80 new of 86 responses total.
The end is AirportBase->WirelessNetworkAdapter->LocalEthernet->G4 (and iMac). There is a neat utility available at http://edge.mcs.drexel.edu/GICL/people/sevy/airport/#Hangup for connecting and disconnecting the Airport modem, from the desktop. I've been unable to connect to Comcast after enrolling, buying the cable modem and connect-kit, and installing everything. After some discussion with Comcast on the phone, and trying various things, they ask about the lights on the modem - and one is not showing cable synchronization. So I agreed for a Comcast house-call. The tech found no digital signal on my cable and, sure enough, he went outside and found a "trap" in place on the incoming cable. He said that at one time traps were installed so people could not steal internet access (but anyone knowing enough to steal internet access would know to climb the pole and switch the cable from the trapped to the untrapped connector on the box....). Anyway, he switched the connector, and I had signal. But... As of right now, I do not yet have high speed cable access. I think it may be that the modem is mis-configured with the IP address I need, so I need to leave it off until the current DHCP lease expires....I hope it is this simple (this heads-up is from Macworld.com/forums).
Broadband! WOW! Instantaneous (nearly...) web pages. I'm sure this is old-hat for a lot of you but I'm excited by staying up with the receding edge (as opposed to the cutting edge). That thing about letting the DHCP lease expire was true. Nothing about that, of course, in the Apple Airport, Linksys cable modem, Comcast installation, or MacWireless literature. I would like to thank those that commented and offered advice while I was setting this up: it all improved by understanding. A couple of amusing footnotes that came out of this: The Comcast tech, while changing the cable to skip the trap, commented that the cable end was "burned". When he brought down the piece of coaxial cable he cut off I saw that the center wire was black. But that is just the result of the oxidation of copper when the weather has access to it: copper oxide is black. While checking my cable modem he had to talk to someone at the Comcast office, and commented that he wasn't going to use the installation disk from Comcast as he "didn't want all that junk on my computer". I then gathered from how he responded to the person on the phone that he was being taken to task for calling what the Installer installs "junk". Anyway, I'll never find out what it does install, as I'm connected without using it.
I have discovered that Comcast has a block on relaying e-mail that I generate in Netscape or IE that will be handled by my e-mail server at another ISP. I have sent them an inquiry about unblocking this relay, but would like to ask here why they do this and whether they are friendly to unblocking it. I've done some web searching, and allegedly it is to cut down on spam. However I did not understant how this would do so. I do not read my e-mail with a browser, and receive none through Comcast. All I want is to be able to click on "send mail" links on web pages and compose a send e-mail. Responses would be through my e-mail ISP.
They don't want you sending spam with forged sender addresses through their mail server, probably. If Comcast won't help you, ask your other ISP if they'll support SMTP authentication so you can send through their servers when you're not on their network. SMTP AUTH has been around a long time but take-up has been kind of slow in the U.S.
Try changine your e-mail composer to use the Comcast SMTP server, but leave everything else alone. That way mail will go out with the same from information but will go: Mac->Comcast->Destination instead of Mac->OtherISP->Destination I think you can make your SMTP server dependant on the Location setting, so you can switch your ISP and SMTP server just by changing the location.
I think that's what he's trying to do, and the Comcast server isn't letting him relay. I could have misunderstood, though. Another thing you can try is making sure that you've checked your Comcast mailbox just before sending. Some servers use "POP before SMTP" for authentication, where it lets you send if you've checked your mail from the same IP address within the last several minutes.
Will using the Comcast SMTP server put comcast.net as the From address?
Re #11: how do I do that, prp? I'm rather "lost" in there protocols. I haven't created an e-mail account with Comcast, yet. Is this the first step?
Re #11: I figured it out myself. I just changed the outgoing mail server to smtp.comcast.net - and it worked. I thought it was going to ask for a password, and it did the first time (my comcast.net PW) but not thereafter.
Ah, okay. I thought you'd already done that and it hadn't worked, that's why I gave somewhat confusing advice.
I shouldn't have said I figured it out "myself" only: I did look at the comcast.net forums and found some comments (full of jargon) that mentioned smtp.comcast.net, so that's what I tried. Funny thing, though, I haven't gotten an answer to my earlier e-mail to Comcast asking about this error message - and their mail response says it "might take a few hours" for a reponse. Maybe everyone was at the beach for Labor Day.
