GWR Journal Image Index

[kevinlms]

6 hours ago, Compound2632 said:

A number of bookshops carry backnumbers of GWRJ and other WSP titles - in other words, a good few issues are not yet out of print - so might be very discontent at digital page images becoming freely available. Furthermore, you have to consider that the copyright in the photographs and some other material will lie with others. 

 

An index to images, however, sounds a useful thing but what information are you going to include in each image caption? Often the thing I'm hunting for is not the subject of the photo but some detail lurking in the margin.

 

 

Yes I agree, a website that displays images will be breaching copyright, unless you have permission. The front cover, I suspect would be acceptable because it helps to identify the specific book/magazine. Beyond that - no way. Even a Contents page is iffy, because someone has compiled it and presumably not given permission to reproduce.

 

For the type of information that is intended to be stored, that is the key part of any index and it needs to be consistent. If you aren't then searches are likely to miss possibly important references.

Obviously spelling is very important and it isn't just typos, but different ways of writing something, i.e. is it 'Kings Cross' or 'King's Cross' station? I don't wish to debate the importance of spelling (I do know which is correct for the London one, but the answer for the one in Sydney, Australia is different!).

 

So some basic information is required. You can always include more, but I think the important thing is accuracy and consistency. These suggestions, probably do sound rather harsh, but I think you will benefit by following.

 

A Title - a brief title in less than 10 words. Can be more if necessary. Essential

 

An author/photographer. Could be a business name, in case of collections/museums etc. Essential Use Anon or unknown if you don't know, but don't leave blank or guess. You can edit it if more info comes to light later.

 

Guessing WILL get you into trouble some time in the future. If you think you know the author/photographer, put that into the notes/description, so you can find it later to edit updates.

 

Some keywords - these is great for a photo that contains more than a photo of a specific item, such as a loco. It could be at a particular station, or there is an interesting signal or whatever. Compile your index in a manner that you can have multiple keywords for an entry. Keywords can be numbers, not just for locos for instance, but create a keyword that means 'a GWR locomotive photo', or 'a GWR signal'. A hint I use is to start such keywords with a 0, or whatever you like, to separate them from a loco number. This way if you create a search on 'a GWR locomotive photo', you can just scroll through all the entries of such a photo.

 

In my index for GWR locomotive photos, I use 0711.  07 means prototype, 1 for locomotives and 1 for GWR.

GWR signal photos would be 07 (prototype), 8 (Signalling and track), 1 GWR. Or 0781.

 

Very handy if you don't know exactly what you are looking for, but you want to see what is available.

 

Not absolutely essential, but rather pointless without.

 

Prototype or it's scale if it's a model. Essential

 

Source of material. Title of book/magazine/other including year/issue number/page number ISBN etc. Essential - rather pointless, if you can't find it later!

 

Who created the entry. I just use the initials and have a separate list of what the initials mean. Makes it somewhat traceable. Not essential.

 

Last Edited Handy to find out when it was last updated.

 

Notes/Description  As much info as you feel is appropriate. Remember that the more information you provide is going to take longer to compile. Sounds obvious, but it's your time spent to write about something, that may not be that great! Maybe your time is better spent starting the next entry? - your call. Essential

[Harlequin]

9 hours ago, GWR_Modeller said:

Hi, Are you aware of this site?

  https://www.steamindex.com/gwrj/gwrj1.htm

All 13 volumes of GWRJ fully indexed, captions etc.  The only issue is each volume is a seperate file.

Regards, Paul

 

Thanks Paul,

 

That site is attempting to cover similar ground to this effort, but with a different rationale behind it.

 

As you point out, the indexes are divided into separate web pages and there doesn't seem to be a search facility. So I used Google's own site search and the acid test is to find references to Small Metros, like this: 'site:steamindex.com "small metro"'. That returns no direct results but then offers results that contain either of the two words, leaving you to pick out results that might be useful.

 

Edited February 19, 2023 by Harlequin

[Andy Keane]

10 hours ago, GWR_Modeller said:

Hi, Are you aware of this site?

  https://www.steamindex.com/gwrj/gwrj1.htm

All 13 volumes of GWRJ fully indexed, captions etc.  The only issue is each volume is a seperate file.

