back to article Support team discovers 'official' vendor paper doesn't rob you blind

Hello, Friday. And hello, therefore, another instalment of On-Call, The Register's week-ending reader-contributed tales of support jobs that occasionally work out for the best. This week, meet “Ben”, who told us that “In the early 'noughts I worked for a large tape/disk vendor.” In his early training some of the tape support …

I find this story very odd.

So the tapes either weren't reading at all, or were reading as different tapes to the ones in the drive - fine.

Now if they weren't reading at all, and to robot thought there was no tape, then surely it would of become very obvious once the robot went to retrieve the tape from the drive, and failed because 'I can't find the tape!'

Terrible debugging and reporting is what caused this problem, exasperated by a customer using unsupported hardware.

2
33
Silver badge

Coupled with the usual "Oh, no, we haven't changed anything" syndrome from the customer, most likely.

Yeah, it's odd that in a device costing quite a lot of money, nobody bothered to put in a routine that said "Hold on, I can't read that barcode, better alert the user in an obvious and sensible fashion".

27
1

Probably because unreadable barcode was resolved into no tape in the slot.

46
0
Silver badge

What a stupid failure mode and assumption that is, then.

Hell, put a barcode at the BACK of the empty slot and if you read that barcode - yeah, it's fair to assume it's empty. Or a 20p micro-switch testing for physical presence per slot to distinguish "no tape / physical obstruction" from "tape has a slightly smudged barcode".

But tapes get handled and modified, and it must have read it once to put it in that location. It's bad design to not distinguish between a bad barcode, and one that doesn't even exist.

21
1
Silver badge

surely it would of become very obvious once the robot went to retrieve the tape from the drive, and failed because 'I can't find the tape!'

The article describes exactly this except that the tapes were missed on the shelves and didn't get as far as the drives.

Even tapes that the team put on shelves by hand weren't being detected.

“The robot sometimes even tried to place other tapes in those 'empty' slots,”

14
0
Silver badge

4 DJ Smiley

"Now if they weren't reading at all, and to robot thought there was no tape, then surely it would of become very obvious once the robot went to retrieve the tape from the drive, and failed because 'I can't find the tape!'

Yes, that is pretty much the article's text. But it also talks about the chaos when a new tape needs to go into a slot already occupied.

Been there, seen the carnage.

"Terrible debugging and reporting is what caused this problem, exasperated by a customer using unsupported hardware."

Nope, wrong in every way possible, including the word exasperated.

What caused this problem was cheap, unfit for purpose labels bought to cut costs.

A situation cannot be exasperated because it does not have emotions. I can be exacerbated though. I imagine you bought the word "exasperated" cheap to save costs and now are ruing the day, much like the boob who bought the shiny label stock that were the same size and orientation and fitted the printer so what's the diff?

8oD

18
4
Silver badge

4 Lee D

"Hell, put a barcode at the BACK of the empty slot and if you read that barcode - yeah, it's fair to assume it's empty. "

So if I understand this suggestion correctly we need to print new labels (on the right sort of paper though we don't know that yet), somehow stick them in the back of each hopper in just the right place for the laser to see it - assuming we don't need to re-engineer the laser assembly to fulfill this new use-case, then rewrite the firmware so the new decision tree is implemented.

How is this easier than buying the right labels? I mean, we work with what we have, right? We can make suggestions to hardware vendors but how many have you ever seen implemented? Post-sales?

As for the tapes already loaded: The issue as I read it wasn't that the arm couldn't find the silo hopper, it was that the firmware couldn't read the barcode of a tape loaded in a hopper. I dunno, but I'm not prepared to call the author a big fat liar on the strength of the tapes already in the silo because been there, seen that. Tapes might have been labelled with old, non-shiny label stock. Might have been loaded the hard way. The article speaks of doing just that.

Sorry. Respectfully disagree with your analysis of the story.

9
3
Silver badge

Or a 20p micro-switch testing for physical presence

But a microswitch per slot isn't just n times 20p, it's also all the associated wiring and the connected logic to read each slot individually.

M.

14
0
Anonymous Coward

Re: 4 Lee D

I have had exactly the same error..

You go to the place you expect to have a tape, stop, read, no read, repeat 3 times, "no tape there, WTF".

From there, your decisions may change, but on a large tape library, probably somebody stopped the machine a took it manually, so you assume "nothing there".

Now, the logical thing would be to take a picture and figure out if there is something there, but in the 90s that was VERY expensive in processing power, etc, way beyond what a micro controller could do.

