2usage(1) if $Help;
# Basic verbosity shows at least the program version to the user.
# By comparing for verbosity > some value each bit of output
# can be fine tuned into different verbosity levels.
print "myprogram version $VERSION\n" if $Verbose > 0;
__END__
=head1 myprogram
Describe your program briefly.
=head1 SYNOPSIS
myprogram [options]
=head1 OPTIONS
=over 8
=item B<--help>
Show this description of the program's use.
=item B<--verbose>
Specifies a positive integer value of zero or greater that controls
the level of verbose detail to be provided by the program. Higher numbers
provide more detail.
The default is zero if the switch is omitted.
Re: A useful perl command line skeleton
Morbus Iff wrote: >As a follow-on to my talk about Perl last month I thought I'd put >together a handy code framework that is very useful for creating command >line or cron based utilities. See also: http://sial.org/code/perl/scripts/blank.pl.html Thanks for sharing the pointer! >#!/usr/bin/perl -w If you're looking for crossplatform fun, "use warnings" is more portable than the -w. Good to know, I never tried using perl for other than experiments on Windows. ># Basic verbosity shows at least the program version to the user. ># By comparing for verbosity > some value each bit of output ># can be fine tuned into different verbosity levels. >print "myprogram version $VERSION\n" if $Verbose > 0; If you use $0 here, you can save yourself some editing. True, but only in the main code body itself. Once you get into the data section after the __END__ indicator this no longer works. So you can't use it in the pod text to auto-substitute the program name there. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: OT: Perl and Javascript
Hmm, you're writing a perl script that is acting as a browser (of sorts) in order to login to a remote site and then do other things there (I"m assuming.) If so then you should be receiving the HTTP data, including the request from the server that they are using to determine whether or not your 'browser' supports javascript. You could examine what they are asking and perhaps determine how to provide the 'proper' response to fake out the other end. Of course it will then probably throw some JavaScript at your code with the expectation that it will do something with it. You will then need to analyze that and see if you can figure out how to throw back a useful result to the server (despite not actually parsing the JavaScript). All this may be possible but you may need to understand (or learn) a fair amount about HTTP (the protocol itself) to pull it off. Of course, you could surf CPAN to see if someone else already did the grunt work for you. http://www.cpan.org/ Kenneth E. Lussier wrote: Since we seem to have an abundance of Perl experts today, I have a perplexing problem... I am trying to write a script that logs into a website. The problem that I am running into is that the website requires javascript. If the site detects that the "browser" doesn't support javascript, then it loads the login form as an "unmodifiable form text field" (as reported by lynx). Does anyone know a way around this? TIA, Kenny ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Detecting root kits?
Last week I uncovered a RedHat box that had been rooted (fortunately it had only recently been installed and nothing important was on it.) Rather than me having to go through a hands-on intensive process of analyzing every other Linux system on the LAN are there tools that I can use to determine whether or not this SOB got into other systems? Any pointers to where I can learn more about the different types of rootkits and how to counter or detect them are also welcome. Thanks! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Detecting root kits?
brian wrote: FWIW, I've also found a lot of rootkits hidden in the /home and games directories on various systems. For starters, I'd also compare the sizes of your various utils, like top, ls, more, etc to known good utils. If you can mount the infected disk on another clean server as RO to analyze it, that would also make diagnosis easier. This is an example of the manual effort that I'm trying to avoid. Doing this with upward of 40 systems takes too long and results in way too much downtime. The chkrootkit package is a quick once over. The best place to look is in /dev, as that's where a lot of rootkits hide their stuff. I find a command like this is pretty useful: find /dev -ls -maxdepth 1|grep d[-r][-w] I'm pulling over the chkrootkit package. Sounds like exactly what I'm looking for! and then make sure those directories that it returns are actually supposed to be there. ls is almost always trojaned, hence the reason to use find. FYI, in this case a lot of utilities were trojaned. The list is: dir, find, locate, md5sum, pstree, slocate, top, lsof, ifconfig, syslogd, login, ls, netstat, and ps. Interestingly I was able to determine this because all of these were owned by a regular user account rather than root! The system also had these odd things: two hidden directories (that I've found, there may be more) /var/nis/..\ \ (that's two dots and two spaces) and /etc/nhm/... (three dots.) Both had binary and data files used to get at things that you'd rather keep private and included a trojaned httpd and a utility called write that was collecting data to be sent elsewhere. There was also a file /etc/cron.daily/sync that sends email to a yahoo.com account with the contents of /etc/.mc In /var/tmp there was a file t.tgz that contained a file called t which had been unpacked in that same directory with owner of root and t was setuid, setgid containing binary data. The file utility did not think it was a normal executable though. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Detecting root kits?
Tom Fogal wrote: Last week I uncovered a RedHat box that had been rooted (fortunately it had only recently been installed and nothing important was on it.) Rather than me having to go through a hands-on intensive process of analyzing every other Linux system on the LAN are there tools that I can use to determine whether or not this SOB got into other systems? Any pointers to where I can learn more about the different types of rootkits and how to counter or detect them are also welcome. Thanks! -- i seem to remember freebsd having a nightly cronjob script that would save the md5sum of every file on the system to a file.. and then compare it with the md5sum of the same file that it generated 24 hours before. differences were mailed to root.. i always thought this to be a good idea, perhaps you could implement it yourself? seems like a simple shell script. anyway, that solution doesnt work well in your current situation. do you have another box w/ the same updates, that you know is clean? you could compare md5's from that one... HTH, -tom Something like this can be done by utilities like tripwire. The catch is that they need to have been setup *before* the breakin occurs. I'm trying to play catchup here with systems that are in an unknown state of security (or lack thereof.) I also know how to detect this particular hack but it involves a lot of manual effort with copying known good utilities (like ls and lsof) and examining a number of different directories and files. Quite time consuming. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: What's the best way to automatically swap smart hosts (sendmail)?
