Southern Coach Sets - Improving The Excel File from SEMG

[Ian J.]

Hi all,

 

I'm starting a project to 'clean up' the Coach Sets Excel file from SEMG (link on this page https://sremg.org.uk/coach/sets.html) to make it easier to interrogate the information.

 

I will eventually put the data into an Access database file, where, if the data can be structured well, it should be possible to enter a set of parameters (one or more of set; dates; locations; services; etc.) that should allow for understanding what ran where at a given date (possibly not to the day, more likely the month).

 

Any help in understanding the data would be gratefully received, as it's in a bit of a mess. I'm currently just looking at the 'Area' column, and there's lots of variants, typos, etc, and abbreviations that aren't explanatory.

 

I'll reply here as and when I make any progress.

 

TIA

[Ian J.]

Due to an issue with NFL GamePass not working properly tonight, I've had a bit extra time on the data.

 

It looks like there are quite a few carriage type conflicts between sources in the spreadsheet, so I'm going to have to resort to checking them against what I have in the Mike King book I have. I have another book on Maunsells that may also provide some extra help. Whatever is left I will post back here for a crowd source check 🙂

[Wickham Green too]

David Gould's SECR book provides loads of information on that railway's sets into SR days and beyond then there's Weddel's series on the LSWR and Ian White's new LBSCR series .................. how many lifetimes were you thinking of spending on this project ?

[Ian J.]

7 hours ago, Wickham Green too said:

David Gould's SECR book provides loads of information on that railway's sets into SR days and beyond then there's Weddel's series on the LSWR and Ian White's new LBSCR series .................. how many lifetimes were you thinking of spending on this project ?

 

Fewer than one... 🤣

[Ian J.]

OK, an example of having data in a better, more structured form, means transcription compromises can be found:

• Mike King's book 'An Illustrated History of Southern Coaches' has a table on page 14, Table 1, noting the stock from pre-grouping as re-numbered by the SR in 1924.
• Mike King's book 'Southern Pull-Push Stock' has a table on page 70, Table 3, noting ex LSWR non corridor Pull-Push sets.
• The SEMG spreadsheet has as its first set number 1, noted as ex LSWR non-corridor Pull-Push, as a BT-C formation with vehicles 2620 and 6488 respectively, matching the Table 3 entry from MK's Pull-Push book.

That Table 3 has column headings for the 'BT' and 'C' vehicles of the sets listed. Taken literally, you'd think 6488 was a 'C', but it's not. The cross reference to Table 1 from MK's Southern Coaches book puts 6488 as a 'BC'. Checking the Pull-Push book in more detail on the chapter for LSWR non corridor sets, it turns out that the diagram for 6488 was 419, the drawing of which is of a 'BC'. So, presuming that the drawing for diagram 419 is correct, the way Table 3 has been laid out causes an interpretation error due to the 'C' column not differentiating between 'C' and 'BC', and leads to a mistake in the SEMG spreadsheet where the listing for Set 1 is as BT-C, when in fact it should be BT-BC.

 

Ideally, the column heading 'C' in Table 3 should have been 'C/BC', then for anyone checking, it would be apparent that it was necessary to do a bit of further reading to know exactly what the type of vehicle 6488 was.

[kevinlms]

My head hurts!

[45655]

12 hours ago, Ian J. said:

OK, an example of having data in a better, more structured form, means transcription compromises can be found:

• Mike King's book 'An Illustrated History of Southern Coaches' has a table on page 14, Table 1, noting the stock from pre-grouping as re-numbered by the SR in 1924.
• Mike King's book 'Southern Pull-Push Stock' has a table on page 70, Table 3, noting ex LSWR non corridor Pull-Push sets.
• The SEMG spreadsheet has as its first set number 1, noted as ex LSWR non-corridor Pull-Push, as a BT-C formation with vehicles 2620 and 6488 respectively, matching the Table 3 entry from MK's Pull-Push book.