What are the relative uses and merits of the various security options available for wireless networks (and Airport wireless networks in particular)? A variety of options are described in Airport Help - WPA, WEP, Closed Network, port settings (SNMP access, etc), and maybe others, but there are no suggestions on choosing them - or choosing them simultaneously (?).
Re 17: I hate those automated messages that say "we got your message and will respond". That might be the right thing for @aol addresses, but not in general.
I have enabled WEP 128-bit encryption on my wireless net as well as "closed network". Although Airport supports both WEP and WPA encryption, my net's wireless Adapter only supports WEP: but that's sufficient for a home net, although not totally secure for a larger net with much more activity. Doing this required a lot of puzzling over Adapter and Airport Help documents (the Airport can be given a security passphrase, but the Adapter requires its hex equivalent - which is available via the Airport Admin Utility, although how is not mentioined in Airport Help - so it went, along with several aborted attempts that required resetting the Airport Base and starting over).
WEP will keep casual hackers out. It's possible to break it, but doing so requires eavesdropping on a fairly large volume of traffic. If you want to stay on the safe side, changing your keys monthly will probably foil any attempts unless you really move a lot of data across the wireless link. Right now there are so many networks around with no encryption at all that it's unlikely anyone will bother, unless there's something special about your network that makes it an attractive target. Hackers tend to go for the low-hanging fruit.
I won't worry until I see a van parked out front sporting antennas.... Since my Adapter has a Site Monitor function, I know there are no open wireless base stations in my vicinity. But still, is there any advice on setting the power level of the Base? The default is the maximum, 32 mw, but it can be dialed down, to be able to be less likely to be "sniffed". The Adapter monitor reported a "52%" power level before I closed the network and couldn't read that any more, but what does that mean? I get the impression that power level and packet transfer efficiency are related, but I wouldn't think that only 52% of the packets being transmitted between the Adapter and the Base are getting through.
I think the 52% is probably the signal level readout from the card, not an indication of packet loss -- kind of analogous to an S-meter reading on a radio receiver. A better indication might be the speed your network card is using. It will fall back to a slower speed if the signal gets too weak compared to the noise level. I'd use the lowest power level that gives you acceptable performance in the most distant location. (But then, I've always believed in the amateur radio credo of using the minimum amount of RF power necessary for communications.)
I've been hunting in these device interfaces for some readout of traffic rates, but can't find any. Is there a (Mac) utility that would show this? I think you are right about 52% being a signal level, although I'm a little surprised they put that function into the Adapter (which is the same thing as a network card, to keep our nomenclature clear, except not inside a computer but just on the network).
It's been years since I used a Mac, so I'm not sure where you'd find that information. Every wireless network device I've seen has some ability to measure signal strength. It's necessary anyway for several functions of the 802.11b protocol -- the ability to choose the closest access point on the network, for example. Since the capability is built into the card's radio modem, reporting it to the user as well isn't too difficult. Most cards I've seen have the ability to measure signal strength, noise level, and "link quality", though not all drivers report all of these values to the user.
Makes sense. Does a computer shift among nearby (open) access points if signal strength varies? (I wrote "open" as a user would have to specify the SSID and WEP Key to access my Base). I'm rather pleased how my network now functions. As I've said, I have two computers on a wired ethernet LAN and this network is hooked wirelessly into the Base Station, which is on Comcast's high speed internet access. For the first time my wife and I don't have to contend for time on the internet! "Modern Living".
I just discovered the "Network Utility" in OSX. This seems to be telling me a lot about the activity on differnt parts of my network - if I can figure out what the jargon means. I start with looking for a meaning for Ethernet Interface (fw0) [or en0], and Network Utility Help comes up with nothing for Interface, fw0 or en0. Who were they writing thost "Help" pages for? (And the Utility says "Link Speed" is 10 mb for en0 - which isn't a *speed*...) Sigh.....
