wheelbuilding question

"jim beam" wrote
Peter Cole wrote:

It's also a misquote. The technique Jobst describes is qualified to work
only
with lightweight (430 g or less), 36 spoke rims.

has he ever said that? i don't have his book in front of me; i can't
recall such a qualification, but i've seen jonsey's kind of statement
here many times.

I wrote that with the book in front of me, it's stated very clearly.


i've read the radial loading argument of high tension [and i know the
difference between strength & stiffness!].

OK, it's really the only argument Jobst makes.

regarding lateral loading,
this adds to the spoke pre tension on one side and subtracts from the
other. radial loads subtract only. if a spoke has a yield strength of
say 300kg, preloading it to 200kg only gives 100kg of lateral load
before yield. if the spokes have 100kg preload, it means they can take
twice as much lateral.

This lateral load argument doesn't really have any practical considerations.

i want to be clear - i'm not advocating "too low tension" - i'm saying
that tension needs to be within spec, not this nebulous unscientific
concept of "as high as the rim can bear". agreed, too low tension can
lead to nipple unscrewing, but i guess that's why spoke manufacturers
sell threadlock & self-locking nipples.

The more serious consequence is that the wheel can become unstable and buckle.

Peter Cole wrote:
"jim beam" wrote

Peter Cole wrote:

It's also a misquote. The technique Jobst describes is qualified to work

only

with lightweight (430 g or less), 36 spoke rims.

has he ever said that? i don't have his book in front of me; i can't
recall such a qualification, but i've seen jonsey's kind of statement
here many times.


I wrote that with the book in front of me, it's stated very clearly.



i've read the radial loading argument of high tension [and i know the
difference between strength & stiffness!].


OK, it's really the only argument Jobst makes.


regarding lateral loading,
this adds to the spoke pre tension on one side and subtracts from the
other. radial loads subtract only. if a spoke has a yield strength of
say 300kg, preloading it to 200kg only gives 100kg of lateral load
before yield. if the spokes have 100kg preload, it means they can take
twice as much lateral.


This lateral load argument doesn't really have any practical considerations.

the only consideration is that if the spoke tension is too high, the rim
is much more prone to taco. the only rim i've ever tacoed [not caused
by a car] was first ride immediately after i'd just built a wheel with
"tension as high as the rim can bear". downhill, bump, pretzel, long
walk home. the bump wasn't even enough to flat the tire.



i want to be clear - i'm not advocating "too low tension" - i'm saying
that tension needs to be within spec, not this nebulous unscientific
concept of "as high as the rim can bear". agreed, too low tension can
lead to nipple unscrewing, but i guess that's why spoke manufacturers
sell threadlock & self-locking nipples.


The more serious consequence is that the wheel can become unstable and buckle.

i would have thought that, but i rode mtb for the best part of a year
with a guy whose rear wheel was always making an irritating grinding
noise. eventually, i pursuaded him to let me take it home for
examination. i was shocked to find that all the spokes were so loose,
they were almost slack - the grinding was the spoke crossings moving
against each other as the hub "sank" relative to center on load - flat
spots on each spoke at that point. and the damnedest thing of all was
that this wheel was as true as i've ever seen!

he rides real hard & loves the fast bumpy downhill stuff, so there was
/plenty/ of opportunity for his wheel to have failed. this one instance
is not enough to demonstrate proof, but it's worth further investigation.

Peter Cole wrote:
"jim beam" wrote

Peter Cole wrote:

It's also a misquote. The technique Jobst describes is qualified to work

only

with lightweight (430 g or less), 36 spoke rims.

has he ever said that? i don't have his book in front of me; i can't
recall such a qualification, but i've seen jonsey's kind of statement
here many times.


I wrote that with the book in front of me, it's stated very clearly.



i've read the radial loading argument of high tension [and i know the
difference between strength & stiffness!].


OK, it's really the only argument Jobst makes.


regarding lateral loading,
this adds to the spoke pre tension on one side and subtracts from the
other. radial loads subtract only. if a spoke has a yield strength of
say 300kg, preloading it to 200kg only gives 100kg of lateral load
before yield. if the spokes have 100kg preload, it means they can take
twice as much lateral.


