The cyclist

In Scotland the cyclist will often look as if they are about to take part in a very dangerous sport. Many cyclists wear helmets, and often hi-visibility clothing. It’s largely culturally accepted that cycling is an activity that requires special cyclist safety equipment, and state cycling information reinforces this. Clothing worn is often cycle-specific,  and tailored for energetic, sporty cycling. Cyclists often are male, and in their 20s to 40s.

Meanwhile, in the Netherlands, cyclists typically dress no differently from anyone else. They will be quite representative of the population as a whole in terms of age, from the young through to the old – dutch children cycle independently to school from age 8 or 9 or so (though, worryingly, this has increased from when I was young, when age 6 or 7 was common).  It’s not unusual to see parents carrying children, e.g. with seats on the rear rack, on the bar, and/or attached to the handlebar (for very young & light children), and sometimes all 3 at the same time. The “bakfiets” (barrow-bicycle) has also become popular as a flexible utility bicycle for young families.

Perhaps this is due to cycling in Scotland being undertaken often as a fitness activity, and not just as a way to get around. Perhaps the fact that Scottish cyclists tend to have to regularly practice “vehicular cycling“, and having to cycle faster to safely mix with nominally 52km/h (30mph) motor traffic, means they need sportier clothing. While in the Netherlands, cycling is something you do to get around – to get to school, to work, to do “boodschappen” (shopping), or nearby friends.

The sports/fitness cyclist of course also exists in the Netherlands. They are a tiny minority of the population in the Netherlands, just as in Scotland, though perhaps a little less so. However, unlike in Scotland, they are also a tiny minority of cyclists. Cycling is simply a normal part of life for most people in the Netherlands.

The bicycle

Bicycles in Scotland tend to be sports-bike derived. They’ll either be road-sports or mountain bikes, or hybrid of these mildly adapted for about-town. They’re nearly always equipped with derailleur gears, which are quite fragile and need constant maintenance. Derailleur gears means chain-guards are impractical, so oily chains are always exposed. Lights are bolt-on after-thoughts, usually battery-driven. Many bikes do not have full mud-guards, e.g. because they are sports-road bikes that have no fittings for them, or because they are mountain bikes on which they are not fashionable. Some mountain bikes may have seatpost mounted half-guards. Many bikes do not have luggage racks, but where they do, it’s usually made from a thin-gauge aluminium which can not carry more than 25kg or so. Brakes are often rim-caliper brakes – which require regular adjusting. Mountain bikes often have suspension, for downhill cycling over rough terrain, which saps pedalling energy.

Bicycles in the Netherlands are usually hub-geared, often 3-speed. This means no derailleurs, which means its more robust and needs much less maintenance, and also that the chain or belt can be fully enclosed. Brakes tend to be low-maintenance drum brakes (the rear often actuated by pedalling backwards). Lights are often dynamo-driven, and even integrated into the bike on some modern designs. Low-drag hub dynamos are common these days, supplanting the rim-driven bottle dynamos still common on older bikes. Rear racks tend to be far more sturdy, as it is common in the Netherlands to use them to give a friend a lift – particularly among teenagers & young adults. For an idea, have a look at the Batavus 2012 city bike line-up (UK site, but the models & specs differ a little, to cater for UK preferences).

The environment

The road environment in the Netherlands is nearly always built to cater as much for the safety and convenience of pedestrians and cyclists, as it is for motorists. Cycling infrastructure is integrated with road design, and has been evolved and improved over decades, with safety for all users engineered in.

Older dutch road & path design

In dense urban areas, speed limits are 30km/h (18.6mph), sometimes even less.  Where ever possible, road users are separated and each given their own, clearly delineated space. So pedestrians will have their foot-path, cyclists their cycle-path, and motorists their road. Where space is constrained, cyclists may have to use the road – in which case the speed limit will practically always be reduced to 30km/h or less. Care is taken to reduce, if not avoid, the impact of conflicts between different road users at junctions and pinch points. On streets where motor traffic may encounter cycle traffic, the junction with side streets may be given raised kerbs, with tight angles, to slow down cars coming in or out of those side-streets. Cycle paths are generally very smooth, clean and well-mantained. You can cycle on them with normal bikes, with no problem.