Regards, Paul

Yes I knew about this index, but still it does not have the picture captions so you cannot for example search for a particular loco number.

there are several overall index files which could usefully be merged but if you want to search for a picture none as yet do that

[Andy Keane]

11 hours ago, longchap said:

 

I received the following message when attempting to download:

 

File not downloaded: Potential security risk.

I think it’s fine. I downloaded fine.

 

[Andy Keane]

I have been out on the bike and pondering table entries irrespective of where and how we store them.

I get to this as a first shot:

 

Most of these are obvious but

all the images from a given page appear as a single entry

all the dates of images on a given page appear as a single entry

the "entries" cols are to allow stuff from other indexes to be pasted in

the three loco cols are to allow entries for specific loco numbers mention in an image or an a page to be recorded (allowing for three different locos)

ditto for carriages and wagons but without extra cols for individual entries

comments for free form input

errors for those identified in the source tex

Does this make sense?

The final plan is for the table to have one row for every page in the set of Journals so about 7000 in total, including front and back covers.

Andy

Edited February 19, 2023 by Andy Keane

[melmoth]

What about a dedicated column for the photographer/image credit? I sometimes remember that Peter Gray/R.C. Riley/A.N. Other took the photo I'm looking for, but have no idea where I've seen it.

[Pteremy]

Am I right in thinking that if every copy of GWJ were available in a digital format then literally anything of potential interest could be searched for with a standard search engine? Perhaps the publishers have digital versions derived from the printing process? 

[martinT]

18 hours ago, Andy Keane said:

By my count  there are 103 issues of GWRJ plus the initial Bumper Preview and Cornish Special.

There are then just shy of 600 articles in them!

Don't forget the 2 special GWR 150 issues of BRJ published in 1985.  168 pages in all. Articles:

No1:

The GWR at Maidenhead (Paul Karau, Mike Clark, Matthew Wells)

'The Great Bear' ( J E Kite)

The 'Iron Duke' class & its ancestry (Andrew Wiles)

GWR Standard Loco boilers (Eric Mountford)

The last outpost - the GWR at Crewe (Don Rowland)

'Metro' Tank controls (Rev D A Tipper)

A GWR Signalwoman (Don Froud)

The Brimscombe Bankers (Mike Fenton)

No2:

The GWR at Goring (Paul Karau, Mike Clark)

Ireland & the GWR (Stanley Jenkins)

A look round Swindon Works (A H Malan)

A rare drawing discovered (Graham Bone)   (for an early 517 0-4-2T)

Box Station (John Froud)

Slough (Charles Jones)

Three West Country scenes (early photos of Durston & Dawlish, & 'Dragon' at Newton Abbot)

GWR passenger train identification (John Copsey)

AEC RAilcar No.2 on trial (L E Copeland, Mike Christensen)

[Andy Keane]

6 minutes ago, Pteremy said:

Am I right in thinking that if every copy of GWJ were available in a digital format then literally anything of potential interest could be searched for with a standard search engine? Perhaps the publishers have digital versions derived from the printing process? 

Yes that would be so but back in 1992 digital copies were not I think taken. The alternative is to scan all 7000 pages and do OCR on them all - a very big ask and also likely to need a full set of magazines to destroy in the process.

Andy

[longchap]

1 minute ago, Pteremy said:

Am I right in thinking that if every copy of GWJ were available in a digital format then literally anything of potential interest could be searched for with a standard search engine? Perhaps the publishers have digital versions derived from the printing process? 

 

That would certainly be the logical conclusion if that condition were to become reality. In the meantime, the idea here is to provide a search facility to help the community find specific data to help with our own modelling projects, e.g. where can I find a photo of the particular loco I am building?

 

It should be worth it 😎

 

[The Stationmaster]

16 hours ago, Neal Ball said:

Plus!

We need to include the issues of Western Times beyond GWRJ as well. As it seems they are continuing in the same vein.

Take care with 'Western Times' as some of its photo captions are incorrect.  Even to the extent of a very obvious place where the photo was taken being misidentified!

 

I'll keep generally quiet about my view of certain items of text because there are still occasional 'howlers' and errors to be found .  One bad example is the stuff in Issue 4 about the Warning Acceptance a lot of it is imaginary and/or total nonsense.   Even the the piece about the relatively recent Reading diesel depot, especially subsequent changes, also contains errors and that is well within the living memory of people a lot younger than me.

