[expert] Weird LPR problem.
Installed Mandrake 7.1 a few days ago, and when I went to
print out a business letter, printing failed, but in the
oddest manner.
The file in question is in my /home/rfg/text directory. It
is called (example), testfile1.txt. I do the following:
prompt $ cat testfile1.txt | lpr
It prints just fine.
A similar file, testfile2.txt will fail to print. I get
the following message printed at the extreme top of the paper
that comes out of the Epson 800N printer:
No way to print this type of input file: fsave (linux)
virus (2570-11-10)
The very last characters, 1-10) actually show up on a second
line.
The only difference between the two files is that testfile2.txt
has a series of carriage returns to start with. I did that so
my text will start below the printed letterhead on my business
letter.
The problem is reproducible; ie, any file, or filename, with
6, 7 or more at the top of the file will fail to print in
that manner. I also tried to start the file with a single
unobtrusive period at the top, followed by 13 's, and
it still failed to print.
This seems to be a bug in LPR, or the input filter for my
printer. Here is a copy of my /etc/printcap file:
##PRINTTOOL3## REMOTE uniprint NAxNA letter {} U_EpsonStylusColor
stcany {}
lp:\
:sd=/var/spool/lpd/lp:\
:mx#0:\
:sh:\
:rm=epson.nook.net:\
:rp=:\
:if=/var/spool/lpd/lp/filter:
**orange[/etc]#
Of course, the cat command (without piping to the printer) shows
up correctly on the screen. I am also able to print pages, email,
etc from Netscape and other applications. Only the piping to lpr
of a text file fails. I never had this problem with RedHat 5.2 or
Mandrake 5.x or 6.x. I have not seen any updates for lpr.
I also took a look at the files with a hex editor, and they seem
perfectly normal. Not for one minute do I believe this is a
linux virus; it seems to me that someone coded this into lpr or
the input filter as a joke. Nevertheless.
--
Ramon Gandia = Sysadmin == Nook Net
http://www.nook.net[EMAIL PROTECTED]
285 West First Avenue tel. 907-443-7575
P.O. Box 970 fax. 907-443-2487
Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Penguin removal - OT!
Joseph S. Gardner wrote: > I used to have an old Teletype with a paper tape reader in the basement > (probably still have some old paper on a roll laying around for it). My father > kept tryin' to teach me basic over an blazin' sub-300 baud modem if memory > serves me (it's been a while). My first microcomputer was an Altair, with an ASR-33 and KSU-33 (teletype with paper punch and paper reader). It ran at 110 bps over a 20 ma serial interface. Later on I upgraded it to an Ann Arbor video terminal (upper case only) at 9600 bps. Ran several accounting programs, some in basic, some in basic with Assembly language modules and some totally in assembler. This was 8080 code which was a pleasure to code in (still is!). In fact, these Altairs all still run here. If you think about it for a while...it may be fun to poke ridicule at those old machines, but. The thing about a computer is that it does not matter if it runs at 2 MHz or 750 mHz. All we want is for the computer to run faster than the operator doing the work at it. In those days, we posted data and payments; the computer printed reports and statements. If I sat in front of it, and I could key in the payments as fast as I could type on the keyboard then the computer gave me 100% satisfaction. Today, we have to take the hands off the keyboard and handle a mouse for the ocassional click. Thus, data entry is not quite as fast as before. Satisfaction is only 90% in this case, even if the computer runs at 333 mHz in my case, vs. 2 mHz for the Altair. If you disagree with me on the mouse click thing, then satisfaction for both systems would be 100%. This means that old computers are not bad at all. True, they did not surf the web or play music...but I am not sure that is an advancement. I have made a lot of money from surfing the web (am an ISP), but I can compare that to making money by selling something addictive. Hehehehehe. We no longer live in the older, gentler days. Things are dizzier today, and if you like computers for their own sake, then I cannot really say that a Pentium II at 333 mHz is more or less satisfying than the old Altair with the hand-soldered circuit boards was. And both seem to do my bookeeping with equal aplomb. Ohthe old Altair was 100% reliable, never crashed and never lost any data either on the paper tape or the 8" floppies that followed. The hard drives in those days, late 70's, were 14" monsters the size of a washing machine and used 1 horsepower motor, about 1 kilowatt of electricity. Large and noisy, but they were reliable too. I never put any accounting on them, just played with the hardware. No longer have those. I bet those 5 MB drives never lost anyone's data either. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Penguin removal - OT!
"John D. Kim" wrote: > > Wow, I've only heard about keypunch. My high school cs teacher used to > tell us about it. I learned COBOL when I was in high school, but we had a > nice VMS machine to code it on. And I thought COBOL just couldn't be any > worse, but I guess I was wrong. > > On Tue, 14 Mar 2000, ibi wrote: > > > And, here I thought I was the only kid on the block old enough to know > > what keypuch is! Shilly me. LOL .. Pj Well, I coded Fortran II and Cobol on an IBM 1620 and 7040 back in 1964 using keypunches. It actually worked quite well. Fortran II is very close to the first Microsoft Basics (as sold by MITS for their Altair). The 1620 was slow, with rotating memory drum; but the 7040 had hand wound core memory and it was FAST. I am not sure that it would not favorably compete with modern micros in term of speed. But you wouldn't want to pay the electric and air conditioning bill! That was in the days before integrated circuits. Each memory bit took two to four transistors and a toroid core with three windings as I recall. Does this make me the old Geezer of the Mandrake List? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] List Bitching
John Aldrich wrote: > > > I think a good compromise would be a WEEKLY posting of the location > of the FAQ. Especially for the first couple months. Perhaps, after a > generous period we could slow it down from a weekly to a semi-monthly > to a monthly posting. > John John think about this for a minute. Just because WE get used to the posts, and change it to semi-monthly, does not mean that newbies don't pop up all the time. I actually think that DAILY, plus sig files are the way to go. Most newcomers will sign up, save the welcome message without reading it (no need to read it until its time to unsubscribe), lurk for a day or even a few hours, then ask the QUESTION. If the alert to the fax is over 24 hours away, that person is just not going to see it. We will then have another WinModem question on the list, etc. The suggestions I hear are: 1. Daily postings of LINKS to the FAQ. 2. Weekly or less frequent LINKS to the FAQ. 3. The actual FAQ itself. 4. Signature files appended to all posts with the LINK to FAQ. The idea being to save bandwidth, improve signal-to-noise ratio for those that remain here, and reduce the aggravation factor to newcomers (newbies as well as users of other distros that have migrated to Mandrake). I am in favor of 1+4, or just 1 if Mandrake cannot add 4. This issue was discussed endlessly in the RedHat list, and the end result was that RedHat Inc would NOT do anything, therefore the idea died on the vine. YOu cannot actually do anything about this issue unless Mandrake gives at least its tacit consent. To do it without their OK would be a good way to Tick the Frogs and get kicked off the list. :-). -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
"Albert E. Whale" wrote: > > John Aldrich > > (www.808hi/56k/trouble2.htm is the site I use now-a-days > > for most 56k modem trouble-shooting.) > > > > > John, > > THanks for the lead, but I cannot seem to quite get the above site. > What is the Web Address? He mistyped it. http://www.808hi.com/56k/trouble2.htm It is a good site. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] List Bitching
Jean-Michel Dault wrote: > > Maybe we can add a signature to the mailing list messages so that everyone > sees mandrakeuser.org on each e-mail? Redhat does/did that. It was at the very end, below the sender's regular signature with the obligatory ASCII-toons, etc. Not too many people paid attention to it. If you frequent Usenet, or othr lists, they have periodic postings with Subjects like Mandrake-FAQ. These are heads-up to readers, specially newcomers. As was mentioned, I think that a daily FAQ will easily save one or more messages a day. If the Subject is carefully chosen, it will be easy to filter out (Mandrake and FAQ are both bad words to filter on. But maybe Mandrake-O-FAQ might work. Procmail users can comment on this. But we need to hear from Mandrake, Inc. on this. It is their list, after all. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] List Bitching
Rich Clark wrote: > > Maybe what the list needs is a FAQ page posted somewhere. Someone > subscribes to the list, they get the nice new welcome note from the list > manager that gives them pointers about how to post, what to post and where > to look for answers *before* they post the same question that's been asked > and answered countless times before. Everyone else has a FAQ, why don't > we? We have a FAQ - sort of - at http://www.mandrakeuser.org What we need is someone that can set up a Cron job to send an email on a daily basis to THIS (and the Newbie) list advertising the FAQ. This same suggestion was made a year and a half ago on the RedHat list, and there was a tremendous outcry against it. The reasoning went that there was too much email messages as it was without cluttering it more with a daily "here is the FAQ" message. Some of us thought that one daily message or maybe three times a week, would reduce the overall number of messages posted asking questions. At any rate, over at RedHat, DJB and the powers that be nixed the idea. Perhaps it would be better received at Mandrake. The cron generated email message is trivial to implement. Whatever is done is going to require the approval of Mandrake as this is THEIR list. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] List Bitching
Ivan Trail wrote: > > On Wed, 08 Mar 2000, you wrote: > > > > > > And what is this long-running modem brand talk? May we please discuss >Linux-related issues? > > Maybe people are uncomfortable to speak of what they really need for fear of being >flamed as > > newbies, a practice of =assholes= who stink pretty bad after a while.(yeah, I >said that) > > I agree. I was on the newbie list for about a month and got tired of having > 80-100 emails to sort through on the subject "installing linux". The list has > it's place, and was a great source of help but after I got it installed I > couldn't get a decent response on a tougher question. Here I have answered > most of my problems by reading what other people have asked. > > I am not trying to start another flame, but here is my suggestion to remedy > this problem: > > Impliment several, more specific, mailing lists. ie. one for installation > help, one for driver development, one for people who actually like too much > caffiene ;^) You get the point. I used to ber on atleast four different > mailing lists from Red Hat at any one time. the redhat place has probably > fifteen to twenty different lists to choose from. So maybe some of you experts > in a specific area of linux could volunteer to host a mailing list. I would do > so but I can't even find actors to help with the scripting language thing, thus > I am far from an expert. > > just my two cents worth, with NO REFUNDS! And my 3 cents will make it a Nickel: These mailing lists have two sorts of people on it. Group 1 are the newcomers that have legitimate questions to ask. Not exacty Newbies, but 'newcomer.' Group 2 are the rest of the grouchos that answer them or seek the Truth. You, sir, probably belong to neither group. You are looking at a social place, and want to see ever expanding level of expertise on this list. This is not the intent. The mailing lists have a place, and it is to help the Mandrake user. By their very nature, the people asking are those that do not know the answers. There is an influx of new people all the time. It is wrong to get angry or be dissapointed just because a Newcomer is having problem with Modems, his WinModem don't work, or his PPP link won't come up. We see these questions asked ALL THE TIME, and they are going to be asked all the time each and every week from now until Hell Freezes Over. That is the way it is. Once the newcomer gets his questions answered, he will move on or become one of the grouchos answering questions. To make it plain, I have received VERY LITTLE help from this list, although I do get a surprise once in a while. I am here to answer some questions. I am an ISP, and what I know is often of help to some newcomers. Others on this list know Perl, or PHP, or Sendmail, oretc. That is what is all about. If you get tired of the Questions and their answers; if the list is repetitive to you, or sounds trivial, then I would say that you have outgrown it. Nothing to get angry about. Hang around if you want, or move on. Maybe to return later. Linux is a long, tall ladder. We all climb this ladder of learning. As we slowly climb, we see those above and ask for help, and at the same time we reach down and help pull up those below us. While we move up this ladder, we have to recognize that new people are getting on the lower rungs all the time. I, and a few others that pitched in, answered and discussed some questions regarding Modemsand Supermount. Frankly, I had posted about this four or five times already. And it will happen again because next week, sure as all get out, some newcomer will want to know how come his "SupraExpress 56i PCI Modem will not be recognized or work with Linux". So, it starts over. Thats the way it was, is, and will be. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