8
0
Silver badge

plus

plus the switch consumes one pin on a microcontroller, unless you start playing with diodes, and use multicircuit switches (ala keyboard).

Now, the switches are relatively unreliable, as anyone who has had to work with them can tell you... so you will put at least two of them and expect the same read.

Nah, the scanner is more reliable.

9
0

Of course it's odd, they did crap trouble shooting

Of course it's odd, the engineer did crap troubleshooting onsite. When we had problems, they always brought their own tape(s) and standard labels to check the robotics with known good stuff. Once they saw that their combo worked, they would berate you if they found out you had gone with sub-standard (theirs!) labels.

Who else remembers printing out DLT barcodes on plain paper with a laser printer since the vendor supplied ones were pure robbery. And of course the hardware at the time wasn't nearly as good and reliable.

The old exabyte EXB-120 libraries we're amazingly finicky about labels too.

5
0
Silver badge

"Or a 20p micro-switch testing for physical presence per slot"

Overcomplicated, unreliable and unnecessary

Some robots check for tapes in slots where they can't see a barcode by poking them with the same arm used to slide in tapes. If they encounter resistance to the poke, there's a tape in that slot. Other robots use an Infrared proximity sensor on the front of the picker.

Both methods work well. I have 2 robots beside me using these methods.

The really fun part is when you have a robotic barcode reader which can only read barcodes along the bottom third of the label and your barcodes were printed with the human-readable numbers on that side (I'm looking at you, Quantum!)

10
0
Silver badge

Re: 4 DJ Smiley

"What caused this problem was cheap, unfit for purpose labels bought to cut costs."

The interesting part is that when it comes to tape labels, the sticky ones are only available from 2 makers in 3 types, none of them are shiny and they're all marked as OK for laser or inkjet use (inkjets get better results)

You can use only sticky labels on LTO/SDLT/SAIT. Older ones could use paper/card inserts - which resulted in a lot of outfits making their own on too flimsy paper and having trouble cutting them out properly

0
1
Anonymous Coward

It's a bar code reader, why or how would you program or set it to read glossy labels? Frosted tape obscuring what it is reading?

I'm not understanding this one.

0
30
Silver badge

The story is ...

... the laser light back-scatter off the glossy tape confused the barcode reader.

Consider the reflectivity range of the materials that the barcode scanner at your local supermarket has no trouble reading. It ranges from from printing on silver Mylar to printing on dark brown cardboard (blue, purple, whathaveyou). The guts of the tape robot's scanner are exactly the same as your supermarket's scanner. It doesn't take a passing grade in Critical Thinking 101 to doubt the anecdote.

6
27
Anonymous Coward

The laser was being confused by the reflection off the glossy labels - so the difference between "black" and "white" was not always clear.

The "frosted" tape was a matt surface so that it reflected light in a different way to the glossy labels. That allowed the laser to differentiate the "black" and "white" areas of the underlying label reliably.

35
0
Anonymous Coward

Ahh, not had my coffee yet, read it as being the other way round. Thanks

But you do have to ask why would a third party vendor supply labels that don't work and if they were then this would have to be the first time the company had found the problem. Also, the third party vendor must have had access to create the labels so surely they tested them. Finally how much were these pieces of sticky label for someone to pirate them in the first place? If I was getting a robotic tape solution I wouldn't be paying extra for labels, I would expect them to be included in the purchase price.

2
2
Anonymous Coward

Re: The story is ...

"It doesn't take a passing grade in Critical Thinking 101 to doubt the anecdote."

It is not unusual for bar codes on products to take several swipes before they register. The assistant knows there is a bar code - so persists by presenting it repeatedly. The story's tape machine apparently does not have the benefit of some other way of sensing that there is a tape available.

My local supermarket prints "last day" reduced prices on yellow sticky labels. Quite often the checkout scanner cannot read them. They appear crisp and legible - and the assistant has no problem reading the printed number below the bar code when they then manually transcribe them.

30
0

Re: The story is ...

..or possibly that barcode scanning tech has improved in the last twenty years?

41
0

Re: The story is ...

My local supermarket prints "last day" reduced prices on yellow sticky labels. Quite often the checkout scanner cannot read them. They appear crisp and legible - and the assistant has no problem reading the printed number below the bar code when they then manually transcribe them.

