The new range of canals and barges is a great addition to the pakset, especially for early years. So is the new collection of steamships. There are a few inconsistencies and issues of balance still remaining though.
- Connecting canals to rivers often does not work as expected, with a section of river being unexpectedly (and sometimes, very expensively) replaced by canal. In some cases I have had to resort to building the canal up to (but not connecting with) the river, and then switching to the public player to build a (zero cost) connection using a river tool instead. While a player should always have the option of converting a river to a canal (via control-drag), this should not happen automatically unless the canal is 'bigger' than the river: nothing should automatically overwrite a large river, only a large ship canal should overwrite a medium river, and only a ship canal or large ship canal should overwrite a small (but navigable) river; of course, any canal should overwrite a non-navigable stream. This may in part be a code rather than pakset issue, but I suspect it arises because in some cases the canal appears 'better' (i.e. either higher speed or greater weight limit) than the corresponding sized river; each river should be at least as good as the matching canal.
- The weight limit for large rivers has reverted to 999 tons. It needs to be much larger than this (at least as big as the 12000 ton limit for a large ship canal).
- Aqueduct and tunnel costs are slightly odd. All three canal aqueducts cost the same, so there is no point building the smaller ones. The narrowboat tunnel is very cheap to build (500) but very expensive to maintain (400); the corresponding figures for the barge tunnel (8000 to build, 256 to maintain) seem much more plausible; I suggest making the narrowboat tunnel costs 50%-60% of the barge ones.
The overall balance between narrowboats and barges feels about right: barges give more capacity for the same running cost, but with substantially increased infrastructure costs. But tub-boats are too good: even a single tub-boat has only slightly less capacity than a narrowboat, with significantly less infrastructure cost and a much faster loading time; multiple tub-boats are even more unbalanced (see the next point). Historically, narrowboats were (and are) overwhelmingly the canal boat of choice in Britain, and the game balance should reflect that; tub-boats should be actively preferred only in a few marginal cases.
Is it really intended that tub-boats can be chained together, with two or three pulled by one horse? This is constructable in the depot, but fully laden is too heavy for a tub-boat canal. On the other hand, it can travel along a narrowboat canal, achieving the capacity of a barge at a discount price, and still having the faster loading time of a tub-boat.
The smaller sailing boats (wherry, Humber keel, etc.) now have a very restricted range of operation, being completely confined to large waterways. Let me again urge that such craft should be allowed to use shallow or coastal open water; I do realise that this would need a code change.
Steamship running costs are wildly erratic, varying from derisory to eye-watering even within ships of the same type (e.g. Clyde puffers).
Passenger loading times for some ferries are unreasonably long, especially compared to their intended journey times. For example, an iron paddle steamer takes 40-45 minutes to load passengers, but its maximum comfortable journey time is only about the same length. It's fair enough to take an hour or more to load trans-Atlantic passengers and their trunks onto the Great Eastern, but 5-10 minutes should be plenty of time for people to get on board for a 45 minute ferry crossing.
Best wishes,
Matthew
Hello MCollett
What for a Simutrans Exp version and what for Pakset version have you?
Quote from: greenling on June 09, 2013, 08:02:05 AM
What for a Simutrans Exp version and what for Pakset version have you?
The latest version of each from Github (11.x commit 0567cc5 and master commit 2716641 respectively).
I've realised that the small sailing boats are even worse off than I thought: although both wherry and keel are theoretically available from 1750, there is no (non-cheating) way to use them before the appearance of the ship canal, because there is no way before then to construct a depot that they can move out of. Ship depots must be either on open water, or at the end of a waterway; the small sailing boats cannot move across open water or on small waterways, and there are no naturally occurring large waterway termini. (The cheating solution is of course to use the public player to build a short spur of medium or large river and then put a depot at the end of it.)
Best wishes,
Matthew
Thank you for your feedback. General balancing has yet to be run, and will be a task of gargantuan proportions, which will need several very substantial code changes before it can be done properly, which code changes are in turn waiting in a long queue behind other challenging work.
As to the loading times, do you have any historical information on how long that it took the coastal paddle steamers to load with passengers, or, for that matter, on the actual construction costs of tub boat canals as opposed to others? Much of the canal prices come directly from the Standard version, and because I have not begun the task of balancing yet, remain unchanged from there.