[Andy Keane]

26 minutes ago, The Stationmaster said:

Take care with 'Western Times' as some of its photo captions are incorrect.  Even to the extent of a very obvious place where the photo was taken being misidentified!

 

I'll keep generally quiet about my view of certain items of text because there are still occasional 'howlers' and errors to be found .  One bad example is the stuff in Issue 4 about the Warning Acceptance a lot of it is imaginary and/or total nonsense.   Even the the piece about the relatively recent Reading diesel depot, especially subsequent changes, also contains errors and that is well within the living memory of people a lot younger than me.

Mike

we plan to include a column for errors to be noted so that people can enter these as and when they are found.

Andy

[Andy Keane]

I now see we already have 

ho hum! maybe we should have found this.

Andy

[The Stationmaster]

4 hours ago, Andy Keane said:

Mike

we plan to include a column for errors to be noted so that people can enter these as and when they are found.

Andy

Nice idea Andy - I don't think it will be needed very often for GWRJ.    It might be useful - if space allows - to adda  'notes'; space which provides for reference to subsequent letters perhaps?  Or will you simply include letters in the overall index?

[Neal Ball]

5 hours ago, The Stationmaster said:

Take care with 'Western Times' as some of its photo captions are incorrect.  Even to the extent of a very obvious place where the photo was taken being misidentified!

 

I'll keep generally quiet about my view of certain items of text because there are still occasional 'howlers' and errors to be found .  One bad example is the stuff in Issue 4 about the Warning Acceptance a lot of it is imaginary and/or total nonsense.   Even the the piece about the relatively recent Reading diesel depot, especially subsequent changes, also contains errors and that is well within the living memory of people a lot younger than me.

Oh! I didn’t realise that, thanks Mike.

[Andy Keane]

30 minutes ago, The Stationmaster said:

Nice idea Andy - I don't think it will be needed very often for GWRJ.    It might be useful - if space allows - to adda  'notes'; space which provides for reference to subsequent letters perhaps?  Or will you simply include letters in the overall index?

I am starting from https://www.steamindex.com/gwrj/gwrj1.htm#219-af which includes a lot of general notes and all stuff on the letters as well.

Its a bit slow but I have copied over the first of the 13 volumes today which gets the backbone there.

After that its a question of slowly grinding through the image captions - but I am hoping that can be a shared project if I set up the basic structure.

Andy

 

 

[Andy Keane]

I have now consolidated the  https://www.steamindex.com/gwrj/gwrj1.htm and Neal Ball data into a single spreadsheet starting from my table of contents.

This results in some 1180 individual lines in the final sheet and these can be search for any text string of interest.

I have also parsed the data for number in the range 1850-2000 for dates and 600-800 plus 2000-10000 for loco numbers.

While not perfect this picks out most of the relevant information - it's just tough if you need to find a loco number outside of these ranges as you get swamped by page or date numbers.

If anyone would like to help out by checking through this for errors please let me know and I will send them a copy.

regards

Andy

 

[longchap]

5 minutes ago, Andy Keane said:

If anyone would like to help out by checking through this for errors please let me know and I will send them a copy.

regards

Andy

 

I'm always happy to help Andy. Let me know if you want any specific areas checked.

 

[Andy Keane]

Mt travails with the GWRJ have made some progress. I have pulled together four different sets of data on the journals into a single data-file.

These are from:

1. The issue contents pages
2. the material on the steamindex.com web page
3. data kindly provided by Neal Ball
4. the spreadsheet by John Dolan on the GWR.org.uk website

I have sorted these into a mega contents file which is ordered by volume, issue and page - this can be text searched for anything of interest.

I have also parsed this data with a computer code looking for numeric strings and added columns that contain

1. strings between 1830 and 2000 which I have taken to be dates (but there will be some overspill because of things like drawing numbers that may contain these as substrings)
2. strings from 720 to 1830 and 2000 to 10000 that are bracket by spaces, commas or semi-colons which I have taken to be engine numbers (this may miss some numbers if they are bracketed by other characters)

Unfortunately there is no point searching for loco numbers outside this range as they get swamped either by references to page numbers (below 720) or dates (1830-2000).

 

Having done this I have processed out an engine loco index which takes the loco numbers from the above table and produces an index for each loco found listing the volume/issue/page triples that have been found. This is really very useful as given a loco of interest you can find all references to it in one line.