That Table 3 has column headings for the 'BT' and 'C' vehicles of the sets listed. Taken literally, you'd think 6488 was a 'C', but it's not. The cross reference to Table 1 from MK's Southern Coaches book puts 6488 as a 'BC'. Checking the Pull-Push book in more detail on the chapter for LSWR non corridor sets, it turns out that the diagram for 6488 was 419, the drawing of which is of a 'BC'. So, presuming that the drawing for diagram 419 is correct, the way Table 3 has been laid out causes an interpretation error due to the 'C' column not differentiating between 'C' and 'BC', and leads to a mistake in the SEMG spreadsheet where the listing for Set 1 is as BT-C, when in fact it should be BT-BC.

 

Ideally, the column heading 'C' in Table 3 should have been 'C/BC', then for anyone checking, it would be apparent that it was necessary to do a bit of further reading to know exactly what the type of vehicle 6488 was.

It’s an easy mistake to make. PP sets 1-6 were I think unique in being converted from existing 2 sets and retaining a van at each end.

 

Pretty much all other PP sets had a single van at the outer (driving) end.

 

Keith

Alton.

[Wickham Green too]

50 minutes ago, 45655 said:

... Pretty much all other PP sets had a single van at the outer (driving) end. ...

A ( the ? ) notable exception being set 659 which had a 'birdcage' brake at the non-driving end.

[Ian J.]

I have the feeling this is going to be quite a big project, as it's looking likely that I'll have to do quite a bit of digging into resources to check and straighten out the data. To some degree, that makes the SEMG file moot, but it's still a good starting point. The next thing on the agenda will be acquiring copies of Gould's Oakwood Press books beyond the one I already have (Maunsell's SR Steam Carriage Stock), so that I at least have the resources to hand.

[Oldddudders]

Do you have the Longworth volume? I can't recommend it as a general intro to Southern coaches, as it lacks the photos and drawings of King, the compactness of Gould, but it has a comprehensive listing of vehicles and sets.

[Ian J.]

2 hours ago, Oldddudders said:

Do you have the Longworth volume? I can't recommend it as a general intro to Southern coaches, as it lacks the photos and drawings of King, the compactness of Gould, but it has a comprehensive listing of vehicles and sets.

 

I don't have it, but it sounds like it could be very useful.

[Wickham Green too]

The new edition of Southern Way ( No.59 ) contains the second half of a copy of Mr.Longworth's listing ( first half in No.58 ) ......... I hope the list includes a lower percentage of errors than the accompanying photo captions : No, Dugald Drummond never designed a railway carriage - he was a Locomotive Engineer ; Set 920 clearly included at least one re-framed LSWR carriage among the SECR vehicles ; Horsham on p.55 ... looks rather like Tonbridge to me ! ; no comment on the conspicuous absence of a 'birdcage' on set 600 .....

 

 

[bécasse]

Access database files can't be opened on a Mac. I doubt whether I am the only RMweb user with Southern interests who only uses Apple products and so wouldn't be able to read your updated file - which isn't to say that I don't think that updating it isn't a worthwhile activity, it clearly is.

[Ian J.]

Due to sources of information coming from the works of Gould and King, etc, and those being in copyright, I don't think it would be wise for me to make the Access file publicly available. It would be more like I'd have the file on my PC, and could use it to look up information based on requests. How much detail could be included in a request output I'm not sure at this point. There's a long way to go before I need to worry about that though.

[kevinlms]

3 hours ago, bécasse said:

Access database files can't be opened on a Mac. I doubt whether I am the only RMweb user with Southern interests who only uses Apple products and so wouldn't be able to read your updated file - which isn't to say that I don't think that updating it isn't a worthwhile activity, it clearly is.

It isn't exactly easy to export properly from an Access file, even within Windows software.

 

Access isn't a bad database, but once it's there you are sort of stuck.

[Graham_Muz]

11 hours ago, Ian J. said:

Due to sources of information coming from the works of Gould and King, etc, and those being in copyright, I don't think it would be wise for me to make the Access file publicly available. It would be more like I'd have the file on my PC, and could use it to look up information based on requests. How much detail could be included in a request output I'm not sure at this point. There's a long way to go before I need to worry about that though.

As the source information in the books you mention is in itself tabulated / represented information from other sources (and much of  information in one book is already an expansion of another)  your subsequent collation, combining  and reformatting from multiple sources would not be subject to copyright. 
 

[Ian J.]

