Talk:M898 motorway

Revocation of 31 March 2013

I have undone this edit for a number of reasons:

• It does not comply with WP:RJL – the style used in the UJK is quite clearly shown at Wikipedia:Manual of Style/Road junction lists#M5 motorway.
• J30 is part of the M8 numbering sequence, not part of the M898 numbering sequence. Streetmap does not show any junction numbers, even though my road atlas does.
• The distance 1.701 km is wrong – see not later
• Can you justify using Junction 1 as the zero point for length measurements. If this motorway followed normal British standards then the marker posts would be part of the M8 sequence.

I checked the distances from the database that User:Fredddie downloaded and compared them to those on Google Earth:

• The 1453 segment appears to run the entire length of the M898, starting at the junction of the eastbound M8 and ending at J1 of the M898.
• The 254 metres segment appears to be the section that runs from the point where the northbound and southbound carriageways of the M898 split to the point where the south-bound carriageway of the M898 splits
• The 264 metre segment appears to be the section that links the M898 southbound to the M8 eastbound
• The other two slip roads are probably linked to the M8 in the database, not the M898.

The 254 and 264 metre segements might be interchanged - I identified them by segment length, not by geographic location.

So you've reverted to a version that violates the general MOS and does not reflect the emerging consensus at WT:RJL? --Rschen7754 08:39, 31 March 2013 (UTC)

Did you read the rest of my comments? What is Fredddie (or you) going to do to address them? Martinvl (talk) 09:40, 31 March 2013 (UTC)

I'm going to have to disagree. I have included this image of the mapping data for reference.

1. The fact that M898 lies within the UK does not preclude it from using the style used for the rest of the world.
2. You are correct. J30 is an M8 junction number. The diamonds on the map are motorway junctions as identified by Ordnance Survey. If it helps the reader find the junction, why in the world would we not include it? But, I have gone ahead and dropped that column entirely.
3. 1.701 km is correct! I'm more inclined to believe a country's mapmaking bureau than Google Maps/Earth. If OS say the 254-meter slip road is part of M898 then I'm going to believe them. Full stop.
4. I will be happy to switch around the junctions (thankfully there are only three lines) such that km 0 is at M8. Thanks for the suggestion. –Fredddie™

A further error - the M898 does not lie within the Erskine Community Area (if that is what Fredddie meant), a large part of it forms the boundary between the Erskine and the Bishopton Areas - see here and here while about 100 metres passes through the Inchinian Community Area (see here). Martinvl (talk) 09:57, 31 March 2013 (UTC)

Fixed. I have taken all the suggestions you gave and worked them into into the article. –Fredddie™ 10:37, 31 March 2013 (UTC)

Just a courtesy note: I'm probably not going to be back online for 12 hours, so I won't be able to reply further until then. –Fredddie™ 10:42, 31 March 2013 (UTC)

You have clearly not read what I wrote.

• There are no railway stations in Inchinnan or Erskine, so the location "Bishopton–Inchinnan line" and "Bishopton–Erskine line" is meaningless. Bishopton is on the Inverclyde Line. An inspection of the map shows that this line does not cross the M898.
• You have not corrected the distances. If you are quoting distances to an accuracy of one metre, you have four different lengths to consider (my measurements from Google Earth in brackets):

1. Start of slip road from M8 Eastbound to the end of the M898 (1453 metres)
2. Start of slip road from M8 Westbound to the end of the M898 (1684 metres)
3. Start of the M898 to the point where the road joins the main carriageway of the M98 Eastbound (1536 metres)
4. Start of the M898 to the point where the road joins the main carriageway of the M898 Westbound (1809 metres)

Given that your distances are meaningless and that your locations are meaningless, I have again reverted. Please explain how you obtained the distance of 1,701 km – your diagram did not explain that to me. What was the exact data that you used? If you calculated them, I draw your attention to WP:CALC – you need to demonstrate that your calculations are "an obvious, correct, and meaningful reflection of the sources".

Have you seen these representations used for the rest of the world (your words, not mine):

• Dutch: nl:A1 (België), nl:Rijksweg 16. Both countries have their RJLs as part of the infobox. Destinations are shown, but no localities
• Italian: it:Autostrada A1 (Italia) The provinces are shown in the RJL, the regions in the infobox.
• French: Both fr:Autoroute A1 (France), fr:Autoroute suisse A1 (Switzerland) give the name & number of the junction, but no locality.
• German - de:Bundesautobahn 1 (Germany), nl:West Autobahn (Austria), de:Autobahn 1 (Schweiz) - RJLs are hidden in the infobox. They can be accessed by clicking "Ausklappen" at the bottom right-hand corner of the infobox. (Again, no locations).