John Aldrich wrote: > > > I have to wonder if this isn't a symptom of the default to > K56Flex by the Rockwell chipset and the default to X2 by > the TI chipset. I seem to recall from past emails that you > (Ramon) use USR products there at your ISP. I know this can > be fixed by turning OFF K56 flex on these modems, just like > people have to do here for USR / 3Com modems. I know that > our customers with 3Com/USR modems who don't turn off the > X2 protocol have a DEVIL of a time connecting and staying > connected. I suspect that the same is true of modems which > default to K56. While these are TECHNICALLY both compatible > modems, due to the fact that they both default to trying > their mutually-exclusive PROPRIETARY protocols FIRST, > there's a conflict and therefore you tend to get lousy > connections, IF you connect at all. No, they both connecting with V.90. My terminal servers are Livingston Portmaster PM3. They use Lucent chipsets. Their protocols are K56Flex/V.90. The server will tell the client first to try V.90. That is the default and is shown in the logs. I tried Cisco 5200 and Ascend 4048, these both have Rockwell chipsets. Neither would negotiate a connection using either V90 or K56Flex. Max speed with them was 31,200. On the client side, the X2/V.90 modems using the TI chipset, such as the US Robotics or Phoebe connect fine to the PM3 at V90 protocol with speeds ranging as high as 53,000, although it is rare to see over 52,000. The PM3 will not do X2. I have never really seen any K56Flex connections. The code for this is rather old now, and the K56Flex and the Robbed Bit signalling/D4/AMI/CT1 used by the Telephone Company in their DMS-10 Nortel switch is not conducive to their working. There are worse switches, I understand from some ISP's that what their phone company uses has never allowed any of the 56K protocols. This all has to do with how the phone company changes the signal from Analog to Digital in their Codecs. My observation has been that the USR/TI chipset client modems have better throughput, and much less susceptibility to noise and adverse line conditions; fewer disconnects. The Rockwells are aggressive in their connection speed and frequently hook up at too high on their speed. In that case, the connection usually fails -- either a disconnect, or worse yet, just pokey throughput regardless of reported connect speed. The best of the WinModems, for what it is worth, is the Lucent. They are now better than the US Robotics WinModem. The Lucent needs a driver of Version 5.66 or better. With the new driver it is a good performer. Keep in mind that the DSP in a WinModem consumes a tremendous amount of CPU power. It requires a Pentium 166 in order to even work. With a Pentium 166, if you need to move the mouse pointer 1/8th of an inch, or scroll a window, or do some other computational operation, the Modem generally has to be told not to transfer data. For this reason older pentiums do not do well at all with a WinModem. On top of that, most older Pentiums have old WinModem drivers, so the problem is compounded. As much as we bitch about Win Modems, the truth is that since the Lucent 5.66 the situation has gotten better. The latest Rockwell drivers are getting passable now. In most cases I can connect a new 450 mHz Pentium to Nook Net via a 56k Rockwell and have it work. The problems is with out of town phone lines. The Rockwells simply fail over longer wires. My laptop with a Phoebe External just connects fine. Most of those Rockewells NEED extra settings to tame the connect speed, or restrict it to V.34 protocol (or have it dial my analog modem bank which I still operate for those with RECALCITRANT modems). Such is the life of an ISP. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
Civileme wrote: > > Of course, with any > modem purchase you are likely to receive by the munificence of > AOL a brilliantly colored coaster suitable for absolutely > anything you can imagine to do with it. Older modems carry the > added benefit of a high-quality 3.5" floppy suitable for holding > data after erasure. Back in the early 90's, those of us on the CP/M groups tested a whole slew of those AOL floppies. What we found out was disturbing. It SEEMS as if they are written to with a high write current. These disks have tracks that either do not format properly (the earlier data 'prints thru') or the track is wider than normal and creates errors when rewritten. The end result was that AOL floppies are fine for what they do: port AOL programs. They should not be formatted, erased or used otherwise. Compuserve floppies were better, but not by much. Lots of people lost lots of data with those! If you think about it for a bit, a floppy that is being mass marketed and probably going to be discarded, is not likely to be of very good quality. I bet the magnetic film was kinda spotty or poor and they made it up by stepping up the webbers on the write heads. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
"James E. Tarvid" wrote: > > The AOpen FM56-ITU/2 works in both Linux, Windows, DOS ... > > There are a few more but this is the best I've found. > > Jim Tarvid > > Aopen Acer 56K V.90 Internal With Voice ISA fax modem (Not aWinmodem) With > VoiceMail #AOPFM56-ITU/2 $ 41 Unknown?? 2/27/00 12:25:03 PM CST Comp-U-Plus > 800-287-8786 > 914-352-8100 > Online Ordering NY FM56 > > Aopen FM56-ITU/2 ,V.90,56K, , ISA, Linux Compatible Internal modem/FM56ITU $ > 47 5.00/FLAT anywhere in cont. USA - no other fee 12/28/99 7:24:00 PM CST > Openlinx Communications > 562-623-9334 > . > Online Ordering CA - Yes, they are ISA and non-Win, but they are Rockwells. Look in http://www.o2.net/~gromitkc/2307a.html for the lowdown. AOPEN makes a Texas Instrument model, which is probably good, its is the FM56PVS-T (the -T is the critical part and makes it a TI chipset modem, hopefully with the V90/X2 protocol. If so, it should be good). Non-Win Rockwells work, but they are not good performers. They have aggressive speeds, usually a notch or two higher than they should for given line conditions. They get LOTS of retrains which I can see on my logs here. They frequently disconnect with "lost carrier" in my logs. Retrains are bad, because it means the modem is spending most of its time resending data that had errors. The Connect speed may be 52,000 vs 46,666 for a USR/TI, but the lower speed modem will pass MORE data bytes in a given time. There are also lots of other bugs that plague Rockwells. On the server end, Rockwell digital modems at the ISP end do not work well with Channelized T1 service using D4/AMI signalling,and that is about 50% of the phone switches in the USA. The US Robotics make great consumer modems but are just terrible for an ISP, although the later revisions seem to be working better. At one time there was a class action suit about it, if I recall. The Lucent digital modems, like I use here at Nook Net seem to be the most robust of the lot. I tried equipment from Ascend and Cisco here and simply could not make it work above 31,200. Cisco tried valiantly with their AS5200no dice. My Lucent Portmasters PM3 do fine. On the Analog end, I use a number of US Robotics couriers, and they do real well too. In one village I was kinda cheap and used US Robotics sporsters 33.6 data fax modems, and they work well too. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
ibi wrote: > > Ah yes, but there is a little caveat. The TI chipset is the key. Ramon > can explain it beautifully and has in prior threads. The modems with the > TI chipset is the only bulletproof unit that is wholly Linux compatible. > I believe Phoebe makes two versions. USR external is the only model with > TI. The internals use a Rockwell chipset that may or may not be winmodem > versions. No, No. Phoebe makes a TI chipset model in both External AND internal versions. I gave the model numbers earlier today. Be careful, Phoebe makes LOTS of junk modems too, you have to pick thru the part numbers to get the right ones, as I did and shared with this list. Because they are the priciest Phoebes, the dealers do not seem to want to stock them. They prefer to sell the $ 9.95 Rockewell WinModems. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: Fwd: [expert] Modem Problem
"Stephen P. Molnar, Ph.D." wrote: > > I fear that I didn't include sufficient information with my original > message for which I apologize. > > I should have elaborated on my statement that it used to work, it did work > on another platform running Linux from which I moved the modem. > > Further isapnpconf found the card and generated the following conf file: > > # $Id: pnpdump.c,v 1.18 1999/02/14 22:47:18 fox Exp $ > # Trying port address 0203 > # Board 1 has serial identifier b6 ff ff ff ff 80 71 93 04 isapnpconf is a program to detect and set up ISA cards, not PCI, and looking over the I/O ports listed in the message, they are all ISA range ports, not PCI. Your modem seems to have been ISA. Find out the model number and we will know for sure. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
Hoyt wrote: > OK, you gave us the "bad" news, now recommend a good > internal modem (based on your experience) that is > affordably priced (http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
things. Better junk. As an ISP, I have seen and worked with thousands of modems. For all practical purposes "I have seen them all" (I do get a surprise now and then, usually not pleasant). Often I get folks that have spent a lot of money on a computer and insist that their modem is of the "best quality". Or had very good luck with a Rockwell and insist that they are "the best" type around. Most, if not all of these fold have been exposed to just one or two modems in their life, and their knowledge is flawed. If you go out and buy a Modem, stick it in a Windows 95 computer, fire it up, and insert a disk when the thing says "new hardware detected". And then use it successfuly to dial up your ISP, I do not think that I would call that as 'experienced with this or that modem'. It is merely anecdotal experience of the type that the manufacturer hopes you have. To know a modem you have to experience horror stories with that type, or conversely, experience nothing but good from this other type. It is almost like cars. You could have bought an Edsel and had nothing but perfect luck with it. 150,000 miles, never even a flat, used no oil and never squeaked. To you, Edsels are great cars. But the vast majority of people had a different experience. Modems are like that. I can see it: About ten people will come into the Mandrake list and start telling us how good their Rockwell modem is. All I can say is "Goody for them!". PS: Most, if not all, current production Diamond Multimedia modems use the Rockwell/Connexant chipset. You pays your money and makes your choice. Cirrus and PCTel chipsets are even worse, but luckily are rare. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem Problem
"Stephen P. Molnar, Ph.D." wrote: > > I've just reinstalled Mandrake Helios on my pentium platform. I have a > Zoom 56K PCI Faxmodem that used to work. The boot process sets up the ISA > PnP devices, but when I look for the modem with kppp I get a message that > the modem is busy. It used to work? In Windows 95/98 maybe. The Zoom PCI modem, as with 99.99% of all PCI modems, is a WinModem. In other words, the parts are stripped out of it, and the functions are in a Win95 "driver program" that is proprietary and does not work in Linux. You need a new modem. It will also improve the performance under Windows. Trust me, I am an ISP. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] (OT) Missed you folks
Civileme wrote: > > I was cut off from internet and long-distance telephone access for the past 21 > hours. GCI switched satellites and didn't tell anyone in advance and it took > quite a while to restore switched routes for internet and telephone. > > Kinda spooky. The resrt of you could have ceased to exist and I would not have > known it. Right. And I am talking to you from the 5th Dimension....! :-) -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Problem Mounting Floppy Disk
uot;bashrc". prompt# umount /mnt/floppy prompt# ls -la /mnt/floppy (now shows "bashrc" oNLY) Ain't that the cat's meow? Now, if you have /etc/fstab set up correctly, you can use the abbreviated command: prompt# mount /mnt/cdrom instead of the longer command prompt# mount -t ext2 /dev/fd0 /mnt/floppy but please note that it depends on what /etc/fstab has in it regarding the floppy device. On thing that you have to be VERY CAREFUL of when you edit /etc/fstab. MAKE CERTAIN that none of the lines wrap. Each line MUST begin in /dev/whatever and end in 0 0 or the two digits for fsck. If they wrap, so you have something like this ANYWHERE: /dev/fd0 /mnt/floppyext2 noauto 0 0 as two lines, your system will NOT BE BOOTABLE. If you are using the pico editor on /etc/fstab, be absolutely certain that you use the -w option and that you check for line wraps with more, less or cat. This is absolutely CRITICAL: prompt# cd /etc prompt# pico -w fstab Another file you don't want to screw up with line wraps is the /etc/inittab file. There are actually many such files in the system, but those two will prevent booting, so don't say that Ramon encouraged experimenting and did not warn you. You can usually get back in with "linux single" at the LILO prompt or with a rescue disk. So I suggest you create a backup of these files first. Then if you screw up, go in as rescue mode and recopy the files. Example: prompt# cd /etc prompt# cp fstab fstab.bak (creates a copy of it as fstab.bak) prompt# pico -w fstab(lets assume you screw fstab up) Reboot. Does not boot. Go into rescue mode. bash# cd /etc bash# pwd (make sure you are in etc) bash# mv fstab.bak fstab or, a lengthier alternative: bash# mv fstab fstab.bad bash# mv fstab.bak fstab Reboot. In the later case, you have fstab.bad that you can peruse to discover the error of your ways. Play around. Worse comes to worse, you will get some training in OS rescue, something totally lacking in Windows 95/98. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] linux printers
"Ronald J. Yacketta" wrote: > I am currently looking to purchase a printer that will sit on my > winblows box (for my son) and have the ability to print to it from linux > (using samba). Perfectly workable, but consider an alternative. You could use a "print server". This is a little box that has ethernet input, and has parallel printer port output. Most support Windows and UNIX LPR printing. It has some advantages over your setup. The main one being that if your Son's computer is off, or crashed, or running a DOS game, etc you cannot print from your Linux computer. These little print servers run in the $80 and up range, and some are not much bigger than an oversized parallel connector. They come with software for Windows. LPR printing is native to Unix so you need nothing there. Just to suggest that you look into it. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] Netpliance
Totally off topic. But has anyone played with the Netpliance units yet? http://www.netpliance.com It runs QNX on 32 MB ram, 16 MB flash, V90 modem, keyboard, 10" color 800x600 LCD display and a $99 price tag. I ordered one today to see what makes it tick. QNX is not Unix or Linux, but it is structured like it. It runs on 8088/8086 type chips, like DOS. But the file system is very familiar to us Linuxers. I fully intend to dissect mine and see how that thing is put together, in particular how the Flash chip is programmed and how does the thing boot up etc. We ought to send one of these boxes to that Johnson kid or whoever he was that broke the DVD Coding. This Netpliance is worth hacking into. Imagine loading a slimmed down Mandrake into it. Hehehe I suspect that the Netpliance is one of those things like Juno or FreeServ that gives you cheap computing and access but plagues you with Commercials and Ads. This is a challenge for all of us to see if we can reprogram the damn thing -- or build a phone line IP filter for it or something. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Auto Log On ... is it possible?
When the computer starts, it runs rc.local When you want something to happen at a set time, use cron. When you want something to happen when someone to log on, use that user's .bash_profile file, or if you want it to happen when ANYONE logs on, use /etc/profile. Assuming you are using the bash shell, of course. Just exec whatever command you want right in the ~/.bash_profile and presto! -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] resetting hosts.allow and deny files
> iTOOL wrote: > > > > If I make changes to the hosts.allow and hosts.deny files, do I have to > > rebbot for these changes to take effect or is there a service (such as > > INETD) that I need to kill and re-start? > Myles Vredenburg answered: > > No need to restart any daemons or reboot for the changes to take effect. > You do not have to restart inetd, but you have to force it to reread the hosts.allow and hosts.deny files. Send it a hangup signal. prompt# killall -HUP inetd will do it. Of course, restarting the inetd daemon or rebooting the computer will also do it, but it's overkill. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Floppy install???