The Scottish sports cyclist might, however, not be entirely happy with the dutch cycling environment, as cycle-paths often are mandatory. Dutch sports road cyclists often complain about being restricted to narrower cycle-paths in places, while ordinary dutch complain about having pelotons of fast roadies flying past at speeds similar to mopeds – which were banned from many cycle-paths in the 90s for the same reason. Various groups, representing both ordinary and sports cyclists, have been campaigning to allow/force such pelotons to use the road instead.

Brand "new" segregated path in Glasgow

In Scotland, cycling infrastructure is more rare. Cycling infrastructure often seems to be designed & built separately from the general road, and does not get a lot of funding. Infrastructure is usually done extremely cheaply, e.g. by designating existing footpaths as “shared use” and putting up a few signs, or by painting an advisory lane on the road. The shared-use paths sometimes have rough surfaces, such as gravel on tow-paths or even very rough cobbles (the Clyde-side section of NCN75 in Glasgow, from Broomielaw to the SECC), making them unpleasant or even completely unsuited for use by normal town bicycles. Which may explain the popularity of very heavy, suspension mountain bikes here in Scotland. On-road cycle lanes often are very narrow and go through the highly dangerous “door zone”, where cyclists can be knocked off their bike into the path of the 52km/h (30mph) traffic.

Even when segregated paths are built, because there is so little investment, very little of it can be built, or , as in the case in the photo here, it’s simply sectioned off from an old & broken road without being properly resurfaced first. Where paths cross side-roads, or where the segregated paths run out (which is often, as so little of it can be built), very little thought is given to reducing the impact of conflicts between cyclists and motor traffic. If thought is given, it will be the cyclist who is required to give way with a “Cyclists Dismount” sign, not the other way around. The local authority may even deliberately require the cyclist to negotiate kerbs, again making life hard for anyone on a normal bicycle.

The lack of separate infrastructure means Scottish cyclists have to choose between cycling on road, amongst fast motor traffic, or often inconvenient shared-use paths where these exist and perhaps illegally cycling on pedestrian footpaths where they do not. Roads in urban areas are often designed to maximise traffic flow. Single lane streets with parking widen out to 2-lanes at junctions. Railings often are installed to segregate the fast, 52km/h traffic from pedestrians – potentially leaving on-road cyclists nowhere to go in any incident with a motor vehicle.

As cyclists are so few, and because they tend to dress distinctly, there seems to be tribalisation between motorists and cyclists. Some motorists view cyclists as obstructive “lycra louts”, getting in their way and slowing them down, even though in urban areas in Glasgow, average journey speeds are often limited to 20 to 30km/h by traffic lights – which is potentially slower than the cyclist. Congestion, caused by cars, not bicycles, can make journey average speeds much slower again. Despite, or perhaps because of, waits at traffic lights equalising average speeds for motorists and cyclists, motorists still will rush impatiently at, and not infrequently above, the 52km/h (30mph) speed limit. Some motorists can be ignorant, a rare few even aggressive. The cyclists in turn feel embattled, and often subjectively quite unsafe, due to this fast and sometimes close passing motor traffic – motor traffic they will often catch anyway at the next lights.

Weather

While the Netherlands might not get quite as much rain as Scotland, it still gets plenty of it outside the summer. That said, it doesn’t rain as often people think, and most cycle journeys in either country will be rain free. Also, the winters can be much much colder in the Netherlands. The dutch simply wear a rain jacket, carry an umbrella and put on more clothes.

Flat land

Much of the Netherlands is quite flat, Holland especially so. However, the south of the Netherlands, near the border with Belgium, can be more hilly and cycling is just as normal there, as anywhere else. The flatter parts can allow strong, continuous winds to blow across, which can be as much of a drag on cycling as any hill!