This lateral load argument doesn't really have any practical considerations.

the only consideration is that if the spoke tension is too high, the rim
is much more prone to taco. the only rim i've ever tacoed [not caused
by a car] was first ride immediately after i'd just built a wheel with
"tension as high as the rim can bear". downhill, bump, pretzel, long
walk home. the bump wasn't even enough to flat the tire.



i want to be clear - i'm not advocating "too low tension" - i'm saying
that tension needs to be within spec, not this nebulous unscientific
concept of "as high as the rim can bear". agreed, too low tension can
lead to nipple unscrewing, but i guess that's why spoke manufacturers
sell threadlock & self-locking nipples.


The more serious consequence is that the wheel can become unstable and buckle.

i would have thought that, but i rode mtb for the best part of a year
with a guy whose rear wheel was always making an irritating grinding
noise. eventually, i pursuaded him to let me take it home for
examination. i was shocked to find that all the spokes were so loose,
they were almost slack - the grinding was the spoke crossings moving
against each other as the hub "sank" relative to center on load - flat
spots on each spoke at that point. and the damnedest thing of all was
that this wheel was as true as i've ever seen!

he rides real hard & loves the fast bumpy downhill stuff, so there was
/plenty/ of opportunity for his wheel to have failed. this one instance
is not enough to demonstrate proof, but it's worth further investigation.

jim beam wrote:
Mark McMaster wrote:

jim beam wrote:

in addition, as can be seen in damon rinard's experiments, increasing
spoke tension makes absolutely no difference to lateral strength. see:

http://www.sheldonbrown.com/rinard/wheel/tension.gif

original page:
http://www.sheldonbrown.com/rinard/wheel/index.htm



Rinard's wheel test did not test (lateral) strength, as he stated in
his first paragraph: "I am measuring stiffness, not strength." His
test of measuring lateral stiffness at varying static spoke tension
mainly serves to confirm Hooke's Law.


hookes law merely states that deformation is directly proportional to
load below yield - the definition of elastic deformation. it's no
predictor of yield or modulus, both of which are measures of "strength".
by that same argument, increasing tension does not increase strength
just the same as it does not increase stiffness.

No argument. Rinard's test just shows that the existence of
a static pre-load doesn't change the wheel stiffness.



However, if you take a closer look at the graph, you'll notice that
the deflection increases dramatically when he backs the tension off
below a certain threshold. This increase in deflection shows a wheel
that is more likely to fail under load.


that's an assumption, not a fact. the deflection increases, for slack
spokes /because/ they're slack. if you're towing a car with a slack
rope, the distance between the two cars will increase until the rope
becomes taught. then the distance between the two cars is essentially
fixed and subject only to minor stretching of the rope - many orders of
magnitude less that slack take-up.

Poor analogy - as the the wheel is continued to be loaded
with slack spokes, there is no point when the spokes
suddenly start supporting the wheel.

Typical shallow section rims are not strong enough on their
own to support the momentary high loads often experienced
when cycling. The rim requires the support of the spokes.
Slackened spokes can no longer contribute support to the
spokes, and the rim must bear the load. Since the
unsupported rim can not bear as much load without damage, a
wheel low static spoke tension is more likely to be damaged
under a high momentary load.


Although Rinard is using a fixed load, it can be inferred from this
data that increased spoke tension can increase the strength of a wheel.


"inferred" how? the material does not change - this material still has
to obey hookes law until it yields. increasing load merely makes it
bend further. if you think about it, pre-tension serves to reduce the
load capacity of a component not increase it.

The graphs shows that at at some minimum tension, some of
the spokes slacken and no longer can contribute to
supporting the rim. It can be inferred that at a higher
static spoke tension, the spokes will not slacken until a
higher applied load. Since a wheel with all spokes still
under tension is a stronger wheel, than a wheel with higher
static spoke tension can support a higher load.


Indeed, you're supposition that "increasing spoke tension makes
absolutely no difference in lateral strength" is directly contradicted
by Rinard's conclusion from his test. From the web page referenced
above:

"A wheel whose spokes become slack while riding is a weak wheel,
because slack spokes cannot support the rim. This can be avoided to a
large extent by building wheels with tighter spokes. If spokes are
tighter initially, then the sudden increase in flexibility shown in
data points 9 and 10 is less likely to occur in use because a tighter
wheel can bear a higher load before spokes become slack."


there's no contradiction. the left part of the graph is essentially a
flat line. leftwards is increasing tension. once you're in the flat
line region, increasing tension is not increasing lateral stiffness.