[EMAIL PROTECTED] wrote: On Tue, 1 Jul 2003, at 9:31am, [EMAIL PROTECTED] wrote: Can anyone think of any cool ways to do this? Use an SMTP relay that supports SMTP authentication, and will allow relay from outside the local network if SMTP AUTH is done. Configure Sendmail to use that relay at all times, and to use SMTP AUTH. That will only not work if you travel between more than one network that blocks TCP 25 outbound. If that's the case, yell. :) Since MV now supports SMTP authentication I think that you have your solution. Use it all the time. ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: [gnhlug-announce] July Mondalug meeting
ray bordt wrote: hi from ray bordt in raleigh, nc please tell me about MySQL tools/apps. i am a Realtor and want to switch from M$ stuff to linux. thanks from the hot South! Ray, you're being pretty vague and general. Given that many books could be written to "tell me about MySQL tools/apps" and given that nobody here is likely to type in that much text, a more specific set of questions would make it more likely that you'll get a useful response. Specifics about just what you are trying to do, how you plan to use it, how much data you are dealing with, etc. provide a much better context for answering questions. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: [gnhlug-announce] SLUG "problem solution" meeting Monday 7/147pm at UNH in Morse 301
On Wed, 2003-07-09 at 21:21, Bruce Dawson wrote: > On Wed, 2003-07-09 at 14:37, Robert Anderson wrote: > > ... > > I'm still open for problems to work on. If you have something you've > > been trying to do and would like to have the groups input, please send > > an email or be ready to explain your problem to the group on Monday. > >... > > Well, something I've been struggling with lately is how to plan a site > for upgrading with minimal downtime (i.e. just the time required to > reboot). > > One thing that we've looked at is having two computers with identical > configurations - software and hardware, and use LVS (or some equivalent) > to switch the web and mail services to the other system when one goes > down. However, this falls apart rather quickly - databases need to be > duplicated, resync-ing things like log files are a royal pain, ... > > I'm mostly wondering how other sites handle this with minimal hardware > and proprietary software investments. > > --Bruce The amount of downtime you'll have is proportional to the amount of money you spend. I'm sure you realize this and were just hoping to get around it somehow. The upgrade operation is rather complex. As you point out things can get sticky. My experience is that the time needed to synchronize databases can be minimized but not completely avoided. Much of the rest of the system can be duplicated in advance. If you roll your logs as part of taking the primary system offline then you can simply copy the old versions to the new system. It is then no different than a normal log rolling operation. The DB resync time I've minimized in the past by using a full DB backup from earlier in the day to populate the standby DB and then updating it with only the newer data from the DB at the time of switchover. This often results in a downtime measured in minutes rather than an hour or more. The specific methods to use for this will of course depend on the database being used. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Graphics blips
On Tue, 2003-07-15 at 22:32, Mark Komarinski wrote: > This is a real hard problem to describe, but it's really annoying. It > seems to happen with only HTML code that I write. > > Take a look at http://www.wayga.org/~mkomarinski/julie.php > > Take a look between the images, in the lower right hand corner outside the > border. > > See a little blue dash? > This is a really old problem. It is, as has been mentioned before, due to white space in your HTML. It is almost always due to a newline just after the tag and before the tag. If you like to make your HTML look neat and tidy then you will almost certainly see this little bugger show up. The trick is to place everything from the anchor start tag () to the anchor end tag () on a single line with absolutely no spaces around the tag that is between them. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Looking for a freeware ecomm solution to run on Linux
The best solution I've found yet is osCommerce. It's a SourceForge project, written in PHP, and using MySQL for the DB. It looks like it will scale well and is pretty flexible. If you've never setup an ecommerce site before you will definitely want to learn as much as you can about the ins and outs of ecommerce as you tackle this. At first glance it seems simple but the reality is that there are a lot of little gotchas that will rear their ugly heads over and over. I've been doing ecommerce stuff since 1997 and have the bruises, war stories, and headaches to prove it! :-) Installing osCommerce is pretty easy, just do *not* install the admin files underneath the catalog files like so many other people do. That's just asking for a cracking. Check out SourceForge for more details. Jason Kern wrote: Anyone know a good solution? I would like to have something I can replicate over several sites on one server, plus have a web interface for clients to administer... Help? Thanks, Jason Kern www.KernBuilt.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: OpenOffice help?
Modify the characteristics of the Default or Text Body style. Or your text may be in some other style. The idea is to modify the style. Right click on the text in question. Select Edit Paragraph Style... Click the Indents and Spacing tab. Change the line spacing and save. Ta da! [EMAIL PROTECTED] wrote: Hi, Anyone know how to change the spacing between lines in OO? I've got a Word doc I need to add some stuff to, and my default spacing is greater than what the original doc was written in. I can't seem to find the right buttons or knobs in OO to tweak. Thanks, -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Mouse swapping on a laptop
My laptop has a touchpad mouse built-in and that works fine with my mouse setting as a generic 2-button ps/2 mouse. I have a Logitech wireless wheel mouse that I attach to it when working with it on the desktop. I do use the touchpad when take the computer elsewhere, no point in dragging around a load of hardware. What I'm trying to determine is whether or not there's a way to enable use of the wheel on the cordless mouse without making a mess of the mouse behavior when it is not plugged in. I tried to set the mouse type to generic wheel mouse and the scroll wheel works like a charm. The problem is that when I then unplug it and try to use the touchpad (after rebooting) the mouse movement goes haywire. Sometimes it tracks okay and sometimes the pointer slams to the lower left corner of the screen and won't come out to play. Urk! I figured maybe disabling GSM would help but it didn't. Do I have to resort to not being able to use the wheel at all unless I reconfigure the mouse every time the system boots with/without the external mouse? Thanks for any bright ideas that may pop up. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Mouse swapping on a laptop
Bill Freeman wrote: Derek Martin writes: > While again, I didn't spend much time on it, the problem with using a > PS/2 mouse seems to be that Linux sees that and the internal mouse > device as the same logical device (/dev/psaux), with the same > configuration. However, obviously that's not true in reality. I know > of no way to handle two different configurations on the same device... I suspect that this is actually a BIOS issue. Probably the same hardware is used to interface the internal mouse and the PS/2 socket, with a wee bit of steering logic. If I'm correct, at boot time the BIOS "probes" for a mouse with the steering set for the external connector, and if it finds none, flips the "switch", and initializes the internal mouse. So either mouse gets talked to by the Linux kernel via the same hardware register set, and /dev/psaux is, in fact, totally correct for both. If the switching control bit is readable, you could conceivably as a module to read it, and adjust the mouse configuration accordingly. That would be chip set, and maybe even brand/model specific, however. Another hope would be for the BIOS to provide a callback interface to query the mouse choice. Bill Well in my case I can use both the external mouse and the touchpad at the same time. So clearly the BIOS is not switching one or the other into place. The idea of plugging the external mouse into the USB port is worth investigating though since this particular mouse can use either USB or PS/2 port. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Mouse swapping on a laptop
On Sat, 2003-08-02 at 18:50, [EMAIL PROTECTED] wrote: > On Fri, 1 Aug 2003, at 2:34pm, [EMAIL PROTECTED] wrote: > > Remember the PS/2 port is not hot-swap safe. > > Most laptops have electronics specifically designed to handle hotswap for > PS/2. I won't say "all", but I'd be surprised to find one that did not. > Ah, that explains why both devices can function simultaneously. > Myself, I have solved this kind of problem in the way that Derek Martin > suggests. There can be only one PS/2 "auxiliary" device, so the laptop's > electronics have to be responsible for multiplexing the PS/2 signals. Not > so with USB. By using only one pointer as a PS/2 pointer, and the rest as > USB pointers, the kernel and/or XFree can handle multiple inputs and > multiple protocols. > I'm trying this approach, it is simplest. I like simple. > I think Bill Mullen's idea of using multiple XFree configuration options > would also work, but the USB method (if the hardware allows for it) avoids > the need to specify X startup options. While I could certainly see how this one would work okay it's just not automated enough for me. Stuff like this should be possible to write code to handle so that I don't have to fool with it. Guess that's why I write software, so I don't have to do stuff manually! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: I HATE SPAM (was Re: Mouse swapping on a laptop)