Overall

Despite the less appealing environment in Scotland, compared to what the Netherlands shows is possible at least, you can still enjoy some dutch style cycling, at least within Glasgow city. So many of the roads are wide and 2-lane, that with a little confidence to take your lane, by moving out away from the gutter and more towards the middle of it, or by riding 2-abreast if with company, you can get the motorist to, perhaps grudgingly, give you the space needed to enjoy nice, relaxed cycling in whatever clothes you’d normally walk out of the door with. Unfortunately though, I suspect many people in Scotland are put off cycling by idea they’d need to develop that kind of confidence.

Mass cycling in Scotland may first require that the roads see more equal investment for cycling. This in turn may require some leadership, and/or a significant change in economic pressures (e.g. fuel costs; public health costs of sedentary lives).

The report concerned had a number of proposals on possible improvements, and motions for action. As the EP (European Parliament) does not have the power to introduce directives, the EP can only suggest that the Commission  should propose to introduce such legislation/directives, for only the Commission (which is appointed by the member states) has such power. Which is what the motion attached to the report asked for, which was approved by the EP.

The Daily Mail article ran with the headline of “Now EU wants to set our speed limits: Brussels plan to enforce 20mph zones and replace Highway Code with European laws“, and a lead-in paragraph as follows:

European Union plans to enforce a mandatory 20mph speed limit in residential zones and replace the Highway Code with European law has sparked outrage from British politicians.

This is clearly bollocks. The EP has no such power. Even if the EP had such power, the motion approved does not call for mandatory limits but “Strongly recommends the responsible authorities to introduce“. As for replacing the highway code, this seems to be a strange perversion of the report’s call to urge EU states to further harmonise various aspects of their road traffic laws. The head-line and lead-in (which the front-page summary was similar to, as I recall) are plainly misleading, and utterly false on at least one count.

The Daily Mail, like many newspapers, is a member of the Press Complaints Commission, which self-regulates the industry. Their Code of Practice has, as its first clause:

1

Accuracy

i) The Press must take care not to publish inaccurate, misleading or distorted information, including pictures. ii) A significant inaccuracy, misleading statement or distortion once recognised must be corrected, promptly and with due prominence, and – where appropriate – an apology published. In cases involving the Commission, prominence should be agreed with the PCC in advance. iii) The Press, whilst free to be partisan, must distinguish clearly between comment, conjecture and fact. iv) A publication must report fairly and accurately the outcome of an action for defamation to which it has been a party, unless an agreed settlement states otherwise, or an agreed statement is published.

So I complained to the PCC:

This article blatantly breaches the “Accuracy” requirement of
the Code of Practice, with its headline and lead-in of :

“European Union plans to enforce a mandatory 20mph speed limit in
residential zones a nd replace the Highway Code with European law has
sparked outrage from British politicians.”

As

1. The EU is not a monolithic entity, but composed of a number of different
entities which represent different interests. The article is actually about
a report from the European Parliament.

2. The claim that the EU plans to enforce limits and replace the Highway
Code is utterly inaccurate and is a misleading distortion of the facts: that
the EP has *endorsed* a report which makes *recommendations* on how
authorities around Europe could improve safety on our roads.

Relevant Clauses : The headline and lead-in breach section 1, clauses i and iii of
the code, by publishing inaccurate information, and by presenting a quite
distorted abstract of the facts to fit the Daily Mail’s anti-EU editorial
stance.

Their decision came back after about a month:

Commission’s decision in the case of

Jakma v Daily Mail

The complainant considered that the headline and opening sentence were
misleading, as the EU could not enforce speed limits or replace the
Highway Code.