There's more to it than simply stiffness. The sharp
decrease in stiffness occurs at the point when some of the
spokes no longer contribute to supporting the wheel. You
claim seem to claim spoke slackening make no difference in
wheel strength - Damon Rinard claims it does (and I agree
with him).

Mark McMaster

jim beam wrote:
Mark McMaster wrote:

jim beam wrote:

in addition, as can be seen in damon rinard's experiments, increasing
spoke tension makes absolutely no difference to lateral strength. see:

http://www.sheldonbrown.com/rinard/wheel/tension.gif

original page:
http://www.sheldonbrown.com/rinard/wheel/index.htm



Rinard's wheel test did not test (lateral) strength, as he stated in
his first paragraph: "I am measuring stiffness, not strength." His
test of measuring lateral stiffness at varying static spoke tension
mainly serves to confirm Hooke's Law.


hookes law merely states that deformation is directly proportional to
load below yield - the definition of elastic deformation. it's no
predictor of yield or modulus, both of which are measures of "strength".
by that same argument, increasing tension does not increase strength
just the same as it does not increase stiffness.

No argument. Rinard's test just shows that the existence of
a static pre-load doesn't change the wheel stiffness.



However, if you take a closer look at the graph, you'll notice that
the deflection increases dramatically when he backs the tension off
below a certain threshold. This increase in deflection shows a wheel
that is more likely to fail under load.


that's an assumption, not a fact. the deflection increases, for slack
spokes /because/ they're slack. if you're towing a car with a slack
rope, the distance between the two cars will increase until the rope
becomes taught. then the distance between the two cars is essentially
fixed and subject only to minor stretching of the rope - many orders of
magnitude less that slack take-up.

Poor analogy - as the the wheel is continued to be loaded
with slack spokes, there is no point when the spokes
suddenly start supporting the wheel.

Typical shallow section rims are not strong enough on their
own to support the momentary high loads often experienced
when cycling. The rim requires the support of the spokes.
Slackened spokes can no longer contribute support to the
spokes, and the rim must bear the load. Since the
unsupported rim can not bear as much load without damage, a
wheel low static spoke tension is more likely to be damaged
under a high momentary load.


Although Rinard is using a fixed load, it can be inferred from this
data that increased spoke tension can increase the strength of a wheel.


"inferred" how? the material does not change - this material still has
to obey hookes law until it yields. increasing load merely makes it
bend further. if you think about it, pre-tension serves to reduce the
load capacity of a component not increase it.

The graphs shows that at at some minimum tension, some of
the spokes slacken and no longer can contribute to
supporting the rim. It can be inferred that at a higher
static spoke tension, the spokes will not slacken until a
higher applied load. Since a wheel with all spokes still
under tension is a stronger wheel, than a wheel with higher
static spoke tension can support a higher load.


Indeed, you're supposition that "increasing spoke tension makes
absolutely no difference in lateral strength" is directly contradicted
by Rinard's conclusion from his test. From the web page referenced
above:

"A wheel whose spokes become slack while riding is a weak wheel,
because slack spokes cannot support the rim. This can be avoided to a
large extent by building wheels with tighter spokes. If spokes are
tighter initially, then the sudden increase in flexibility shown in
data points 9 and 10 is less likely to occur in use because a tighter
wheel can bear a higher load before spokes become slack."


there's no contradiction. the left part of the graph is essentially a
flat line. leftwards is increasing tension. once you're in the flat
line region, increasing tension is not increasing lateral stiffness.

There's more to it than simply stiffness. The sharp
decrease in stiffness occurs at the point when some of the
spokes no longer contribute to supporting the wheel. You
claim seem to claim spoke slackening make no difference in
wheel strength - Damon Rinard claims it does (and I agree
with him).

Mark McMaster

Jonesy wrote:
jim beam wrote in message ...

Peter Cole wrote:

"jim beam" wrote


Jonesy wrote:


That's true - the more tension you can bring to bear (before
destroying the rim) the better. Read "The Bicycle Wheel" by Jobst
Brandt.

i know that "high tension" recommendation is "in the book" and often
repeated here, but it's a fundamentally flawed piece of advice.


