Talk:Prestel

I seem to remember program downloads (at least on Micronet) were charged not by having the first few frames being 99p and the remainder, but the *last* frames - so if the connection died halfway through, usually it was before you had been charged and therefore easier to sort out. Or am I just remembering incorrectly?

Ali0th 15:48, 3 August 2006 (UTC)

That's correct, and for those exact reasons. For a time I used to work in Micronet THD (Technical Help Desk) and one of my jobs was putting up the telesoftware. (If connection did die afer charging commenced, the procedure was to ask the customer to send in a tape/disk with what they had managed to download saved on it, as there was no way to actually track whom had paid for what!)  robert@irrelevant.com. 29th January 2007. (Sorry don't have an account.) I do now: Robirrelevant (talk) 21:24, 18 October 2009 (UTC)

Post Office Girobank
As the Post Office owned Telecommunications, thus also Prestel: plus also owning the 'Bank' called Girobank, why didn't the Post Office sell on-line purchasing sooner? —Preceding unsigned comment added by Acb58 (talk • contribs) 19:26, 28 October 2008 (UTC)

IRC
"data was entered on a central Update Computer, "Duke", located in London, and then mirrored onto a number of satellites (mirrored computers known as IRCs)" - what did IRC stand for, in this context? Google leaves me none the wiser. -Ashley Pomeroy (talk) 15:14, 18 November 2007 (UTC)

Information Retrieval Computer -- they cached retrieved pages to reduce the load on the main database. 81.187.162.106 (talk) 09:35, 19 July 2008 (UTC)

Hate to be pendantic but according to the Post Office publication "Prestel. The Technology." [see below] the term IRC actually meant Information Retrieval Centre, where (typically) a pair of retrieval computers were housed. Inspeximus (talk) 18:03, 31 October 2009 (UTC)

Sorry - just noticed that the article itself now has this correct definition of IRC. Inspeximus (talk) 18:05, 31 October 2009 (UTC)

There were no computers (IRC or Update) at St Alphage House during the years I was technical planning manager. This housed the operations and planning teams. The update machine was located at Ironmonger Row telephone exchange.91.104.114.184 (talk) 17:24, 1 November 2010 (UTC)

Teletext
Is it correct to have Prestel in the category Teletext? I think Teletext refers to a one-way service (as on TV sets) while Prestel was a two-way service and should be categorized as Videotex. Biscuittin (talk) 16:03, 22 December 2007 (UTC)

CUGs
I've slightly adjusted the edit made by 81.149.213.8 with regard to CUGs. Not all CUGs were for paid subscriptions. Some, e.g. 15500 were for Micronet800 and were indeed (usually) granted to paid up subscribers, but there were definitely CUGs such as 7500 (ClubSpot 810 Editors) which were granted to Prestel users for no payment whatsoever, merely to give them access to pages relating IP operations. Robirrelevant (talk) 21:24, 18 October 2009 (UTC)

The difference seems to be that there were (according to the source 'Prestel. The Technology' see below) two types of CUG. CUG 1's were rented by the IP in order that he could manage which users had access to certain pages that he wished to restrict access to, e.g. those which required membership via a subscription to view. On the other hand CUG 2's were freely available to all IP's and seem to have been a mechanism simply to bar all access to frames, e.g. while they were in process of being created or edited. If this Post Office document is generally regarded as a valid source then I'm happy to make the appropriate edits. Inspeximus (talk) 08:47, 29 October 2009 (UTC)

I think this may have been overstated - I've not got a dump of an actual Prestel edit screen to hand, but according to the (Prestel Microcomputing) Dynamic Frame Editor handbook, and also the Bulk Update specification, which I do have, the access controls available are the CUG field as described (valid values 0-32768, assuming you've rented one) and a simple "Access? Y/N" flag. This latter was used to disable access to frames from normal users. (Browsing using the IP account would ignore this, so allowing the IP to view their own hidden pages.) This would seem to be the final implementation of the "CUG 2" you refer to. Robirrelevant (talk) 16:43, 1 November 2009 (UTC)

This suonds like a good explanation of how CUG 2's ("CUG's 2"?) operated - but still leaves open the question of the 7500 CUG for "ClubSpot 810 Editors". Perhaps they were in effect given sub-IP status by the main IP - without the usual payment for passing on CUG access?