On Sat, 2003-08-02 at 23:52, Derek Martin wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > > I suspect the problem Dan Coutu <[EMAIL PROTECTED] deleted]> is seeing arises > > I must publicly and strenuously object to the practice of including > posters' e-mail addresses in the body of your message, including in > attribution lines[1], on the basis that it is tantamount to asking > spammers to spam your friends. > I'm with Derek on this one! Ben, thanks for trying to acknowledge attribution but it could readily be done without the email address detail. The realization of spammer behavior that Derek describes is why I removed by email address from my signature. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: internationalization and localization
On Fri, 2003-08-08 at 11:25, Derek Martin wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Anyone know of any /good/ references to i18n and the problems (and > solutions) which lie therein? > > Let me state categorically that source code is NOT, IMO, a good > reference to anything, other than perhaps a specific implementation to > the solution of some problem (i.e. itself). There are MUCH better > ways to learn about things (assuming they're available) than to read > source code, despite what Linus T. says. > The best source of this information I've found is the Tru64 book called "Writing Software for the International Market". It used to be clearly available on the web. The closest I can come now to a valid link for it is this chapter: http://h30097.www3.hp.com/docs/base_doc/DOCUMENTATION/V51B_HTML/ARH9YCTE/CHSSFTWR.HTM There are also some good man pages on the Tru64 system, but you probably don't have access to one. I also have a reasonably good paper book on the subject but it's at home. Will have to look it up later. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: internationalization and localization
On Fri, 2003-08-08 at 11:25, Derek Martin wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Anyone know of any /good/ references to i18n and the problems (and > solutions) which lie therein? > A book that I've found very helpful is: "Programming for the World" by Sandra Martin O'Donnell -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: gnome startup errors
On Sun, 2003-08-10 at 00:34, Derek Doucette wrote: > I have somehow broken something in gnome. Whenever I start X as gnome > is loading gnome-panel, magicdev, metacity, and nautalis all segfault > and then sometimes work again. I had this happen just a couple days ago after I used Ximian's updater to install Evolution 1.4 on my system. The problem turned out to be that for some reason orbit was not getting started. Orbit is the Corba server that almost all Gnome stuff uses. I was able to determine this by starting applications from the command line and seeing the error messages that normally go into the bit bucket when you startup things from the graphic interface. I did find that if I logged with Gnome instead of KDE then all worked correctly. I never fully resolved the issue since I reinstalled Red Hat yesterday onto a new bigger drive and just migrated my old files off the old drive. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Booting Debian to single user with SCSI Raid
I'm having difficulties trying to boot a Debian 3.0 system to single user on a Compaq ProLiant 380 system. This has a SCSI Raid array and so drives are named /dev/ida/c0d0p1, /dev/ida/c0d0p2, etc. I've determined that boot is on c0d0p0 and root is c0d0p1. I've tried to boot with the Debian install cd (disk 1) and entered rescue root=/dev/ida/c0d0p1 1 at the boot: prompt but it gives me a kernel panic every time, claiming that it can't find the device. I've tried to see if I can stop the lilo boot process to kick it into single user mode but that doesn't fly either. At one point in the boot process it allows me to hit Enter to get a shell prompt. But this turns out to be pretty useless. I can see what files are available in the ramdisk by using "echo *" and there is a mount command but of course it also will not mount the drive. The basic problem I've got is that the system seems to be getting wedged when it tries to start X. I cannot get an alternate console window on the thing and the network doesn't startup either. So I have no way in to modify the init config so that it doesn't try to start X anymore. Any ideas are welcome! Thanks, Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Web Hosting
> I'm looking for a new hosting service as I don't have the time or > bandwidth to support my current setup anymore (and it's only a matter of > time until Comcast realizes I'm operating a high traffic commercial > website from my basement). I naturally Googled and came up with > Dotservant.com I've researched hosting options and found a local company in Nashua that offers competitive pricing and sound UNIX know-how. They offer FreeBSD systems and Windows systems. Access to MySQL is included in the pricing. Take a look and see for yourself if they meet your needs. http://www.itza.net/ Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Question about printing PCL files
I just caught wind of someone having trouble with printing PCL files (that's HP's printer control language, I guess PostScript wasn't good enough for them.) They are using RH 9, standard CUPS printer server setup. They're using lp to submit print jobs and trying to specify options for controlling orientation (landscape/portrait) and pitch (I'm assuming cpi). The options they specify to lp are being ignored. Naturally one wonders if it might be a printer driver issue or a filter issue. Any thoughts about what to look into to resolve it? I'd like to be able to help them through this knothole so that they'll be happy with with their conversion from SCO Unix to Linux. Thanks in advance! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Question about printing PCL files
On Fri, 2003-09-26 at 18:33, mike ledoux wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Fri, Sep 26, 2003 at 06:07:13PM -0400, Dan Coutu wrote: > > I just caught wind of someone having trouble with printing PCL > > files (that's HP's printer control language, I guess PostScript wasn't > > good enough for them.) > > > They are using RH 9, standard CUPS printer server setup. They're > > using lp to submit print jobs and trying to specify options > > for controlling orientation (landscape/portrait) and pitch > > (I'm assuming cpi). The options they specify to lp are being > > ignored. > > IME, most PCL files start with a reset sequence, which would cause > this behaviour. If that isn't it, they should also ensure that > their PCL files aren't setting those options explicitly. > While that might be case a little testing shows that it is not the only reason that lp would ignore options. I tried to print a gif file with no options at all. It printed in landscape mode. I have no idea why. When I tried to use the options to lp, specifically this: lp -o landscape image.gif It still printed in landscape mode. Before I get flamed you should check the manpages, they specify that the 'landscape' option will rotate the printout by 90 degrees. This is important, it doesn't say it will print the page in landscape mode! So if the page normally prints in landscape mode you would have to use the landscape option to print it in portrait mode. Also note that there is *no* 'portrait' option! Interesting no? So there would seem to be other things to consider. Other ideas? -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Best In Class
On Wed, 2003-10-08 at 07:39, Jon maddog Hall wrote: > O.K., here is the situation: > > Mixture of Windows, Unix and Linux systems, the person needs to do these > things over the net, preferably with a graphical interface. > > Workload Monitoring > User Activity Monitoring > Automated Monitoring of Log Files > Process Monitoring > > ideally you could also do this: > > Application Resource Management > Performance Management > Availability Management > Directory and File Management > > Are there any packages out there, either freeware or proprietary that do > much of these things? One that I've used a lot over the years that the first half (the monitoring) of the functions you describe is Big Brother. It has both a server and a client component. The server runs on Linux, various Unixen (I've used it on Solaris), and on Windows NT/2K. The client runs on lots of systems including Netware, MacOS, Windows, Linux, UNIX, etc. It is *very* extensible in that you can write your own extensions to monitor whatever you want and then plug them in to the server or client as appropriate. Extensions can be written as shell scripts, Perl, PHP, C, C++, etc. It isn't explicitly setup to the administration stuff but perhaps the extension mechanism plus sufficient cleverness will get you that. It can notify you of problems by email, pager, or cell-phone SMS messages. Of course you can write your own custom notifier mechanism too. Their web site is: http://www.bb4.com/ You can download their demo package for free, it has no timebomb and is not 'crippleware'. It's fully functional. A business would most likely find it useful to 'purchase' the product. The purchase is actually more like an annual support contract than a product purchase. The support from the vendor is outstanding, they're a great bunch. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: apache config for pw protected web pages.