It's also a misquote. The technique Jobst describes is qualified to work only
with lightweight (430 g or less), 36 spoke rims.

has he ever said that? i don't have his book in front of me; i can't
recall such a qualification, but i've seen jonsey's kind of statement
here many times.


It was a gross generalization. It was not meant to be definitive,
which is why I suggested further reading.

[snip]


i want to be clear - i'm not advocating "too low tension" - i'm saying
that tension needs to be within spec, not this nebulous unscientific
concept of "as high as the rim can bear".


If one was pedantic, they would say that "within spec" is "as high as
the rim can bear." But I was not being specific.


agreed, too low tension can
lead to nipple unscrewing, but i guess that's why spoke manufacturers
sell threadlock & self-locking nipples.


It seems to me that, on a properly-tensioned spoke with no wind-up,
threadlocking materials are completely unnecessary.

My apologies for not being more specific. After being berated for
being long-winded, I'm trying to cut down. I guess being specific has
some cost.

My intent was to steer the OP toward some reading materials. Mr.
Beam, if you would be so kind, why don't you assemble the tension
specs for rims such that you may in the future warn folks about
over-tensioning (past rim manufacturer's specs). That would be a
fantastic resource, and easier than arguing the theoretical materials
science behind your commentary - something that might not help a poor
wheelbuilder like myself.

funny you should say that!... thing is, i did start something like that
a while back, but i'm not "in the trade" so don't have the same easy
access to this info that a bike shop would. maybe appropriate tension
data for each can be added to damon's spoke calculator spreadsheet along
with rim erd's?

community project?

Jonesy wrote:
jim beam wrote in message ...

Peter Cole wrote:

"jim beam" wrote


Jonesy wrote:


That's true - the more tension you can bring to bear (before
destroying the rim) the better. Read "The Bicycle Wheel" by Jobst
Brandt.

i know that "high tension" recommendation is "in the book" and often
repeated here, but it's a fundamentally flawed piece of advice.


It's also a misquote. The technique Jobst describes is qualified to work only
with lightweight (430 g or less), 36 spoke rims.

has he ever said that? i don't have his book in front of me; i can't
recall such a qualification, but i've seen jonsey's kind of statement
here many times.


It was a gross generalization. It was not meant to be definitive,
which is why I suggested further reading.

[snip]


i want to be clear - i'm not advocating "too low tension" - i'm saying
that tension needs to be within spec, not this nebulous unscientific
concept of "as high as the rim can bear".


If one was pedantic, they would say that "within spec" is "as high as
the rim can bear." But I was not being specific.


agreed, too low tension can
lead to nipple unscrewing, but i guess that's why spoke manufacturers
sell threadlock & self-locking nipples.


It seems to me that, on a properly-tensioned spoke with no wind-up,
threadlocking materials are completely unnecessary.

My apologies for not being more specific. After being berated for
being long-winded, I'm trying to cut down. I guess being specific has
some cost.

My intent was to steer the OP toward some reading materials. Mr.
Beam, if you would be so kind, why don't you assemble the tension
specs for rims such that you may in the future warn folks about
over-tensioning (past rim manufacturer's specs). That would be a
fantastic resource, and easier than arguing the theoretical materials
science behind your commentary - something that might not help a poor
wheelbuilder like myself.

funny you should say that!... thing is, i did start something like that
a while back, but i'm not "in the trade" so don't have the same easy
access to this info that a bike shop would. maybe appropriate tension
data for each can be added to damon's spoke calculator spreadsheet along
with rim erd's?

community project?

Mark McMaster wrote:
jim beam wrote:

Mark McMaster wrote:

jim beam wrote:

in addition, as can be seen in damon rinard's experiments,
increasing spoke tension makes absolutely no difference to lateral
strength. see:

http://www.sheldonbrown.com/rinard/wheel/tension.gif

original page:
http://www.sheldonbrown.com/rinard/wheel/index.htm




Rinard's wheel test did not test (lateral) strength, as he stated in
his first paragraph: "I am measuring stiffness, not strength." His
test of measuring lateral stiffness at varying static spoke tension
mainly serves to confirm Hooke's Law.



hookes law merely states that deformation is directly proportional to
load below yield - the definition of elastic deformation. it's no
predictor of yield or modulus, both of which are measures of
"strength". by that same argument, increasing tension does not
increase strength just the same as it does not increase stiffness.


No argument. Rinard's test just shows that the existence of a static
pre-load doesn't change the wheel stiffness.