Re resp:26: It depends on the network setup being done properly, but yes, it's possible to roam around and have the computer switch from one access point to another. This relies on the access points all being part of the same network, though. Automatic, seamless hand-offs from one network to another aren't possible because of the IP address and routing changes. Re resp:27: Okay, I don't know MacOS X, but I'll see if I can help you. fw0 and en0 are the names of interfaces. In BSD (which OS X is based on) devices traditionally get names in that format, where the two letters mean something about the type of the advice. For example, my FreeBSD system has two DEC ethernet cards, de0 and de1. (The de stands for DEC Ethernet, apparently.) A little Googling suggests that en0 is the Mac's ethernet interface, and fw0 is FireWire. "10 mb" probably means 10 megabits per second.
en0 is the built-in ehternet, and en1 is the Airport card. Assuming you have an airport card.
"ifconfig -a" is a useful command, in Terminal.
Does OS X have any equivalent of Linux's 'iwconfig' command, for displaying wireless-specific information?
From the System Profiler: AirPort: Interface: en1 Type: AirPort IP Address: 207.75.135.18 Subnet Mask: 255.255.255.0 Broadcast Address: 207.75.135.255 Router Address: 207.75.135.5 DNS Servers: 172.22.22.27 Domain: wccnet.org Ethernet Address: 00:30:65:2a:16:7c From Network Utility: Hardware Address 00:30:65"2a:16:7c IP Address(es): 207.75.135.18 Link Speed 11Mb Link Status Active Vendor Apple Model Wireless Network Adapter (802.11) Sent Packets 3181 Send Errors 0 Recv Packets 54381 Recv Errors 0 Collisions 0
That computer has an Airport card. Mine doesn't.
You can get the same information for en0, the built-in ethernet, as for en1, the airport card. re 51: What sort of information does iwconfig provide?. The command iwconfig is not there, but the information may be somewhere.
Back to Comcast Modems, I heard that they they have problems with more than 15 computers on the ethernet network. Motorola says the modem works with up to 32 computers.
Re resp:33: I thought we were talking about a wireless network?
Right - my ethernet LAN (2 computers) connects wirelessly to the internet. Neither computer has an Airport card - the LAN has a wireless adapter, but its not an Airport card. The router, however, is an Airport base.
Oh, okay. In that case, you're not going to get signal strength information from your computer directly. You'll have to get it from the wireless bridge on the LAN. Maybe it has a built-in webpage you can look at?
It has a built-in webpage for configuring the SSID and security parameters but it only shows field strength if one configures the Airport base to transmit its SSID - which isn't adviseable. But I think you've put your finger on it - I can't have my pie and eat it too.
(Further to the above stuff....) For security on my wi-fi Internet access I have made it a Closed Network and implemented 128 bit WEP. I now think I can also implement Access Control to the Apple Extreme Base Station by listing there the permitted MAC addresses. As described previously, I have a wired LAN with a PowerMac G4 and an iMac, connecting to the Base Station with a wireless ethernet Adapter. I can find a MAC address on the G4 but not on the iMac (neither has an Airport card installed), while the Adapter itself has a MAC address. Is the latter all I need to list on the Base Station to permit access for both of the computers on the LAN? I realize that if I have a visitor for whom I would like to permit access to my Internet connection from their laptop, I would have to enter their MAC address into the Base Station as well as the security parameters into their computer. This doesn't happen often, however. (The reason I don't just try it is that if I lock myself out of the Base Station by something I do - which I did several times in setting up the system - , it takes a while to reset the Base Station to its default settings and the Adapter for that, so I can access it to reset the security options.)
That's why I like having a wireless router that has an ethernet switch built in -- if I lock the wireless portion, I can adjust the configuration with a wired system. (Right now I'm using a Netgear 802.11b router; I hope to upgrade to a Linksys 802.11g router soon. The reason I want to switch to Linksys is that their firmware is open source, so I can add things like IPv6 support.)
The reason I have my wireless setup is so I would not have to run an ethernet cable from the LAN, which is in a different part of the house from the cable termination point.
Further to my #40 (and, more specifically): Question was: do I have to enter both the computer MAC address and the Adapter MAC address into the Apple Extreme Base Station Access Control list? New Question: what formats of the MAC (hardware) address are allowable. I've seen them written as 04:05:a6:..., as 04.05.a6...., and as 04 05 a6.... Does it, then, not matter what the delimiter is, even just a space? (I gather that case is not important - correct?) (I found the iMac MAC (hardware) address.)