> Hello all, > > Problem: Trying to get the contents of a specific location (e.g. > www.domain.com/semiprivate) to be password protected. I have set the > .htpassword file in the "semiprivate" directory. > I have added the following: > > AllowOverride Options > > > Now I am confused and have some mixed information. Should I.. > > a. go to the .htaccess file in the /var/www/html/domain/semiprivate and add > the following: > > AuthType Basic > AuthName "My secret pages" > AuthUserFile /path/to/directory/.htpassword > require valid-user > order deny,allow > allow from all > > if so, where? Just drop em in any ole place? > > --OR-- > > b. Edit the apache directives for the virtual host and add previous among > ServerAlias, DomainRoot and the like. > It depends on how you like to maintain your systems. You could go either way. If you like having your settings close to your files then put it in the .htpasswd file. If you like your settings all in one place then put it in the httpd.conf. Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Allowing remote root login
First of all, I know this isn't a great idea, but it is required by a specific scenario. Here's the situation: Got a RedHat 9 system that I need to allow remote telnet logins to root from the LAN. I had thought that an entry in /etc/security/access.conf would do the trick but it didn't. I also went to /etc/xinetd.d and edited the telnet file to enable telnet. Still no joy. I must be missing something else but don't know what. A search of the mail archives turned up nothing. Ideas? -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Allowing remote root login
On Mon, 2003-10-13 at 16:23, Jared Watkins wrote: > Dan Coutu wrote: > > >Got a RedHat 9 system that I need to allow remote telnet logins > >to root from the LAN. I had thought that an entry in > > > > > What about using sssh? If you can then simply enable root logins in > /etc/ssh/sshd_config > Using telnet is a bad idea... it continues a bad habit. SSH with > passwords is good... with keys is better. > > Jared I thought I had made myself perfectly clear that I understand ssh is better than telnet. Due to circumstances way beyond my control I *require* a telnet based solution. I've already tried to find a way to do the same thing with ssh and it just doesn't work due to termcap and terminfo handling. I'm working with some antique application packages that are not ssh savvy. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Allowing remote root login
On Mon, 2003-10-13 at 17:10, Jerry Feldman wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Mon, 13 Oct 2003 15:58:45 -0400 > Dan Coutu <[EMAIL PROTECTED]> wrote: > > > First of all, I know this isn't a great idea, but it is required > > by a specific scenario. Here's the situation: > > > > Got a RedHat 9 system that I need to allow remote telnet logins > > to root from the LAN. I had thought that an entry in > > /etc/security/access.conf would do the trick but it didn't. I also > > went to /etc/xinetd.d and edited the telnet file to enable telnet. > > Still no joy. > > > > I must be missing something else but don't know what. A search of > > the mail archives turned up nothing. Ideas? > Dan, > WRT: The other posts, have you been able to allow for non-root logins > over telnet. Then, once logged in, then use su, sudo or sux to become > root. > > You may need to add /dev/pts to /etc/securetty. > But, I think you are opening up a can of worms. > Yes, I know it is a can of worms. Unfortunately I need to first solve this and then work on finding a way to solve it better. The problem is that the client has an application which allows administrative actions only when logged in as root. It uses an ancient curses based interface that uses strange terminal mapping that simple doesn't work correctly with any of the conventional termcap entries that I've tried. Either the keyboard mapping is wrong or the display gets unreadable. Everything works correctly though when connecting with a terminal emulator that only supports telnet connections. Sigh. Thanks for the help folks, sorry for the impatient reply to Jared earlier. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Allowing remote root login
On Mon, 2003-10-13 at 17:27, Jerry Feldman wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Yes, but can you first log in as a regular user then either su or sudo > or sux? Ah, sorry, forgot to answer that. No, the application runs as a captive login. (When you login the application starts, you can't get to a shell.) -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: A parable on portability (was: Allowing remote root login)
On Mon, 2003-10-13 at 19:32, [EMAIL PROTECTED] wrote: > I'd be willing to bet the software in question has hard-coded terminal > control sequences embedded in it, or, at best, that it has its own, private > database of terminal control sequences, and ignores termcap/terminfo > entirely. The terminal control sequences it uses likely came from some > [expletive] programmer who didn't bother understanding how things were > supposed to work, but just hacked at it until it worked on his terminal. > So, now you need to use his terminal to use his program. I'm sure you're close to the target. From what I've been able to determine the bad programming is less likely in the application itself as in the programmer who customized the application for this particular client. I've already found that he bypassed a very elegant and flexible printing system in order to hardwire in values he liked because it was 'easier'. Hmmm. Given that the UI is fully customizable I'm betting that he did the same with the UI. Sigh. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
[OT] Need linux printing help
Okay, I've tried to post a note to the gnhlug-jobs list but it seems to be in moderator limbo. I figure 24 hours is long enough to wait for it to either show up or for me to be notified that it was killed. I need help with a client on the seacoast (they're actually in Salisbury, a stone's throw out of NH) who has just switched to Linux. The do a *lot* of printing and are having trouble getting their printouts to look right. I'm not asking for volunteer help, this would pay cash money. I've flagged this as off topic because it truly belongs in the other mailing list as a job posting. Note that this is very short term stuff. I don't know if if it will turn into more work or not. If you're interested and *very* printer savvy please contact me offline rather than reply back to the group. It's bad enough that I had to even post this note here. The early bird gets the worm on this one. Thanks, and please forgive the non-technical content. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Calendaring (server side)?
