Now be kind, dear reader, - this old fool is on a vertical learning-curve...!
Just over a year ago, when I 'accidentally' embarked (Put it down to a lack of self-control...!) on this project, I had no locos whatsoever - absolutely zip - not even the twinkling of an eye. I now have well over a hundred, and it would be hard to imagine a more eclectic collection - and no, they are not all expensive locos. Many were dirt cheap.

OO;- Most, perhaps 80%, of the OO locos are RTR, albeit modified in some cases.

P4;- The P4 locos are either RTR suitable for conversion (Few.), or, more usually, kits converted or 'to convert' to P4, a few are EM with already compensated chassis, so I hope relatively easy conversions to P4.

With the exception of the GW Diesel Railcars, they are all steam, since I'm only interested in pre-1947.

Diesel;- There will be GWR Diesel Railcars in both gauges. However, I have, for entirely practical reasons, acquired a small fleet of diesel-prototypes from the BR era. These are my 'Mules' for track testing. I chose these for a number of mainly practical reasons;-

1. They do not have delicate details to be easily broken.
2. To save wear & tear on the steam types.
3. They have lots of wheels and therefore adhesion. (More anon....)
4. They are very simple mechanically.
5. They are easy to maintain & modify.
6. They are cheap as chips.
   

I chose the Mainline early green Type 4, 1Co-Co1 / Type 40/45 (I'm a bit confused on that...!), as I was lucky enough to find several that were NOS (new-old-stock). This was primarily because they have a lot of wheels - (......and of all the diesels, these were a type that I often used to travel behind many years ago, on late night mail trains. I think the tickets must have been cheaper...!). However - I don't actually have any affinity with non-steam whatsoever - these 'Mules' will have to work hard to 'prove' the trackwork. Now - the more modern/diesel-oriented reader will immediately realise that despite appearances, these models really only have a small four-wheel driven part of only one bogie. Typically, I didn't realise this..! (Doooh..!) so there may well be much more suitable 'Mules' out there, waiting for me to discover them. My plan for my 'Mk.I Mules', was, if necessary, to put two motor-bogies in each shell and fill the rest of the shell with lead-shot. Two ran OK, but a third used example ran poorly, so I took it apart to service the motor. It ran better afterwards, but it did reveal what a Heath-Robinson design they are.

Generally, I have been very impressed with the modern RTR - and this was really what originally drew me back into railway modelling. Apart from the obvious improvement in scale-appearance over nearly fifty years, it's the vastly improved running that has been most noticeable. The modern Hornby locos seems pretty good, the Bachmann - even better. Both are generallysmooth and quiet - and that's even before adding DCC.

Then there are what I term as 'Middle-Aged' locos, such as older Hornby, Lima & Mainline etc. I suppose these were produced very roughly between the late 1970's to around 2005. These seem to be quite good scale-wise - especially the Mainline. However, these latter locos do seem to run very very weakly (-and noisily...!) for some reason - and I may have to look at the possibility of re-motoring these if practicable, as they usually look pretty good otherwise. Types with congenital problems will need to be weeded-out and disposed-of.

Finally there are the real oldies.... Now some may think it rather odd that I would even countenance running these when I have modern OO, let alone P4 running too. Here I would refer folks back to the original concepts that I'm striving to meet;- The OO is primarily for outdoors - where running scale-length trains and entertainment are the prime objectives. - ie; The 'impression' and practicality are uppermost. For these locos, power and reliability are paramount, but they will be variously upgraded insofar as it is practical to do so - and they'll all be DCC - fitted.
It'll be tough on the locos outside too, and the old ones are dead-easy to maintain. It's dirtier outside - and then there will be the unavoidable gradient. Poor scale attributes aside - and the fact that some are rather noisy (Less of an issue outside.), many of those old locos are real sloggers. For example, I have an old Hornby Dublo 8F, re-wheeled with Romfords and upscaled with added detail. That's ideal, it's got power, weight and lots of wheels, - so lots of real grip. Some of the OO are metal kits, and those too have the weight to get some decent grip. Weight is a biggie, after all, and it was the only reason the old 'singles' worked at all. (Don't get me going on Singles...).