I believe that Slackware is distributed in "Disk Sets". The newer one comes on CD, but is or should still be broken down into individual disk sets comprised of a stack of 1.44's. Example, the base system including the kernel, login etc. is disk set A, and used to be about 4 floppies. You load those in, then disk set C, E, N whatever you want. Just keep feeding it floppies. :-) Slackware, like all other distros, can be obtained on $1.99 CD from Cheapbytes and their ilk. I'd spend the bucks and investigate this. I used to use Slackware back in 1994. I do not think anything has changed in this regards. It actually is a good system that has stood the test of time. Be aware Slackware is BSD-flavored Linux and not Sys-V flavored Linux like RedHat/Mandrake/SuSE. If you do not know what I am talking about, you will soon find out... hehehehehehe. :-) The home of Slackware is http://www.cdrom.com It is also the home of FreeBSD. Nice site. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Network Question:
> > I have 2 computers . > > > > 1: A mandrake 7.0 new install, 256 meg ram Ethernet Card ( System at > > Work) (T-1 Line) > > 2: A windows 2000 box at home which also has 256 meg ram 600 Mhz > > (Modem Line For Kids) > > > > Is it possible that system 1 be connected to system 2 . So system 2 > > can benifuit from the work T1 Line. You can do a mini-ISP thing. Get a Web Ramp. The model with external modems is fine, like the 300e or 310e. use a single modem. Your home computer then dials into the webramp via its regular modem/phone line. The webramp answers and puts you on the ethernet at the office. How you go from there is up to you. http://www.rampnet.com Its a small box, size of a paperback or cigar box. It has 3 serial ports for up to three modems. Other end has a small ethernet hub. It is a complete dial in/dial out and IP Masquerade solution in a small box with no moving parts. About $300, far cheaper than a computer. From the ethernet side it is managed via a built-in web server interface. Out of the box it is 192.168.1.1, but that can be changed easily. There are several models; be sure you get one that allows dial-in. You can do the same thing with an older 486 or Pentium box running IPMasquerade, Coyote Linux, Linux Router Project or FreeSCO (Linux based) or IPRoute (MSDOS based). Of those, it seems Coyote has its act put together the best. I am playing with one here right now. Coyote does ether-ether and does not do PPP, but the others all do ether-ether, ether-PPP or PPP-PPP. In my experience, using a dedicated box like the WebRamp is a better solution if it fits your requirements. Cisco, Bay, Ascend and others make similar equipment. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] resetting hosts.allow and deny files
iTOOL wrote: > > If I make changes to the hosts.allow and hosts.deny files, do I have to > rebbot for these changes to take effect or is there a service (such as > INETD) that I need to kill and re-start? Yes, you have to restart or hangup inetd. Like this: prompt # killall -HUP inetd That makes inetd reread its configuration files, including hosts.allow and hosts.deny. One thing you need to watch for, is that not all daemons use inetd. Look in /etc/inetd.conf and you will see what runs what. If you see something like /etc/bin/inetd in.telnetd then you know that telnet is running under inet. The way this works is that when the request comes in for a telnet login, what is started is NOT telnet, but inetd. Inetd then checks its configuration files to see if the connection is allowed for that port (hosts.deny hosts.allow), and if so, then it spawns in.telnetd and hands things over to it. Another gotcha is "in.telnetd". This is probably a link or a startup file for the telnet program. In some cases, due to modifications to the Linux system, the link is dead and some other program actually listens on the port and responds and inetd is ignored. It is best to check and double check these things. Here on my servers I use qmail and ncftpd, neither of which runs under inetd although the ports are listed there as well (commented out in my case). Play with things and verify them. What works well to do this is to get a friend to give you a telnet account somewhere and you can probe, poke and telnet into YOUR computer from HIS. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] ftp
Dan Swartzendruber wrote: > > At 07:24 PM 01/28/2000 +, lee binkley wrote: > >I am having problems uploading a file to my ftp site. > > And we should do what? Well, for instance, you ought to check to see if the computer is plugged in .. :-) -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] problems with masquerading and NT services
Dan Swartzendruber wrote: > > At 10:43 PM 2/27/00 +, James Lewis wrote: > > >I have 3 NT workstations on an internal LAN, going through ipmasq (on > >2.2.13) which is providing NAT. General services such as web browsing, FTP > >etc, are working fine but we are experiencing very annoying problems with > >Microsoft Outlook 2000 (running in 'Corporate Workgroup mode) running on > >the 'Internal' machines connecting to an External Exchange server with a > >public IP. > > > >The problems seem to be because the exchange server is seeing requests > >from 3 different machines coming from the single IP address of the NAT > >machine > > This is peculiar. I have a client who is doing the same thing (well, the > NAT is being done by a cisco router between internal and external LANs), > and they have no problems getting to the exchange server. Can you be more > specific about what the problems are? NAT includes IP-Masquerade, but IP-Masquerade is not all of NAT. In other words, IPMasq is a subset of NAT. Cisco routers can do a Many-One IPMasquerade type of translation. They can also do Many-Many translation. This provides firewall services, but each internal machine maps to a separate real IP address. It does not help with conservation of IP's, but would certainly help with the problem described above. Provided, of course, you can obtain 3 IP addresses from your supplier instead of the present one. I am not certain, but full fledged NAT may be available for Linux. Does anyone know for certain? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Freakin' Netscape...
Alex V Flinsch wrote: > > Personally I find that just installing navigator (the browser only) is quite a > bit more stable than installing communicator (browser/email/newsreader/html > editor). Besides there are other options for the rest of the software that > makes up communicator. True enough. But it is aggravating to have some other email program that does not integrate with a browser. You click on a link in the email, and usually nothing happens. The URL has to be manually entered or cut/pasted. You also have to be careful on the emailer you use, unless you want to trash mailKMail in particular will do this for you. totally useless. Netscape mail never loses any messages. I have been playing with Spruce...the latest seems nice. There is also an xpine that may be good a couple releases down the road. But neither calls the browser of your choice to display HTML links, or GIF/PNG/JPG images. Bummer. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Freakin' Netscape...
Rich Clark wrote: > > I'm pissed. I downloaded the v4.72 rpm's from the cooker archives on > rpmfind and attempted to do an upgrade. I started out doing a simple rpm > -Uvh netscape-common*rpm, rpm -Uvh netscape-navigator*rpm, then rpm -Uvh > netscape-communicator*rpm. Rpm will bitch that the navigator and > communicator packages conflict with each other, As well it should. They are two different products. Navigator = Browser only. does not have email/news Communicator = Includes email, which Navigator does not. Common = You need this for either of the above. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] can't mount ext2 floppy
Guillermo Belli wrote: > > I got a strange problem I can format a floppy with the ext2 filesystem, but > the I try to mount it and I get this error: > > 'mount: wrong fs type, bad option, bad superblock on dev/fd0 or too many > mounted file systems' > > It's not because of a bad floppy, because this happens with every floppy I > format with ext2, and soem are new floppies. If I format 'em with dos I can > mount the floppyes without problems. > any suggestions? Several. We need to know what command you are using to format the floppy and put a filesystem on it, and we also need to know what command you use for mounting it. Assuming you have /mnt/floppy as an existing hard drive directory, try this: prompt# fdformat /dev/fd0H1440 prompt# mke2fs -c /dev/fd0 prompt# mount -t ext2 /dev/fd0 /mnt/floppy Then see if it works. If it DOES, then you may not have the mount point defined in /etc/fstab (at least for ext2). Many folks simply issue "mount /mnt/floppy" but that is a shortcut that assumes things not in evidence (proper fstab entry). Let us know. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] no route to host
Guillermo Belli wrote: > > Hello everyone: > > I was trying to connect to the internet with another ISP. I'm sure i've > configured it correctly, but after it connects I try to enter an address an I > get this error in netscape: TCP error: no route to host. I've tryed with > different DNS from the same ISP, and from another ISPs, but I get the same > error. I tryed to connect with Kppp, linuxconf, netcfg, but always the same > situation. The strange thing is that in windoze it works! Any ideas? > Thank you. Sure. It means you have no routing set up, probably lacking a default route or gateway. Send us the output of the following command and we can tell you instantly: prompt# route -n We will be awaiting your reply. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Getting the eepro100 pci to work
I think all he has to do is to bring in the module first (from his email it sounds as if he is not doing this), then activate the interface. # insmod eepro100.o # ifconfig eth0 up 192.168.44.56 That ought to do it. Use your address. Netmask and broadcast should be generated automatically. To see the results, use ifconfig with no arguments and you should see if the card is up. If you get error messages, investigate them. # ifconfig eth0 Link encap:Ethernet HWaddr 00:06:29:05:8B:23 inet addr:192.168.44.56 Bcast:192.168.44.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:107492 errors:0 dropped:0 overruns:0 frame:0 TX packets:51862 errors:0 dropped:0 overruns:0 carrier:0 collisions:792 txqueuelen:100 Interrupt:14 Base address:0x5100 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:3924 Metric:1 RX packets:1282 errors:0 dropped:0 overruns:0 frame:0 TX packets:1282 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 [The above is for a pcnet32 AMD pci ethernet 10/100 card running at 10 mbps). If the above works, check your routes. Then you will have to figure out how to have Linux do it automatically at boot time. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Re: timeout sending Config-Requests [Incident:mcpaid000219-0017]
Jean-Michel Dault wrote: > > The log said your line is not 8-bit clean. That probably means your modem > speed is not set correctly. Try with 19200, 38400 and 115200 to see if it > changes something. Common enough assumption. It means nothing of the sort. It simply means the pppd deamon did not start. The error message is misleading. This is a "well known behaviour." In his case, the modems had already negotiated and passed some data back and forth. So its not a serial port problem, its a protocol problem. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Security with cable
Fred Frigerio wrote: > > What about DSL? I think it is similar to a PPP connection but havent > seen one yet and am thinking about getting it. Does anybody have good > info about it? Running together with Linux? DSL systems have a router port per customer, at least those that I have seen. The difference is that Cable daisy chains everybody down the street, its a bus topology. DSL, or dialup PPP, have a star topology with a router port per hookup. The "secure" systems that are being talked about in Cable use what is basically a VPN to each customer. That is a virtual private network where the IP data is encrypted to each customer with a different key, and the Cable modem has the encruption key. This system iswelllike the CSS fiasco proved, open to persistent attacks and hacking. It is easyh to do because the encrypted data stream STILL appears at every computer on the Cable modem segment. If you have cable, I would get some sort of router between the Cable modem and your computer(s). You then run a private IP on your computer(s) and use Network Address Transpation to go out on the internet. Most Cable users are clueless Windows or Mac users and expect security out of the box with their cable modem. It is a hacker's paradise. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Security with cable
Axalon Bloodstone wrote: > it's not different from being on a dialup, your just as open then. > There is a big difference, actually. On a cable modem, your home computer is part of a large ethernet segment. Any user can sniff your packets because everyone's data appears on the other user's ethernet port. A dialup system, however, has a separate router port for every dialup. The administrator, or someone upstream of the dialup server router can sniff packets, but a USER on another dialup cannot. Each dialup port sees only the data routed to it. To give an example. Let us say two systems, one dialup and one cable. Both have identical POP3 mail servers. Further, the administrator has a computer on the same ethernet as the mail server. Question, who can see what? A user checks mail. This is a cleartext POP3 mail function with username and passwords sent in the clear, INCLUDING the administrator when he checks his mail. A user on a cable system can sniff all packets and grab anyone else's username and password, including that of the adminstrator. God help the admin if he uses a root password! Home networks are particularly vulnerable as the passwords etc sent among their computers also appear on the entire cable system segment. On a dialup system, however, the administrator can sniff out all user packets because he is on the common ethernet part. However, the dialup users, being on separate router ports cannot sniff the administrator's password nor that of the other users. This has been a fantastic screwup problem on cable systems and all sorts of esoteric security methods are being devised or implemented with varying degrees of success. In the meantime, its a hacker's paradise! -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] telnet not working...
Lyle wrote: > > I am running LM 6.0. When I first set it up, I could telnet into the > machine. It, of course came with the KDE desktop and has the auto update > feature. I launched that and added almost all the recommended updates. Now > I cann't telnet into the computer. > > If I run 'ps as | grep telnetd', it shows telnet as a running process with > the following response; > > 9648 pts/0S 0:00 grep telnetd > > But when I try to telnet, I get an almost immediate response "Connection to > Host lost". You cannot 'ps' for the telnetd process, because it does not run unless it is started by inetd. That will not happen until you open up a telnet link. Since your telnet connection is not working, you cannot ps for it. I think that netstat is what you would like to use As I recall, Mandrake does not install the telnet server package by default. All other distros, 'telnet' includes the client and the server. Mandrake did until 6.0. After that, the client and server are separate packages. Check it out. A Assuming it is installed, check /etc/inetd.conf and find out which binary gets called for telnet. Check to see that it is there in the location called for. Test it from withing the machine; ie, telnet to yourself. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] M13
Benjamin Sher wrote: > > Dear Ramon and friends: > > My thanks to Ramon and other gurus for your evaluation of Mozilla 13. As > a non-techie, it is all too easy for me to be won over by the lovely > design and shiny chrome. It's good to know that what's under the hood > still needs a lot of work. I am, of course, disappointed to know that > Mozilla 13 is not yet ready for prime time, but truth, in the long run, > never hurt anybody. Let's hope the Mozilla folks continue to improve > their new Communicator. That is the whole problem, Ben. Communicator was never a good product, for either Windows or Linux. The code on Communicator has been in the 13MB to 20 MB depending on version. Mozilla right now has a fairly light footprint, about 4 MB, but that is not the Netscape 5.0 they talk about. Netscape 5.0 will use Mozilla as the BASIS. then they will add all of the Shopping stuff to it and get it up to 15MB just watch. A much better approach would have been to base Mozilla on the old Netscape Navigator 3.x. Just change the Java engine on it and aways to go. Unfortunately, bloatware and commercialism killed that idea. Netscape Corporation directed the Mozilla project and told it what the end result needed to be. This looks to me like a project that is not going to be very good. Specially for Linux. Does anyone know if the Mozilla source code database contain the Netscape 3.x source code? I haven't found it, and I doubt it is there. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
[expert] M13