Inspeximus (talk) 18:03, 9 November 2009 (UTC)

ClubSpot operated by having a main IP account (for 810) and sub-ip accounts (for 8100, 8102, 8104 etc) the login details for which were given to the individual editors. The editors' personal accounts were added to the CUG so that they could view pages relating to the operation of the IP without having to use the IP acconuts. Editing could only be done with the actual IP or Sub-IP accounts. The CUG merely allowed hiding of information from general view. Robirrelevant (talk) 19:04, 9 January 2010 (UTC)

Prestel. The Technology - Post Office publicity brochure dated 1980.
Having kept a copy of the above mentioned document for the past 30 years or so, I have started to use it in order to enhance this article. Today I've added one additional detail of the Prestel Service, namely the requirement for installation of a Jack 96A. I have already confirmed with Reliable sources/Noticeboard that this document would in principle be considered suitable as a source material in this area. I propose to add further information, such as an outline of the Prestel Organisation, some Operational aspects, and extra Computer Systems information. Before going much further I would be interested to hear whether such additions from this source meet with general approval by the editors active on this page.

Inspeximus (talk) 19:42, 28 October 2009 (UTC)

Sounds fair enough to me, although with the proviso that a 1980 document would have been produced right at the very start of service; Prestel did evolve over it's existence, and some requirements such as the Jack 96A were obsolete by, say, the mid 1980s when the 'current' phone sockets were starting to become common. (the cost of rental for it, incidentally, is shown on the price list already linked as a referance.) I've a few other technical docs which I'll scan and add to the brochures already available at www.viewdata.org.uk, when I get the chance. Somewhere I've got a published book, "viewdata and teletext" which has a lot of detail, which I'll dig out and see if it's relevant to cite. Robirrelevant (talk) —Preceding undated comment added 18:52, 1 November 2009 (UTC).

Good point about the date of the document. I'm hoping to get access to other original documents shortly and will review likewise to see what they may have to add to what I've already been working on in draft.

Inspeximus (talk) 18:14, 9 November 2009 (UTC)

Subscriber numbers?
Have we got any sources for the numbers of subscribers? 90,000 ?

I've got an original save of their publicity factframe from 1987 here: http://www.viewdata.org.uk/galerien/Prestel/65659a.gif which shows 77,000 users, 61% business, 310,000 frames available to view, with 9.1M accesses and 139,000 messages sent per week.

Conversely, this frame: http://www.viewdata.org.uk/galerien/Prestel/9961a.gif from around 1988 saved from the non-public-access side of things, says less than 19K users. This is probably for the single IRC that it relates to - as there were three pairs of these, and most users were only registered on their local pair, the total figure is likely to be around three times this - about 60K users. —Preceding unsigned comment added by Robirrelevant (talk • contribs) 18:58, 9 January 2010 (UTC)

Yes, system generated figures would relate only to the individual host IRC computer, but I believe that total numbers eventually reached nearer 77k as peak user base. Were there only 3 pairs of IRC machines left by 1985 - from a peak of around 18 or so I think - I should remember but it's so long ago. I've been collecting a set of Prestel related articles from the Times which include numbers of users, pages and or IP's at various points over the period 1979 to 1985 - will put up the summary data shortly. Trying to collect post 1985 data via another database of newspapers at the moment. Have also finally managed to get hold of the 1981 article in the Post Office Electrical Engineer's Journal from which I plan to extract any useful stuff. Inspeximus (talk) 19:46, 20 January 2010 (UTC)

What do people think about these figures for Prestel Growth - all extracted from articles in the Times dated as indicated in the table:

Hopefully I can improve on the table formatting given a bit more practice!

I'm also trying to track down more recent figures but seem to have lost free access to the later (1985-2005) Times Archive database.

Inspeximus (talk) 17:50, 21 January 2010 (UTC)

I found a slightly later Factframe, hosted temporarily here: http://www.viewdata.org.uk/rob/vv.php?gal=frames&page=65659b This says: TO END OF DECEMBER 1988 Terminals attached to Prestel... 95,460 Proportion in businesses........ 57%   Proportion in homes............. 43%   Frames available for access..... 258,120 Frame accesses per week......... 7.4M Messages sent per week.......... 165,000

I've got some magazines, which may have figures in, but think they are all earlier than this. Looks like they finally got a bit of growth in users, though... Robirrelevant (talk) 17:56, 22 January 2010 (UTC)

18 IRC machines? The only one I encountered were Dickens/Keats, Dryden/Kipling, Derwent/Enterprise. They added (back?) Austen/Bronte at some point. Duke was the edit computer, Pandora handled mail but I don't know if there was any general remote access. There were also apparently other non-public development machines.