In view of the above, please explain why your layout meets the layout for the rest of the world?

Martinvl (talk) 16:45, 31 March 2013 (UTC)

We are not bound by the guidelines of other Wikipedias. We are, however, bound by the guidelines of the English Wikipedia. --Rschen7754 17:39, 31 March 2013 (UTC)

Other Wikipedias are useful to see whether or not we are on the right lines. They also show up User:Fredddie's comment "for the rest of the world" as being ill-founded. Martinvl (talk) 17:51, 31 March 2013 (UTC)

Again, we are not bound by them. Moreover, you may want to take a look at es:California State Route 52 then, because that format is definitely not universal. --Rschen7754 17:55, 31 March 2013 (UTC)

See also de:M-134#Hauptkreuzungen for a German article that uses the English Wikipedia format for the junction list. The other language editions are free to use different stylistic choices. California Department of Transportation is translated to Departamento de Transporte de California in Spanish, while others leave the name intact, like California Department of Transportation in French and California Department of Transportation in Dutch. The controlling factor here is the Manual of Style for the English Wikipedia. Imzadi 1979 → 18:35, 31 March 2013 (UTC)

Martin, you are correct in that we can be inspired by what works on the other languages. However, Rschen and Imzadi are correct in interpreting what I meant. We at the English Wikipedia are not bound by the German Wikipedia MoS in the same way dewp is not bound by enwp and so on and so forth. When I said "the rest of the world", I meant on the English Wikipedia.

Now, when I was editing this morning, it was around 5:30am my time. Hardly the best time for me to dig in to the wiki. So, I have to clarify a few things. "Bishopton–Inchinnan line" and "Bishopton–Erskine line" refers to city limits, corporate limits, whatever you want to call them. It's the boundary between two population areas. In the context of a highway, it should have been clear (probably was clear to USRD editors) that it was the dividing line between villages. Had I said "Bishopton–Inchinnan village boundary" or "Bishopton–Erskine town boundary", we would not be discussing this.

According to Ordnance Survey, M898 is 2.061 km. (OK, I admit that earlier I was wrong. I originally claimed it was 1.707 and did not include the slip road south of J30.) Why are we holding Google Maps data in higher esteem than OS data?! Anyway, here is the attribute table looking for all segments numbered "M898" There are clearly three segments and we can add the lengths together (1453+254+264) and get 2061. This follows WP:CALC in that the OS data unambiguously shows three segments and 2061 meters is the sum of the segments' lengths, and more importantly, it passes WP:V since any one of us can download the data and check it out for ourselves. –Fredddie™ 00:05, 1 April 2013 (UTC)

Hi Fredddie,

I knew that you meant the English-language version of Wikipedia. If you go back to the vey first version of WP:RJL, you will see that it was an attempt to standardise the US road junction lists. If you want to export that format to Europe, first check what the local editors have developed before you try to impose a US-developed version onto everybody.

I knew exactly what you meant by the text “Bishopton–Erskine line”, but I wanted you to realise that you were using a US colloquialism in a UK article. Beware of doing this. Have you actually checked the areas or populations of Bishopton or Erskine? Do you honestly think that they add anything to the road junction list? Quite frankly I don’t care which community area the junctions are located in – the community area councils have no jurisdiction over the road. If you read Local government in Scotland you will see that they have no jurisdiction over anything! You are also technically wrong in writing “Erskine”, you should have written “Erskine Community Area” to distinguish it from the town of Erskine. In short, is it even sensible to mention Erskine or Bishopton in the location column? Erskine is already mention in the destination list and if I was going to Bishopton I would avoid the M898 – look at the map and convince yourself. If you look at the map, you will also notice that the Erskine Golf Course lies within the Bishopton Community Area.

As regards distances, I summed all the sectors in the OS database that make up the M3 motorway, a motorway that is not too far from where I live. The totals distance was 100112 metres yet the lowest numbered driver location signs is 24.8 (here)and the highest numbered location sign is 119.5 (here). The DLSs have a difference of 95.7 km – what happened to the other 5.2 km? About 1 km can be accounted for by the bridge over J1 and the last bit of slip road before joining the M27, but that still leaves about 4 km unaccounted for. I suspect that this is made up of slip roads that links the M3 with other motorways – the M25 and the M27(E). It is probably the same with the M898.