On Mon, 2003-11-03 at 17:20, [EMAIL PROTECTED] wrote: > So: I'm looking for something, server-side, that's RFC 2445 (iCal) > compliant, that someone's actually -used-, preferably with multiple > RFC-2445 clients (eg. Outlook, Mozilla). OSS would be preferable, but I'm > willing to pay if need be -- though I'd -vastly- prefer a Linux-based > solution over a Windows-based one. > > Any suggestions? I ran across this company last year while looking for new clients. Their product is written in Perl and will run on Linux. It's not an open source solution though. I'm not certain that they do the iCal compliance. You'd have to ask. The nice thing is that once you get the product the source code comes with it! http://www.webevent.com/ -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Novell to acquire Suse
On Tue, 2003-11-04 at 09:40, Michael ODonnell wrote: > http://www.suse.com/us/company/press/press_releases/archive03/novell_suse.html This is pretty interesting. Now couple this with Red Hat's announcement yesterday that they were discontinuing the ordinary Red Hat Linux distribution and only selling the enterprise package. That particular announcement had left me thinking that maybe SuSe would be a good choice for the distribution to switch to. It seems like the only distribution not in turmoil lately is Debian and they are just really really slow. (I know, they like to say they're 'stable'.) I guess we live in interesting times... -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Novell to acquire Suse
Paul Lussier wrote: In a message dated: Tue, 04 Nov 2003 09:51:32 EST Dan Coutu said: It seems like the only distribution not in turmoil lately is Debian and they are just really really slow. (I know, they like to say they're 'stable'.) I've never been bothered by Debian's release cycle. I run Debian/Testing on my desktop with Debian/Stable on my servers. If I need a newer version of package on my stable servers, it's usually available from backports.org. I've run 'unstable' on my home machine for a few years now with no problems. I recommend you try Debian, it's like Green Eggs and Ham. You might just find you'll like it, Sam-I-am! :) Just to be clear, I've been managing a set of about 7 Debian production servers since May and while I do see how solid it is I seem to be often frustrated trying to figure out how to properly update modules on the system in a precise manner that won't impact the production code that is running while allowing addition of new features that depend on newer modules. The biggest burr in my butt so far has been trying to get the systems so that I can use X applications (like xemacs) on the servers that display back to my desktop via an SSH tunnel. The servers do not have X installed on them because they're not workstations. But it sure would be nice to be able to start X applications there and display them on my X server. Perhaps I'm just encountering a severe shortage of information, knowledge, or both. Given that I've been working with UNIX systems for over 15 years the knowledge shortfall would have to be something specific to Debian though. ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Novell to acquire Suse
Michael ODonnell wrote: The biggest burr in my butt so far has been trying to get the systems so that I can use X applications (like xemacs) on the servers that display back to my desktop via an SSH tunnel. The servers do not have X installed on them because they're not workstations. But it sure would be nice to be able to start X applications there and display them on my X server. Perhaps I'm just encountering a severe shortage of information, knowledge, or both. Given that I've been working with UNIX systems for over 15 years the knowledge shortfall would have to be something specific to Debian though. It's certainly possible that you can legitimately ascribe your problems to Debian. Another explanation might be that you've overlooked something that's applicable to *any* situation involving X. Would you care to tell us what you've done so far in your attempts to get those X clients to use the remote (desktop) displays? Somebody here might be able to help you, and in any case such an exchange in a forum like this might be interesting to others, as well. Just for starters (for example) have you verified that the X servers on the desktop boxes don't have their "nolisten" options selected, via either their command lines or their config files? The servers won't even be willing to consider inbound connection attempts from external hosts unless instructed that it's OK... Well, I have had no problem with this when interacting with other flavors of Linux or even FreeBSD. So I'm certain it is not my local system that's the problem. I'm using ssh to connect remotely like this: ssh -XCA -l mylogin remote.system.name It is only Debian (3.0) server systems that do not define a DISPLAY environment variable on the remote end (ssh normally does this automatically when you use the switches I've specified above) and so X applications will not display on my local system's X server. I tried Cole's suggestion of installing xbase-clients but that didn't do it. I had been thinking that perhaps I needed to start with a remote environment that already had a valid DISPLAY defined because it was a workstation. Can anyone verify or deny this theory? Thanks, Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: X with ssh, (was Novell to acquire Suse)
Michael ODonnell wrote: Well, I have had no problem with this when interacting with other flavors of Linux or even FreeBSD. So I'm certain it is not my local system that's the problem. I'm using ssh to connect remotely like this: ssh -XCA -l mylogin remote.system.name OK - so you're instructing the SSH clients (via their command lines) to do X forwarding, compression and to allow X authentication traffic to pass. That's a good start. Your ~/.ssh/config files should also have lines in them like this: ForwardX11 true Okay, added it. No difference. What about on the desktop boxes? Have you verified that /etc/ssh/sshd_config and /etc/default on the desktop boxes don't have any entries in them that could be causing trouble? You want entries like these in your /etc/ssh/sshd_config: X11Forwardingyes X11DisplayOffset 10 I forgot to mention that I had also done this step that Cole had pointed out. This was on the server that I'm trying to run the applications on. So I just tried also insuring that the workstation has the same settings. They were okay. ...and you want to be sure that those capabilities aren't being overridden in the /etc/default/ssh files, which are read by the SSH daemons on startup. There are no /etc/default/ssh files on either end. It is only Debian (3.0) server systems that do not define a DISPLAY environment variable on the remote end (ssh normally does this automatically when you use the switches I've specified above) and so X applications will not display on my local system's X server. Since a number of us have Debian boxes that allow remote X connections just fine, I can say with a high degree of confidence that it's much more likely that the problem here is simply that your local config needs tweaking. Oh I'm sure that's the problem. I just don't know what the missing piece is. :-( I tried Cole's suggestion of installing xbase-clients but that didn't do it. I had been thinking that perhaps I needed to start with a remote environment that already had a valid DISPLAY defined because it was a workstation. Can anyone verify or deny this theory? BTW, the SSH clients and servers emit very useful debug info when instructed to be verbose - have you tried that? I tried that. It nicely told me that it was requesting X forwarding with authentication but still there is no DISPLAY variable defined and thus no X forwarding is possible. Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: X with ssh, (was Novell to acquire Suse)
On Wed, 2003-11-05 at 08:03, Cole Tuininga wrote: > Dan - A potentially silly question, but that's usually the kind of thing > that trips one up in the first place. 8) > > On your local workstation, what is your display variable set to? If it > is unset, ssh -X 'ing won't work. I'm aware of that. The DISPLAY variable is set on my workstation. Remember that this box works just fine using ssh and remote X applications on some systems. The key here seems to be that remote X display works fine if the remote system is a workstation that is already running X. If the remote system is configured as a server *without X* (has a text mode console) then there is no DISPLAY variable defined on the remote system in the first place and ssh doesn't seem to define one no matter what switch settings I specify. This is very likely not a Debian specific issue, BTW. My original gripe was that I had no idea how to go about figuring out what package I would need to install on the remote server so that I could successfully do remote X displays. Cole answered this by pointing out the apt-cache search xauth command that I had not previously known about. So far nobody else seems to have figured it out either. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: X with ssh, (was Novell to acquire Suse)