One major UK supermarket had a problem in the past. They decided to encode the product price in the reduction barcode as well as the product code. Unfortunately they chose a symbology - a method of encoding the data - that their till system did not actually support. So the assistants at the tills had to peel off the new label stuck over the old label, scan the regular product code, hit the reduction button and type in the price.

The new till software was sufficiently delayed that the reduction labels were optimised - the centre horizontal strip of the label, carrying the reduction barcode, was deliberately made less sticky and gained perforations, so that the operator could more easily tear it off to get to the product code underneath!

They did eventually get tills able to read the reduction barcode directly, but it took several years, as I recall.

17
0
Silver badge

Re: The story is ...

"The assistant knows there is a bar code - so persists by presenting it repeatedly. "

And is utterly convinced that a barcode has to be in motion in order to be read, just like a mag stripe.

22
2
Anonymous Coward

Re: convinced that a barcode has to be in motion

Surely all they are doing is moving it across the approximate field of view so as to give the reader a good change of spotting it..?

21
0
Silver badge

Re: "Has to be in motion"...

Well, it improves your odds. Chance you put it in EXACTLY the right place, angle and unfold all those creases and sit there for 60 seconds thinking it will read?

No thanks, I'll swipe and give it a MASSIVE scanning area to hopefully succeed in. :D

(I play a lot of computer games, sweeping motion gives you a better chance of a successful attempt in many things, as you then only have to get timing/angle correct, instead of trying to hit a bulls eye with a shaky cam from 50 feet away)

14
0

Re: The story is ...

I think you'll find the scanner in the supermarket is way bulkier and more expensive than the scanner in the tape library.

The supermarket scanners have multiple beams coming in from multiple angles, with both a horizontal and vertical scanner.

17
1
Silver badge

Re: The story is ...

I take it you've never seen a hand-held barcode scanner, bigiain. Are you old enough to remember the CueCat?

6
4
Silver badge

Re: The story is ...

It is not unusual for bar codes on products to take several swipes before they register. The assistant knows there is a bar code - so persists by presenting it repeatedly.

And at different angles, usually. The robot doesn't have that option.

10
0
Silver badge

Re: The story is ...

The robot doesn't need that option; in fact the option isn't actually an option. The items (tapes in this case) are always presented in exactly the same way.

The human doesn't need the option, either. In fact, if they'd just stop the bloody item from moving for a split second, the machine would read it on the first pass.

3
6
Silver badge

Re: The story is ...

I think you'll find the scanner in the supermarket is way bulkier and more expensive than the scanner in the tape library.

Bulkier? Yes, but only because the handheld ones need a grip and a battery, and the till-mounted ones contain both rotating and stationary mirrors to move a laser beam over the area where a barcode might show up. More expensive? See previous remark. The library reader needs to deal with only one size of label in one position (respective to the orientation of the tape cartridge), so can be much simpler mechanically/optically. The optoelectronics and the software are not that different.

One customer I visited occasionally had a large StorageTek carousel installed that read the tape barcodes using video cameras (this was the second half of the 1980's), and for shits and giggles they had a set of monitors hooked up in parallel with the decoding units. Fascinating, especially when the robot arms had to pass tapes between them (they had two storage carousels but only one had drives installed, the other was just for shelf space expansion).

2
0
Silver badge

Re: The story is ...

"And is utterly convinced that a barcode has to be in motion in order to be read, just like a mag stripe."

These days it's unlikely, but not that long ago barcode readers used a photo-diode to pick up the stripes, so the reader had to be moved relative to the barcode in a steady motion.

Not to mention that not many people understand how a modern barcode reader works, and those people that do understand probably don't work on a checkout.

14
0
Bronze badge

Re: The story is ...

I remember printing lots of them when I worked in a supermarket - the label printers often have worn or damaged heating heads (the barcodes are actually "burned" into the label coating, the same as receipts) and feed rollers, which make the lines the incorrect width.

The problem of course is that the width of the lines is what determines what number they are, so by having these minute errors the barcode can easily be unreadable yet still look OK.

4
0

You're assuming that the third party vendor was selling the labels for the specific purpose of labelling tapes for the specified machine. Much more likely they were generic labels that happened to be the correct dimension and compatible with the customer's printer. The third-party vendor is unlikely to have the least inkling about the tape library, much less tested its labels on it.

12
0
Silver badge

Re: The story is ...

"Consider the reflectivity range of the materials that the barcode scanner at your local supermarket has no trouble reading."

I see barcode read failures due to media problems every time I'm on line to buy, whether at a supermarket, my local hobby shop or Home Despot.

Sorry Jake, judging by my own real-world everyday experience your initial premise is flawed.

"The guts of the tape robot's scanner are exactly the same as your supermarket's scanner. It doesn't take a passing grade in Critical Thinking 101 to doubt the anecdote."

Guts: not so much really. I think what's needed is more actual engineering knowledge of supermarket scanner works and robot scanner works. Critical thinking needs to be based on hard facts, not "I guess this is true". The robot arm could never support the mass of a supermarket scanner, nor does it need to.

As I see it, this is a case of two very different machines that happen to use the same scientific principle to get their very different jobs done.Comparisons should be done with care.

11
2
Silver badge

4 coffee-less AC

"But you do have to ask why would a third party vendor supply labels that don't work and if they were then this would have to be the first time the company had found the problem. "

I've seen the third party vendor thing first hand, and for me it is easier to believe that a purchaser would buy generic label stock that just happens to be incompatible with the reader than that a vendor would manufacture paper tape that was so shiny the tape-reader's pinch-roller couldn't grip it, yet I saw that very thing happen in 1979 to the embarrassment of the ops manager still fresh from his back-patting for cost-cutting. Cost a pretty penny to rush the Right Stuff to the plant so the unionised and militant workforce could be paid on time.

And why couldn't this be the first recorded incident for that robot vendor? There has to be a first time for each problem.

11
0
Anonymous Coward

Re: The story is ...

No much.

The 2D was fine, and a modern 2D one is just a tad better, mainly faster

the new fangled thing is 3d ones... or should I say cameras?

0
0
Silver badge

Glossy labels vs matt labels - and the "frosted tape" is 3M's magic mending tape - it has a matt top surface.

I've resorted to this fix when dealing with optical problems and overly reflective surfaces in the past, along with using Tippex (liquid paper) to narrow the viewing window on an optical sensor so that the chopper wheel it was shining through could work - the manufacturer one had a narrow window compared to the replacement unit, slots in the chopper were wide enough to allow light pulses through that were long/smeary enough to screw up a centrifuge as it hit 15,000 RPM on the way up to 45,000 - resulting in emergency shutdown being invoked. You really _don't_ want that to happen often. Closing up the optical viewing window solved the problem and got nice clean pulses at silly high speeds

6
0
Silver badge

Re: The story is ...

Older barcode readers in libraries used a scanned laser to pick up the light/dark areas.

Newer readers effectively use a camera and OCR technology because it's much cheaper and eliminates the 'unreliable' mechanical deflection bits.

Guess which one works better?

4
0
Bronze badge

Re: barcode in motion

I use the self-scan gadgets in our local supermarket. I have found that shiny bar code labels should be read at an angle, so that the specular reflection doesn't blind the scanner to the diffuse reflection from the bar code.

I have also found that some bar codes can only be read in motion. It shouldn't be true but it is a repeatable observation.

7
0
Silver badge

Re: The story is ...

I suspect you mean 1D and 2D barcodes. 1D barcodes are the more traditional types that are a series of stripes and you'd find on books and supermarket products. 2D barcodes are often square (but don't have to be), for example QR codes and are often rotationally symetical but don't have to be. While technically there are 3D barcode scanners these are often just video cameras because they are usually 2D barcodes with the additional "dimension" of colour or, occasionally, relative shading, as well.

3
0

Re: The story is ...

@ Mike Dimmick

"They decided to encode the product price in the reduction barcode as well as the product code."

Not a good idea for another reason. It would have taken me about 30 seconds to figure out how to get big discounts on stuff from that store. Print my own bar-code labels!

I write code for printing bar-codes as part of my job. I can actually read several symbologies just by looking at them (sad isn't it).

7
0
Silver badge
Joke

Re: "Has to be in motion"...

"sweeping motion gives you a better chance of a successful attempt in many things, as you then only have to get timing/angle correct, instead of trying to hit a bulls eye"

You're not married, are you.

3
0
Silver badge

Re: The story is ...

"It would have taken me about 30 seconds to figure out how to get big discounts on stuff from that store. Print my own bar-code labels!"

Which is exactly what happened and resulted in a number of prosecutions for fraud.

0
0
Silver badge

Re: The story is ...

<e."It would have taken me about 30 seconds to figure out how to get big discounts on stuff from that store. Print my own bar-code labels!"