Working on this theory and matching up sector lengths with what I measured on Google Earth, I believe that the 1453 metre sector applies to the whole of the M898 as traversed from the M8 (Eastbound) to the end of the M898 and the other two sectors belong to sections of slip road that link the M898 to the M8(E) when joining the M8(E). Go to Google Earth and convince yourself.

In short, you have fallen foul of WP:ENGVAR and you have indulged in WP:OR by not floowing the limitations of WP:CALC. Martinvl (talk) 05:37, 1 April 2013 (UTC)

Grasping at straws, again, I see. I will reply in length, again, when I get home from work. –Fredddie™ 12:09, 1 April 2013 (UTC)

File:M8-M898 Temporary for discussion.svg

M8-M898 layout

Hardly - My measurements taken from Google Earth suggest the following:

• Sector AB - 1450 metres
• Sector CD - 250 metres
• Sector DE - 260 metres
• Sector XZ - classified as part of the M8, not M898
• Sector YD - classified as part of the M8, not M898

Please check these measurements for yourself (BTW, in the diagram, north is downwards).

Why do you choose A as your reference point rather than E, X or Y? That is why I prefer the use of marker posts - they are unambiguous. Martinvl (talk) 07:48, 2 April 2013 (UTC)

I am continuing part of this discussion on the WP:RJL Talk page as I feel that a number of fundemntal issues were raised that should be aired there. Martinvl (talk) 11:14, 2 April 2013 (UTC)

The OS data being used here is not perfect, and I wouldn't consider it an ideal source as to the length of the road. The Meridian product is a generalized data set for 1:50,000-scale maps, which means its somewhat simplified. The coastline paradox will apply here to an extent and higher quality data (ie the non-free OS MasterMap) would show a subtly different length.

That said "The road is Xm long according to OS Meridian" is information from a reliable source. "The road is Xm long according to my interpretation of Google Maps and the signs by the side of the road" is original research. The ideal would be a source that directly states the road length. Transport Scotland almost gives that data here. That is the sort of thing that will resolve this, not looking at images of the actual road.--Nilfanion (talk) 12:31, 4 April 2013 (UTC)

A number of points:

• The OS database did not state that the road is "X km long" – it catalogued three sections of 1453&nbspm, 254 m and 264 m, all of which were part of the M898
• I do not believe that the Coastline paradox applies here – I identified the section of the M898 shown as AB on the above diagram as being 1450 metres (straight line distance as per Google Earth is about 1180 metres). The OS figures cited that a section is 1453 metres. I believe that we were looking at one and the same thing.
• Although the OS database might be correct, linking the various sectors to real pieces of road is not a trivial task – Fredddie wrongly (IMHO) just added them up.
• In this instance I used Google Maps to cross-check Fredddie’s results and I was able to correlate my reading from Google Maps with the OS reading and get a different result to Fredddie. In order to verify things further, I cross-checked against the M3. Again there was a discrepancy which could be reconciled if I were to remove a distance of about 5 km – a distance that is commensurate with the link roads to the M27 and the M25.
• The report that you identified was "nearly" there. Is this type of detail normal for Scottish road reports? Are there any reprots that are actuasll;y usable? I could not find any.

Martinvl (talk) 15:06, 4 April 2013 (UTC)

My point is the OS data is somewhat iffy for this purpose and cannot be directly quoted from a table (a look at Google to verify it is perfectly acceptable). Its best to viewed as what it is designed to be, a map, and interpreted accordingly. Also, the different map projections may make a significant impact to the length. In this case:

1. The 1453m feature is from the bridge at J1 to where Old Greenock Road crosses the centreline of the split near J30. This is along the centreline of the M898, but goes too far north.
2. The 254m feature is a straight line from where Old Greenock Road crosses the centreline of the split to near where the bridge taking the M898 slip to the westbound M8 crosses the hard shoulder of the eastbound carriagway - it doesn't correlate to any real feature.
3. The 264m feature is from that point, to the centreline of the M8 at the end of the sliproads south of the junction. It vaguely tracks the slip from the westbound M8 to the M898.

As you can see, only that 1,453m feature actually corresponds to a length of road - and thats not the M898, but the M898 plus a bit of A898. Its sheer coincidence that the 1,453m happens to correspond to the 1,450m you traced yourself. None of the values from Meridian are suitable as a result of this.

As a comparison, I have checked using the higher quality OS VectorMap District product. The segments labelled M898 cover: Both carriageways of the mainline up to the point the slips from J1 towards the Erskine Bridge join, the full slips at J1 to/from the M8, the full entrance slips from both M8 carriageways and part of the M898 exit slip from the carriageway split to the point where it crosses westbound carriageway of the M8.

Using these segments I get the following results (all based on OSGB, distances measured on Google map may differ substantially due to the different projection):

1. Northbound mainline from the point where entrance slips merge to end of motorway: 1,177m
2. Southbound mainline from the start of motorway to point where exit slips split: 1,206m
3. M898-M8(W) slip: 654m
4. M898-M8(E) slip: 359m
5. M8(W)-M898 slip: 350m
6. M8(E)-M898 slip: 537m

This illustrates the pitfalls of map tracing. Not one figure quoted here is labelled "length of the M898" or "length of the M898 (northbound)". The Transport Scotland document does quote lengths, but unfortunately not only for the M898.--Nilfanion (talk) 21:53, 4 April 2013 (UTC)

I am quite happy to accept these distances (they appear to be properly reasearched) but am ambivalent as to how we should interpret them. What, therefore is the length of the M898 - 1177 m or 1206 m or should we include slip roads giving four different lengths ranging from 1527 m (1177 + 350) to 1860 m (1206+654)? If one looks at the Northern terminus of the M1, one sees a difference of a little over 1 km (I measured 1050 m on Google Maps) between the northern ends of the northbound and southbound carriageways. Martinvl (talk) 06:05, 5 April 2013 (UTC)

I don't see how we can use the numbers above without original research at this time. If a reliable source quoting the length can be found, we can use that. If a reliable source can be found describing the route (eg the M898 starts at the split with westbound M8 and runs to the A898 at junction 1) then we can use that to interpret the numbers.

What those figures do confirm what I already stated with regards to Meridian: Its a generalized product, prone to the coastline paradox and ultimately unusable as a primary source for road lengths. They also strongly indicate the 1/2 mile / 800 m length quoted in article is wrong too. Changing the length to "under 2km" or just "very short" would be justified.--Nilfanion (talk) 21:34, 5 April 2013 (UTC)

That is the reason that I prefer citing values from driver location signs - they show exactly what is there, can easily be veruified using Google Earth and if the reader wants to quickly check where he is (for example is caught up in a traffic jam and want to plan an alternative route), he can call up Wikipedia on his iPad and cross-check his position. (I have actually done that over the phone when my wife was in such a position with a number of girl-friends - they wanted an alternative destination for their sat-nav, but weren't too sure which junction they had just passed). Martinvl (talk) 06:58, 6 April 2013 (UTC)

The DLS are no good either for this sort of data, even if they are accurately spaced. Their zero point is notional, so you need a source to tell you a DLS saying "M898 A 0.0" is actually at the start of the M898 northbound. Likewise "M898 A 1.5" immediately before the end-of-motorway sign doesn't imply the road is 1.5 km long. The M898 is particularly problematic, as it doesn't have a distinct southern end and as its such a short route the exact start point matters. In contrast, both ends of the M4 can be easily be precisely located (as both ends are onto A roads).--Nilfanion (talk) 11:19, 6 April 2013 (UTC)

On the other hand, if you want to know the distance from Junction A to Junction B, rather than the length of the whoel road, the notional start of the road does not matter - you subtract the two. Moreover, if you are using DLSs you can check you progress as you drive - something that I do all the time. I have alkso been in the situation where I have had to make a decision - do I use the next services that are coming up or do I drive on to the following services. Letting the navigator collate the DLS that you have just passed with the DLS that is next to the relevant services allows you to make an informed decision. Martinvl (talk) 12:06, 6 April 2013 (UTC)

Wikipedia is an encyclopedia, it is not an aid to navigation. What matters is providing the encyclopedic info to people who want to learn about the road, not telling someone how far they have to go to get to J28. If we do supply usable navigation info, that's a serendipitous benefit not a designed feature. Likewise, if improving the quality of the encyclopedic info means losing the navigational utility, then that's what happens too.

This is a case in point - the encyclopedic info is how far from start to Junction X, something that rarely matters when you are using the road. The distance between junction X and junction Y isn't normally that important in encyclopedic terms, but is the critical factor for navigation.--Nilfanion (talk) 12:23, 6 April 2013 (UTC)

If we have a distance column in the RJL, we need to make it as useful as possible to as many people as possible? What purposes do you see in the distance column of the RJL? I want to use it for navigation. What benefit are you bringing to Wikipedia by depriving me of that use. The column is already there, so it can't be a space issue. Martinvl (talk) 13:04, 6 April 2013 (UTC)