As to the keels and wherries issue, that is a difficult one to solve: they cannot sensibly be let upon barge canals, as that would unbalance them signficantly, yet allowing them in shallow waters (did they in fact sail in shallow coastal waters?) would require a substantial code change in circumstances were there is a mounting queue of important and challenging coding work.
As to the river weight limits, the max_weight=12000 line is already set in the large river, so I am not sure why you report that it has "reverted" to the default of 999 tonnes.
Would it be possible, or desirable, to ignore weight and way-type restrictions, on any single tile that holds a depot? It does seem pointless to be able to build a vehicle and then not be able to move it onto an adjacent tile.
That is an interesting thought - but should people really be able to build railway depots on very light unelectrified track, and have heavy electric locomotives emerge from them; or build a large ship on a depot built on a tub boat canal?
Supposing the suspension of waytype restrictions for movement within depot tiles -- the "what vehicles can you build" logic should stay unchanged, so electric locomotives would not even appear in the dialog; but yes if you built a large ship in a depot that had been built on any canal, you could at least move it to the neighboring tile; obviously a large ship would not move past the depot onto a tub boat canal.
Quote from: MCollett on June 09, 2013, 02:11:25 AM
The new range of canals and barges is a great addition to the pakset, especially for early years. So is the new collection of steamships. There are a few inconsistencies and issues of balance still remaining though.
- Connecting canals to rivers often does not work as expected, with a section of river being unexpectedly (and sometimes, very expensively) replaced by canal.
This is because of the way in which Simutrans determines which way is "better", which is entirely based on speed limit. So it thinks the canal is "better" than the river, which it often *isn't*. In the case of canals, it should be based on weight limit or restriction types instead. That's a program bug and we should fix it.
Regarding a number of the other problems, they would most easily be solved by allowing ship canals earlier. Short ship canals *were* dug earlier, they were just ferociously expensive and nobody could build locks.[/list]
The point about ship canals is interesting - when were the early ones built? As to locks, I do not think that there is currently any way to have a canal that does not allow locks, although this would be a useful feature to have.
By 1750, marine engineering had advanced to enlarge a river by digging; in Holland, polders to reclaim land from the sea had already been being built for centuries.
For example, a section of the River Dee in Wales was enlarged by 1736 to permit shipping to the city of Chester -- see: http://en.wikipedia.org/wiki/River_Dee_(Wales)#Canalised_section (http://en.wikipedia.org/wiki/River_Dee_(Wales)#Canalised_section) . Perhaps we could add an "Early Ship Canal" with extremely high cost, available starting 1750.
It's worth noting that when Experimental incorporates new landscapes ship canals only real function will be to change level. Adding water climate to a tile should make it unrestricted...
W. Lindley - enlarging a river is a different operation to building a ship canal from scratch, not least because no surveying of the terrain for a suitable route is involved.
Kieron - it does not seem sensible to let players add a "water" climate to a tile for free. If this were incorporated into Standard, I should have to restrict it to the public player in Experimental. It would have no basis in reality and would seriously interfere with the basic economics of canal building. It does not make much sense that players other than the public player should be able to make arbitrary changes to the climates of any given tile.
Indeed - at the moment I made it free to use for purposes of testing but it is something I'm going to have to consider.
The 1732-1736 River Dee canal was indeed a ship canal with an entirely new alignment. From Wikipedia:
QuoteWest of Chester, the river flows along an artificial channel excavated between 1732 and 1736. The work was planned and undertaken by engineers from the Netherlands and paid for by local merchants and Chester Corporation. It was an attempt to improve navigation for shipping and reduce silting. Chester's trade had declined steadily since the end of the 17th century as sediment had prevented larger craft reaching the city.
After four years' work, the river was diverted from its meandering natural course which passed Blacon (http://en.wikipedia.org/wiki/Blacon), Saughall (http://en.wikipedia.org/wiki/Saughall), Shotwick Castle (http://en.wikipedia.org/wiki/Shotwick_Castle), Burton (http://en.wikipedia.org/wiki/Burton,_Ellesmere_Port_and_Neston) and Parkgate (http://en.wikipedia.org/wiki/Parkgate,_Cheshire) and up the west shore of Wirral (http://en.wikipedia.org/wiki/Wirral_Peninsula). Instead the new canalised section followed the coast along North East Wales. During this time, Sealand (http://en.wikipedia.org/wiki/Sealand,_Flintshire) and Shotton (http://en.wikipedia.org/wiki/Shotton,_Flintshire) were reclaimed from the estuary (http://en.wikipedia.org/wiki/Estuary). Land reclamation in this area continued until 1916. The river's natural course can still be determined by following the bank and low bluffs that mark the western edge of the Wirral Peninsula.