The Commission made clear that it considers headlines in the context of
the article as a whole rather than as a standalone statement. This is
due to their brevity – they can represent only a limited summary of a
more complex set of circumstances. The Commission understood the
concerns raised by the complainant; however, it noted that the body of
the article made the situation clear – namely that a list of
recommendations had been made in a report to the European Parliament on
the improvement of road safety across Europe, which had been endorsed by
a number of MEPs. It made clear that this did not equate to legislation
and it was unlikely that it would do so in future. The Commission did
not consider that the headline was a significantly inaccurate summary of
the situation given that a report including recommendations on Europe
wide safety measures had been endorsed by the European Parliament, the
parliamentary institution of the European Union. Furthermore, when the
headline was read in conjunction with the body of the article, readers
would not be misled as to the situation. In the circumstances, the
Commission could not establish a breach of Clause 1 (Accuracy) of the
Code.

Reference No. 114462

In short, because the Daily Mail were more honest about the facts later in the story, it doesn’t matter how much they blatantly lied in their headline and lead-in (which is all some readers will have read)!

(TabloidWatch have some other examples of shamefully inaccurate EU reporting, by The Express – no doubt their stories were also still PCC Code of Practice compliant.)

While amongst cyclists who suffer injury, helmets of themselves do reduce head injuries significantly, they also increase neck and facial injuries, so that there appears to be negligible benefit overall (Accident Analysis & Prevention: … meta-analysis of bicycle helmet efficacy). Study of bicycle injury rates in Australia around the time of introduction of mandatory helmet laws suggests that, though there is a noticeable dip in injury rates around the introduction of the law itself (not necessarily attributable to the helmet itself in my opinion) that injury rates then started increasing again, to the point rates were nearly the same at the end of the study period as before the helmet law, and trending to surpass it! (My Blog: study-shows-australian-cyclist-helmet-law-leads-to-increasing-head-injury-rates). Helmet use also appears to induce risk-compensation behaviour in motor vehicle drivers – they make closer passes (Accident Analysis & Prevention: Drivers overtaking bicyclists…). No doubt the cyclists themselves also are subject to risk compensation. Thus, by wearing a helmet there may be an increased risk of getting into an accident.

There may be further population wide psychological effects caused by a culture of “Must be wearing a helmet to be safe!”. It is sending the message that cycling needs safety equipment, and hence must be dangerous, which surely will put off many – certainly where mandatory helmet use laws are introduced rates of cycling then significantly decrease. The reverse is of course true: the overall health benefits of cycling greatly outweigh the quite tiny risks – risks which are not greatly changed by helmet wearing, the studies appear to say. In other words, by advocating helmet use, one may be harming the rates of cycling by sending the wrong message on safety, and hence harming public health overall. Further, as cyclist safety on the roads correlates strongly with rates of cycling – more cyclists leads to more awareness & safer roads, and similarly fewer cyclists means less safe roads – this means a culture of helmet use may well lead to increased injury rates amongst cyclists (in addition to the general adverse public health effects of fewer people cycling). This would be very hard to categorically prove or disprove in causal terms, however the Australian experience certainly suggests a correlation, as I think would a comparison of the UK and Netherlands.

Finally, in the Netherlands, one of the safest places for cycling in the world, cyclists almost universally do not wear helmets, including very young cyclists. Thus, we can be quite certain that helmet usage is not a pre-requisite for safe cycling. Indeed, it is in places like the UK and USA, with some of the worst cycling safety in the developed world, where the focus on safety equipment for the cyclist seems to be greatest.

In short, the focus needs to be on those things around the cyclist (e.g. default legal liability to influence motorists’ behaviour, safer road infrastructure, etc) – not what is on cyclists. Focusing on cyclist safety equipment to me seems futile at best, and perhaps even detrimental to the cause of mass, safe cycling, if that’s a cause you believe worthwhile.

NB: Helmet use should always be a personal choice. The issue is complex, the trade-offs may differ greatly in different scenarios – helmets may be very beneficial in some settings, e.g. some kinds of racing. The choice should be your own. However, general advocacy of cycling helmets seems inappropriate and probably harmful, to me.

