By what the cable is doing at that point - passing through, or stopping.

A suspension clamp sits where the cable continues in both directions. It carries the cable's weight into the pole and locates it, but the cable is meant to remain a continuous span: free to expand, contract and swing, with tension broadly equal on both sides.

A dead-end grips the cable where it terminates or where a section of route ends. It must hold the entire tension of the span indefinitely without slipping, because there is nothing on the other side to balance it.

So the clamping force, the gripping length and the way load is distributed are quite different, and the two are not interchangeable in either direction.

Using a suspension clamp where a dead-end is needed lets the cable pull through. Using a dead-end where a suspension clamp belongs over-grips a cable that needs to move, which stresses it at one point and, on fibre, shows up as attenuation.

Because an aerial cable changes length constantly and swings in wind, and preventing that concentrates the stress at the fitting.

A cable strung between poles expands as it warms and contracts as it cools, across a range of tens of degrees between a winter night and a summer afternoon in sun. It also swings and vibrates in wind - aeolian vibration on long spans is a continuous low-amplitude flexing.

If the cable is rigidly gripped, all of that movement is absorbed at the clamp instead of being distributed along the span. The result is fatigue at the clamp edge, jacket damage, and on fibre a rise in loss from micro-bending.

Articulated clamps swing to follow the cable's direction, and resilient liners let the cable move slightly within the clamp while still carrying its weight.

So 'grip firmly' is the wrong instinct here. The correct clamping is enough to carry the vertical load and locate the cable, and no more - which is why the manufacturer's tightening figure matters.

Spreading the clamping load along the cable instead of concentrating it at the clamp edges - and it is the component that protects a fibre cable.

A bare metal clamp closing on a cable applies its force over a short length and, worse, produces a sharp pressure gradient at each end of the clamp where the cable enters and leaves. That edge is exactly where a jacket splits and where fibre micro-bending occurs.

A resilient liner - elastomeric, or a moulded plastic insert shaped to the cable - deforms under the clamping force so pressure is distributed along the whole clamp length and tapers gently at the ends.

The liner is also what makes one clamp body suit several cable diameters: different liners cover different ranges.

So the liner is not packing. It is the functional part, and using the wrong one, or reusing a perished one on a refit, is a common cause of jacket damage and of unexplained optical loss.

From the cable's actual outside diameter, checked against the clamp's stated range - and with the middle of the range preferred to either end.

Every clamp and liner combination covers a diameter band. A cable at the bottom of the band is gripped by a clamp that has to close a long way, which can distort the liner; one at the top may not allow the clamp to close properly at all.

Measure the cable rather than working from its nominal designation - manufacturing tolerance and different constructions of the same nominal cable vary more than people expect, and a route often carries cable from more than one batch or supplier.

Also confirm the construction. Round cable, figure-of-eight with an integral messenger, and ADSS all need different clamps even at the same overall diameter, because what the clamp should grip differs - on figure-of-eight it is the messenger, not the cable.

Where a route carries mixed cable, order clamps per section rather than one type throughout.

The vertical weight of the cable it supports, plus whatever ice and wind add - and the ice case usually governs.

The static load is the weight of half the span on each side, which for a telecoms cable is modest. What changes the picture is radial ice, which can multiply the effective weight of a cable several times over, and wind acting on the increased diameter.

So the design load comes from the local loading standard's combination of ice thickness and wind speed rather than from the cable's own weight.

A suspension clamp also has to cope with unequal tension either side - at a slight change of direction, or where one span is longer than the next - which produces a horizontal component the fitting must resist without the cable pulling through.

Where the angle becomes significant, the correct fitting is an angle suspension clamp or, beyond a defined angle, a dead-end arrangement on each side. Using a plain suspension clamp at a sharp angle is a common route defect.

To the manufacturer's tightening figure, aligned with the cable's direction, and with the liner seated properly - none of which can be judged by feel.

Overtightening is the commonest error, because it feels safer. It crushes the liner, concentrates pressure on the cable and defeats the clamp's ability to let the cable move. Undertightening lets the cable creep. Both are avoided by using the stated torque rather than pulling until it seems tight.

The clamp must align with the resultant direction of the cable. On an articulated clamp that happens naturally; on a fixed one it has to be set, and a clamp mounted square to the pole while the cable arrives at an angle bends the cable at the fitting.

Check the liner is fully seated and the correct one for the cable before closing the clamp - a liner half in place is not visible once the clamp is closed.

Record the cable type and clamp size fitted, so a later repair uses the same combination.

Cable creep, corrosion of the fixings, perished liners and clamps that have rotated out of alignment.

Creep shows as changed sag - one span deeper than its neighbours - and means the cable has pulled through the clamp. The cause is usually undertightening or a liner that has hardened and lost its grip, and the consequence is lost ground clearance somewhere along the route.

Corrosion attacks the bolt and the bracket rather than the clamp body, and a seized bolt is a problem for the next person who has to open it as much as a strength concern.

Liners perish with ultraviolet exposure and cold, becoming hard and cracked. A perished liner no longer distributes load, so the cable is being pinched at the clamp edges without anything looking wrong.

Rotation happens where the route has moved - a leaning pole, a re-tensioned span - and leaves the cable bending at the fitting. All four are visual findings, which is why a walked survey is worth more than it appears.