And nobody I can find seems to have held onto a single backup tape or hard disc... Robirrelevant (talk) 18:09, 22 January 2010 (UTC)

Subscriber number table corrected to 1985 and updated to 1991 (as far as The Times has reported data) - some numbers inconsistent but that's as reported by the paper. Also Micronet800 figures broken out in new final column, but they all appear to be ballpark estimates. Unfortunately, pulling this altogether all seems to be original research to me, so I will leave here until we can find some independent summary sources. Inspeximus (talk) 12:28, 1 February 2010 (UTC)

Prestel Computer Names
Clearly there is much more the be found about those Prestel machine names and numbers - and indeed locations. I'm struggling to find sources but an article in InfoWorld 25 May 1981 on Google Books refers to 18 GEC 4082 "computer installations" - which might of course include development servers. So far we have named "Duke" for IP updates, "Pandora" for mailbox and “Dryden/Kipling”, “Derwent/Enterprise”, “Dickens/Keats”, “Bronte”, “Eliot” and “Austen” as IRC machines, but I'm not sure about the pairing of the last three. Machines were sited in Baynard House, London (Duke, Pandora, Dryden/Kipling), and also at Croydon (Derwent/Enterprise?) and places like: Birmingham, Leeds, Cardiff, Edinburgh, Norwich (Ipswich?), Glasgow?, Manchester?. Names were largely based on local literary figures. Perhaps the residence of D.H.Lawrence in Croydon explains the divergence to Derwent/Enterprise! Hogarth was a singleton International machine (in Birmingham?), with Burns located somewhere in Scotland and presumably paired with another. Will see what I can dig up from reputable sources. Inspeximus (talk) 19:19, 22 January 2010 (UTC)

There is also Eliot - I've just been digging around in an old Filofax and found my list of access numbers for the IRCs. They are grouped as Dryden, Kipling and Bronte, (all on 01-248-xxxx), Dickens, Keats and Austen (also all 01-248-xxxx) and Derwent, Enterprise and Eliot (all on 01-489-xxxx). Duke was on 01-583-xxxx and 01-248-xxxx, Eliot also had a 1200 (full duplex?) baud port on 01-680-xxxx. Maybe an old BT exchange location list will help identify the areas. (Prestel Operations were also on an 01-248 number, but IP services and other staff direct numbers were in 01-822 - that would have been Telephone House in Temple.) Robirrelevant (talk) 16:36, 23 January 2010 (UTC)

As regards names I think we had already spotted Eliot, but I also recall Constable (again plus another) in East Anglia - probably Norwich. Hogarth which started off life as an international machine later became used as the Gateway test machine for IP's. So the "famous local author" basis for machine naming must have been extended to "famous local artist". As regards access codes, according to http://www.ukphoneinfo.com/old_std_list.php both 01-583 and 01-822 were dialling codes for Fleet Street, hence Prestel Headquarters at Telephone House. 01-248 was City of London St Paul's so probably Baynard House and 01-680 was Croydon. Derwent & Enterprise were based in Ryland House, Croydon. I think that the numbering scheme you mention may have been based on some kind of fall back policy with two local IRC machines plus a remote one for each of the codes you list - perhaps they were for IP access rather than Prestel user access. On the other hand perhaps machines were concentrated into sets of 3 in fewer locations at some point towards the end of public service in order to further cut costs. Initially users had needed to dial ordinary phone numbers to access their nearest machine, which wasn't always a local code. I think for example an 01 686 xxxx number was used for Croydon. When "Lo-Call" numbers finally became available the two Prestel short codes 618 and 918 became universal for all users across the country. Inspeximus (talk) 10:50, 24 January 2010 (UTC)

Just to confirm at least one of the groupings; I've found an application confirmation letter, (the one with ID and password) dated 15/9/89 that confirms I may use Dryden, Kipling and Bronte, all on telephone number 618. (This was the same as when I was in Manchester previously.) There's a PS obviously added to the end of the standard letter saying I've been registered on Dickens Keats Enterprise Derwent Austen and Eliot also. The numbers I gave above were indeed given for IP use, although they did work for anybody who wanted to get at a particular IRC. 01-618-1111 also worked from outside London for accessing Dryden/Kipling.. Robirrelevant (talk) 17:24, 24 January 2010 (UTC)