The man-made channel, which runs in a straight line for 5 miles (8.0 km), passes beneath three road bridges...
Google maps view here. (https://maps.google.com/maps?q=chester,+wales&ll=53.197807,-2.969055&spn=0.170911,0.308647&oe=utf-8&client=firefox-a&hnear=Chester,+United+Kingdom&t=m&z=12)
As a purely practical matter of gameplay, it is currently impossible to use any of the wherries or the Humber Keel until 1822; it is impossible to build shipyard on any tile which would let them operate onto an adjoining river tile.
Duplicating the existing entry in canal_ship.dat, with an introduction date of 1750 and retire date of 1822, and a cost of 6000 (above the cost of an artificial slope) would solve the gameplay problem in an historically accurate way.
That is a very interesting article, and some very useful research - thank you.
Thank you everyone for all of your feedback and suggestions on this. Now that I have fixed some of the major code bugs and released the latest RC, I have had the chance to do some work on the pakset, starting with merging Nathaneal's latest changes (and thank you Nathaneal for all the work on that), and continuing to canals. Turning to the original poster's issues in numerical order:
(1) this is indeed a code issue, as Nathaneal has pointed out, and will need to be addressed separately;
(2) this turned out to be a code issue: this has been fixed in my 11.x branch ready for the next release;
(3) I have overhauled the canal tunnel and aqueduct costs as suggested;
(4) I have (hopefully) fixed this by greatly reducing the capacity per tub boat, which was formerly far too high, to what is, based on research, a realistic level;
(5) this is definitely intended - this is exactly how tub boats worked in reality: see here (http://en.wikipedia.org/wiki/Tub_boat) (the weight issues should be solved by reducing their capacity);
(6) this would indeed require a code change; however, I have added, as suggested, an earlier (but more expensive) ship canal to deal with the inability to build these boats;
(7) all but one of the Clyde puffers did not have their costs properly set; I shall have to leave the more general balancing to a later time when balancing is covered more comprehensively, however;
(8) I should still be interested in historical information about this before deciding whether to make any changes - remember, this is not just the embarkation time, but also the time for the ship to reload with coal, water, food for the on-board catering, be cleaned, etc..
I like the idea of not being restricted by the way type beneath the depot.
1) Once the depot is built, how do you know / upgrade the way type?
2) What does it matter if you built a cheap way type at first? You're going to pay for the depot and its infrastructure anyway.
3) Practically speaking, your depot (enginehouse / shipyard) should have/adapt internal facilities adequate to the vehicles being served.
4) Seems frustrating to users who want to see what electrical vehicles are available, but can't see any, unless they electrify ... then you can see all electrics, not just those supported by the electrical type you just ran.
2c. Thanks!
-danny
I just noticed that my passenger docks handle mail, but my passenger canal quays do not ... there's also a big discrepancy in the CapEx and OpsEx between these two. 1813. I guess I have to build a post office or goods storage in some towns ...
-danny
Hello
I can those problem confirm.
Quote from: dannyman on July 14, 2013, 01:19:04 AM
I like the idea of not being restricted by the way type beneath the depot.
1) Once the depot is built, how do you know / upgrade the way type?
2) What does it matter if you built a cheap way type at first? You're going to pay for the depot and its infrastructure anyway.
3) Practically speaking, your depot (enginehouse / shipyard) should have/adapt internal facilities adequate to the vehicles being served.
4) Seems frustrating to users who want to see what electrical vehicles are available, but can't see any, unless they electrify ... then you can see all electrics, not just those supported by the electrical type you just ran.
2c. Thanks!
-danny
The way type beneath a depot can be checked by clicking on the depot when the depot window is already open. It can be upgraded by dragging the newer way type from a neighbouring tile to underneath the depot (and can be electrified in the same way).
As to seeing electric vehicles, that is an interesting point - perhaps the "show all" behaviour could be adjusted in this respect.