However, if you take a closer look at the graph, you'll notice that
the deflection increases dramatically when he backs the tension off
below a certain threshold. This increase in deflection shows a wheel
that is more likely to fail under load.



that's an assumption, not a fact. the deflection increases, for slack
spokes /because/ they're slack. if you're towing a car with a slack
rope, the distance between the two cars will increase until the rope
becomes taught. then the distance between the two cars is essentially
fixed and subject only to minor stretching of the rope - many orders
of magnitude less that slack take-up.


Poor analogy - as the the wheel is continued to be loaded with slack
spokes, there is no point when the spokes suddenly start supporting the
wheel.

Typical shallow section rims are not strong enough on their own to
support the momentary high loads often experienced when cycling. The
rim requires the support of the spokes. Slackened spokes can no longer
contribute support to the spokes, and the rim must bear the load. Since
the unsupported rim can not bear as much load without damage, a wheel
low static spoke tension is more likely to be damaged under a high
momentary load.

check out my friend's slack spoked mtb wheel experience - with peter
cole in this same thread. heavy mtb use did _not_ damage a loose spoked
wheel.



Although Rinard is using a fixed load, it can be inferred from this
data that increased spoke tension can increase the strength of a wheel.



"inferred" how? the material does not change - this material still
has to obey hookes law until it yields. increasing load merely makes
it bend further. if you think about it, pre-tension serves to reduce
the load capacity of a component not increase it.


The graphs shows that at at some minimum tension, some of the spokes
slacken and no longer can contribute to supporting the rim. It can be
inferred that at a higher static spoke tension, the spokes will not
slacken until a higher applied load. Since a wheel with all spokes
still under tension is a stronger wheel, than a wheel with higher static
spoke tension can support a higher load.

how? wheels do not experience purely radial load. and with respect,
i'm not convinced you understand the graph. the high tension part is
essentially a flat line - there's no evidence of increasing tension
affecting wheel stiffness whatsoever, which is what you would expect
being as the material has not changed. is a compressed spring stiffer
than an uncompressed spring?

and stiffness is commonly regarded as a measure of perceived load
capacity, e.g. a wheel built with revos [which are very elastic] is
considered unsuitable for loaded touring.



Indeed, you're supposition that "increasing spoke tension makes
absolutely no difference in lateral strength" is directly
contradicted by Rinard's conclusion from his test. From the web page
referenced above:

"A wheel whose spokes become slack while riding is a weak wheel,
because slack spokes cannot support the rim. This can be avoided to a
large extent by building wheels with tighter spokes. If spokes are
tighter initially, then the sudden increase in flexibility shown in
data points 9 and 10 is less likely to occur in use because a tighter
wheel can bear a higher load before spokes become slack."



there's no contradiction. the left part of the graph is essentially a
flat line. leftwards is increasing tension. once you're in the flat
line region, increasing tension is not increasing lateral stiffness.


There's more to it than simply stiffness. The sharp decrease in
stiffness occurs at the point when some of the spokes no longer
contribute to supporting the wheel. You claim seem to claim spoke
slackening make no difference in wheel strength - Damon Rinard claims it
does (and I agree with him).

it makes no difference to lateral deflection while the spokes still have
tension - the flat line part of the graph!!! it's only when the spokes
are slack that any difference in lateral deflection is observed - just
like the tow rope analogy. that's why the graph has two distinct regions.

Mark McMaster wrote:
jim beam wrote:

Mark McMaster wrote:

jim beam wrote:

in addition, as can be seen in damon rinard's experiments,
increasing spoke tension makes absolutely no difference to lateral
strength. see:

http://www.sheldonbrown.com/rinard/wheel/tension.gif

original page:
http://www.sheldonbrown.com/rinard/wheel/index.htm




Rinard's wheel test did not test (lateral) strength, as he stated in
his first paragraph: "I am measuring stiffness, not strength." His
test of measuring lateral stiffness at varying static spoke tension
mainly serves to confirm Hooke's Law.



hookes law merely states that deformation is directly proportional to
load below yield - the definition of elastic deformation. it's no
predictor of yield or modulus, both of which are measures of
"strength". by that same argument, increasing tension does not
increase strength just the same as it does not increase stiffness.


No argument. Rinard's test just shows that the existence of a static
pre-load doesn't change the wheel stiffness.