Is it me, or is it M13? I downloaded it, and it worksbut. It sure is quirky. Loading pages, it jerks around. It does not seem to allocate space for images, and then jerks the page around to accomodate them. Like IE in Windows does. It is also very slow on my Pentium II-300 running MDK 6.1. This thing is nowhere near release as far as I can tell. My other impression is that the Mozilla crew has made a bad mistake by following the Communicator idea. Netscape 3.0 would have been a better role model, specially for Linux. We do not need "Profiles" in Linux, that is what user accounts are for in Linux. I suppose there is some use for it, but nothing like in Windows. The email in Communicator/M13 is dreadfully slow, not like in Netscape 3.x. Try it with 1000 emails in a folder and it just bogs down. 3.x never did that even with over 8000 in a folder. I hope Opera and/or KFM does well, because I think we have a disaster coming with the Mozilla-for-Linux thing as being the browser of choice for Linux. I'd use 3.x if it wasn't the browser crashes often on Java. Its beautiful, simple and fast. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] Re: Misbehaving "ls" in console -- update
Sounds like your xterm is not behaving correctly. Be careful here, are you saying 'xterm' in the generic meaning, or are you using the real /usr/X11R6/bin/xterm program? I am pretty sure what you mean here that you are using either KTerm or Konsole, which is what is launched when you click on the KDE terminal button - the X Terminal. X Terminal and xterm are not the same thing. My experience with KTerm has been that it is a very buggy program. The lines ending in 31m or 20m or whatever mean that the line endings have not been properly processed. Usually this is a representation of ^M or control-M which is the carriage return symbol. MOST linux files do not have lines ending with a ^M. This is an MSDOS thing. MSDOS ends lines in CR LF, or ^M^L, whereas Unix lines end in LF only. When a Unix terminal looks at an MSDOS file it will see this ^M and sometimes not process it correctly and display it as a funny control character. However, in these X Terminals, I have even seen Unix files with improper line endings, specially in the output of the ls command. I think what you need to do here is twofold. One, do not use KTerm. I think Konsole is okay, but xterm is the better one. You can assign the new program to the KDE button. that is what I do. The second thing is that you can redefine your terminal emulation in the /etc/profile. I usually have these lines in there: TERM=vt100 export TERM However, that said, I think that you can also define an XTERM environment variable, but I am not sure about that. In any case, if you define the TERM variable correctly, some of these problems may go away. Your choices are vt100, vt220 etc. There are also possibilities for PC compatible consoles. You can experiment with these. For the most part, anything you pick will become effective at the next login. So log out and log back in and you get the new TERM. Not much you can screw up here, but you may prefer to play with it in ~/.bash_profile rather than in /etc/profile. That way if something locks you out of a login as 'sher' you can log in as root and edit it. Look in /etc/termcap for a listing of your possible TERM settings. I would start with vt100. vt100 and vt52 are also understood to Windows 95/98 telnet sessions, although Windows does not handle them too well and most people think of Unix/Linux to be kludgy when in fact it is the terminal emulation in the Windows telnet that is screwed up. I should mention that the REAL vt100 terminal has a keyboard unlike the PC, which means that the emulation is not without holes, and some keys map over in strange ways. You may be better off setting TERM=pcsomething. If you experiment like this, please report back to the Mandrake list what you found out. In particular after you test it not only within Mandrake but also from a telnet session from a Windows 95/98 machine. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] managing numerous passwords
Harondel J. Sibble wrote: > > (curious) what methods > > people use to keep the password list from growing out of hand. Also > > especially when you are admining/maintaining a network of co-located > > machines. Do you use say a few passwords that are randomly distributed > > between machines, a different password for each machine, or what. > > Phil Edwards replied: > Although I can't claim to know a great deal about it, the usual solution in > cases like this is to designate one of your machines as an NIS server I think the low-tech method is best. Get too fancy, NIS, rsync and other stuff, and if something breaks, it all falls down like a string of dominoes. I think that passwords should be kept in one of three places: your head, a 3x5 card in the safe, or a laminated card in your wallet. It also helps if YOUR non-root accounts, regardless of the username have the same password. YOu may wish to have the root passwords all the same, or follow some scheme that is obvious only to you. Howver, all this is moot if you have a wallet-size card. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Kppp in Air
Alan Shoemaker wrote: > > HiI've installed Mandrake 7 on three different machines now > and this problem exists on them all. The third install was on > my masquerade machine (I call it my modem server). My isp uses > PAP authenication. > Jan 20 19:24:22 obi-wan pppd[1605]: The remote system is > required to authenticate itself but I > Jan 20 19:24:22 obi-wan pppd[1605]: couldn't find any secret > (password) which would let it use an IP address. > > The isp tech says that their logs show me signing on > successfully and then immediately hanging up. The Kppp settings > are the same on the 6.1 setup that's working right now as they > were on the 7.0 setup that didn't work. Make sure your /etc/ppp/pap-secrets file has your username and password. This sounds like you have not enabled PAP authentication or have it misconfigured somehow. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] O.T. personal internet access via employer connection
"Joseph S. Gardner" wrote: > > Greetings all, > > I recently proposed to our human resources dept. that our company > consider allowing it's employee's to use it's internet backbone > connection form home during off hours as an employee benefit, some thing > along the lines of becoming a small scale ISP for it's employees. If it is a real benefit, ie, part of the employee compensation package, then the company in offering this benefit will be required to perform. If this is not what you intend, then I would just let it be known that there are modems available to dial in. But if you take steps like give out accounts and passwords, I think that you create a defacto employee benefit. If it is a benefit, that has intrinsic value, it would need to be reported as income to the employee. Question: Does every employee get allocated a part of the value, or only those that actually USE it? It makes a difference come W2 and income tax time. By offering this service, you create a layer of administration. If you are not accounting or billing it, then the onus of administration eases a bit, but you still running web/mail/etc. If the company only releases the modems certain times of the day, and some employees find that unacceptable, then those people will have to retain their present ISP. Question is, they would feel somewhat angry that others are getting free ISP, while they have to pay because of scheduling conflict. It gets more interesting if the benefit is taxed and computed in the income. It matters not that you "give it away". A gift is taxable at its fair market value. You should evaluate all of these factors before you decide on doing it. Consider too that most modem banks bypass all firewalls. You need to evaluate the security implications of having users on the same network. Their login names and passwords may become compromised, and you do not know who might show up logged into your LAN. Most ISP's take this into account as part of the nature of the business. But a corporation may not find this possible security breach acceptable. Only yout company can make this decision. Give some thought as to the level of service you will offer. If you are going to offer ISDN or V90/56k service it is going to cost you more than just a few dollars. To handle incoming 56k calls you need digital modems with ISDN PRI, ISDN BRI or Channelized T1. On an analog line no user will exceed 33.6K. PRI has 23 channels, BRI has two channels, and CT1 has 24 channels. None of this gear is cheap. A terminal server to handle one CT1 line or ISDN PRI with 24 modems will set you back about $8,500 or more, plus the cost of the line installation etc. BRI can be enabled two lines at a time per BRI interface. It is somewhat cheaper to set up if you have well under 24 lines. If you approach 24 lines, the PRI or CT1 will be cheaper. Analog modems are easy to conceive of. I can assure you that with 12 lines into a bunch of modems you will be pulling your hair out keeping the damned things going. Channelized digital equipment is MUCH easier to administer. Good luck. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modems (long)
d with that later. This is great news. Now all we have to do is knock some heads at Mandrake, Slackware, RedHat and have someone come out with a PPP wizard that is identical to Win95 or Mac and we are in business. In fact, I would go so far as to say that what we need can be simpler: If your ISP needs scripted Login, do NOT use this wizard, period. Another thing is that windows 95 has a protocol to get the DNS numbers from the server. This is NOT dhcp. Its a simple part of the ppp protocol that is not implemenmted in the Mac or Linux o/s's. It adds a layer of complexity to Linux as well. This needs fixing. The Win95 default is to use server assigned DNS, and that works fine here at Nook Net. However, you can alwauys in a Mac or Win box write in the DNS numbers. Anyhow, that is my rant today, one day after the Ides of January. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] Re: "Speed Connection"
eing GPL, the guy that did the rewriting went ahead and changed the meaning of it. Possible? Yes. Probable? Yes too. Now, I was writing KPPP, I would call the setting not "Connect Speed", but "Modem's Serial Port Speed Modem to Computer." That is very specific and obvious, as compared to "Computer Serial Port Speed, Computer to Modem." 11. To confuse the issue, most serial ports are "Auto Sensing", so if the Computer is autosensing, and the modem is set to 115.2K, you get a 115.2K link. I can assure you that this is NOT A GIVEN. For instance, with the Multitech ZDXb unit, you can set things to 115.2K all day long and end up with a 57.6K serial cable link. this is because both ends are autosensing and the Multitech then says:"Gee Whiz, one of us has to go first, so I will try 57.6K and see what happens." A few seconds later the two communicate, so the modem says "Gee, 57.6 worked, I better stay here." INfact, there are a LOT of modems that will come up on 57.6K and not on 115.2K. YOu have to be careful because using 57.6K on a 33.6K or 56K modem will have terrible effect on thruput. How do you fix the problem? You do it with the proper AT command to the modem to tell it to start at 115.2K. It then will work okay if the computer is autosensing. But beats me what the AT command is, and it will be different for different brands. windows 95/98 has this ALL FIGURED out in the .INF files that reside in C:\WINDOWS\INF (a hidden directory but you can CD there). There you will see hundreds of MDM*.INF files pertaining to modems, and each one has had its quirks figured out for Windows 95/98 and it will come up with the best settings. With Linuxwell, you have to figure this out on your own. This is one of the reasons that Linux, in 90% of instances will transfer PPP slower than windows. There are all sorts of nuances to this. A lot of it requires proprietary manufacturer info. In fact, not so much proprietary, but simply that no one asked the manufacturer for the details, and no one has written a KPPP or other PPP daemon that has such a database. I think that it would be fairly easy to write such a PPP or KPPP program that could use the INF files from Windows 98. Its a bunch of text files, after all. The problem here is that they are probably copyright by Microsoft or the Modem maker and you cannot distribute them. However, I suppose that the PPP program can be Linux distributed with the instructions to "Take a Win95 or Win98 cd and copy over all the MDM*.INF files". I should mention that the INF files are basically unreadable by humans. They are text, but go try and figure things out. What I have found that works is to set your modem properties so you write a modem log, C:\MODEM.LOG as I recall, and make your connection to the Internet via Windows 95/98. Then log out. If the connection was satisfactory, you can then look at the modem.log file and see exactly what AT commands were sent to the modem by Windows 98. These are probably exactly what you want for your Linux box. Take these commands and plug them in by manually editing the pppd dialog files. This is all kinda tedious for us Linux users, but it is all part of the fight against the Evil Empire and their machinations and attempts at world domination. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modems (long)
tions. The PPP logfile of Linux is unique; but there are corresponding log files at the ISP and at your Windows 98 box. Any or all of these may provide clues as to what is wrong, as will long phone cords, voltmeters, desk telephone sets and spare modems. A good ISP that actually provides technical support is armed with all of these tools. For instance in recalcitrant cases I go to the customer house with my laptop. It runs Windows 95, and has a PCMCIA modem that I know works very well. I can plop it down next to the customer computer and fire up using HIS login and password and demonstrate to him tht his phone connection and Nook Net is fine, and that the problem is inside his computer. As an example. Or, I could also have trouble, in which case it could be a phone line. Lastly, assuming the problem is in his computer, I can use my external USR 56K (the most bulletproof modem I know!) and show him he has a poor modem. Rockwell WinModems are really, really bad. the new .160 and later drivers for them make them barely acceptable. The Lucent LT WinModem with the new 566 driver seems to do well too, but earlier versions have lots of problems. Poor thruput regardless of reported speed with WinModems is endemic. A new driver can sometimes help. It takes a Pentium 166 minimum to handle a WinModem. Less than that, the WinModem is paused every time the mouse cursor needs to move or the disk drive is accessed. WinModems, as a rule, do not handle poor line conditions. They disconnect, or spend all their time in retraining and changing speeds and do not handle data for 15 or more seconds at a time. Then they resume for a while, and then again the same thing. Very bad. And the Rockwells are the worse. Actually, the Cirrus Logic are the worse, but they are discontinued and never were very popular. Alas, all new computers come with WinModems and most are Rockwells. Try and tell an owner his brand new Hewlett Packard $1,500 computer has a piece of shit modem and see what happens. With some customers, what I do is LOAN them the USR external 56K for a week or so, and after they use it a week their eyes are opened and understanding settles in their souls and they spring for a better modem - after consulting with me first so they do not get another piece of junk. Or, they can buy the one I left there. In most cases, that is what happens and I get to sell a USR external 56K Sportster at List price. :-) You already have the best modem made, for all practical purposes. Most modems, the big thing is the Chipset. Alas, most externals come with Diamonds. How can you tell, here is the clue: 1. On a 33.6K connection, V34 analog, all sound the same. 2. On a Rockwell, when you dial out and the modems connect, yhou will hear first the Eh-O, Eh-O of the 33.6K initial connection. If you hear more than one repetition of this double h-ohhh, it means that you are retraining for a lower speed. Bad. 3. Then there is much hissing, now comes the important part. (a) The rockwells will growl like a dog, followed by a beep and you get silence and you are in. (b) The Lucent will go click click click click like a ratchet being turned, then the beep and then silence. (c) The USR/3Com/TI will go BongBoong like Big Ben in the fog. Then the beep and silence. Of course, these noises can change depending on what the ISP has at his end. But its a starting point. At the ISP end, you can have USR, Rockewell or Lucent equipment. I have Lucent. No one could make the Rockwell work here (Cisco, Ascend), but the Lucent Portmaster works well. Like I said, Rockwell really does not have their shit together. A lot depends on intervening Telco equipment. 4. If the noises are no clue, in Windows 95 go to Control Panel, Modems. There you will see your modems listed. Go to Diagnostics. In Diagnostics highlight the appropiate COM port, then hit more info. That sends the modem a series of ATI commands. ATI3 and ATI4 will reveal the modem type. Usually you will see things like Rockwell HCF or Rockwell HSP. Or there is no clue but you will see version numbers with embedded digits like .538. Basically embedded digits in the .100-.200 range are Rockwell, and digits in the 400-600 range are Lucent. Robotics and TI will state so. Some of this information is not appropiate to your problem, but as long as was giving a speech here on the list I thought that I would spiel out the propaganda. You can find the final authoritative answer as to which modem is WinModem or not, along with a host of data on ALL and EVERY modem ever made, at http://www.o2.net/~gromitkc/winmodem.html this is the final authority on modems and should be a bookmark for everyone on this list that has a modem, or will ever help a friend with a modem. Good luck! -- Ramon Gandia = Sysadmin == Nook Net http://www.