On Wed, 2003-11-05 at 11:07, [EMAIL PROTECTED] wrote: > On Tue, 4 Nov 2003, at 3:56pm, [EMAIL PROTECTED] wrote: > >>> ssh -XCA -l mylogin remote.system.name > > What if you turn off compression? I've seen it cause weird compatibility > problems before. Turning off compression makes no difference at all. > > It nicely told me that it was requesting X forwarding with authentication > > but still there is no DISPLAY variable defined and thus no X forwarding is > > possible. > > Could you please post the exact error message, along with anything else > that looks interesting? For example, here is what I see if I try to SSH > into a system with no X11 support installed, but requesting X11 forwarding > anyway: Here you go: debug1: ssh-userauth2 successful: method password debug1: channel 0: new [client-session] debug1: send channel open 0 debug1: Entering interactive session. debug1: ssh_session2_setup: id 0 debug1: channel request 0: pty-req debug1: Requesting X11 forwarding with authentication spoofing. debug1: channel request 0: x11-req debug1: Requesting authentication agent forwarding. debug1: channel request 0: [EMAIL PROTECTED] debug1: channel request 0: shell debug1: fd 3 setting TCP_NODELAY debug1: channel 0: open confirm rwindow 0 rmax 32768 As you'll note I don't get the xauth error message that you got. > While I'm not sure where that error message about the DISPLAY variable is > coming from, I don't think the lack of a DISPLAY variable on the server is > causing your problem. The sshd daemon is going to be running without a > DISPLAY variable anyway -- indeed, it won't even have a TTY. I do think > you're on to something about some needed package being required but not > present. I respectfully disagree. Time and time again I've noted that ssh will define a DISPLAY variable that points to the 'virtual display socket' (or whatever the right term is) that gets X applications connected through the ssh tunnel back to my workstation. An X application will not work if there is no DISPLAY environment variable defined and there is no value for display specified in the command line when you start it. An X display is required. Here's an example from an ssh connection that does the remote X display thing correctly: DISPLAY=localhost:10.0 You'll note that instead of the conventional value of '0.0' or '0' the display is offset by the amount specified in the sshd config file (typically 10). > When you SSH in, check the output of "netstat". Specifically, look for a > TCP connection listening on port 6010 or so. That will at least tell us if > the SSH server is even trying to proxy the X11 connection. I suspect not. It shows no network connections near 6000. I only see the ssh connection between the two machines. Experimentation shows that when a successful X connection is established (by actually running an application) I see a network connection whose local address is 'localhost:x11-ssh-offset'. With no live X connection this address is not listed by netstat. > Can you run a test on a failing system with the sshd daemon in debug mode > (foreground, verbose output to console, single connection only, no forking)? > While you pretty much need local access (since you're testing your remote > access method), the output can be very informative. This is a little more involved. I'll have to see if I can identify a system that I can do this with without impacting production work. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: X with ssh, (was Novell to acquire Suse)
Thanks for the useful ideas, unfortunately I was traveling around yesterday and couldn't test them. More below: On Thu, 2003-11-06 at 18:26, [EMAIL PROTECTED] wrote: > >> Can you run a test on a failing system with the sshd daemon in debug mode > >> (foreground, verbose output to console, single connection only, no > >> forking)? While you pretty much need local access (since you're testing > >> your remote access method), the output can be very informative. > > > > This is a little more involved. > > Yes, it is. Unfortunately, the more we look at this, the more I think > you're going to have to do this. > > If SSH is in production use, you can try creating another instance of sshd > on a different port, and using that for debugging. For example: > > sshd -d -e -p 26 > > That asks sshd to run in debug mode (-d), to send debug output to stderr > instead of syslog (-e), and to bind and listen on port 26 (-p 26). Specify > "-d" multiple times to increase verboseness. In debug mode, sshd does not > fork, runs in the foreground, generates debugging output, allows only one > session, and exits when the session closes. I'll give this a try on Monday to see what it reveals. At this point I'm becoming more and more convinced that the core of the problem is just missing package(s) on the system that has no X server. Hopefully the test will provide a clue as to what might be missing. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Convenient way to send email from a script
I worked this out yesterday and thought others might find it useful. I needed to be able to send email via a shell script where the script was operating on behalf of any number of users on the system. The script was running under a login different from that of the individuals using it. As a consequence all email sent using 'mail' or 'Mail' was seen as coming from the user under which the script ran, the individual users had no proof that the mail ever got sent, and the recipients were confused as to who the mail was really from. Here's the no-code solution to the dilemna, use mutt like this: Mutt allows you to define an EMAIL environment variable to specify the From: address. If you define it as just a user like [EMAIL PROTECTED] then it will format the from address (asuming the account running the script is called 'sharedapp') like this: From: sharedapp <[EMAIL PROTECTED]> Well, this is probably not what you want. But if you do this: EMAIL="Foo Bar <[EMAIL PROTECTED]>" Then the from address will be a thing of beauty like this: From: Foo Bar <[EMAIL PROTECTED]> So that established the sender stuff. Now how about the other pieces? Well we can tell mutt to send a CC to whoever we like. The subject can also be specified on the command line, and it nicely reads from STDIN or from a file. So commands like this are useful: cat mymessage | mutt -c [EMAIL PROTECTED] -s "This is a test message" \ [EMAIL PROTECTED] mutt -c [EMAIL PROTECTED] -s "This is a test message" [EMAIL PROTECTED] < mymessage Or you can use the -f switch on mutt to specify a file instead of using input redirection. Hope others find this useful too! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Recommendations for Commercial Backup packages?
I've got a client with a lame backup package that just isn't proving to be able to do disaster recovery. They have a RH Linux 9 server setup as the backup server and need to be able to backup another Linux server and a Windows server all onto a single SCSI tape drive. The solution must be capable of fully restoring all three systems to a brand new empty disk drive without having to first install the O/S. I'm investigating Arkeia and NovaNet so far, other suggestions are welcome. BRU has been ruled out because it can't handle the Windows backup. Also note that remote backups over the net are also not going to fly. The backup data has to go onto tape. Thanks for any hints, tips, war stories, etc. I'll summarize the findings and the choice along with the reasoning for the choice so that everyone else can benefit from this process. Thanks, Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: CUPS help
Kenneth E. Lussier wrote: Hi All, I have started playing around with CUPS (finally). I have one small question about (who wouold have guessed it), Windows clinets. I can point them to http://server:631/printers/foo and everything works fine. The problem starts when I enable basic authentication on the CUPS server. I can add authentication to an existing printer on the Windows side, but I can't install a new printer because access is denied. Does anyone know how to specify the username and password as part of the ipp/http URL? TIA, Kenny Well first the direct answer to your last question about username and password in the URL. Do this: http://username:[EMAIL PROTECTED]/path/file I'd assume that the ipp syntax would be the same. Secondly, you may find it easier to setup Samba to deal with the printer installation because it can be setup so that when a Windows system adds the shared printer the proper printer drivers can be downloaded to the Windows system. Pretty cool trick. Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Recommendations for Commercial Backup packages?