Here’s the bollard:

Jagged, rusty metal on bollard, at side-road access on A814

Part of the problem with it is that the row of bollards here run parallel with the A814, rather than being more square across the entrance and to the A814. This means that any cyclist who is coming down the A814 (e.g. because they don’t fancy the shared-use path that forces cyclists to slow to a crawl for deliberate obstacles, which is there because of a potentially dangerous car-park entrance), which usually has to be at speed to placate the many terribly impatient motor vehicle drivers out there, and who wishes to enter that quiet side-road has to try negotiate a very tight angle between this rusty, jagged bollard and the light post:

View of the angle between the jagged metal bollards and traffic-light-post, on approach from A814

Every time I go through there, I worry that if I brake too hard on the A814 to prepare for the turn, that the motorists behind will plough into me. Then I worry I’m going to get caught on that nasty bollard as I go through there.

So, is the council right to leave this apparently broken & disused bollard in place?

The contractor did re-surface the worst part:

Resurfaced section of new cycle-path, London Rd near Glasgow Cross

However, they did a half-arsed job:

The remainder of the new path, left with the original, old & broken surface.

Closer view of the "new" cycle-path surface.

However, regarding the Clydeport car-park entrance on the Broomielaw and how it interrupts and inconveniences the national cycle route that crosses its entrance with a dropped kerb, their response was:

At the Broomielaw, National Cycle Route 75 is located on the north side of the Casino, using a shared use pedestrian/ cycle track. Drivers exiting the Clydeport car park have poor sightlines, hence the kerbs nearest the wall on the cycle track have not been dropped, thus encouraging cyclists to stay away from this area, where they are harder to see.

Now, bear in mind this is an almost unused car-park – I’ve never seen a car going in or out there, even though I commute past there every morning and evening. The cycle-path there on the other hand is heavily used. Further, the kerb, though lower toward the outside, is still a raised kerb and an impediment to cycling! Never mind the road surface is also broken up to at least one side.

So the message is, when it comes down to the convenience of a large number of cyclists, versus that of a property holding company and its under-utilised, city centre car park for a couple of employees, well the cyclists can more or less go and get stuffed.

Update: Went past the Clydeport car-park later in the day today and it does seem full.

Entrance to the Clydeport car-park on the Broomielaw, with the National Cycle Route 75 crossing it. The cycle-path has a dropped kerb, very inconvenient for cyclists, particularly any carrying luggarge, and any on normal, non-offroad bikes. Note the cyclists having to stand up to negotiate the kerbs, even though they are out from entrance. Still looks like urban mountain-biking, sadly.

Another view of the car park entrance. Note that cars actually have plenty of visibility to this direction. It is only the wall on the other side that is impeding visibility. Note also the broken surface of the road, making the far kerb even harder to negotiate.

It is actually quite possible to engineer minor side-streets and entrances that must cross cycle-paths so that cyclists progress and safety is not  compromised. They manage it all the time in the Netherlands. The line-of-sight issue could be better resolved by modifying the wall, and/or installing a lowish, progressive projection/kerb around the entrance wall to ensure cyclists move further out (indeed, this would be a good idea either way). Further, it should be Clydeports’ responsibility to ensure that its property does not impose undue risks on the passing public. At least, I would hope there is legislation and/or by-laws in place, to that effect.

So how do the dutch do it?

Dutch cycle-lane, with a minor-side-street junction. The junction is engineered to safely allow cyclists on the busier street the right of way over the quieter joining entrances. Note the conspicuous signage to remind drivers of the cycle lane, and note the concrete segregation barriers immediately before the junction, to prevent car drivers from turning in fast and early.

Note how the road is further engineered to slow down any cars before they cross the cycle lane, and so reduce conflicts. The cycle lane is raised up slightly with a smooth gradient, while the roads have a steeper kerb to negotiate, to force them to slow down and further help make them aware of the cycle lane.