Which is exactly what happened and resulted in a number of prosecutions for fraud.</em>

I can vaguely remember, back in the days when price labels were human readable and the checkout operator had to key them in, being presented with a case of alleged attempted fraud. IIRC it was alleged a store detective had observed a customer trying to pick the top label off a can with multiple labels, presumably on the basis that the price had been marked up and the first label would have been cheaper. Carefully peeling off the top label I found the original price was higher. I never got called to court on that one.

0
0

Re: The story is ...

Certain supermarket round here still does it to this day. Last few digits (excluding the final checksum) of the barcode are the new price in pence.

Of course, EAN checksums being as they are, will treat 0100 the same as 0001... Now, when you can just key in the barcode at the self-scan, it's easy to be a little dyslexic, and no need to print your own labels at all.

0
0
Silver badge

I first heard this story back in the early 1990s.

It's always "I heard it from a friend", or a friend of a friend, never first-hand.

I'm not saying it never happened. I am saying it's anecdotal at best.

5
2
Anonymous Coward

Re: I first heard this story back in the early 1990s.

So it could be an urban myth created by vendors to ensure customers always buy official stock.

That ties in with the cdr disks that could kill your writer urban myth as well back in the days of 2x writing.

I think you may be onto something.

18
0
Anonymous Coward

Re: I first heard this story back in the early 1990s.

There are always constraints in technology that might not be apparent to the uninitiated observer.

After the Iraq War some enterprising local metal workers started to make satellite dishes. As they didn't understand the need for a mathematically precise shape - their products looked right but didn't work.

There are many different types of glue. They all usually feel sticky to the touch. However each type is formulated for a limited set of materials and conditions. Use the wrong one and it may not stick the pieces together at all. Worse still - it may appear to work then fails unexpectedly under some force or environmental stress. IIRC one glue worked by creating a vacuum. It was useless for spacecraft applications.

Different types of manufactured iron have different strength and flexibility properties. You trust your supplier and engineers to know which is which.

My sculpture tutor put a house brick in the pottery kiln to support a student's clay model while it had its second firing. The brick melted into a glassy mess over everything at that temperature. Josiah Wedgwood invented a system of clay cones that deformed at different temperatures so that the temperature in coal-fired bottle kilns could be more accurately judged.

34
0
Anonymous Coward

Re: I first heard this story back in the early 1990s.

"Worse still - it may appear to work then fails unexpectedly under some force or environmental stress."

My friend's father bought an old farmhouse to renovate as the family home. In the middle of nowhere, he decided to have a large swimming pool - partly justified as a fire precaution water reservoir.

The large fibreglass pool was delivered and positioned into the large hole. He then connected up the large diameter grey plastic pipes to circulate the water through a small pump house. Great fun when it was first filled - until someone blocked the bottom outlet for a few seconds with their foot. At which point there was a dull thud and water exploded from the pump house. Two pipes had disconnected themselves. This happened a few times before the penny dropped.

The grey plastic pipes had been glued together - but it wasn't the specific Osma-weld glue that dissolved the plastic to make a strong "welded" joint

21
0
Bronze badge

Re: I first heard this story back in the early 1990s.

In a similar vein to Wedgewood to fit (metal) tyres on train wheels (for they do have them) they heat the tyres to a set temperature to expand them relative to the actual wheel. In the old days the foreman used to spit on the tyre to see if his spit balled up - if it did it was hot enough, if not it needed more heat.

These days, they have wax pens which melt and ball up at different temperatures - they merely mark the tyre and watch the wax.

10
0
Silver badge

Re: I first heard this story back in the early 1990s.

The grey plastic pipes had been glued together

In the days when I was a self-employed electrician I once did a job at a house where the tradesman before me had been a plumber, installing an en-suite into the master bedroom. Said bedroom was at the front of the house while the plumbing was at the back, so he installed a Saniflo macerator / pump. (note to readers, if there's any way at all that you can avoid a Saniflo, please, please, please do so, the things are nothing but trouble. Horrid, messy, smelly trouble).

Anyway, the twit had presumably not read the instructions and instead of using solvent-weld pipe, he used push-fit. A couple of flushes later and the elbow where the horizontal pipe turned vertical, fortunately outside the house, popped right off and the contents of the Saniflo were sprayed all over the rear yard.

M.

12
0
Anonymous Coward

Re: I first heard this story back in the early 1990s.

Could be true even if it is now lore and sung by a technobard.

3
0

Page:

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Forums

Biting the hand that feeds IT © 1998–2017