On Mon, 2003-12-15 at 20:56, [EMAIL PROTECTED] wrote: > In a message dated: Mon, 08 Dec 2003 13:10:44 EST > Dan Coutu said: > > >I've got a client with a lame backup package that just isn't proving > >to be able to do disaster recovery. They have a RH Linux 9 server > >setup as the backup server and need to be able to backup another > >Linux server and a Windows server all onto a single SCSI tape drive. > > This part is easy. As for 'commercial', I stay away from all of them. > For small-medium shops you can't beat the price/performance of AMANDA. > It's free, the suppport mailing list is awesome, and it works with > every SCSI tape drive/stacker I've ever tried it with. I've been > using it for close to a decade now, and I've never lost a backup yet! > (I've also never seen a commercial package as easy to use, with as > many features as AMANDA has!) Interesting to know, I'll investigate it. > > >The solution must be capable of fully restoring all three systems to a brand > >new empty disk drive without having to first install the O/S. > > Now, this is the part I'm stumped on. If you don't have an O/S on > the system, and your backups are stored on tape, the drive for which > is on a separate system, how do you propose the system to be > recovered get it's O/S restored to it's drives if it can't boot? Well, you sort of answered your own question! The real deal is that the solutions do use a boot disk, either floppy or CD, to do exactly this. Nowadays a boot CD is more common because it can load enough to have network drivers that let you connect to the backup server and download the drive image that you need to restore things from scratch. Some solutions require a special 'bare-metal' backup tape for this while others can use an ordinary full backup. > >I'm investigating Arkeia and NovaNet so far, other suggestions are > >welcome. BRU has been ruled out because it can't handle the Windows backup. > >Also note that remote backups over the net are also not > >going to fly. The backup data has to go onto tape. > > Errr, what does that mean? Above you said: > > >They have a RH Linux 9 server setup as the backup server and > >need to be able to backup another Linux server and a Windows > >server all onto a single SCSI tape drive. > > How do you propose to back up 3 systems to a single tape drive if not > over the network? Okay, I wasn't clear. I meant a solution that involves sending the backup data to another computer at some other company (i.e. a hosted backup solution.) > > >Thanks for any hints, tips, war stories, etc. > > Have you read W. Curtis Preston's O'Reilly book on backups? I highly > recommend you do if not. There's some great information in there, > and he's widely recognized as _the_ backup guru. Never heard of him, thanks for the pointer! I hope to wrap up the investigation this week and will post more once I get all my facts together. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: stty erase?
Charlie Farinella wrote: Hi, Probably a rudimentary question and maybe off topic, but it's making me nutty. This relates to the backspace key using vim in a terminal. I find no consistency from installation to installation. Some of these machines work as expected, others print ^? when I strike the backspace key, but work if I hold down the shift key. One machine (Slackware) works ok when I login as a user using xterm or rxvt, but not if I su to root. If I use aterm then it doesn't work no matter who I am. Most of these are RedHat installations various versions, two Slackware, one OpenBSD. How is this controlled, and what do I need to set it to? --charlie The mapping of your keyboard keys to characters is done using different rules. That's the source of all the variation you're seeing. You can use stty to change the character to use for different operations. For example to setup the traditional backspace as the character that erases the character to the left of the cursor do this: stty erase ^h On the other hand if your backspace key mapped to send an ASCII DEL character (delete) then do this: stty erase ^? Now you see what that ^? means. It represents the DEL character (ASCII 127). You can use stty to set things up as you choose. I'm betting though that there are even more elegant ways to do keyboard mapping and undoubtedly someone will explain how just as soon as I post this... (It always seems to work that way!) Hope this helps, Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Templated text to PS conversion
I'm searching the net for this and thought I'd also ask here in case someone knows of such a beast already. I'm trying to find something that's like an enhanced txt2ps converter. It would not only convert text to PostScript, retaining the same layout, but also allow inserting of graphic elements, such as a logo, into the output in a pre-defined place. The idea is to be able to use textual reports, such as you get from ordinary SQL, and print them so that they look like they were typeset. The requirement to be able to add in at least one graphic is key. Ring any bells? Does anyone know of something like this out there? Thanks in advance, -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Templated text to PS conversion
On Thu, 2004-01-29 at 14:59, Michael Costolo wrote: > --- Dan Coutu <[EMAIL PROTECTED]> wrote: > > I'm searching the net for this and thought I'd also ask here in case > > someone knows of such a beast already. > > > > I'm trying to find something that's like an enhanced txt2ps converter. > > It would not only convert text to PostScript, retaining the same layout, > > but also allow inserting of graphic elements, such as a logo, into the > > output in a pre-defined place. > > > > The idea is to be able to use textual reports, such as you get from > > ordinary SQL, and print them so that they look like they were typeset. > > The requirement to be able to add in at least one graphic is key. > > > > Ring any bells? Does anyone know of something like this out there? > > If you can make it so that you output some some LaTeX prior to your text output (and > after), you can add all the graphics you want and convert to PS or PDF: > Hmm, this might work. I'd have to make the filter to add in the Latex, do the conversion to PS and send it to the printer but that shouldn't be too hard. It would be harder to document how to use it so that a non-programmer could use the thing effectively, which is what I'll have to do. Thanks! -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: sound card
On Thu, 2004-01-29 at 08:15, micheal kasuba wrote: > The sound card is via ac97, its an onthe mother board sound card. > Running linux 8 as I could not get linux 9 to install that I got from > this club at the Hosstraderers Swapfest. Can one one help me get the > sound card working. Strictly speaking ac97 is not a sound card but rather a standard for sound cards. Chances are your motherboard's built in sound is using the chips from a conventional sound card such as Creative makes. The fact that RH9 would not install is also a puzzle. Can you say more about the motherboard itself? Who made it? What CPU does it support? -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: OT: Replacing CPU fan
Greg, I'll have to contradict Randy a little here in that you may in fact need a screwdriver to remove the old fan. But you do not use the screwdriver in a rotary fashion. (In other words you are not using it to operate on a screw.) A lot of CPU fans nowadays use a fan clip that has a section which stands out to the side of the clip. It is designed such that if you use a flat blade screwdriver (not a Philips head) it will slide nicely into the slot (not really but I can't think of a better name) made by that out standing clip section. You then push down about a quarter inch at which point you can then push the screwdriver handle sideways (toward the CPU fan) and the bottom of the clip will move sideways (away from the CPU in an equal and opposite reaction). (Rocket Science! Oh No!) Then let up on the pressure and the the whole thing comes apart. When putting a new fan on BE SURE to obtain either some 'phase transition material' (which is visually like double sided tape) or some heat transfer paste (I know that CompUSA sells it). Be sure to clean the old stuff off the top of the CPU. Put on the new heat transfer stuff, then put the new fan on. If you leave out the heat transfer stuff you may slowly cook your CPU and that would be bad. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux Webpage tools
Sharpe, Richard wrote: Hi Does anyone know of a Linux webpage design tool, something like Dreamweaver ? I am trying to get my wife who is a website designer to convert over to LINUX. Thanks Rich I've used Quanta and like it fairly well. Having not used Dreamweaver I have no idea how they compare. Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: How do *REAL* programmers work?