The new path starts off at Saltmarket, then joins London Rd, where it’s still under construction. It’s being built by sectioning off some of the old road surface with kerbs and concrete.

Start of new segregated cycle path at Saltmarket

New segregated path under construction, London Rd

The problem of course is that the roads in Glasgow at the moment are in a terrible state, especially on busier roads (like London Rd) and especially towards the sides of the roads, for some reason.

Cycling down this new London Rd section on a normal town bike is not comfortable. It’s even less comfortable when you’re carrying luggage, e.g. loaded panniers as many commuters do. It’d be even more uncomfortable again on a road bike with narrow tyres (I’d stick to the main road in that case). Basically, to make use of this path you need to either cycle very slowly or else you need a heavy, full-suspension mountain bike – not exactly ideal for town!

Surface of new, segragated cycle path, London Rd by Moir St

As this path is only just built, it’s possible there’s still surfacing work to be done. I can hope! However, it’s also possible that the council do not have surfacing equipment narrow enough to fit down these paths and that, once the concrete segregation barriers are down, these paths can no longer be properly resurfaced – that the council will be limited to making ad-hoc patchwork repairs with men with buckets of hot asphalt and spades.

Surface of new, segragated cycle path, London Rd, looking toward Glasgow Cross

It’s not just this path though. The existing NCR75 Clyde-side path is surfaced with brick cobbles in several sections! Beyond uncomfortable, it’s near unusable for all but mountain bikes. The section of NCR75 along the SECC is also cobbled, as well as being narrow and shared with pedestrians.

NCR75, deep gapped brick cobble where it goes under the M8. Note the cyclist using the narrow strip of smoother brick cobble to the right.

NCR75, more cobble on Lancefield Quay. Note the cyclist ahead sticking to the road.

While not wishing to seem ungrateful, what is the point of building cycling paths that are bone-rattlingly, bike-to-bits-shakingly uncomfortable to use to all but those on off-road bikes? This is unlikely to help promote cycling to anyone, but already fairly dedicated cyclists.

A quite old and obsolete, dutch cycle-path. Brick-tile surface, but still manages to be smooth and enjoyable to cycle on.

Hi,

I’m writing about your recent press release at:

http://www.prweb.com/releases/truck-insurance/accident-camera/prweb8592980.htm

In the press release you quote the truck firm, who appears to think his driver is completely innocent. This is a view which Towergate Insurance appears to endorse to some degree, as you have put the video on Youtube with the title “Death Wish Cyclist“. As a car driver and regular cyclist, I think it’s worth pointing out that the video shows the truck driver’s behaviour is far from perfect. Regardless of whatever blame may be attached to the cyclist, there are a number of problems with how the truck is being driven:

a) The Highway Code advises that cyclists should be left plenty of room when passed by motorists (rule 163 and 212). The road concerned is single-lane in each direction, and there is oncoming traffic, so the truck driver could not have changed lane. Yet they were still preparing to overtake, and even accelerated to 44mph as part of this. Had they passed the cylist, it would have been far too close and at too high a speed. The lorry driver was not intending to leave any margin of safety, needed in case the cyclist had to weave or move for some reason, e.g. from a mechanical problem or because of potholes. Motorists who close pass cyclists are driving irresponsibly, especially so when done with heavy goods vehicles which are almost to certain to kill cyclists if there is any contact.

b) The road is through Southampton Common and appears to have a 40mph speed limit. Yet the video  shows the driver reaching 44mph as they prepare to overtake. They appear to be committing a quite obvious offence with their aborted overtake, presuming the data from your system is accurate.

c) There is a side junction ahead, which presumably is easily visible given that the cyclist was manoeuvring to take it. The Highway Code rule 167 states “DO NOT overtake where you might come into conflict with other road users. For example: * approaching or at a road junction on either side of the road“. The Highway Code says this precisely because road users are likely to be manoeuvring to leave or enter the road ahead – as the cyclist is.

d) The Highway Code also states that motorists should take extra care around vulnerable road users, such as motorcyclists, cyclists. Further, though it’s impossible to be certain because the camera lens is extremely wide-angle and doesn’t show enough detail, the cyclist looks like they perhaps may be young – a teenager. The highway code specifically advises motorists to be particularly careful around young and inexperienced road users (rule 204).

Road safety requires that we not only avoid making mistakes, but also that we take care to allow for the inevitability that mistakes will sometimes be made by others. On this basis alone, the driving shown by the truck driver was less than perfect. Further, the driving shown appears to directly contravene a number of rules given in the Highway Code, with which all drivers should be familiar. Finally, contingent on the accuracy of your system, the driver may even have been speeding.

All together, it’s very hard to see how this driver is “blameless” for the narrowly averted accident shown in this incident. Without intending to absolve the cyclist in any way of blame for their part, I think it’s important that anyone watching that video should also realise that the driving shown falls far below what is expected of motorists.

We believe the long range growth and development of internetwork communication would be seriously inhibited by specifying how much larger than the minimum a packet size can be, for the following reasons.

1. If a maximum permitted packet size is specified then it becomes impossible to completely isolate the internal packet size parameters of one network from the internal packet size parameters of all other networks.
2. It would be very difficult to increase the maximum permitted packet size in response to new technology (e.g. large memory systems, higher data rate communication facilities, etc.) since this would require the agreement and then implementation by all participating networks.
3. Associative addressing and packet encryption may require the size of a particular packet to expand during transit for incorporation of new information.

Fragmentation generally is undesirable if it can be avoided, as it has a performance cost. The fragmenting router may do so on a slow-path, for example; and re-assembly at the end-host may introduce delay. As a consequence, end hosts have for a long while generally performed path-MTU-discovery (PMTUD) to discover the right overall MTU to a destination, thus allowing them to generate IP packets of just the right size (if the upper-level protocol doesn’t support some kind of segmentation, like TCP, this may still require it to generate IP fragments) and so set the “Don’t Fragment” bit on all packets and generally avoid intermediary fragmentation.  Unfortunately however PMTUD relies on ICMP messages which are sent out-of-band, and unfortunately as the internet became bigger, more and more less-than-clueful people became involved in the design and administration of the equipment needed to route IP packets. Routers started to either ignore over-size packets and (even more commonly) firewalls started to stupidly filter out nearly all ICMP – including the important “Destination Unreachable: Fragmentation Needed” ICMP message needed for PMTUD. As a consequence, end-host path-MTU discovery can be fragile. When it fails to work, the end-result is a “Path MTU blackhole”: packets get dropped for being too big at a router while the ICMP messages sent back to the host get dropped (usually elsewhere), meaning it never learns to drop its packet sizes. Where with IP fragmentation communication may be slow, but with PMTU blackholing it becomes impossible.

As a consequence of this, some upper-level applications protocols actually implement their own blackhole detection, on top of any lower-layer PMTU/segmentation support. An example being EDNS0, which specifies that EDNS0 implementations must take path-MTU into account (above the transport layer!).

So now the internet is crippled by an effective 1500 MTU. Though our equipment generally is capable of sending much larger datagrams, we have collectively failed to heed Cerf & Kahn’s wise words. The internet can not use the handy tool of encapsulation to encrypt packets, or to reroute them to mobile users. Possibly the worst aspect is that IPv6 completely removed fragmentation support. While there’s a good argument that end-end level packet resizing may be more ideal than intermediary fragmentation, as IPv6 still relies on out-of-band signalling of over-size packets, without addressing that mechanism’s fragility problem, it likely means IPv6 has cast the MTU-mess into stone for the next generation of inter-networking.

Updated: Some clarifications. Added consequence of how PMTU breaks due to ICMP filtering. Added how ULPs now have to work around these transport layer failings. Added why fragmentation was removed from IPv6, and word-smithed the conclusion a bit.