Re: [expert] 336 Supra Express
D HOPP wrote: > > I have an ISA Diamon Supra Express 336 PNP Voice modem. They do not have such a model. They have models like Supra Express 336i V+ and so on. You have to be a bit careful. Supras come in both WinModems and Non-Win. However, looking at all the variants of the 336's did not show any of them to be WinModems. However, some could be RPI's which are just as bad. > I',m running > Mandrake 7.0 and trying to get this modem to work (I didn't get it working > under 6.1 either). This isn't a winmodem (to my knowledge) but Lothar > doesn't detect it (I know it's in beta). I'm wondering how I can get this > working if Lothar doesn't think it's there. > > Thanks, > Dennis Most, if not all Diamond Supra modems are PnP. Often when you exit windows or reset the computer, the PnP settings are set back to default mode. this may be non-functional in your computer. Most Diamonds do NOT have Jumpers. This makes them tough to set up. There is such a thing as PnPTools, but I have never gotten the thing to work for me. Sometimes you can do this: fire up the computer in Windows95 and take a look at the modem settings in Control Panel, System, Device Manager. IN particular look at the resources, such as the IO Port and the IRQ it uses. >From there you can determine if that is a normal port, and which one (cua0, cua1 etc). YOu can then plug those settings into Linux and see if the modem shows up. ISA PnP was a kludge that never worked well, even for Windows. I have serveral modems in here that are ISA PnP and my computer would never pick up from Windows 95. Usually because there was a serial port or other already on the computer that was "blocking" the modem, so PnP would not alert Win95. Ergo, no modem recognized. My feeling is that you will get this thing working, but you may be in for a rough road for a bit. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Listening port for Sendmail
R_Yeo wrote: > > Hi, > At work, we have a Class C address with it's own mail server. > My Linux box is configured with sendmail to reply to mail addressed to > me via my ISP. Corporate IT is clamping down on smtp servers due to > spammers. Is there anyway to close down my smtp port (25) so that it > will not be listening or only listening to the local host? Corporate > IT does a periodic scan of the port to look out for unauthorised > servers. > OTOH, if someone has a suggestion for qmail or postfix, I'm all > ears. Normally, most people handle mail thru an ISP or an intranet mail server, like your company's. In either case, you use a mail program such as PINE, or the mail part of Netscape. Perhaps KMail. In those cases, the mail program is set up to talk to a mail server. POP3 or IMAP for incoming mail, and an SMTP server for outgoing mail. You do NOT need a mail server in your desktop computer, no more than a Windows 95 computer would. Therefore, what I would do in your case is to shut down Sendmail, or remove it entirely. Most Linux distributions, however, come preconfigured with Sendmail active, and have their mail programs such as Netscape or Pine talk to the built-in server. This, as you have seen, adds a layer of security and complexity to the whole operation. Just get rid of Sendmail and you'll be fine. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] splitting /home dir's
Ian Douglas wrote: > > The beauty of using QNX is the transparent file system it has in place. The way > you can configure them allows you to mount multiple file systems on the same > "branch" of the tree. > > So /dev/hda3 could be mounted as /home/ and so could /dev/hda4, /dev/hdb1-5, > etc. > > So looking at each device itself, you would see: > /dev/hda3/bob > /dev/hda4/allen > /dev/hdb1/susie > > But all mounted into /home you would see > > /home/bob > /home/allen > /home/susie > > Trust me, there are MANY time I've wished I could do this under Linux. I > believe this file system stacking is done under Neutrino, a small embedded > version of their OS. I do not understand. You are taking a partition, like hdb1 and putting the susie stuff on it and mounting it to /home/susie. This is straightforward in Linux. Or am I missing something in what you said? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] splitting /home dir's
Bug Hunter wrote: > > I believe the md driver enables fairly much what you want. It > "bridges" drives together to allow concatenation of drives under a > directory. It is part of the RAID package. The only drawback to this sort of thing (raid striping, concatenation, etc) is that it adds another layer of sophistication and complexity. If you have concatenated drives, and the raid package sofware dies for some reason, you may end up with a system that cannot even be mounted single-user or booted. I am not saying that this cannot be configured around, but it adds a learning curve, and another program to administer. Frankly, in an office environment where workers depend on having their files and programs available day in and day out, reliability and ease of use take precedence. The Raid solution is viable, but it does add a layer of administration. The so-called-administrator in this case is a person that is not yet very Linux experienced and whose time may be best spent in doing other office work or running a business rather than spending days or hours trying to figure out some obscure or arcane piece of software of dubious reliability. For this reason I suggested either X windows terminals (diskless if possible), or the Snap! Server which will save the person some time rather than add another burden. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] splitting /home dir's
ounted ones, check it out at http://www.cobaltmicro.com Again, they run Linux. Costs a bit more, though, but they are very good. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] SOHO System Administration
does it. The cost of administering a bunch of computers in the office vs administering a single central one starts making tremendous inroads into the cost of running an office. You can start working on something like this fairly easily. Displaying programs on each other computer is something for you to experiment with, as is 100MB ethernet or even Fiber Connected Network. Each user would have his own direcotry area on the central computer, and you know how that works. For instance, running Netscape, each user has his own preferences, files, colors, bookmarks, attributes, etc. The same applies to virtually any other program such as Star Office or Applix. The Central computer, incredibly, does not have to run X windows or even have X installed. Because of this, you do not generally need it to the aforementioned 700 mHz whiz box mentioned above Once you have figured out how to display stuff in other computers, get the network working properly, you can then figure out how to make the desktops into diskless work stations. While getting all this figured out puts you into a learning curve, it has the advantage that your USERS really need learn nothing. They turn the computer on and in a minute or two they are working at their station just as if they were sitting at a monitor and keyboard on the central computer. For the administrator of the central computer, his tasks are much easier. If he has to upgrade, say, Netscape. He does this ONCE in the central computer. There are really no issues concerning users; it adds at most just a little overhead to his chores. On the other hand, a network of stand alone computers, with drive and each with ever-diverging Linux (or Windows95) installs becomes a real headache in short order. The central disk drive idea fits about halfway in between the two, but is still much more labor intensive than the central computer and satellite X display stations. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] OTZ
Well, any word, thinking about it, or did they laugh me out? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Upgrade problem
You really ought NOT to send HTML messages. It makes it hard for some of us to read your email. I am sure your question is important to you, but you cut off half or more of your potential replies by sending them in HTML. At 12:49 PM 12/30/1999 -0600, Alejandro Arredondo wrote: Hello I Upgraded my kernel from 2.2.13-7mdk to 2.2.13-22mdk and it can initialize my scsi device, and it worked in the last kernel version. It drops me an error message that says that the aha152x.o module was compiled for kernel 2.2.13-7mdk and it can not run in 2.2.13-22mdk. How can I compile it for kernel 2.2.13-22mdk. I followed the next steps 1. Install new RPM (kernel 2.2.13-22mdk) 2. Edit lilo.conf 3. Run lilo 4. make dependencies again make dep 5. make module to compile my modules again 6. make module_install to put the compiled modules where they should be 7. Shutdown 8. Reboot my system And it didn't worked. Did I miss something? or maybe I did everything wrong? The problem is that the complaint is coming from the loading of the scsi module that is in the initrd image. If you go to /boot and do 'file xxx' on the ramdisk image, it will tell you it is gzipped. I fixed it this way: mv initrdxxx xxx.gz gunzip xxx mkdir /mnt/foo mount -o xxx /mnt/foo (now copy the new scsi module to /mnt/foo/lib, overwriting the old one) umount /mnt/foo gzip xxx mv xxx initrdxxx /sbin/lilo (and now reboot) -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] Radius
Right now I have an overloaded server, running Radius, my accounting program, Web, FTP and mail. I need to separate these functions, but of course they all need authentication. I have thought of rsync or NIS, but would perhaps like to see if RADIUS can be made to work. Here is what I want, as an example. My Web server, www.nook.net, requires that my users be able to FTP to and from it to load their web pages. This is no problem as they are on the /etc/passwd and /etc/shadow files. Instead, I would like to remove their entries from these files, and have the authentication be done against a remote Radius server, which will contain their passwords (either on the /etc/passwd and /etc/shadow files, or on the /etc/raddb/users file). In other words, if the Radius server can authenticate remote logins from the terminal servers, it should also be able to authenticate logins to other computers. What I need is a Radius authentication CLIENT module that would work with RedHat/Mandrake linux. I have not been able to locate any such. All I see is Radius servers, but no clients. Can anyone help me in locating something like this? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Linux Distributions
Toyswins wrote: > > I'm not an expert by any means but a couple of years ago I had a copy of > Red Hat 5.1 which installed just fine on a 486/66 with 8 Mb RAM. It was > awfully slow and the hard drive was small, don't remember how small. No > GUI loaded nor much else, but it loaded. Sorta' like DOS, you can load > a minimal install and it'll work, but not much else. Linux is like horses. There are the slow, strong, workhorses that pull the wagons and plows. Then there are the show horses, racers and the like. The former will make you a lot of money, but generate no interest or hype. The latter get all the attention but are generally expensive toys or hobbies. A linux server without GUI or even X windows, works like a plow horse. My boxes handle email, web, cache, DNS, ftp, Radius and the accounting and database servers. They do this without fanfare, for hundreds of days at a time. They are rackmounted, have no video, keyboards, sound, modems or any fancy stuff. Just an ethernet and serial ports. They run RedHat 5.2 and take care of Nook Net business in a reliable way. I can always count on these boxes to run without reboots, memory leaks, freezeups or other problems. They need a bit of minding, trimming of log files, and a few little things, but generally just run in the background. They make $buck$ for Nook Net. On the other hand, I sit at a workstation running Mandrake 6.1, X Windows, flat panel display, with all the goodies on it. It makes me no money to speak of, but is the box I stare at all evening. During the day, at my office, I have a RH 5.2 with XFCE2 and I spend all day in front of that one and its flat panel display. My secretary Sara spends her day in front of a Mandrake 5.3 box with KDE and Applix. My wife spends evenings with another Mandrake 5.3 box. These desktop boxes get all the attention, and is what visitors see when "they want to see what Linux looks like". But in reality, its the headless boxes at the Net Center that are the heart and soul of Nook Net. God Bless Linux this Christmas, it made Nook Net possible. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Mandrake 6.5 and ZOOM FaxModem 688
John Aldrich wrote: > > On Thu, 23 Dec 1999, you wrote: > > > Jack > > -- > > Linux: because you can't do a sig like this in Windows. > > 8:46am up 19 days, 15:29, 2 users, load average: 0.53, 1.39, 1.26 > > > Kewldo you run this script EVERY time you post a > message or daily, or what? > John assuming you have your .signature file in /home/rfg you first create the first part of the signature file: /home/rfg/sig.part.one Ramon Gandia Linux: because you can't do a sig like this in windows Then you generate a second file, called /home/rfg/sig.part.two: In cron: uptime > /home/rfg/sig.part.two do this every so many minutes, maybe 5 minutes or 1 minute or whatever. "cron" or "at" will do it. Third you concatenate the files, as part of the SAME cron or at script. cat sig.part.one sig.part.two > .signature You can combine step two and three on the same cron script: cat sig.part.one > .signature uptime >> .signature This should work. My uptime is 192 days here. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Opera-BETA really an ALPHA
WH Bouterse wrote: > > Yes it is almost as buggy as Mozilla releases, > but at least they got it out the door > for public testing. !!! Well, it might work with Mandrake 6.1, but it definetely fails to run on Mandrake 5.3 and RedHat 5.2 When I get home tonite I will try it on my LM 6.1 box and see what it does. Public Betas are good, that is how they get feedback. Like my problem. They can either supply libc6.1 or libc6.2 versions, or they can statically link the libraries and bloat their code by beacoup megabytes. I sure wish that Netscape had just upgraded and fixed their 3.04 version. It was trim, slim, fast, and the email is still a joy. Communicator was -and still is- a bloated thing geared chiefly to put advertising on your desktop. Lets give Opera some slack and see how they shape up in the next few months. PS: the KDE browser as well as KMail simply Suck. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Mandrake 6.5 and ZOOM FaxModem 688
. It is cheap and easy to administer but ISP's have problems with it. The Analog-CODEC-digital-CODEC-Analog path to my 33.6k couriers works perfectly and the connect speeds with them is still 33.6k with no problems. The Analog-CODEC-digital-DSP-Multiplex-Nook.Net path is not. We surmise that the bandwidth on the digital path to NookNet has dropped from 64K per channel to about 56K. In fact, this is what Nortel specs it out as. It simply is very marginal. We have found timing problems and over 70% of the customer modems simply became unusable. Modems that were previously connecting at 50,666 or so would only do 26,400 or so. Worse, dialing to the digital pool with a customer 33.6K modem only gave 31.2K max. No amount of work has corrected the problem at the telco end. However, by tweaking things at the customer end ... extra init strings, commas after the phone number, upgrading drivers etc; we have satisfied about 90% of our 56k customer base. The biggest source of headaches are the Rockwell chipset modems. These were bad with the DMS-5 equipment, and virtual dogs with the DMS-10. As an ISP, my hands are tied. I believe that PRI ISDN at 23 channels per T1 will solve the problem, but the Telco does not want to file for the needed Tariff or get the equipment just for me (there is no use for ISDN otherwise in a small town). My dismay has grown exponentially with the virtual takeover of the Rockwell modems. 18 months ago, virtually all computers had either analog modems, or a healthy mix of 56k modems of which the Robotics X2 (standard or WinModem) was the leader. Rockwell came out with its Host Signal Processing model, the HSP and later the HCF chipset. These chipsets are simply a phone interface that dump the phone noise into the Pentium and let your software do the work. All modem functions are done by the CPU. The chips cost the OEM about $3 to $4 per computer and largely account for the price drop in computers. But it is simply a terrible device, and the performance and quirks of this modem depend on many things like CPU speed, whether Office 97 is running in the background, whether you have IE and Netscape running simultanously, etc. etc. This is not an issue with regular modems, but it is with the HSP/HCF. The moral of the story is: get a real modem, and expect diffeernces in performance and reliability if you move from one area to another that has a different Telco switch. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Mandrake 6.5 and ZOOM FaxModem 688
Jan Rocho wrote: > > Hi! > > Im kinda afraid Microsoft is really taking over the PC. When I got this PC the > Soundcard wasn't working on Linux, the modem wasn't and the network card isn't. > Also Windows somehow automagicly removed my Linux partitions after I installed > that the first time, I don't know why. Windows 95 never did this, but Windows 98 has been widely reported to do this. It must be something specific that MS did on purpose. I do not think they like Linux, can't figure out why. -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] Mandrake 6.5 and ZOOM FaxModem 688
Jan Rocho wrote: > > Ok, I could exchange it but would the ISA version of the Zoom 2925 work? They > also had that modem when I got this one. > No, its still a WinModem model. You need a real modem. Keep this in mind: 1. With two minor exceptions (very hard to find), ALL PCI modems are WinModems. This should be a red flag for you. PCI=WINMODEM. 2. There are plenty of ISA WinModems. 3. ALL Externals are non-Winmodems; ie, they will work with Linux. 4. Rockwell chipset modems are the junkiest, the cheapest, and the most likely to give poor performance. The Rockwell chipset is the ones that manufacturers put in bargain PC's. Another name for the Rockwell chipset is CONNEXANT. There are plenty of modems that have Rockwell chipsets, but are not WinModems. Thse work under Linux, but my comments on Rockwell stands. They are not very good. 5. You will find that the vast majority of Modems manufactured today are Rockwell WinModems. There are two excpetions to this general rule. a. US Robotics modems. Use X2/V90 technology. This chipset is also made by Texas Instruments. There are both WinModems and conventional internal/externals in this product line, so be careful you do not end up with the WinModem. The largest user of the USR WinModem is Gateway 2000. I only know of two manufacturers of good modems with this chipset, although there may be others. US Robotics (3COM) Makes internal and external Sporster in 33.6K and 56K, both in data/fax or data/fax/voice. The other is PhoebeMicro. Their modems are sold by www.hitech-usa.com. The modems in question are the CM1456VQH-X (internal, ISA) and the CM1456VQE-X (external). These retail for about $43 and $51 respectively, and are essentially identical in performance and software to the U.S.R. In fact, the USR Windows 95 drivers work fine with them. They use the TI Chipset, and are licensed by USR. One third the price of USR. Recommended. But be careful if you get the Phoebe, and make sure you get the proper part number. They also make a lot of Rockwell Junk and WinModems. They also sell Cirrus chipset modems, which are even Junkier. But those two above are EXCELLENT modems. b. Lucent Technologies makes chipsets. They also make an LT WinModem which you want to avoid. Also, the LT WinModem has had a driver written for it under Linux. It does not work very well yet, nor does anyone expect it to, but it is there. One nice thing about the LT modems is that there is only one driver involved, regardless of whom the modem OEM is. Unlike Rockwell, for which you need your particular driver from Compaq, HP, IBM etc., LT just has one driver fits all. A year ago they were junk, but nowadays, they are not too bad for WinModems. There are VERY FEW Lucent chipset modems that are not WinModem, so my advice is to avoid them unless you are sure, like if it is external. It is getting very hard to find modems that are not WinModems. In a typical retail store, your only choice seems to be the USR/3COM Sporster, internal or external at prices ranging from $110 - $170. The advent of Junky Rockewell WinModems has created a tremendous demand for decent modems, and the law of supply and demand has jacked the price on the USR's up. This is why I recommend the Phoebe. As far as I can tell, its the same modem for one third the dollars. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] FTP port range
Nick Kay wrote: > > At 04:37 20/12/99 -0500, you wrote: > >Hi, > > > >I'm setting up a firewall on my mandrake 6.1 box, and am trying to lock > >down as many ports as possible. Does anybody know what range of ports > >*active* ftp connections use? FTP listens on port 21. When an outside request comes in for ftp, the ftp daemon and the outside client negotiate a port. The firewall detects this desire on the part of the ftp server and opens up the port. When the use of the port is finished, the port is closed. That is the way that firewalls work. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Mandrake SSH ...
Axalon Bloodstone wrote: > Didn't have the problem last time i built it on a 5.x box, the source is > definatly there, or there'd be no ssh-askpass in the glibc binary. > redirecti it to a file and see why it doesn't build.. The complaint is that ssh-askpass does not exist. I would think that ssh-askpass is installed by the ssh src.rpm; I can't envision it being part of the standard glibc library. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Mandrake SSH ...
Denis Havlik (and Axalon Bloodstone) wrote: > > Have you tried rebuilding the srpm? Should work. OK fellows, I take the hint. I didn't do it before because I did not know the ssh-mdk version was rsaref free. I checked it on my Mandrake 6.1 workstation, and indeed it is compiled without the RSAREF module. I am downloading the src.rpm right now. TAkes a while over a satellite link, but when done I will rebuild it and test out the resulting binary on Mandrake 5.3 and RedHat 5.2 systems and report back here to the list. One thing I like about the Mandrake ssh is that there is only one rpm to deal with. Redhat has ssh, ssh-extras, ssh-client, and ssh-server. The mandrake one does it all. Nice. In the 5.x systems I will save my /etc/ssh/*conf files and put them back in. Let you guys know in a few hours. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Mandrake SSH ...
OK, I tried the source rpm and my attempt failed. On a Mandrake 5.3 system, there was no support for bzip2, so I downloaded the bzip2 source rpm and rebuilt it. It installed correctly and now I have bzip2. Then I issued the rpm --rebuild ssh-1.2.27-2mdk.src.rpm it compiled for quite a while, then failed here: + ln -s scp1 /var/tmp/build.ssh-1.2.27-2mdk/usr/bin/scp + install -s -m 0755 ssh-askpass /var/tmp/build.ssh-1.2.27-2mdk/usr/bin/ssh-askpass1 install: ssh-askpass: No such file or directory Bad exit status from /var/tmp/rpm-tmp.19664 (%install) **root@gray[/usr/src/other]# apparently, something like ssh-askpass is missing from the source. How about it, Axalon? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Mandrake SSH ...
"Ronald J. Yacketta" wrote: > > Michael Flaig wrote: > > > > Hello ! > > How can I get SSH for Mandrake 6.1 ? > > It is needed for Servers ... to get Access from some remote stations ... > > > > It is not delivered with the Mandrake GPL Distribution ... > > I can´t find it ... The versions of ssh currently on the ftp and download sites is the old 1.2.27. This version has a newly discovered bug in the RSAREF2 module which leaves your computer wide open to hackers to execute root code. The 1.2.28 version should be released soon to cover this security hole. It would also be possible to get the 1.2.27 version in source form and do the security patch and then reinstall. ssh 2.x is not affected. I am not sure what the effect would be of getting the present 1.2.27 source RPM and rebuilding it with rsaref disabled. It may not work, but I would like to hear from people about this. I think if you are going to rebuild rpm's, it should be fairly trivial to get the source rpm, do the patch, and then rebuild the rpm to something like ssh-1.2.27-rsaref-good-mdk-i586.rpm or something like that. I have not investigated this too deeply yet. The CERT advisory on ssh 1.2.27 and prior just came out this week on Wednesday. Its a critical security hole if you have your /etc/hosts.deny and /etc/hosts.allow files set up to allow global ssh logins (most are, and is the default in that ssh is not mentioned in those files, leaving them wide open). I was just on the Freshmeat and the www.ssh.org sites today, Saturday Dec 18, and so far ALL versions there are the vulnerable ones with no patches done yet. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Qmail
On Tue, 07 Dec 1999, Rickard Åberg wrote: > Sure it's stable... but for how long? :) Qmail is stable. On my first server I had an uptime of over 400 days before I took it down to make room for a different server. The new server is running RedHat 5.2, Qmail, and the uptime is 174 days. It handles thousands of emails a day here at Nook Net. It has NEVER burped. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Re:
On Tue, 07 Dec 1999, Civileme wrote: > > Software modems do not perform as well as hardware modems for any given speed. > > That said, it appears that your beta software is a bit too beta as yet. How do you get it to even work? Give the driver a NICE setting of -13? Or is this thing built into the kernel or module? I just can't really picture this modem doing any good in Linux. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] Re:
On Tue, 07 Dec 1999, Stephen P. Molnar, Ph.D. wrote: > I replaced the slower isa modem on my Mandrake/Venus system with a pci > modem with a Lucent Technologies chip set. I'm happy with the performance > in Win95, but am having trouble in Linux ( PCI=WinModem=Only Works with Win95. Get a real modem. If I may suggest you can get the Phoebe CMV1456VQH-X internal or the CMV1456VQE-X external. For all practical purposes thse modems are US Robotics workalikes. They have Texas Instrument Chip Set (X2/V.90) just like the USR's, and they make all the USR Noises. Work fine Linux/DOS/Win95/98. In fact, you do not even need the Phoebe Drivers unless you want their voice features; they work fine with the USR/3COM Sporster driver. I get 52,000 connect speed with them. But be SURE you get the right model, Phoebe has a lot of junk out there as well. I paid $41 and $53 respectively for internal and externals. Cable not included. http://www.phoebemicro.com (manufacturer) http://www.hitech-usa.com 1-888-868-8778. You can ask for Phoebe Gin (Name is a coincidence!!) and she will take your order for the Phoebe modems. Hi-Tech is a dealer. This last summer, two manufacturers came out with PCI modems that are not WinModems. However, they still need special drivers, and one of them is vaporware. There is one LinModem on the market, not recommended. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Re: modem
On Mon, 06 Dec 1999, ibi wrote: > I have an external USRobotics Sportster 56k that isn't being recognized. > Does anyone have a suggestion? External modems aren't "recognized". What is happening to you is that your external serial port is not working or is not properly set up. Start by going to the BIOS and making sure your serial port is enabled on the proper i/o and irq. Generally speaking, you want to use COM2 (cua1), on irq=3, and you want to use a hard setting rather than "auto", and do not use PNP. Different BIOSes will call these things by different names. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] making bootdisk on LS120 drive?
Civileme wrote: > > Correct, now that I have tested mkbootdisk. It looks like it's close, though. I'll > check the source when I get a chance. > > Never tried that particular utility before. I don't see any burning need for >that > particular support (LS-120 to make a boot floppy) but now my curiosity has been >picqued. If I had to do it quick and dirty, I would write the boot floppy IMAGE to the hard drive, then dd that to the boot LS-120 floppy (not cartridge, I mean floppy). I think that would work. But I'd like to see Civileme's proposed hack of mkbootdisk. Are you going to add an option to mkbootdisk, something like mkbootdisk --ls120 linux-2.2.14-3 or are you going to make an ls120 program for it like "mkls120boot" or you going to put in some autodetection in the program so that running mkbootdisk will determine whether it has a floppy, a cartridge or whatever and behave accordingly? -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] making bootdisk on LS120 drive?
On Fri, 03 Dec 1999, Alan Shoemaker wrote: > Ramonok, after checking out your URL I see where we're differing > here. Let me re-state myself. A while back there was a discussion here > (or in newbie, I'm not sure which) about mkbootdisk and the ls-120. > Nobody came up with a way to use mkbootdisk to create a boot floppy on > an ls-120 drive, either on a standard 1.44 meg floppy or on an ls-120 > disc. In fact Axalon indicated that support needed to be added to > mkbootdisk for this very purpose. ;-) mkbootdisk is a fairly new command I never done a bootdisk this way until about a year ago, and then I found that mkbootdisk is very strict about what you name your kernel in /boot, and a few other things. mkbootdisk is really just a script. It only writes to /dev/fd0 which an LS-120 is definetely not, etc etc. The man page on mkbootdisk is messed up too. RedHat offered to have me write up a new man page for it, but about that time I had a dispute with RedHat and decided to take myself to Mandrake. As I no longer have an LS-120, I better quit babbling too much, other than to say that the linuxrouter people have it 100% correct in there and their procedure will work. You can then create an LS-120 cartridge "bootdisk". I have no idea if you can create a bootdisk on a floppy in the LS-120 device using mkbootdisk. I would tend to use dd in this case...but -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] making bootdisk on LS120 drive?
On Fri, 03 Dec 1999, Alan Shoemaker wrote: > Ramonthanks, I'll check it out. Where were you a few months back > when we discussed this? ;-) I was on the RedHat list for three yearsuntil a few things happened there and I moved to Mandrake. :-) -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] making bootdisk on LS120 drive?
On Thu, 02 Dec 1999, Alan Shoemaker wrote: > DanAFAIK you can't. This was discussed a month or so ago and nobody > could get it to work. > Nonsense. The boys at the Linux Router Project do it regularly. See http://www.linuxrouter.orgThey have specific directions there on using the LS-120 like that. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] making bootdisk on LS120 drive?
On Thu, 02 Dec 1999, Dan Swartzendruber wrote: > How do I do it? The vanilla incantation fails, as it tries to do a format > operation, which doesn't work. If I try giving it /dev/hdb instead, that > hangs until I interrupt it. I believe you set the LS120 to be the boot device in your BIOS. Then you have to make the LS-120 whatever device it is in the IDE chain. If it is connected as slave on the primary interface, it will be hdb. Then you write LILO to it, since you want it to boot. There may be other issues, but the main one is to set it to be the boot device in the BIOS. If your BIOS has no support for LS-120 as a boot disk, then you have to get a circuit card that will tell the bios that you have an LS-120 boot device. Those generally do not come with the LS-120. Else, you are out of luck and you treat the LS-120 as merely a removable hard drive device. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Lynx -- the color browser ?
On Thu, 02 Dec 1999, John Aldrich wrote: > On Thu, 02 Dec 1999, you wrote: > > Dear friends: > > > > I noticed some time ago that Lynx, which is available on Mandrake, > > features a lovely spectrum of colors when used in the console. We all > > know how fast it is. But the colors make it a very attractive browser, > > too. it's all text, but sometimes, you wouldn't know it. At any rate, in > > xterm its colors are the old grim-looking, dull black and white. Is > > there any way to use Lynx in xterm with the wonderful colors that you > > see in the console? > > > Sure...run it under Konsole instead of Xterm. ;-) The problem with Konsole is that its buggier than a fly trap. Try cut and paste between two Konsoles. Ugh. xterm works fine that way. It was enough for me to junk Konsole and put xterm on my toolbar. There IS a color-xterm. I think it is called just that, but I haven't played with it. It should do what you want. I also believe that color-xterm can be invoked as an argument to xterm, like prompt# xterm --colour (notice British spelling of the word) Do a bit of research on docs and man pages, I think you will be pleased. Konsole....better wait for KDE 2.0 on that. It will corrupt your files. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Rescue.img
On Tue, 30 Nov 1999, Axalon Bloodstone wrote: > > if you don't have a images/rescue/, you have a faulty disk. Well, I don't have it on mine either. Its the Mandrake PowerPack. My guess is, you intended for it to be on 6.1 but it didn't make it to the CD pressing. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Microsoft's "Vulnerability" ;-)
On Mon, 29 Nov 1999, Sergio Korlowsky wrote: > Check this one... hehe > > Microsoft has released a version upgrade that eliminates a vulnerability in > Microsoft(r) Internet Explorer 5. "A vulnerability in an optional component" > could allow a malicious user to gain additional privileges on a Windows NT > machine that allowed him or her to create or change files. > > -- > > How about that... a 'vulnerability is an "optional" component... > > soon they are going to claim that a "Vulnerability" is an 'additional Feature' > > in Winsucks products! he he.. Sorry I couldn't resist! ;-) You guys should have been around in the late 70's and early 80's. Those were the heyday of CP/M, Altairs and lots of S-100 computers14" drivesand 64K memory was a rich man's toy. Back then, there were some pretty zany expressions out there. Lets take a Quiz: 1. "That's not a bug, that's a FEATURE!" 2. The dBASE-II ad with the rule bilge pump. "Some databases...well...suck." 3. "Vaporware" was coined. 4. Jerry Pournelle was ranting about some new software or hardware product that was going to be released "Real_Soon_Now! tm" It WAS trademarked, and is now in the lexicon as RSNin the same context as Vaporware. 5. The IBM PC at 4.77 mHz was soundly beat in computing speed by an Altair or any 2 mHz Z-80. A good selling point is that it came with TWO 160K double sided floppy drives! You could buy the XT with an incredible 5 MB hard drive, but nobody could figure out what you would need that vast amount of storage for! 6. "The bill is $1,000. $5 for the time it took to tap it with the screwdriver handleand $995 for knowing WHERE to tap it." IBM circa 7040/7044 computers in 1960's. Of course, when Microsoft came strong on the scene in 1992all rules were cast aside. ANYTHING goes now. we have: "Freedom to Innovate" "Microsoft did not do anything wrong because it is the top selling and most wanted operating system...". "I didn't say thatoh, this email? I don't remember it...". -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] NTP will not update!
On Mon, 29 Nov 1999, Jason Antonacci wrote: > I believe Network Time Protocol (NTP) only uses UDP port 123. > The utility ntptrace times out (ntp equivanet of traceroute). > SO, I open a tunnel from EXT to the NTP server for UDP 123, > update the filters and it still times out. What is the problem > here? The ntp.conf is the bare minimum directed to an open > public server (tick.gatech.edu) I don't know about NTP and port 123, but I can use regular time protocol port 37 on tick.gatech.edu and it works OK rdate -p tick.gatech.edu a bit slow in responding, I use time.nist.gov myself. Are you sure that tick.gatech.edu has an NTP server in addition to the regular time server? NTP is not a very common protocol. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem configuration Problem
On Sun, 28 Nov 1999, Civileme wrote: > ACK, it's happening! > > The PCTel HSP is a Host Signal Processing Modem, one of the first Winmodems to > become a linmodem > > You NEED a DRIVER for Linux which PCTel does provide. GO to their Website or to > www.pcchips.com where your board was made. One of the two will have the driver. If this modem performs as poorly in Linux as it does under Windows, God Help Us all. Rockwell based WinModems are the worse of the worse. Hundreds of ISP's are screaming about them and pulling their hair out. The newest one is the HCF. That one at least does not need commas after the dialed phone number, but about a mile or so from the telco switch is about it, and then it acts up. Their worse feature is that while they may show a connect speed of 52,000, they simply do not have the data throughput that a real modem does at 52,000. I really wish a different chipset had been chosen. Of the WinModems, the lucent and USR do an order of magnitude better. Texas Instruments makes a nice chipset that uses the 3COM/USR protocols for X2 and V.90. They are not WinModems, but seem compact enough to fit onto a motherboard. The Rockwell chipset costs the manufacturer about $3 to put into the motherboard All I can sayis too bad. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Can't save or backup in SO51a
On Sun, 28 Nov 1999, Benjamin Sher wrote: > Dear friends: > > 1) Can't SAVE: > > I can't figure this out: I can't save any new documents: > > When I open a new window in StarWriter and type in text and try to save > it I get the following error message: > > "Error while saving document Untitled 1: > Object not accessible > The object cannot be accessed > due to insufficient user rights." Hmmmpf. Didn't StarOffice come with a 30 or 60 day trial license period, after which it will not save documents unless you register it (and maybe even PAY for it)? -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Basic Install
On Sun, 28 Nov 1999, Steve Philp wrote: > Here's a nasty little secret I've never seen documented anywhere. To > get the absolute smallest installation, just unselect ALL of the group > choices (like GNOME Desktop, KDE Desktop, Network Management, Kernel > Development, etc) during installation. You will get a minimum booting > system that works very nicely for building onto. I needed a remote DNS server, and I did a similar trick with RedHat 4.2. After I got done, I went thru it and zapped all sorts of stuff uneeded libraries, docs, etc. Rather than deleting, it is better to rename it so it can be put back together if you need to. Sometimes something vital is taken out and it won't work, not even to fix it. No problem. Take the drive out and hook it as a second drive in another Linux machine, mount it and do the editing to fix it. Lucky for me, on that DNS, I did not encounter THAT problem, but with hardly any work at all (maybe two hours from start of the 4.2 install to having it working as a DNS) I got it shrunk to about 28 MB, and put the whole thing on an old 40 MB hard drive. Way I did that, once I got it working on a large drive, I just copied the drive over to the 40 MB unit, made the 40 the master drive, booted from floppy, ran LILO and bingo, its done. There are several IDE interfaced solid-state drives out there, and although I have not tried one yet, it should make for a very reliable server! No moving parts! See http://www.sandisk.com -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Modem configuration Problem
On Sun, 28 Nov 1999, Miranda Heinz-AHM008 wrote: > Greeting from Peru. > Perhaps my problem isn't a big deal, but I couldn't get the solution in the > newbie list. > I work with a Pentium III that has all the peripherals integrated in the > mainboard, but I can't configure either the modem or the sound system. > The modem is a PCTel HSP56 , I found the technical specs of this and it says > that this modem works with windows and linux, Hate to tell you this, but it is a Winmodem regardless of what the brochure says. The HSP56 is the Rockwell Host Signal Processing chipset. That said, they may have a Linux driver for itif they don't, get your money back. Buena suerte... -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Networking and hubs...
On Sat, 27 Nov 1999, Brian R. Thacker wrote: > I am a graduate student and in my office I have > access to one RJ-45 network port. I have two computers > I would like to use a simple web server, and my own > workstation), and was trying to figure out how to > share the single port. This all depends what is pouring out of the RJ-45 jack! There are two possibilities here. (1) This jack goes to a regular hub or switch, part of the school network, and there are no IP restrictions. (2) Only ONE IP address can come out of that port. If it is situation (2), then you will either have to use IP Masquerade or Network Address Translation to enable more than one computer on this jack, or talk to the school's computing administration to ease off. In the case of situation (1), just use a HUB to split off. When connecting a hub, like the Encore mini-hub, you need to remember that Hub jacks are either "inlet" or "outlet". You cannot connect an outlet to an outlet or it will not work. The RJ-45 in your room is an "outlet" type. Your Encore hub may have one of its ports labeled as "IN", "LINK" or "REVERSE". If so, then a regular RJ-45 ethernet cord from that port to the room's outlet will work fine. The other 7 ports connect to your computers. If there is no special port on the hub (read the docs often the first or last port is like that, or there is a switch to convert one of the ports to the other variety), then you have to use a CROSSOVER cable. with such a cable you can connect two jacks of the "outlet" type. Crossover cables are sold, but they are not a bit ticket item, and you will have to look a bit to find one. All major computer stores and catalogs have them. A computer that presently works directly off the wall jack will work via the hub regardless of whether your wall jack is type 1 or type 2. Once you have gotten to this point, you know your hub and wiring are okay. Now you plug in your SECOND computer to the hub and see what happens. Most school networks use DHCP to assign your computer your IP address. Make sure that your computers can connect properly one at a time. Linux, in particular, has been a bit hostile to setting up DHCP. For a web server to work, its IP address must be visible and routable to the computers that need to access it. You need to get such an IP if you expect the web server to be visible to other networks. For instance, many schools use IP addresses in the 10.x.x.x range. None of those, or the 172.x or 192.168.x.x. addresses will be visible outside the school. Depending on the school's internal routing, using those IP addresses may even make the server invisible to parts of the school. In Linux, use the ifconfig command to find out yor IP address. In Windows 95/98, use the winipcfg command to see your network info (START, RUN, type "winipcfg" without the quotes). What you want to do may not be possible unless you have the cooperation of the schools MIS department. Most of them are protective of their network and paranoid, or simply do not want to be bothered. Check out what comes out of your jack before you contact them. You may well find a policy that prohibits "servers" from operating within their student network. This policy is in effect in most major Universities. In the past, Linux machines were considered "servers" and were banned; however, in recent months this policy has eased and now most schools allow Linux computers on their networks as long as they are not explicitly set up as servers. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] ppp problem
On Fri, 26 Nov 1999, David Hart wrote: > On Fri, 26 Nov 1999, Civileme wrote: > > First, see if you can ping the DNS servers you have listed. > > Actually, I can, but that seems to be the only thing I can ping through > this connection. Hit a wall after that; tried using several known working DNS > servers, but I got the same results. 100% packet loss no matter what I ping, > except the ISP's DNS servers. Oh, if you can ping anything, then your PPP is up. This sounds like a routing problem. HINT: Compare routing in Windows 95 to routing in the Linux box. Here are the commands: In LINUX: ifconfig route -n In Windows 95/98 (from an MSDOS prompt screen): winipcfg (look deeper here) route print I bet you have a lack of routing or your hops are set one short of what you need. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] ppp problem
On Fri, 26 Nov 1999, David Hart wrote: > On Fri, 26 Nov 1999, Ramon Gandia wrote: > > It is > > "server assigned DNS" in dial up networking, and in > > the networking icon it is set up to "disable DNS". > > Thanks for the ideas. Problem is that this ISP doesn't use Server > Assigned DNS. I've got the numbers, 4 of them actually, and they all work > through Dial-Up Networking. > Tried pinging some known IP's anyway, 100% packet loss. Went back to my > secondary dial-up and pinged the DNS numbers for the one that doesn't work and > they came back quite fast (200 ms). > Got any other ideas? It then has to be authentication. Lets see if we can figure it out. In Windows 95: (a) You are using a script file? This is a scripted login. (b) You do not use a script. You enter the username and password in the DialUp box (the one that has the CONNECT button on it). This is PAP authentication. Now you have to match your authentication in Linux to what worked in Windows 95. A lot of people, when they set up KPPP or Linux PPP, immediately put in a username and password in the script or chat script. Alas, in KPPP and Linux PPP that is the default. When you set it up, it brings up this nice screen that has a place for username and password. Using chat scripts in Linux is the equivalent of Method (a) in Windows 95. If your Windows 95 uses Method (b), the PAP authentication, you must leave the username and password areas BLANK in KPPP or Linux PPP. Yes, I said leave them BLANK. Then find the "Use PAP authentication" checkbox, and set up your PAP. Basically all that does is set up your /etc/ppp/pap-secrets file. That file will have a single line that will contain your username and password. This may all seem trivial, and there is a temptation to fill out BOTH the chat script and the PAP authentication both. But I can assure you that in many cases it will not work if you do both, or use chat instead of PAP. Most ISP's today use PAP authentication because of the ease of use with Windows 95 clients. In the case of Nook Net, if you supply any sort of Chat script, the server will take your username, password and then go into LA-LA land. You have a connection but it does not take you anywhere. This is pretty standard with most servers using RADIUS authentication, probably over 90% of all terminal servers nationwide. A way to test it is with minicom. If you fire up Minicom, dial into Nook Net, you will get a username prompt. Answer that, give it the password and it will go into nevermore. Even with PAP set. On the other hand, if when the username prompt comes up in Minicom, you terminate minicom (but stay connected, which is the ALT-Q option), then PAP will start automatically and you are in. Assuming, of course, that /etc/ppp/pap-secrets is properly set up. Read a bit on the PPP HOWTO, specially the part on PAP for clients. However, the KPPP setup wizard will set it up properly for you if you know what you are looking for. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] A new problem with linux (all of 'em)
On Fri, 26 Nov 1999, Civileme wrote: > It doesn't stop me from using LILO, but it does say LILO is a virus. > > > I can boot into Linux just fine. I still get this message and two screens to > negotiate every time I start up that warns me I have a virus in my boot sector. > I think that the biggest problem is that this whole thing makes the computer useless as a server. Imagine you have a power failure. The UPS signals a power failure and the computer then shut downs orderly to save the batteries. But when the power comes back up, the computer halts at the stupid screen and does not boot. This makes it useless as a server. This is as aggravating as some of those that have power supplies that do not kick in when the power comes back on, or that halt on lack of a keyboard. When a manufacturer takes out options on the BIOS setup, it detracts from the usefulness of the machine. I do not think it would take too many complains about your virus detection behaviour to have the company issue a flash bios upgrade. :-) -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] A new problem with linux (all of 'em)
On Fri, 26 Nov 1999, Denis Havlik wrote: > :>Try turning off the virus detection in your BIOS. (helll!!) ;) > > On the other hand, he IS right - noone should cry VIRUS for LILO. Making > some BAD publicity for those who do it could help. What most of these BIOS boot sector antiviruses do is do a checksum of the bootsector. They can detect if there has been a change. If it has changed, a virus is assumed. I am surprised there is no way to allow for a boot sector change. A lot of things can go in the bootsector besides LILO or the Win95 boot: Win NT boot is different, so is System Commander and other boot utilities. I would be quite surprised if this BIOS did not have a way to cope with it. Maybe you ought to look at the BIOS code yourself (and modify it with the flash utility). 8086 BIOS code is not hard to figure out. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] ppp problem
On Thu, 25 Nov 1999, David Hart wrote: > This is kind of a shot in the dark, but maybe somebody out there will > have a clue. One of the local dial-up ISP's I use recently mergered with a > larger organization, and part of the process involved making the old DNS > servers and other local hardware redundant, as well as "upgrading" their lines > (and possibly other hardware). Before these changes, I had no problems using > KPPP to establish a connection. Just plug in the DNS and phone numbers and off > we go. But since the changes were made, I get nothing. To be precise, I dial > up, it logs in, everything seems fine, but Netscape goes nowhere. The > connection seems to be there, but addresses never resolve Microsoft Windows has a protocol that can get the proper DNS server IP from the terminal server. That is how I provide DNS numbers to my Win95/NT customers. I never tell them the DNS address or enter it. It is "server assigned DNS" in dial up networking, and in the networking icon it is set up to "disable DNS". This will never work in Linux. You have to put in the DNS address in /etc/resolv.conf Fire up your KPPP connection and let it log in. Then try to PING by IP address instead of by name. Bring up your Konsole or Xterm and do this prompt$ ping 206.28.142.2 and see if you get an answer. If you do, then your PPP link is working. You need a DNS. Use one of the ones assigned to vincitydesign.com: 206.28.142.2 216.42.24.50 just put the above two lines in /etc/resolv.conf If you ping those two hosts, one will give you better ping time, and that is the one should be listed FIRST. I checkd it from here. The first one replied to my pings, the 216 one did not, but both responded as nameservers. The 206 one seemed more lively. Your mileage may vary. Your ISP may have better DNS numbers for you. Don't tell them you are running Linux, tell them you are using a Mac. Macs do not resolve server assigned DNS's, they need to have the actual numbers entered in them too. Some ISP's do not have Linux-savvy tech support people, so they refuse to help Linux people. Just tell them its a Mac and they will give you the numbers. But you have to be able to ping first or its not going to work for you, maybe something is wrong with the PPP link. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Security
On Tue, 23 Nov 1999, Devin Vo wrote: > I think I have someone attempting to break into my Linux box. I see these > messages in /var/log/messages > > "Nov 22 19:03:54 telnetd[2786]: ttloop: read: Connection reset by peer" > > I don't think he got in but I like to know what this message means. > > Thanks I also have here a Linux box made by Cobalt Microsystems. It has a monitoring software that tests every so many minutes for the working of the various components, such as the ftp server, the telnet server, the POP3 server, IMAP and so on. In other words, the monitor program attempts to log into those services every so many minutes. When the login prompt is received by the monitor, the monitor assumes the service is working and it terminates the connection. The Log file then shows the message you quoted above. So what it means is that someone tried to log in, and got the login prompt, but he then terminated the connection (reset by peer) without attempting to supply a username and password. This could have been done by an automatic program to find open ports, or it could have been done manually by someone for whatever reason which I leave to your imagination. Often it is legit, like someone trying to get POP3 mail but forgetting to supply the port 110;, ie, telnet mail.nook.net instead of telnet mail.nook.net 110. Get the idea? It also means that your telnet port; 23, is open in ways that you probably do not intend. In your case it means your tcp wrapper (inetd) program is allowing the telnet port 23 to be open to possibly unintended persons. If your port is closed outside your network, then the attempt at login came from WITHIN your network. Check your /etc/hosts.allow and /etc/hosts.deny files to see if things are properly set up. If you think you are properly set up and want it tested, you can send me a private email with your host name and IP and I can probe it for you from outside your network and see what results I get. If indeed the port is closed and does not result in the message "reset by peer" then the cause if from within your network, or even within your machine from some rampant program. You are doing the right thing by reading your log files and finding things like this. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Setting up linux box to connect to MS network
On Tue, 23 Nov 1999, Denis Havlik wrote: > > MS Exchange. Can anyone explain me what is this program good for? > Is it true that you have to pay per-seat licences for it? > And what is a reason for actually doing so. > > Please - this is a serious question. I have never seen any MS-Exchange > server, and it really makes me wonder why would anyone pay for a mail > server, especially on per-seat basis. How much does it actually cost? I would speculate that the owner already has an NT server in the network, and needs to add Mail capability. He does not want to add a Linux box because he neither wants the expense of an extra box, nor does he want to learn UNIX to run it. These are legitimate reasons. Having said that, I would like to know the costs and other reasons for choosing the MS Exchange server software package, and what it can do that Linux mail servers can't or don't. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
[expert] Re: Mozilla/Netscape 5.0
Hi Ben, I thought I would share my answer to your messages with the Mandrake crowd and see what they make of it. Here it goes. Bear in mind that Mozilla is not Netscape 5.0. I am not sure if they share much if any code at all. I suspect that Netscape 5.0 will have SOME of Mozilla's work on it. The goals are totally different. Mozilla's goal is to have a slim, fast, efficient, standards compliant browser/email/news/java platform with a good set of API's for applications that want to hook into their code. This is the end result of the Open-Source/GPL crew that is working on it. Netscape/AOL's goal is a Browser that features shopping sites, channel bars (for advertising and shopping), and all sorts of bells, whistles and plug ins. For instance, things like Netscape Radio with its advertising, the built in MP3 decoder with sites that will SELL you the MP3 music, and all sorts of active content for businesses to put stuff on your screen, yours speakers and your brain. The difference will be about 2 MB vs 23 MB size. Make no mistake, Netscape 5.0 is NOT Mozilla. What begs the question is to what extent are the two camps committed to each other. I doubt that Netscape/AOL has any interest in the Mozilla project other than it can save them from paying to develop future Netscape versions or functionality. I think that AOL's vision is to have the core functionality of their future browsers be the Mozilla engine. For us Linux users, I seriously doubt that we are interested much in the additional 21 MB of code designed to promote the interests of advertising and marketing companies. The squabbles between AOL and the Mozilla project are rooted on this. You will hear a lot of protests that my words are off-base or not true, but every one of those voices has vested interest in masking the truth. Opera, with its price tag of about $40 can afford to develop a browser that is free of all the marketing stuff, which is why so far it fits on a floppy. While its not open source, and fully commercial, it is a stand-alone product that does not depend on free give away. Both Netscape and Microsoft give their browsers away for free. When did this all start? Answer: When the browsers became a gateway for advertising. Like "search buttons" that lead to advertising sites. Or lead to search engines that prioritize and order the search results based on the amount of money they are paid for. Or bookmarks that are predefined for you and is nothing but a listing of advertisers. Or for the Netcenter page showing on the email page when no message is selected, etc. etc. ad-nauseam. As both Mozilla, now at milestone 10, and Netscape 5.0 approach release date, you will see a lot of this come to a head. You will see either a bloated Mozilla, or a Netscape 5.0 that takes over their coding and embelishes it, or the two camps split and Mozilla becomes a force of its own. Then they lose the AOL money and who knows what happens next. My take on it is that ALL of us will be losers on this whole fiasco. Stay tuned, we will see this all work out by year's end most likely. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] A script to discover IP?