I'd guess just the MAC address of your wireless adapter. I'm assuming the "computer MAC address" is the MAC address of its ethernet adapter. I'm not absolutely sure, though; if you're doing some kind of bridging, you may need to enter both. MAC addresses are written out with various delimiters, but the wireless access points I've used seem to want them entered with the delimiters left out, as a single string of hexadecimal digits.
Sorry to muddy the waters, but mine wants hex digits separated into pairs by colons.
Are they case sensitive, too?
I doubt it, but God only knows what someone out there is programmed. All of my wireless access points have had odd, sometimes amusing firmware bugs. (But then, I buy cheap crap.)
What someone out there *has* programmed, rather.
http://www.erg.abdn.ac.uk/users/gorry/course/lan-pages/mac-vendor-codes.htm l has some info about the MAC address, viz: "Ethernet MAC "Ethernet hardware addresses are 48 bits, expressed as 12 hexadecimal digits (0-9, plus A-F, capitalized). They might be written unhyphenated (e.g., 123456789ABC), or with one hyphen (e.g., 123456-789ABC), but should be written hyphenated by octets (e.g., 12:34:56:78:9A:BC). This gives a theoretical 281,474,976,710,656 addresses. This is more than 56,000 MAC addresses for each person on the planet!" Note that it says "A-F, capitalized". Yet in both my iMac and G4 their MAC addresses are shown with the letters in lower case, while the iMac has the address hyphenated with dots and the G4 with colons (and the Adapter uses the format stated in the quote above).
As far as I know, capitalizing hex digits is just a convention. I've seen them both ways, and of course from a mathmatical point of view it doesn't matter at all.
Of course, the software has to recognize an "e" as an "E", which have different hex representations when considered as ASCII characters. But I understand what you (and the cited URL) are saying: the representation is just straight hex. I think I got myself confused on this point, in part because the 128 bit WEP password is the hex equivalent of an ASCII password. So, what about whether the "hyphenation" must be by colons, or that hyphenation is acutally ignored when the string is read?
On some access points, the 128 bit WEP password is just the hex equivalent of the ASCII password. On *most* access points, though, it's actually an MD5 hash of the ASCII password. There's not much standardization on this point, though, which means if you use wireless hardware from different manufacturers, you may have to enter the keys in hex. The reason for not just using a straight ASCII -> hex conversion is it would artificially limit the keyspace.
I have enabled Base Station Logging on my Wi-Fi network, but have some questions: 1. What does the "Logging Level" mean"? It goes from 0 = emergency to 7 = debug. Airport Help is no help at all - nothing about the Logging Level. 2. Where do I read the logging, and what should I be looking for? It is apparently sent to the IP address of this computer, and I can find a mare's-nest of logs in Console, but nothing that looks like logging of Base Station activity. Will it log logins to the Base Station (I'm looking, in particular, to find if anyone else is accessing it)?
The logging level probably sets how verbose the log is. Usually with these sorts of things a setting of 0 will log only critical errors, and a log level of 7 logs a huge quantity of debugging information. Often the most useful setting is somewhere in the lower half of the range, but you'll have to experiment. I'm not familiar with OS X's logging, so I can't tell you what file to look in. You could temporarily turn up the logging level to something fairly high, then search the log files for the base station's IP address. I also don't know if OS X is set, by default, to accept log messages from remote devices. That feature is often turned off on other OS's, because it can easily be used as an unauthenticated, remotely-accessable disk filling service. ;)
OS X does text-based logging in (I think) /var/logs. Basically the Unix approach, and when I was first trying to figure out a modem problem on my iBook it was easy to open a terminal window and find the logs.
I upped the Logging Level to "Informational" (6). In /var/log are many files (and 6 directories). Those that are not empty are: CDIS.custom, cups(dir), daily.out, system.log, windowserver.log, and windowserver_last.log. I can't find an IP address (10.0.1.1 for the Base Station) in any of them. I directed the Base Station to log to this computer, which has IP 10.0.1.2.