The P4 locos are nearly all kits with just a few converted/convertable RTR thrown-in. They will not have to contend with the dirt, gradients and long trains that the OO locos will, and the Branch will also be much slower paced, with - probably - mechanical point operation and signalling. All the rest - the OO, indoors and outside, will be electronically-controlled, via a PC. I have some very interesting locos that I'd really like to make P4, but there is no historical precedent on the Branchline for doing so. That said - I have heard, anecdotally, that an errant Castle used the Branch during the war - but I have not seen any documentary evidence for that unusual event....!

Hardly any of the non-RTR are actually truly scratch-built. They are either kits or what I term my 'Chimera', which are built from a mixture of kit, RTR and scratch assemblies. One example of the latter for example is an LMS Crab, which has an etched chassis, a Bachmann 'body' and a Hornby tender. It sounds odd, but is actually an excellent loco.

Of course, my original intent was that one lonesome GWR loco for the mantlepiece...... Now things have broadened-out a lot, fuelled by the freedom the garden railway offers - and my own unbridled capriciousness. There will now be many trains that could in no way be genuinely justified in the scenic area, but make a very nice rolling display of their own in the garden. This would include Caledonian, GNR, LNER and Southern - although the latter Grouping hasn't materialised - yet. I can foresee other colourful arrivals such as Chatam Wainrights and some LBSC perhaps.
This sudden plurality of interest has quite surprised me. This more diverse strand seems to have one primary limitation, and that is the lack of realistic appropriate RTR stock for some of the old Pre-Grouping companies. Of course I can built kits or even scratchbuilt, but we are back in conflict with the big bogie of time again..... So whilst I am building (some.) locos and stock for the OO part of this scheme, I will always go for RTR - or ready-built kits if I possibly can.
I have to say that the average built-up kits that I have acquired are pretty poorly made. Some of the more irredeemable locos will be sold-on. I will only keep those requiring modest remedial work.
In some cases, knuckling-down to building is really the only option - but I have to ration these. For example, I have a DJH kit for a Midland Fowler 0-10-0 'Big Bertha' for that garden 'Lickey Incline', as well as a number of otherwise unobtainable Singles of the GWR, LNWR & MR etc.

I'm as yet too inexperienced with current parts and assemblies to make any serious assessment of motors and gearboxes etc. There are a lot of older kit-built locos out there. Perhaps around a third run terribly, if at all, perhaps a similar percentage run well-enough for pulling, but have very poor starting/slow-running performance. Less than a third might be describes as 'good', and none thus far as 'excellent'. Clearly, a good motor, a well built gearbox and a suitable gear ratio are paramount before one even looks at fangled electrickery.
One thing that has puzzled me is that I haven't seen any mention of the use of Coreless motors. I wonder why....? These are commonplace in model aircraft, drones and boats.

I have alluded earlier to Singles. These are a particular peccadillo that I have developed with my rediscovered interest in model railways. I have a couple of dozen of one sort or another. These range from the dear old Triang types, through modified versions, through kits, to the latest - yet to be delivered - Locomotion Stirling RTR. This latter gains additional traction from a powered rear axle-set. What it will pull remains to be seen, but the single-axle drivers are very limited in what they will pull. I'm planning to add as much weight to the locos as possible to ameliorate this issue. I have examples for the GWR, LMS, LNWR, Midland and Caledonian and await the GNR Stirling with great anticipation.

Finally, - DCC. This is all new to me, and as I've touched on elsewhere, I have collected some Lenz gear. This on the basis that it was the originator, it's German, - so it'll be well made, and it generally also has a good reputation.
The decoders - at least basic non-sound examples, now seem to be quite cheap. The sound-decoders are quite another matter. Whilst the small speakers seem basically cheap, the sound-decoders themselves are crazy prices. For this reason, I have been speaking to some contacts with regard to - just perhaps - producing my own. I'm certainly not going to spend £10,000 or more just on decoders. Of course, this may well be a complete non-starter, but don't bet on it..... Can they be produced..? Of course. Can they be made small-enough to a price..? ..ummmm maybe. Can they be produced economically..? No idea whatsoever....yet. My target is to get them under 50% of the cost of the commercially-available examples - but I'm in no mad rush. I'm still hoping that someone else will get some cheaper ones made and save me the hassle....!
The P4 Branch, may just possibly, remain analogue, as operation is so simple, but the rest is intended to be automated DCC.
I have downloaded an App' that I have on all my Apple devices, which is called 'TouchCab'. I have Wi-Fi boosters anyway, so all of my property, including all of the garden, falls well within good Wi-Fi coverage. This means that, in theory at least, I can use my Mac, iPad or iPhone anywhere on site to effect basic control via the interface on the Lenz System which I have already acquired. As yet, I do not know if the TouchCab can cope with the automation itself - or whether I will need additional software. At the moment my plan is to built a section of the upper, scenic level of the layout for testing locos and the Lenz DCC system to get a feel for what is achievable. As I have mentioned previously, I'd originally fancied the very well-presented Bachman system, but there was a plethora of comments online to the effect that the Bachman's control range was very limited, even indoors, which would be of no use to me whatsoever, sadly, as the system looks very user-friendly.
How then, will the final design operate? My intention is that the laptop will run the OO network rather like one uses iTunes or similar. That is to say there will be various playlists. Each train will be called-out of its dedicated loop in turn, and returned there afterwards, just like a tune being played. The trains themselves will be set-pieces, and will stay coupled normally. The P4 will probably be three/screw-link for simplicity and aesthetics. I haven't finally decided for the OO yet. I'd ideally like to go three-link, but that's a lot of very fiddly work, and messing with all the stock will devalue it, especially if the toy couplings will no longer fit back on. I'll probably just leave most as they are. The intention is that the train and it's constituents will reflect the typical load for that loco in that period, as it is liveried.
The P4 Branch is fairly simple. I'm hoping that I can arrange some basic interlocking so that when the signals are changed to allow a branch train to cross the main-line, this will trigger the main-line trains to stop automatically - or prevent the signals moving if a train is already in the block etc. With relatively few trains in the main OO lines on the upper, scenic level, I doubt that this will present a major issue.
The main gyrus on the lower level is four-tracks. There will be no points outside whatsoever, and about 95% of all the points will be associated with the storage-loops on the lower, gyrus level. The main running lines will have only three junction areas. One will be the exit & return for the garden lines. The second and third will be to launch & recovery points from the storage loops inside the main gyrus.
However, to keep it entertaining in the garden, since it's quite long, the garden-loop ideally needs to be able to cope with several trains running on any loop at any given time. Let's say, for the sake of discussion, there is an outbound, a return, and one inside the building, so three. That means the PC-based program will have to keep them separate. This might be via speed modulation or stopping - depending on how clever the PC program is. Clearly, if some sort of simple block is in operation, then trains cannot enter the block ahead if the preceding train has not cleared it.
There is only one pair of lines up to and down from, the scenic section. This will be fed from a separate inner gyrus on the lower level, still with access to the main storage loops, so that trains called-off for the scenic section do not have to cross the main gyrus which runs around the outside, so less conflict and less pointwork. One side of this inner gyrus will have 180-degree loops to the scenic section to ape two-way running, but in reality, all the gyrus lines will run clockwise. This keeps both trackwork and control much simpler, as electrically, there is no reversal.
How the DCC will cope with the garden loops I have no idea. I'm working on the presumption that I will need a regular boost-points to keep-up the energy supply. Clearly - this is an issue that others, imbued with greater wisdom than I, will have faced and resolved, as many can be seen running on You Tube, so I am hoping that some kindly souls will remedy my paucity of knowledge.
There seem to be many unknowns. For example;- Would a rake of lit coaches noticeably reduce the performance of the loco....? Would this require more closely-spaced boost-points......surely it must.....? Likewise, the use of sound and lights on the loco must use more energy, albeit quite small.

I've basically got all the locos I want now, and all the faffing-about with the kits and rebuilds will have to wait until I have got the basic set-up operational. In the meantime, I am setting-about a couple of boards set into a loop to allow setting-up and testing of the fangled electronics. More as and when I get that running....