On Sat, 13 Nov 1999, Tim Howell wrote: > Thanks for your help, Ramon. I now have a script that emails me only the > assigned IP address - took a while to get the parsing right! A couple > more questions, if you don't mind: > > 1. Is there a way to pipe the contents of a file into the subject line > when using the "mail" command? I would like the designated IP address to > show up as the subject, if possible. Well, you can do it even more crudely than the first one. Let us say that you wanted to have a message sent this way: cat somefile | mail -s subject [EMAIL PROTECTED] In the command above, "somefile" will be the text of the message, "subject" (preceded by -s) is the subject of the message, and [EMAIL PROTECTED] is the intended recipient (victim). Now if the command above is inside a script file, the scriptfile itself can be concatenated from several files. Here is an example file 1 cat somefile | mail -s file 2 this is your IP address that you want to report file 3 [EMAIL PROTECTED] Then them master script would simply have something like this cat file1 file2 file3 > scriptfile2 and when scriptfile2 runs, it contains the command cat somefile | mail -s 192.168.97.45 [EMAIL PROTECTED] This trick is used extensively in HTML programming to put content into the middle of HTML files. You have basically three files: the HTML headers, the part you want inserted and the HTML footers. They are concatenated together by a script and assembled into a complete HTML file. I your case, you would use the same technique, but the final assembled product is actually executed so the mail message can be sent. Be sure your permissions are right. The use of the redirector > might give you problems here, you may want to use something like this: cat bumfile > script if the file "script" has been created with correct permissions and is blank 0 bytes, then the script keeps its ownership and permissions. You can then use this to assemble it: cat file1 file2 file3 >> script This will append the 3 files to the (blank) file, so the permissions and ownerships stays the same. This trick is often used to erase a file but leaving it in existence: cat /dev/null > file in which case the "file" keeps its ownerships & permissions. This is handy to erase log files because the > replaces the contents, whereas >> appends contents. I am sure there are more elegant ways to do this, but hey, I am getting too old to learn fancy scripting and sometimes the cheap and dirty works just fine. > 2. I am using "pump" to retrieve my DHCP info, and am doing so on the > @HOME network. I have read documentation stating that pump will contact > the DHCP server every 3 hours to try to renew the lease - any idea where > this gets called from? I would like the script to run every time the > lease is renewed. No idea. But commands can usually be located this way: prompt# which pump /usr/sbin/pump NOTE: the above is ficticious, I have no idea if there is a /usr/sbin/pump. I do not have that program here. Anyway, if "pump" shows up, then see if its a binary or a script and use the method I outlined in my first email to append or prepend something to it. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] A script to discover IP?
On Fri, 12 Nov 1999, Tim Howell wrote: > Hi, > > Just wondering if anyone more familiar with shell programming than I had > any ideas how to write a script that would take a client's IP address (as > assigned by a DHCP server) and email to a user? I have recently > configured my system as a DHCP client, and would like such a script, to > avoid having to log on to my firewall machine and run ifconfig. Any > ideas? I think you can write this yourself. If you look at the output of the command /sbin/ifconfig you will see that the IP address is there. All you need now is for it to be emailed. Here is a CRUDE way to do this. 1. Redirect the output of the command to a file. 2. Email the file. 1. /sbin/ifconfig > /home/rfg/myip.address 2. cat /home/rfg/myip.address | mail -s IPAddress [EMAIL PROTECTED] YOu can combine the above two statements into a script file. How you "trigger" the thing is up to you. I am going to make a suggestion. If the connection is as a result of a dialup connection, the you can include the script in the dialup startup, but give it enough time to make a connection and negotiate an IP address. The command "sleep 60" will do that. Just add the three lines; ie, sleep60, then #1 then #2 to the end of your ppp startup script. You can add more, like to use the directive "cut" to parse your text file prior to emailing it, etc. But I think the above will get you started. If you are on an ethernet, I am not sure what would trigger it, but you could look at the dhcp command and see what calls it, then append it. If dhcp is a binary file, you can write a wrapper script. dhcp binary renamed to dhcp.bin write a script called dhcp and put it in the same directory as dhcp. The script has the following sequence dhcp.binary sleep statement line 1 line 2 That ought to get you started in becoming a script writer. It is crude, but its a start. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Switch Hub
On Thu, 11 Nov 1999, Singer XJ Wang wrote: > I have a 10BaseT Dumb Hub > > what's a swtich?and what does it have that's better > then an 8 hub? In a hub, all the ethernet traffic appears at all ports of the hub. Even those not meant for your computer. In a switch, this is segregated. It reduces collissions and improves performance. With a switch, only your computer gets traffic intended for it, and the other computers on the switch do not get YOUR data on their ethernet cable. They are expensive devices. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] multiple nics
On Thu, 11 Nov 1999, Pat Mc wrote: > Having some trouble setting up multiple nics on my 486. I'm using ne > clones with Mandrake 5.3 kernel 2.0.36. The cards work fine > individually, but with both configured I can only get one to work at a > time. I have tried netconf but that is not doing the trick. > > At this point I have done the configuration by hand. Both cards are > recognized at boot and ifconfig shows both up and running. But only one > works..ie ping etc You probably do not have routing set up for any cards but the first. No route, no pings. With one card, it will assume the proper routes. With two cards, only one will route unless you set it up specifically. I do not know about the new Kernels, but in the 2.0.36 and prior kernels, if you had more than one NIC of the same type, you could not use modules. Modules would work if they needed different drivers. Otherwise, with identical cards you needed to compile the driver directly into the kernel. I have no idea about the 2.2.x kernels, but you should check this out. -- Ramon Gandia = Sysadmin == Nook Net http://www.nook.net[EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970 fax. 907-443-2487 Nome, Alaska 99762-0970 Alaska Toll Free. 888-443-7525
Re: [expert] Two Questions
On Sat, 06 Nov 1999, Stephen P. Molnar, Ph.D. wrote: > After an abortive two weeks trying to implement v6.1 I have reverted (with > sighs of relief) to v6.0, I'm facing two boot messages at the present time > that I don't understand. or know what to do about. > > The first message is: > > [mntent]: warning: no final new line at end of /etc/fstab > This means that the last line in your /etc/fstab file does not have a terminating new line character. To fix it, put it into the editor, take your cursor to the end of the last line and hit the enter key. Then save the file. Presto. ALL Unix configuration files are supposed to end in a new line character. Its standard, and a lot of programs will complain or go weewee on you if you do not do this. As you found out. Ramon Gandia = Sysadmin Nook Net http://www.nook.net [EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970fax. 907-443-2487 Nome, Alaska 99762-0970 == Alaska Toll Free. 888-443-7525
Re: [expert] Slackware Vs Mandrake - Who is better?
On Thu, 28 Oct 1999, you wrote: > Hello, > > The first distribution I ever used was Slackware. It's a great > distribution, but Mandrake is better. This could be the start of another religious war but I will say this: if you are running a SERVER, specially on the internet, you are crazy to run X-windows. Run the whole thing from the command line, don't even install X. Print remotely if you need to, and make sure all services and things are stripped down. This is common sense, Slackware, RedHat or Mandrake. As for which is betterwell, one is chocolate and the other strawberry. Slackware is popular with academia, and most books about Linux, specially the older ones up to about 1997...1998... are geared towards Slackware. Why? Because academics and long-time-Linux users dating from the days when Slackware was the only game in town, wrote those books. Newer books are coming out every day, and most of them bypass Slackware. Slackware's great forte is as a teaching tool. However, that said, if you know what you are doing, it works well. Remember that there is very little difference among the distributions as far as what gets installed. Its the installation and management processes that are different. RPM vs Tarballs vs YAST etc. You pays your money and takes your choice. I ran Slackware in 1994-1995. Then RedHat came out and I tried it. At first I did not like it. Then I did; then Mandrake came out and seems to be a more responsive company without their noses high and their heads in the clouds. So I am using Mandrake now, first 5.3 then 6.1. It is optimized for Pentiums, which is nice. Once in a while I need a stripped Linux to run on a 386, a 486 or a MIPS and I go ahead and get something more appropiate, usually one of the slim distros that fit on a floppy or two. That is what is nice about Linux. You can get the flavor you want for your whim of the day. With Microsoft, there are no choices. Try run NT Server on a 386 with 8 MB RAM. Heeheheheheh. -- Ramon Gandia = Sysadmin Nook Net http://www.nook.net [EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970fax. 907-443-2487 Nome, Alaska 99762-0970 == Alaska Toll Free. 888-443-7525
Re: [expert] mounting cdrw
Larry Sword wrote: > > Try /dev/sdc0 My ATAPI cd burner mounts as /dev/sga -- Ramon Gandia ---Sysadmin --- http://www.nook.net 285 West 1st Avenue ISP for Western Alaska P.O. Box 970 tel. 907-443-7575 Nome, Alaska 99762fax. 907-443-2487 ===
Re: [expert] boot up -- syntax error
Axalon Bloodstone wrote: > > BTW, you only have to rm -f /bin/linux_logo to be rid of that.. My custom has been to rename these things rather than delete them. Once deleted, they are gone for good. If you rename things, you can change your mind later. Good names would be: linux_logo.bak (traditional .bak for backup file) linux_logo.rfg (my initials. they remind me this is something that I did) -- Ramon Gandia = Sysadmin Nook Net http://www.nook.net [EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970fax. 907-443-2487 Nome, Alaska 99762-0970 == Alaska Toll Free. 888-443-7525
Re: [expert] XFree86-xfs problem
Benjamin Sher wrote: > > Dear friends: > > I have a strange problem that apparently was caused by my foolish > tampering with the ttfonts directory. > > For whatever reason it broke the X Font Server. I have corrected it > since, but I get the following strange result. I wanted to uninstall > XFree86-xfs and then reinstall it. This is what I get: > > [root@adsl-77-232-172 rpm]# rpm -e XFree86-xfs-3.3.5-3mdk.i586.rpm > error: package XFree86-xfs-3.3.5-3mdk.i586.rpm is not installed > [root@adsl-77-232-172 rpm]# rpm -Uvh XFree86-xfs-3.3.5-3mdk.i586.rpm > package XFree86-xfs-3.3.5-3mdk is already installed > [root@adsl-77-232-172 rpm]# When you install a package, you give it the full name, but when you remove/uninstall a package, do not include the version part or it will bomb out. What you should have done is this: prompt# rpm -e XFree86-xfs prompt# rpm -Uvh XFree86-xfs-3.3.5-3mdk.i586.rpm Think about it. The remove or -e option will remove whichever package is the one currently installed and running. There is a possibility that the version is not what you think it is, so the remove option without the version numbers will always do the right thing. It says "remove XFree86-xfs-whatever". On the INSTALL or upgrad, the -Uvh option, the exact package name has to be specified. AFter all, you may have more than one rpm in there. Lets say you have, over the months or years been downloading XFree86-xfs packages, and you have four or five RPM's in there. How is rpm going to know which one you want to install? It cant,so you have to specify the -3.3.5-3mdk.i586.rpm part. OK? -- Ramon Gandia = Sysadmin Nook Net http://www.nook.net [EMAIL PROTECTED] 285 West First Avenue tel. 907-443-7575 P.O. Box 970fax. 907-443-2487 Nome, Alaska 99762-0970 == Alaska Toll Free. 888-443-7525