You'll probably need to activate a feature on 10.0.1.2 to allow other machines to write to it's logs. 'man syslogd' might have the info you need. 'apropos syslog' may be a longer route to the info. Once it is ready to receive log messages it will be listening on port 514 udp.
Do you know that to be a feature of OS X? There is a syslogd command, but I don't know enough to set it up to log from the Base Station (if that can even be done).
Your syslogd should have inherited the feature from its UNIX roots.
Refer to the included manual page for syslogd from a Linux system.
-r This option will enable the facility to receive message from the
network using an internet domain socket with the syslog service
(see services(5)). The default is to not receive any messages
from the network.
This option is introduced in version 1.3 of the sysklogd pack-
age. Please note that the default behavior is the opposite of
how older versions behave, so you might have to turn this on.
I don't know for sure how OS X sets options for services. Linux would
use /etc/sysconfig/syslogd as a text file containing startup options for
syslogd. It may be the same in OS X. I think they use similar rc scripts.
I would appreciate some information about DHCP in IP/TCP. In the Network dialog in OS X I can implement automatic DHCP, but have no other control over it except to manually renew the DHCP lease. However in the Airport Base Station I'm using I can choose the DHCP lease term from minutes to days. What are the consequences of choosing a lease term? What happens at the time a lease "expires" (and does its expiration depend upon whether the computer, or Airport Base, is in use, idle, or off)? The default lease term on the Airport Base was 4 hours, and on some occasions I observed I no longer had internet access after my system had been idle for more than 4 hours. Is this related, or a coincidence, or what? If the former, how do I get a new lease immediately? Mac Help is pretty useless in regard to these questions.
The DHCP lease is something that the DHCP server sets, not the client. For your application I'd set the lease either off or for many days - a shorter lease is only useful if you have other clients connecting and needing a lease from a limited pool of IP addresses.
Makes sense. Now what might I experience at lease expiration if I set the term for several days (and how does this depend on whether the computers and/or the Airport router is always on and/or used intermittently)? Or maybe it is better to ask, what if any conditions of system operation and lease term setting will lead to any observable consequence of setting a finite lease term? It doesn't look like I can set the lease term to zero.
A lease term of 0 wouldn't make any sense. The lease term sets how long a client gets to hold onto an IP address before it has to ask to have it renewed. Setting the lease shorter means the IP will be available again sooner if the client goes away; setting it longer means that a client that isn't connected all the time will have a more stable IP address. It doesn't generally affect how often the client asks to have the lease renewed, though. My experience is that Windows machines, for example, will always ask for a renewal once every five minutes as long as they're online.
I'm trying to understand how this applies in my system, where I have two computers on a wired LAN that connects to a wireless router. I set the DHCP lease term on the *router*, which is on most of the time. The computers are shut down often. When the lease term was set to 4 hours, on occasion I could not connect from the computers until I had shut everything down and restarted. Can this behavior be explained with how the lease system works?
I don't think you should be losing connection that way.
I don't think that's related, unless there's a bug somewhere. All a lease term of 4 hours means is if the router doesn't hear from a particular computer for 4 hours, it releases that machine's IP address and may hand it out to another machine if a new request comes in. The original computer will still get an IP address when it asks for one again, it just may not be the one it had originally.
Here's what I get for the IP addresses of the various devices:
iMac G4 Bridge Router
10.0.1.2 10.0.1.3 192.168.30.1 10.0.1.1
69.137.222.90
I think that Bridge IP address is for accessing it's software to configure
it (?). The second one for the router is its "Public (WAN) IP address".
The router has a default setting of
"Share a single IP address (using DHCP and NAT)"
with an alternative, not being used of
"Share a range of IP addresses (using only DHCP)"
Does this tell anyone anything about how this system is working (and if
it might lead to a disconnection if the lease term in the router is 4
hours)?
It looks all right to me. It's a little odd that the bridge's address isn't in the same subnet, but for bridging purposes that really shouldn't matter.
How long does it take for the router to get a new IP address at the end of a lease term? If this can be a significant time, then I would have an increased probability of being able to connect anew while it is negotiating for a new lease if the lease term is short (like it was, at 4 hours). I have increased it to 4 days, and I have not encountered the inability to connect since.