However, if you take a closer look at the graph, you'll notice that
the deflection increases dramatically when he backs the tension off
below a certain threshold. This increase in deflection shows a wheel
that is more likely to fail under load.



that's an assumption, not a fact. the deflection increases, for slack
spokes /because/ they're slack. if you're towing a car with a slack
rope, the distance between the two cars will increase until the rope
becomes taught. then the distance between the two cars is essentially
fixed and subject only to minor stretching of the rope - many orders
of magnitude less that slack take-up.


Poor analogy - as the the wheel is continued to be loaded with slack
spokes, there is no point when the spokes suddenly start supporting the
wheel.

Typical shallow section rims are not strong enough on their own to
support the momentary high loads often experienced when cycling. The
rim requires the support of the spokes. Slackened spokes can no longer
contribute support to the spokes, and the rim must bear the load. Since
the unsupported rim can not bear as much load without damage, a wheel
low static spoke tension is more likely to be damaged under a high
momentary load.

check out my friend's slack spoked mtb wheel experience - with peter
cole in this same thread. heavy mtb use did _not_ damage a loose spoked
wheel.



Although Rinard is using a fixed load, it can be inferred from this
data that increased spoke tension can increase the strength of a wheel.



"inferred" how? the material does not change - this material still
has to obey hookes law until it yields. increasing load merely makes
it bend further. if you think about it, pre-tension serves to reduce
the load capacity of a component not increase it.


The graphs shows that at at some minimum tension, some of the spokes
slacken and no longer can contribute to supporting the rim. It can be
inferred that at a higher static spoke tension, the spokes will not
slacken until a higher applied load. Since a wheel with all spokes
still under tension is a stronger wheel, than a wheel with higher static
spoke tension can support a higher load.

how? wheels do not experience purely radial load. and with respect,
i'm not convinced you understand the graph. the high tension part is
essentially a flat line - there's no evidence of increasing tension
affecting wheel stiffness whatsoever, which is what you would expect
being as the material has not changed. is a compressed spring stiffer
than an uncompressed spring?

and stiffness is commonly regarded as a measure of perceived load
capacity, e.g. a wheel built with revos [which are very elastic] is
considered unsuitable for loaded touring.



Indeed, you're supposition that "increasing spoke tension makes
absolutely no difference in lateral strength" is directly
contradicted by Rinard's conclusion from his test. From the web page
referenced above:

"A wheel whose spokes become slack while riding is a weak wheel,
because slack spokes cannot support the rim. This can be avoided to a
large extent by building wheels with tighter spokes. If spokes are
tighter initially, then the sudden increase in flexibility shown in
data points 9 and 10 is less likely to occur in use because a tighter
wheel can bear a higher load before spokes become slack."



there's no contradiction. the left part of the graph is essentially a
flat line. leftwards is increasing tension. once you're in the flat
line region, increasing tension is not increasing lateral stiffness.


There's more to it than simply stiffness. The sharp decrease in
stiffness occurs at the point when some of the spokes no longer
contribute to supporting the wheel. You claim seem to claim spoke
slackening make no difference in wheel strength - Damon Rinard claims it
does (and I agree with him).

it makes no difference to lateral deflection while the spokes still have
tension - the flat line part of the graph!!! it's only when the spokes
are slack that any difference in lateral deflection is observed - just
like the tow rope analogy. that's why the graph has two distinct regions.

jim beam wrote:

Jonesy wrote:
My intent was to steer the OP toward some reading materials. Mr.
Beam, if you would be so kind, why don't you assemble the tension
specs for rims such that you may in the future warn folks about
over-tensioning (past rim manufacturer's specs). That would be a
fantastic resource, and easier than arguing the theoretical materials
science behind your commentary - something that might not help a poor
wheelbuilder like myself.

funny you should say that!... thing is, i did start something like that
a while back, but i'm not "in the trade" so don't have the same easy
access to this info that a bike shop would. maybe appropriate tension
data for each can be added to damon's spoke calculator spreadsheet along
with rim erd's?

community project?

I think it would be a good idea, whether or not the theory of "building
as tightly as the rim can stand it" is correct. In fact, I think it would
be interesting to compare the tension arrived at by the latter method to
the "official spec".

--
Benjamin Lewis

Seeing is deceiving. It's eating that's believing.
-- James Thurber