Worst Explanation From Tech Support? 1907
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?"
No wonder... (Score:3, Informative)
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....
hmm... (Score:3, Informative)
Not to mention the submitter has it backwards (Score:5, Informative)
Shouldn't it be bits/sec = bytes/sec * 8?
Re:"it's a feature, not a bug." (Score:5, Informative)
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.
The Bits/Bytes Multiplier (Score:5, Informative)
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
(bytes / sec) != (bits / sec) * 8 ? (Score:1, Informative)
Re:My ISP is retarted (Score:4, Informative)
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.
Re:Not to mention the submitter has it backwards (Score:1, Informative)
* the term was more liquid in the past
Re:Worst reply i've GIVEN.... (Score:4, Informative)
Claims to own. Novell has other opinions.
Re:Not to mention the submitter has it backwards (Score:4, Informative)
You can read a bit more about it here [reference.com]
neh, Fry's (Score:4, Informative)
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!
Re:Iomega support sucks ass! (Score:5, Informative)
Details here:
http://news.com.com/2100-1023-208214.html?
DSL and network overhead. (Score:2, Informative)
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
Re:You said it... (Score:5, Informative)
Thank god it wasn't me, I would have cussed him out in Christianity.
Hindi is the language. Hindu is the religion.
Re:CompUSA (Score:5, Informative)
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?
Bits about Bytes (Score:5, Informative)
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.
Re:CompUSA (Score:4, Informative)
The LISA daemon could be your problem (Score:5, Informative)
Re:I work in tech support.... (Score:5, Informative)
A 1-port on the other hand...
Re:Not to mention the submitter has it backwards (Score:3, Informative)
Now, if this person had set a tiny MTU (68 bytes is minimum, IIRC), then it would be possible to get a really high overhead. Most likely he 'tweaked' his TCP configuration and tuned it for peak modem performance, which limited him to modem performance when he connected to a higher speed connection.
Re:My ISP is retarted (Score:2, Informative)
Plenum Rated vs Normal Cat5 (Score:5, Informative)
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.
Want some funny stories? Here you go! (Score:3, Informative)
http://rinkworks.com/stupid/ [rinkworks.com]
Re:Some of my best lines : (Score:5, Informative)
Re:Overheard at Best Buy (Score:1, Informative)
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.
Re:no, not in this decade. (Score:5, Informative)
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.
Re:Sunspots (Score:2, Informative)
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
Didn't happen to me but.... (Score:2, Informative)
Re:Worst Explanation? (Score:5, Informative)
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?
Re: Mebibytes and Megabytes (Score:3, Informative)
Re:CompUSA (Score:3, Informative)
Re: Disconnect the printer please (Score:3, Informative)
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.
Through XP Interface... (Score:3, Informative)
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.
Re:no, not in this decade. (Score:3, Informative)
Re:Not to mention the submitter has it backwards (Score:3, Informative)
Byte was *never* variable. You're thinking of "word" which represents a natural unit for a system. Thus a PC has a 32 bit word, a Unisys Mainframe has 48 bit words, and an embedded processor has a 16 bit word.
Re:CompUSA (Score:1, Informative)
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.
Re:Worst Explanation? (Score:5, Informative)
Re:Not to mention the submitter has it backwards (Score:3, Informative)
For example: a dial-up's best speed in Kb/s is 4.5. 4.5 Kb/s x 1024 = 4608 bytes/s. 4608 bytes/s* 8 = 36864 bits/s or 36.8 Kbps But if you use 11 you get: 4608 bytes/s * 11 = 50688 or 50.6 Kbps, the best speed you can get on a 56 kbps modem.
A 256 kbps broadband connection should be: 256000 / 11 = 23272 bytes/s. 23272 / 1024 = 22.72 KB/s. Does 22.72 KB/s sound familiar for a 256 connection?
Ernesto.
Re:Not to mention the submitter has it backwards (Score:4, Informative)
byte:
[techspeak] A unit of memory or data equal to the amount used to represent one character; on modern architectures this is invariably 8 bits. Some older architectures used byte for quantities of 6, 7, or (especially) 9 bits, and the PDP-10 supported bytes that were actually bitfields of 1 to 36 bits! These usages are now obsolete, killed off by universal adoption of power-of-2 word sizes.
Historical note: The term was coined by Werner Buchholz in 1956 during the early design phase for the IBM Stretch computer; originally it was described as 1 to 6 bits (typical I/O equipment of the period used 6-bit chunks of information). The move to an 8-bit byte happened in late 1956, and this size was later adopted and promulgated as a standard by the System/360. The word was coined by mutating the word 'bite' so it would not be accidentally misspelled as bit. See also nybble.
Re:no, not in this decade. (Score:3, Informative)
Don't let them confuse you... (Score:5, Informative)
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
wakarimashita (Score:2, Informative)
unfortunately, i think it causes the expression to lose meaning to them when translated to english. they use it much more loosely.
Re:Overheard at Best Buy (Score:2, Informative)
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.
Re:Widescreen idiocy (Score:3, Informative)
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.
Re:Not to mention the submitter has it backwards (Score:1, Informative)
Re:wakarimashita (Score:2, Informative)
A Japanese speaker might say something like:
"Imi|setsumei ga tsuujiteiru deshou ka?" ("Is my meaning|explanation coming across clearly?")
Or more politely, especially when dealing with a customer:
("Go fumei na tokoro ga gozaimasu ka?" - "Are there any points (in what I've said) that are unclear?")
Disclaimer: I'm not a native Japanese speaker, but have studied it for 15 years.
Re:Worst Explanation? (Score:3, Informative)
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