POW and Evaders Database Work

**CZ_RAF** · 21st December 2020, 13:29

I have started check of Czech and Slovak PoWs, those already checked will have cell in column J with green background and my nick CZ_RAF. When I will finish this first check I will try to search for missing info which I have not in my DB yet.

Pavel

**Resmoroh** · 29th December 2020, 14:30

Hello All,

In the db for Halifax BIII LW621 of 192 Sqn lost on 1944-07-05 there is an inconsistency.

Plt Off Aidan AISTON (176489), Air Bomber, is shown as having died on 1944-07-05 and being buried in Fort Mahon Plage Cemetery (which I can’t locate).
Also shown – as a POW - is:-
A ALSTON (no rank or number) who became a POW on 1944-07-04.

These are one and the same. ALSTON does not exist, and if he did he would have become a POW on 1944-07-05.
LW621 came down (for whatever reason) in the sea near the shore. Most of the crew are on Runnymede. Some of the bodies must have drifted ashore and are buried in French coastal cemeteries. There is a note on the POW s/sheet that ALSTON “died in captivity 1944-07-16” which indicates that he was alive when he drifted ashore (or was rescued by the Germans) and actually died (of Wounds?) 11 days later in a military/civil hospital. Where this note originated I know not. CWGC record AISTON's DoD as 1944-07-05

This is not an unknown occurrence. Surnames often have one, or even two, wrong letters (MURI turned out to be MUIR!!!). Service (and other) Numbers often have errors (3/5, 6/8, etc). It is possible that the original documents from which the lists were compiled had some dictation/typing errors. These typewritten sheets then went to the printer where the hot-metal type-setter made the odd error – which was not picked up at the proofing stage. And then the printed documents/books were (many yrs later!) subjected to OCR programs which were in their very early days. I’m not saying this was what happened, but I suspect that a number of such errors are likely to have been included on the way from the original document/card to the rafcommands database!!

Makes for interesting researching!!!

Peter Davies

**Ross_McNeill** · 30th December 2020, 09:56

Hi Peter,

Some background on the source document and initial link to loss that I did for the first generation RAFCommands listing.

When first transcribed from the AIR document only partial published RAF, Commonwealth and Dominion PoW listings were available eg Chorley, Franks etc.

At the time I was writing Coastal losses V1 and wanted to step up the crew information to include service numbers/personal numbers. Where possible primary or good quality secondary sources were used but PoW details were hard to source.

On visits to PRO/TNA a primary source was found consisting of typed list of Aircrew/Groundcrew names and service numbers/personal numbers. It was updated periodically as war progressed so I selected the last update to use, considering that, like the AFL, Authority would mostly correct previous errors in the typed list leaving only latest entries as error possible.

The list did not contain ranks only names, numbers and where presently held, split into RAF/Commonwealth/Dominion etc service arm. The original format was typed with a few margin hand annotations but for archive from RAFAHB to PRO it had been microfilmed with some loss of resolution. It was still a microfilmed, typed original and not typeset or content recognised. The most common visual error was as you found, bleed of i to l etc.

Conversion from A3 PRO supplied paper prints of the AIR document was by Mk1 eyeball as I had tried early OCR for the work I did for RNLI and found it sadly lacking as a workflow alternative. Once in Access Database that I was using for CCL record keeping an additional field for Serial was added.

To aid my attributing men to Coastal losses I did a first pass of linking names to aircraft serial from the available versions of Chorley and Franks to eliminate known BC and FC and started to populate for early Coastal Command. All this was for my own purposes of CCL and not as a definitive listing of RAF PoW.

Some additional Coastal and BC PoW details came from Clutton-Brock in correspondence after publication of CCL Vol 1.

While working on CCL V2 I decided to aid others by doing a current state PoW database dump from Access and formatting into Alpha pages with Camp Location detail onto RAFCommands web pages.

As with the original AIR document it remained an snapshot of time at publication date and was not subject to further updates.

I hope this will give guidance to you on how and from what the original RAFCommands PoW pages were created and clarify methods used to avoid doubt on how errors could be generated.

In the specific case you cite - my transcript error between i and l is generated by the microfilm process, error in subsequent checks is mine as I was not rigorous at the time as it was a BC PoW, not CC, along with confirmation of death detail from C-B.

Ross

**Jagan** · 30th December 2020, 14:59

Ross,

You mentioned in the older version of the site that this was based on the AIR20/2336 document . I have a downloaded copy and it looks like a Printed version (like that of WO392/18) but annotated with hand written corrections as you mentioned.

The file has 86 pages, titled "Germany and German Occupied Territories, Imperial Prisoners of War, Alphabetical List, Section 8 Air Forces. Officers RAF RAF RAAF RCAF RNZAF SAAF"

It is essentially two sets of 43 pages. the first set has a lot of hand written entries..
second set has "CORRECTED GENERALLY TO 30th MARCH 1945" written on it - but is all printed. no hand written entries/corrections. Infact the hand-written entries from the first set of pages seem to have been added to the second set and its in print form.

I think its an early version of WO392-08 and WO392-18 final printed volumes on Air Force POWs - the 1944 and 1945 versions.

Surprisingly i dont see AISTON, ALSTON in it (maybe because he had already died?) so was there another document?

Peter, I hope you have the WO392-18 file? if not, just holler!

-Jagan

**Resmoroh** · 30th December 2020, 15:47

Ross. Mni tks for that. I was not trying to be critical, just to understand the process(es). There are, now, several sets of data incorporated into the experimental POW Google-sheet. They will have arrived by different routes/means (some of it photographed printed pages (at about 6pt!!) which were then subjected to a variable OCR!). The Google-sheet is – as I understand it – an attempt to rationalise/collate all this data into a single operable format. A task, I suspect, analogous to “cleaning the Augean Stables”!!! It is only when all the data are sorted/equalised that any sort of statistical investigations can be realistically pursued?
Many years ago I was tasked with transcribing some of the weather observations made by the monks in the monasteries on the northern Greek borders. We were moving from the ‘quill pen’ to the ‘Amstrad’ in one fell swoop – so I know the problems (and I didn’t read the Cyrillic script/numerals too well either!!). The project was soon scrubbed (whether the data – or the transcriber – were not good enough was not made known to me!!)
Tks again.

Jagan, Not got WO392-18. I might well be getting out of my depth/expertise here?. While you're on, can I have a Google-sheet re-order, and can I have a filter ordered by alphabetic a/c serial numbers. I can then begin to fill in some of the "Type" column stuff. With the Google-sheet, and my ancient desk-top, it takes 17 secs to start a Filter (and 17 secs to shut it down) - I should live so long!!

Peace be upon your cattle and upon your cooking pots for 2021!

Peter Davies