1 hour ago, Graham_Muz said:

As the source information in the books you mention is in itself tabulated / represented information from other sources (and much of  information in one book is already an expansion of another)  your subsequent collation, combining  and reformatting from multiple sources would not be subject to copyright. 
 

 

Really? A lot of research work goes into authoring books the likes of Gould's and King's, and I'm sure they (or their estates/publishers/etc.,) wouldn't want to feel that their effort could be negated by the same information appearing in online form for free. If indeed it isn't subject to copyright, I think I would need some kind of confirmation that if I made the Access file available I couldn't be sued.

Edited October 30, 2022 by Ian J.

[Wickham Green too]

It's not the information that's copyrighted in those sources as that, itself, has come from origins elsewhere ...... copying any comments, text or specific formatting might be problematic.

[Ian J.]

2 minutes ago, Wickham Green too said:

It's not the information that's copyrighted in those sources as that, itself, has come from origins elsewhere ...... copying any comments, text or specific formatting might be problematic.

 

Agreed, but aren't there examples where occasionally information is deliberately in error in a given source so that it can be identified as having come from a particular source? Also, notes are sometimes where information is given, and knowing how to decipher the information from the note isn't always obvious.

 

Lots of 'traps' to consider...

[Wickham Green too]

Indeed, I believe A2Z maps always included a deliberate error - but that's an error in information they've gathered, themselves, on the ground ...... If Hugh Longworth ( etc.) had been around at the time and personally collected all the information he's tabulated you might have a point about copyright - unless he was everywhere on the Southern and is now very old, he's assembled this information from disparate official & non-official sources which he, himself, didn't consider to be subject to copyright.

[Ian J.]

From some internet searching this morning, it appears that U.S. copyright law is quite clear that facts can't have copyright, but I'm not so sure yet about U.K. law. There is an article that I can't view the whole of as it requires a log in (https://www.jstor.org/stable/24866738) that discusses the issue. I obviously can't tell if it has a resolution, but I haven't been able to find anything regarding the issue as it pertains to U.K. law.

[drduncan]

You can register for a free jstor account - as I have and many other researchers do…

Duncan

[Ian J.]

OK, so I've read through the jstor article, and it's saying, unlike U.S. law, it's unclear just how U.K. courts decide on 'sweat of the brow'. It was written nearly thirty years ago in 1995, so there may well have been changes and clarifications in the meantime, not least of which is various revisions of copyright law in the E.U. (which may still apply here) and internationally with things like the Berne Convention, so I can't take its conclusions as a given.

 

For now, without any further knowledge of what may have changed in the meantime, I'll take it that, even though the facts themselves may not have copyright, there may be issues over 'sweat of the brow' in U.K. law on the work to collect facts together, and as such won't publish the Access database itself. If further clarification of the legal position comes to light that could show all would be OK, I'll reconsider then.

[PeterStiles]

The data cannot be copyrighted - because "any" person could compile that data themselves from Other Sources, especially if you need to augment the data by sourcing extra information from other sources. Plus as the format of the spreadsheet is not really loadable into a relational database without much rejigging I doubt anyone's going to come after you for *using* the data. 

 

There's an open source dude who releases a database with all the Formula One race stats in., and if anyone was going to go after someone on copyright reasons it'd be F1! All you're doing is the same as him - taking publically available _data_ and presenting it in a fresh way .

 

 

OTOH MS Access is a horrible format but unfortunately one that _can_ be read on virtually any platform (yes you can read the data on a mac - if I can read an MS Access on Linux you can read it on  mac....

[Ian J.]

I think the problem is that while the facts of the data cannot be copyrighted, there may be an issue with U.K. courts on the subject of the time and effort ('sweat of the brow') to gather and collate the data, and that's as against its presentation in the Gould and King books (which is covered by copyright). Of course, I could go and get the data again myself and that bypasses the issue, but one of the points made in the jstor article was that it might be unreasonable and perhaps even unwanted that new interpretations of facts have to be re-discovered by later users of said data. However, until there is legal clarification, I stand by not making the data file publicly available.

 

An Access file as a pure data file is openly readable by pretty much any system designed to connect to databases. What isn't compatible on other OSes are the interface elements (forms, reports, queries, etc). Their design and structure might be readable via DDL statements, but they can't run without the Access program which I think only runs on Windows.