Ted Roche wrote: In a Former Development Environment That Need Not Be Named, I would use UML design tools, CASE and Entity-Relationship Diagram (ERD) tools to design this app, and GUI project management, source code control, and a rich-client GUI IDE with immediate windows, color-coded editor with macro capabilities, compiler, object browser and integrated documentation. This is where the key difference is. If you're doing FOSS then chances are you can't afford things like Rational Rose and such. So it requires doing things 'the old way'. More time goes into process and methods because you don't have those labor-saving tools to help you along. As a result of writing code for over 20 years I find that the GUI type tools with graphic representations (such as UML) don't always make a lot of sense. This is due to lots of years where textual descriptions of the requirements, specifications, and design are all we had to work with. Those methods happen to work wonderfully in the FOSS world because they're cheap and mostly dependent on my brain rather than some external thing that costs beaucoup bucks.. Emacs is still one of my most heavily used tools. I've taken to liking Cervisia more than the command line CVS interface because it lets me do more from one place. For Java development I find that Oracle's JDeveloper is pretty nice. I hear that Eclipse might be better so it is on my list of things to check out. For things like database work I still rely on a command line interface that processes plain old SQL. While ER diagrams are really useful for reverse engineering and explaining a design to someone else I find that they always leave out important design details about the database that just can't be captured in an ER diagram. Well that should provide at least a glimpse into things in my world for you. Hope it helps. Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Not a fibonacci series
Greg Rundlett wrote: If I have four options: a) b) c) d) and none are mutually exclusive, then there are 15 possible combinations. Assigning a value to each: a = 1, b=2, c=4, d=8, the unique combination can be assigned a code by summing the values No. Combination Code 1 a 1 2 b 2 3 c 4 4 d 8 5 a, b 3 6 a, c 5 7 a, d 9 8 b, c 6 9 b, d 10 10 c, d 12 11 a, b, c 7 12 a, b, d 11 13 a, c, d 13 14 b, c, d 14 15 a, b, c, d 15 What is this type of problem referred to? I know it's not a fibonacci series. Anyone know of good examples of processing this info programmatically? You may ask: What the hell is a fibonacci series? http://www.textism.com/bucket/fib.html Thanks, Greg What you're talking about here is a combination problem. Generally permutations and combinations are studied at the same time in Math courses. If I recall correctly the key difference is that permutations see the position in a sequence as consequential. So a,c,d is not the same as c,d,a. In a combination, on the other hand, the order does not matter. So your example above is clearly a combination. A pointer to more detail on the math behind the scenes is: http://engineering.uow.edu.au/Courses/Stats/File2417.html Hope this helps! Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: help
Ken wrote: Hi to all, I hope this has not been a recently discussed issue I am running RH6.2 with Postfix, Apache and Mailman. I currently have over 500 email accounts. I have built up new box (more horsepower, ram, drive space, etc) and do not wish to enter these users manually. Please point me to a solution if one exists. Thanks in advance, Ken One of the things that I really like about UNIX based systems is that once data is in the system somehow you pretty much never need to re-enter it. Just transform it or move it. You should be able to copy most of your existing setup's key files from one system to another and get the new one up and running just like the old. If you already have Postfix, Apache, and Mailman installed on the new system try copying the config files from the old system to the new. Likewise you can use tar to migrate entire directories such as user's home directories and the mail spool directories. You probably want to copy /var/spool/mail (incoming mail) and /var/spool/mqueue (outgoing mail) also. Just be sure to turn off sendmail on both systems before doing the transfer and enable sendmail only on the new system once you're done. There are probably other details to tend to but hopefully this will cover things that other people miss. Dan ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Shiny webserver control panels
On Fri, 2004-03-05 at 06:47, Brian wrote: > Looking for any reviews/recommendations/etc on end-user geared > website/mailbox control panel packages for linux servers running many > sites in a shared environment. > > I've looked a bit (a very short bit) at Plesk, which on the surface > seems okay. I have used Plesk quite a lot in conjunction with a client of mine that is entirely non-technical. It does a really good job of letting you do most of what you'd like to regarding management of your web site and mailboxes. The one thing I've griped about is not being able to see things like the PHP error log or system messages log but it makes sense that since those are shared resources they would not be visible. My client was able to pretty quickly get a grip on how to use Plesk as well so it passed the naive user test. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
RE: F/OSS Database experiences & recommendations
> Ted Roche writes: > > > I would welcome experiences and opinions (like I have to ask!) on the > > various database backends available. I've used both MySQL and PostgreSQL extensively. My take on them is that if you need something lightweight and fast to implement that non-DBA types can handle then use MySQL. If you need a full-fledged production database use PostgreSQL. Some of the really nice features of PostgreSQL are: - You can create your own functions use either the pgsql 'language' or are written in some other language such as C, C++, and so on. I'm using this capability to be able to have a function within the DB that generates GUID values. The function is actually written in C. (Note: The much ballyhooed MS SQL Server does *not* let you create functions of any sort. Bah!) - You can create stored procedures (MySQL claims this is coming). - You can define triggers (MySQL can't do this). - The time handling is exceptional. One place where I'm using the DB requires being able to display dates and times to users anywhere in the world using their own timezone. The time handling in the DB really helps a lot in making this job easier. (Note: not all timezones are a full hour apart from each other!) I know of no other database that can deal with time as well. Note that you can get GUI tools for administering it. The only one I've fooled with much runs on Windows though. It is the EMS PostgreSQL Manager. I believe there are a number of other tools out there that provide a GUI for you. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: F/OSS Database experiences & recommendations
On Wed, 2004-03-17 at 16:46, Travis Roy wrote: > > - You can define triggers (MySQL can't do this). > > From: http://www.mysql.com/products/mysql/ > > "Stored procedures and triggers > > Stored procedures allow you to create functions and subroutines > that run on the server. This makes it possible to grant access to > specific queries without granting carte blanche access to the underlying > data, or validate data in the database before it is stored. Triggers can > be configured to fire when certain conditions are fulfilled. > > The MySQL database server will provide hooks for implementing > stored procedures in multiple languages, as well as including support > for the Persistent Stored Modules syntax defined as part of ANSI SQL-99. > > Support for stored procedures was added in version 5.0, and support > for triggers will be added in version 5.1." > > Plus they have some documentation for it already. > > http://www.mysql.com/documentation/maxdb/a7/41ee0b605911d3a98800a0c9449261/content.htm > > A friend of mine that uses MySQL says that it's in the dev version already. Interesting! I had not heard that they were planning to support triggers too. Thanks for pointing that out. -- Dan Coutu Managing Director Snowy Owl Internet Consulting, LLC http://www.snowy-owl.com/ ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