Usually it doesn't take more than a second or two.
Is it possible for the ISP (Comcast, in this case) to run out of leases?
I get the impression that, with Comcast, *any*thing is possible. . . . But it should be improbable.
Given the 'always on' nature of cable internet, I would call it unethical to sell more accounts than they have IPs. So in correct operation it should not be possible for them to run out of leases. But as juicy says... I've had to debug and solve problems for them on a couple occasions.
I have a wired Ethernet LAN with two computers (iMac and G4) and a printer (Brother 2070N), connected with a wireless adapter to an Airport base station (and from that to a cable router). I can also connect to the internet through the base station from a Mac laptop. I would like to *print* with the printer on the LAN from the laptop. Is this possible and, if so, how? The Airport Extreme manuals only have information about printing to a USB printer connected to the base station.
Assuming the airport base station is like other wireless router/access points, the laptop should already be on the same network as the workstations and printer. Have you tried pinging the printer from the laptop?
From termimal? Using its DHCP IP address? I thought this was all built into OSX. The Printer Browser on the laptop shows the printer, but gives no location for it. (I'm in the dummy class for a lot of these networking details so I need help.)
I would think you should be able to access the printer like you would from the PCs. What kind of location information do the PCs give?
I got it to work with the DHCP IP address. The Mac Help helped but the procedure is not intuitive. What led me astray is that when I installed the printer software, it showed up in the printer selection menu, which I thought meant that it saw the printer. But no: one has still to Add the printer and give it an arbitrary name and location. There is then an obscure menu, which is hard to find, where one chooses the make and model of the printer (there are hundreds listed!).
Well, as unintuitive as it might have seemed, I can assure you it would have been much more confusing in Windows. :)
Case in point: My networked laser printer is in the basement while my windows PC is on the 2nd floor. If I add the printer to the windows PC, the "add printer wizard" asks if it is a local or network printer. You have to answer "local" in order to install it.
Another problem I had while trying to get to print across the network was that my wife was watching streaming video on the wired LAN while I was trying to print wirelessly from the laptop, which so jammed he pipe that none of my packets got to the printer until she shut down. I'm now wondering if I even had to install the printer driver on the laotop, since one of the steps in adding the printer was selecting a make and model from a dialog. Does OS 10.4 already have printer drivers?
Often there is a list of printers for you to install. And often the printer you have is not in that list until you install the software that came with it. Just to add some complexity here, if you can figure out how to force the base station to assign the same IP every time, or how to assign to the printer a static IP on the same network segment but outside the DHCP range, you may save yourself some trouble later on if the base station should ever decide to assign a different IP to the printer. Or you could just remember that this is a possibility and if the printer every mysteriously stops working start by checking what IP it is using.
I'll keep that in mind. So far, the IP assignments have been stable. I don't know where they come from, though.
The success I had in printing mentioned above was a one line test phrase. Now I am been trying to print a two page WORD document. I had two types of failures on repeated attempts: 1. After an hour or so *trying* to print, I got "Print file was was not accepted (client-error-not-found)! or 2. Error dialog "printer busy, will retry in 10 sec". This every now and then printed either a blank page or a page with just the first line of the document. What's gumming it up? Could it be the driver I installed when I really did not have to?
Re resp:80: Yeah, the Windows idea of "local printers" and "network printers'"is utterly confusing until you realize it's really asking whether it's a local *queue* or not. I suspect what Microsoft expects is that you'll tie all your network printers to one central Windows server, and then let all your desktop systems print through that machine, making the difference moot for typical users.
I wrote Brother (the printer maker) about the problem and they sent me the standard procedure for installing a printer uner OS X 10.4 on a network - but not on a composite wired and wireless network like I have. Anyway, I tried what they suggested, and it didn't work. Now I'm inquiring on an apple.com forum. It's a problem, though, in getting answers when the system involves a Mac computer and an Apple Airport router, a Brother printer, and a MacWireless adapter. Probably as bad as a Windows system.... Yes, the usual assumption in printing with an Airport WiFi system is that the printer is plugged in a USB or Ethernet port on the router. The Airport manual doesn't address the option of a LAN connected to the router wirelessly.
You have several choices: