Slashdot Mirror
Worst Explanation From Tech Support?
Disgruntled-with-Tech-Support asks: "Let's face it: At some point or another, we've had to deal with some form of tech support. Quite often, it's a hit-or-miss experience depending on the level of support required. Occasionally, strange, bizarre, or nonsensical explanations result from the problems reported, such as this one: I had just had DSL installed, only to find it much slower than the 56K line I was looking to get rid of. On calling the provider, I was told (by someone who likely reading off cue cards) to visit one of their internal websites for measuring bandwidth. While there, I observed that they had both bytes per second and bits per second listed, and that the number of bytes/sec != bits/sec * 8, rather a factor around 13 or 14. I pointed this out as a possible problem, and the guy's reasoning: 'Uh, it looks like the bytes are getting through to you ok, but the bits are getting stuck someplace.' What was your worst explanation from tech support?"
bytes/sec != bits/sec * 8
It should be bits/sec*(1/8), since you're getting one byte per every 8 bits. And you probably knew that, but I'm anal.
On the other hand, who knows what's happening when the bits are getting stuck someplace....
webpage
...it could be their admin system hadn't yet updated your e-status, and the isp tsr said what he knew...
at the number of bytes/sec != bits/sec * 8, rather a factor around 13 or 14.
;)
Shouldn't it be bits/sec = bytes/sec * 8?
I almost went through the phone to choke the bastard.
Uh, only problem is, he was mostly right. While LCDs do in fact have scan rates and frequency settings, no one cares, since they're mostly fixed. Almost all LCDs (at least in the home user market) have a 60 Hz vertical refresh rate. And most LCDs have a fixed resolution, so the scan rate is fixed (it is derived from vertical refresh and resolution). So he mostly knew what he was talking about, assuming the question was "How do I configure XFree86".
Now, if the question was "Can I install Linux on a laptop?" and the answer was "No, because LCDs don't have scan rates", then that's pretty stupid. But that's not clear from the post. Also, how long ago was this? It wasn't that long ago that Linux on a laptop required a lot of kludging, especially to get X running.
There is no sig, there is only Zuul.
The canonical multiplier to go from bytes/sec to bits/sec is ten (10): One start bit, eight data bits, one stop bit. This is how things were over serial/modem connections not so very long ago.
I find it still remains a reasonable rule of thumb. DSL and Ethernet frame data packets differently, of course. There are no start or stop bits surrounding each byte, but there is a multi-byte packet header and trailer. IP framing, of course, adds more overhead, but I find the 10:1 rule is close enough for most purposes. Besides, it's really easy to calculate in your head.
Schwab
Editor, A1-AAA AmeriCaptions
I assume he means (bytes / sec)*8 != (bits / sec). I would not expect a 56 kilobits per second modem to give me 56*8 kilobytes per second.
I had that a while back - after a significant amount of time, a technician came out to the house. It turned out that a line filter was a bit corroded on the outside of the house, and a quick replacement fixed it.
After reading several cable installer manuals, I found that a lot of cable installers will staple the coax too close to the rain gutters, and the ensuing rain tends to saturate the filter. Asking the cable installer to staple (or place hook-snaps) near-flush against the eaves usually reduces the likelihood.
System 5, I think. . . Which SCO owns.
Claims to own. Novell has other opinions.
The only reason we have the rights we have is that people just like us died to gain those rights. -- Cheerio Boy
You see, Fry's goes one step further than just having a horde of ill-trained customer service people roaming the store. They assign a person to each section, and go as far as to post a picture of them at the end of the aisles they're in charge of.
One day, upon needing some cable ends for some ethernet I was running, I decided to go to Fry's. They do have a good selection of networking hardware, so I figured I should have no problem getting the connectors. While I'm trying to find the RJ45s for rounded solid cable amongst the RJ11s, MMJs, and cable boots I get accosted by the salesdude, wanting to know if I need help. This is the same guy whose picture is pasted to the shelf. So I says to him, I says, "Could you help me find some RJ45s for plenum cable?" Reasonable request, right? I mean, there were routers to the left of me and telco racks to the right, and big spools of CAT5 behind me, so somewhere in that vicinity should be cable ends. His response: "I'm sorry, sir, I'm not sure what you are talking about."
I eventually found them on my own.
The moral of this story? Don't ask a customer if they need any help if you don't even know what products you sell!
YES they would charge the card. I had a Zip drive and had to call for tech support because Windows 95 wouldn't recognize it. I had to pay $14.95 just to speak to a person. A few years later I received a letter in the mail saying that I was able to join a class action lawsuit against Iomega.
l egacy=c net
Details here:
http://news.com.com/2100-1023-208214.html?
While I can't speak to the obvious idiotic answer the poster received I can say that I used to support DSL. While the web page was obviously wrong you cannot assume that you will get what you pay for on the invoice. Because of the way that most telecommunications providers provision DSL there is a possibility that up to 20% of the bandwidth you pay for can be taken up by network overhead. Not necessarily your network overhead. It actually depends on the number of people on the local loop.
As for actually troubleshooting it, pings don't mean crap. A traceroute will give you much more information. A continuous trace using something like mtr (Matt's TraceRoute) or some other similar program will help you narrow down where the problem is occuring, additionally, if its a router and not a bridge then the CPE configs should be checked. If its a bridge then an ATM ping from the edge router (ERX) can tell you if its actually on the circuit as there should only be the one hop between the ERX and the CPE. The tech support reps should be able to do this last one for you.
Regards,
FreeBSD Knight
Of course it's company policy never to, imply ownership in the event of a dildo... always use the indefinite article a d
"had him cuss the tech support guy in Hindu"
Thank god it wasn't me, I would have cussed him out in Christianity.
Hindi is the language. Hindu is the religion.
now supporting:
cmdrTaco for president '04
michael for oval office intern summer '05
but, uhm... isn't that the case? On a hub, every device must "dumb down" to the slowest link, whereas on a switch, every port can have its own speed settings (duplex, 10 vs 100mbps, etc)
How is this not so?
The Free desktop that Just Works
actually in asynch serial comms (such as your 56K) a byte is data bits, plus start, stop, parity bits
so the old standard 8,n,1
is really 1 start bit, plus 8 data bits, no parity bit, and 1 stop bit
so 10 bits in this case
the largest commmon byte would be something like
8,e,2 (1 start+8+1(for even)+2 stop bits, thus
12 bits in that byte as transmitted.
I'm glad that someone else noticed that the store reps explanation was fundamentally correct. Now it's not 100% correct, and it's not the answer that anyone who has taken a course in networking would answer, but it's correct enough that a customer would not have been misinformed by the rep's answer.
-Millions of Monkeys, Millions of typewriters, 6 hours of sorting through faeces encrusted pages to find: This post
I had this EXACT same problem with my ISP. It turned out that the LISA daemon that comes standard with Mandrake (dunno about other distros...) burps out ICMP pings over your network. My ISP took his ICMP ping traffic to be port scanning and/or some MS-Blast virus, and disconnected my connection. The bastards finally turned it back on once I tracked down exactly what was generating this (very minor) ping traffic.
They are more commmonly called repeaters, but they do exist. They regenerate the signal to allow a longer run. Much like you can find a two port switch. Usually they are called a bridge and are used for media conversion or monitoring (and are quire rare these days) but they are really 2-port switches.
A 1-port on the other hand...
Yes, the lines the run to your modem need to be taken care of much more carefully than your tv lines. The most important thing is to have your modem connected to the original splitter. Two splitters will work sometimes, three will almost never work. Perfect signal for a modem is 0dB. Anything under -5dB wont work; and anything over +10dB is bad for the modem. The second most common problem with modems are bad fittings. All of the fittings in the line from the tap to the modem need to be perfect. Otherwise your modem is going to drop packets and possibly lose sync with the head end. Other problems can be caused by nicks, cuts, kinks, or twists in the line. Significant amounts of signal can be lost even if the line is stapled or screw-clipped in equal segments. You can avoid these things by having your installer run your cable in the walls. It might cost you a little extra, but it looks nicer and your cable will last longer. The best tip I can give you is make sure your installer knows what the fuck is going on. If you see him wandering back and forth countless times, ask him whats wrong. Don't get in his hair though, because then he WILL do a shitty job on your install. Check his work after he leaves though. Good day all.
Plenums are defined to be any compartment or chamber which is connected to or a part of the air distribution system of a structure. Think things like ducts, flow shafts, and sometimes even the void above a dropped ceiling. The outer PVC jacket on normal Cat5 cable burns at a relatively low temperature and produces large quantities of highly toxic black smoke. Plenum rated Cat5 has a much higher combustion temperature and produces smaller quantites of smoke. The National Electric Code specifies that only Plenum Rated Cat5 can be run through any space connected to the air distribution system. Since air ducts are handy ways to run cable, a lot of Plenum Rated gets sold.
http://rinkworks.com/stupid/
Sometimes you have to do things like that because the caller wouldn't understand you if you tried to explain. Sometimes, it's not worth your while to try. If you really want to see what it's like "on the other side of the phone," check out this book of tech support horror stories. You'll have a little more sympathy for the techs once you've finished.
Good, inexpensive web hosting
* "This [less expensive] camera can only hold 15 seconds of video because of the 'cache overflow'" - about a Sony Cybershot P7 whose video length is limited only by Memory Stick size
That's entirely possible, actually. I know my camera temporarily stores pictures and movies in ram before writing them to the flash, the result is that there's a maximum movie size regardless of the actual flash size. Thus I can only record 3 minute movies, but I can record ten 3 minute movies.
In the 1960's, yes. Now, no, not really
Tell that to Unisys. Their mainframes (at least the ones I have to use) still have their 36 bit architecture, hence a 9 bit byte. Unusual? Yep.
Damn, so many fools around.
... and had to throw out the data. .. and complain to my friends about it on staticy phone lines from the same effect.
The 11 year solar cycle peaked a couple years ago. Sunspot activity peaks with the cycle.
The activity can cause huge electromagnetic disturbances that can very much be picked up by phone lines, cable lines etc, any antenna like structure.
Go ask your local geo-physicist about the hastles he/she had trying to perform exploration surveys on mining properties through the period of the peak. Like the phone lines, their long cables laid out in the bush also act like antennae, picking up the solar activity and wrecking any data you are trying to collect.
Been there. Done that
George Bush + Linux = "I will not let information get in the way of the fight against Windows"
I knew a guy who worked at one of the local Best Buy's. A lady, in her lat 30's came into the store with a computer that wouldn't boot. She asked the guy I knew to have a look at it. He took it back and realized the power supply wasn't working. One of his co-workers convinced him it would be rather funny if instead of telling her what the problem is and fixing it they tell her a BS story and see if she buys it. So they take the computer back to her and inform her that her computer is in need of a new flux capacitor. She apparently has no idea what they are talking about and they decide to run with it, the guy tells her that they do not have any flux capacitors in stock at the moment, but the Circuit City across the street usually does. Well the lady takes her computer over to the Circuit City and apparently tells them what she was told. About 20 to 30 minutes later the lady comes back into the Best Buy with her computer and tells the guy and his co-worker that Circuit City didn't know what she was talking about. They tell her they'll take another look at her computer and take it back and about 10 minutes later come out and say that they were right and she does in fact need a new flux capacitor. They suggest she take it to the Best Buy in the town just north of where their location and ask them, visibly frustrated at this point she leaves the store. About an hour later they get a call from the lady, who is furious and screams over the phone something to the effect of "YOU LITTLE ****ERS! A FLUX CAPACITOR ISN'T ****ING REAL!!! IT'S FROM A ****ING MICHAEL J FOX MOVIE!!!! I will NEVER shop at Best Buy again!!!!"
Bungo!
There's a self-defeating statment if I've ever seen one. Regardless, a quick search reveals that many mechanic services do indeed bill for diagnostics. Those who don't either pay their technicians less or charge you a higher hourly rate. The general reason why diagnositc fees are either all or nothing is because it is common to spend differing amounts of time diagnosing the same symptoms. Even a doctor will tell you that (who do, in fact, charge for s/office visits/diagnostic fees/).
Well, a former professional mechanic (me) says different. 99% of problems people have with cars can be diagnosed in 5 minutes or less, usually less. Building a relationship with a customer is worth spending those 5 minutes working for *free* to diagnose their car. I can't even think of how many stupid GM AC pressure sensors I sold just because they always looked the same on the gauges, and it literally took 2 seconds to hook up the gauges.
IN some specific areas, like exhaust and brakes, the free-looky is standard practice.
Besides the dealer (you know, the greediest little fuck on the block), most mechanics will only charge for diagnostics when they can't tell within 3-5 minutes what's wrong. That's the rule of thumb generally applied, actually. In the meantime, though, *every* mechanic shop posts something somewhere that says "We charge *this whole ton of money* for diagnostics", knowing that 99% of their diagnostics will be done for free.
Think about it. You're a customer, and you see a sign that says "Pay us $60 to tell us why your car is fucked up" and the mechanic just walks out and does it without billing you. Now how do you feel? How much does it increase the likelihood that you'll buy from these people who are obviously dedicated to serving the customer rather than bleeding him?
Like what I said? You might like my music
but what do i know, i'm just a model.
Actually, I remember a case where disconnecting the printer actually helped find the problem.
A customer called up because his computer would lock up frequently. We troubleshot the hell out of the system. We found that when the printer was not connected, the system was stable. So, he returned the printer, thinking that he had a bad printer.
Next day, new printer, same model... Same problem.
So, we went into SERIOUS troubleshooting mode.
It turns out that he recently fixed the ongoing moisture problem in his basement, and the grounding rod (Yes, this was a very old house) was not grounded. He poured some water over the rod, and found that the system was stable. I strongly advised him to get the house re-grounded.
He figured that he'd just give his house a cup of coffee every morning until he could have someone come out.
Try this (works on most every install, as it doesn't require 'My Computer' to be visible - which it isn't by default in XP :
1. Go to the Start menu
2. Choose "Help and Support"
3. Search for "Diagnostics"
4. In the results click on "Network Diagnostics"
5. Click "Scan your system"
Windows XP will now test various settings until the results are displayed.
6. Under "Modems and Network Adapters" expand (click on the + besides) "Network Adapters"
7. Expand the pertinent card (there may be more than one)
8. Read MACAddress line.
Much less daunting, a few less steps, and in the end the tech support person has a wealth of information available to them through the user. In addition, once things are setup, the scan can easily be performed again to make sure things work.
Have a look at the Wikipedia entry for Byte.
-- The plural of 'anecdote' is not 'data'.
there is absolutely no difference in security between a hub and a switch.
Right.
Well, except for the fact that there is a major difference in security between a hub and a switch, but I could never figure out why people like you chime in with brief little sentences that happen to be completely wrong.
If you put your NIC into promiscuous mode, on a hub, you'll get every packet meant for every station attached to the hub. On a well running switch, you'll only get the packets meant for you. That makes it much harder to sniff for NT password hashes, for example, or non-SSH telnet sessions.
On a switch, it's possible to do a promiscuous attack, by doing a rolling series of MAC address spoofs... fill the MAC table on the switch, and many of them will failover to being a simple hub, just to keep the traffic flowing, at which point you can sniff away. Of course, if you tried that on one of my switches, I'd get an SNMP trap with your port number on it sent to my Blackberry, and if I was in the building at the time, I'd be at your desk with a baseball bat in under five minutes.
So uh... yeah. No difference at all, sunny jim.
By the way - I'd rather deal with 10 clueless people who call a monitor, "the computer", and install spyware on their systems left and right than deal with one user like yourself. If you know you don't know what you're talking about, that's one thing. But a little "knowledge" is far more dangerous.
Serious answer: they probably had a number of small planes in the queue and your plane would have caused too much wake turbulence for them. They could hold you and let a bunch of small guys take off, or let you go and hold up a half dozen planes.
"Seven Deadly Sins? I thought it was a to-do list!"
Even on a 64-bit machine such as the Alpha, a WORD is DEFINEd as 16 bits, a LONGword is 32, and a QUADword is 64. Your way would force redefinitions when a 32-bit or 16-bit program is ported to a 64-bit platform, making the port even more tedious. Because DEC maintained the deinition of LONG and WORD, porting from VAX to Alpha was that much easier.
Gamingmuseum.com: Give your 3D accelerator a rest.
They have no idea what they are talking about. Bytes/sec and Bits/sec do not have any special meaning beyond the number of bits or bytes that can be passed in a given second. .00015 seconds!
First, it's totally and completely moronic to strip the TCP Header off for one measurement but not for the other, it's not some sort of standard practice.
Second, the story says that instead of Bytes = 8*Bits, it was mroe like Bytes=13 or 14*Bits. That means that by their logic the BYTES were getting stuck as it was taking more bits to make a single byte, not the BITS.
Third, the speed test is probably hosted on a simple webserver with a moderately sized file, whose download is timed, because this makes sense, it's how everyone does it. With that being said, why and how would they see the TCP headers from that layer?
Fourth, in order to code a working speed test, you couldn't use packets small enough for a TCP header to matter it to matter, and I've never seen a speed test that tried. A standard TCP header is 120bits (15 Bytes) IIRC. Let's figure out how small it would have to be: (x*13)-(x*8)=15, solve for X and find that the packet would have to be 3, do the same for 14 and you'll find that it's between 2.5Bytes and 3Bytes of data in order for the difference to be the size of the TCP header. What kind of a speed tests measures accurately to anywhere near modern DSL line abilities (let's say 100,000Bytes/sec to 400,000Bytes/sec) with just 2.5Bytes? Even for just 100K/sec DSL that's a 400,000th of one second of bandwidth. That's like checking someone's hearbeat by touching their wrist for
Parent posts with this theory are merely flamebait and should be ignored.
Jamon
I can count to 1023 on my hands. Ask me about #132.
in japan, they say "i understand" (wakarimashita) constantly during conversation; the meaning is pretty much synonymous with "i am listening." this is called aizuchi and is required to show you are paying attention. it is why japanese people might make "hmm" and "ah" noises and nod a lot while you are speaking. it's just a custom of politeness.
unfortunately, i think it causes the expression to lose meaning to them when translated to english. they use it much more loosely.
"This [less expensive] camera can only hold 15 seconds of video because of the 'cache overflow'" - about a Sony Cybershot P7 whose video length is limited only by Memory Stick size"
Erm, that's completely true. Cheap cameras can't encode video realtime, and also it can't write it to the flash chip fast enough, and therefore runs out of cache in 15 secs. Sony Cybershot, however (pretty damned good cameras, btw), can encode it realtime, and write to the fairly fast Memory Stick devices.
toresbe
Want to know what really happened with the problem?
First, the DVD player back then didn't know much about wide aspect ratio or high-resolution displays. So...
First it switched your display to a lower resolution it could handle. The lower resolution modes have a 4:3 aspect ratio. Without 'stretching', or widening the pixels in the horizontal direction, which would look odd, this results in a display mode of operation in which the display produces black bands on the left and right of the display.
Next, now that it had a display resolution it knew about, the player looked at the aspect ratio of the movie, and determined that when it filled the active display width (which was matted, but it didn't know that), that there wouldn't be enough pixels of height to fill the display vertically. The program then generated a set of horizontal matte bars.
The vertical matte bars came from the display mode of operation. The horizontal bars came from the program's need to generate a matte.
Since then, the player has been substantially rewritten, and now knows much more about display hardware. I'd have been very surprised if the support tech understood details of display hardware and the DVD player internals.
I actually recommend Tight VNC as it's a bit faster than standard VNC and an extension (fork) of the original source IIRC. I use it here at work.
That said, if you're ISP tech support, getting a user to install a VNC client is not only painfully difficult, but a security risk as well. In the case of VNC coming with the ISP software, I'd actually be extremely ticked off (e.g. cancel my service) if a VNC client was installed with any ISP software I installed.
Cheers
~Dalcius
Rome wasn't burnt in a day.