(But I should note that the steamindex search engine does much the same thing for just their index which is always a good place to start and it covers BackTrack and other things as well.)

 

This GWRJ process is still under test and I may have some coding glitches but if anyone wants to post here a loco number they are interested in I will fire back the data I think is relevant so we can see if it works. If this is successful I will then see about posting my tables some place sensible.

 

Crucially, beyond that we would need fresh data - I am planning to try a digital pen scanner on picture captions to see how painful that is.

 

regards

Andy

Edited February 25, 2023 by Andy Keane

[kevinlms]

A suggestion for loco numbers outside the range.

 

Instead of just the number - for example '517', why not format them as 'No. 517' or similar and do a search that way? Obviously, that field becomes a text one, rather than numeric.

 

Probably easier to format ALL loco numbers as 'No. ####' as required to keep things consistent so they can't be confused as page numbers or something irrelevant. Probably best as 4 digits 'i.e. ' No. 0101' for the Churchward experimental 0-4-0ST, of which Hornby has ruined!

[Andy Keane]

2 minutes ago, kevinlms said:

A suggestion for loco numbers outside the range.

 

Instead of just the number - for example '517', why not format them as 'No. 517' or similar and do a search that way? Obviously, that field becomes a text one, rather than numeric.

 

Probably easier to format ALL loco numbers as 'No. ####' as required to keep things consistent so they can't be confused as page numbers or something irrelevant. Probably best as 4 digits 'i.e. ' No. 0101' for the Churchward experimental 0-4-0ST, of which Hornby has ruined!

The problem is I am using somebody else's pre-existing data and they have not done this. Thus when I search the four sets of pre-existing data I have combined I cannot distinguish between a loco 517 and a pag517 or a loco 1896 and a date 1896. To fix this I would have to manually go through some 5000 lines of text and edit the change in. Its all in the context - no doubt some boffin at Google could get an AI system to do that but it is beyond me.

[Andy M]

On 19/02/2023 at 13:00, The Stationmaster said:

Take care with 'Western Times' as some of its photo captions are incorrect.  Even to the extent of a very obvious place where the photo was taken being misidentified!

 

I'll keep generally quiet about my view of certain items of text because there are still occasional 'howlers' and errors to be found .  One bad example is the stuff in Issue 4 about the Warning Acceptance a lot of it is imaginary and/or total nonsense.   Even the the piece about the relatively recent Reading diesel depot, especially subsequent changes, also contains errors and that is well within the living memory of people a lot younger than me.

 

Mike,

If you would be kind enough to quantify and substantiate these errors, I will be most willing to put them in print for the benefit of the wider readership of Western Times and indeed the education of the authors of the relevant articles.

 

Please don't worry about correcting the location you refer to in your first point, as that was corrected in the very next issue. Your challenge to 'Signalling Questions & Some Answers' (Issue 4) and 'Reading Diesel Depot' (Issue 5), I await with interest. Please email me at WesternTimes@mail.com

 

Kind regards,

Andy Malthouse (Editor - Western Times).

[Andy Keane]

1 hour ago, Andy M said:

Andy Malthouse (Editor - Western Times).

Andy

do you have a digital index of references to locos or locations in the western times? is it available online?

[Andy M]

5 minutes ago, Andy Keane said:

Andy

do you have a digital index of references to locos or locations in the western times? is it available online?

 

Hi Andy,

 

An Index for Western Times, both article topics and images, has been proposed and is currently being compiled. It is my intention to make this available in both printed form (within the publication as an insert flyer) and digitally. I will keep in mind your project when this is compiled. 

 

Regards,

 

Andy.

[Andy Keane]

41 minutes ago, Andy M said:

 

Hi Andy,

 

An Index for Western Times, both article topics and images, has been proposed and is currently being compiled. It is my intention to make this available in both printed form (within the publication as an insert flyer) and digitally. I will keep in mind your project when this is compiled. 

 

Regards,

 

Andy.

Sounds great - the problem with the GWRJ is there are about 10,000 photos in its 103 issues and no obvious way of finding all the photos of a given loco or location. I think the Railway Modeller did a great job with their archives allowing text searches over their entire run. Hopefully the WT will do something similar.

Andy