Buried in the opening paragraphs of the article "Prestel: The First Year of Public Service" (W.R.Broadhead, July 1981, Post Office Electrical Engineers Journal, Vol. 74, p129) is the following "...the Prestel computer network which grew from these 3 computers serving 30% of the telephone population to 18 retrieval computers serving 62%...". The three original machine referred to are described as two information retrieval computers plus one update machine. I reckon that we have named 11 IRC machines so far, (ignoring the special machines, Duke, Pandora and Hogarth):


 * 1) Dryden
 * 2) Kipling
 * 3) Derwent
 * 4) Enterprise
 * 5) Dickens
 * 6) Keats
 * 7) Bronte
 * 8) Eliot
 * 9) Austen
 * 10) Burns
 * 11) Constable

so we are looking for the names of 7 more machines. Inspeximus (talk) 12:19, 25 January 2010 (UTC)

Certainly in/by April 1985, only the first six of those were in use, as I have a pair of form letters from Prestel in that month giving my passwords on each IRC, listing those six and with my details and passwords filled in for each. I only registered in October 1984, when I was granted access to Dryden & Kipling only (on the 300 baud port). The next three were definitely in use by 1989 based on the other letter above. Robirrelevant (talk) 17:27, 25 January 2010 (UTC)

I think that it is a case of changing policies and to some extent changing technology which is causing some confusion here. In the early 1980's there was no centralised pool of user data, i.e. no Prestel Admin computer, so all user registrations had to be carried out by hand separately on to each machine which a user could potentially access. The key access code was hard wired into each TV set and had to be extracted over the phone by an operator before being keyed into the user file on each computer. It was therefore practice only to register ordinary users on two computers in the early 1980's and so any individual user would only be given access details, including direct phone lines, to two machines. The registration activity was carried out by the local regional Prestel Centre who each managed at least one IRC. IP's and sub-IP's required access rights to every machine in order to retrieve response frames, see billing data etc. but were registered centrally by the National Operations Centre. By the mid 1980's the registration process had been automated such that user details were added once into the Prestel Admin machine which distributed them as required to relevant IRC machines. The Prestel ID was no longer extracted but in the reverse process was issued to users by Prestel Centres by means of a letter. It was then up to the user to programme this number into his adaptor or more likely micro-computer. At this point it became practicable to enable users to access more than just their local computers and with the roll out of the 618/918 codes wider registration was introduced and eventually universal registration became the norm. It was no longer really necessary for users to be aware of which individual machines were local to them: they just dialled a short code and knew that could logon wherever they ended up. I suspect that even later some IRC's were closed and the remaining machines concentrated into fewer centres - but I was no longer around by then! Inspeximus (talk) 20:04, 25 January 2010 (UTC)

I found a saved frame that refers to a "Juniper" computer: http://www.viewdata.org.uk/galleries/vv.php?gal=prestel&page=9001a I don't know if this is a genuine name, though, as the saved image has (most of) a "MNET Contributors" IP header - this was mainly used for Gallery pages I think (and the flashing cursor is definitely an addition). It is a copy of the real password page, but I can't say for sure any more than that. So, take with a pinch of salt.. Robirrelevant (talk) 12:30, 30 January 2010 (UTC)

Yes, I share your scepticism, the name doesn't ring a bell and it seems out of kilter with the other known naming schemes. It could have been a test machine I suppose but I can't remember how they were named. Inspeximus (talk) 18:25, 30 January 2010 (UTC)

Just found this unusually informative description in a Prestel Travel brochure, undated but came with a 11/1985 price list and 11/1986 newsletter. It's in the to-be-scanned pile now. "Prestel consists of 9 GEC 4190 Computers. 1. 6 computers are located in pairs:- Dickens & Keats in Birmingham serve N. Ireland, Scotland, Wales, N. England and overseas customers. Derwent and Enterprise in Croydon serve S. England. Dryden and Kipling in London serve London. These computers are used for retrieving information.  If one of the computers is faulty or engaged the user is automatically switched to the other computer in the pair, this providing the user with a very reliable service. "2. The computer used for updating information is located in London and known as DUKE. (The names of the Prestel computers derive from the names given to the Victorian telephone exchanges, these were based on poets and adventurers.) "3. Another pair of computers located in London is used for Prestel Gateway. These computers are known as X25 switches. They collect all the calls made through users through the 3 pairs of retrieval computers and switch the call through to the external computer.  The second computer in the pair is a reserve and is only used in the event of the other machine failing. Under such circumstances all Gateway links that were connected to the computer are switched to the reserve computer." And a very nice diagram to show all this too. Of note - two more computer accounted for, albeit unnamed, as X25 switches. And an explanation of the naming scheme.

The brochure also indicates a 98% local coverage of the UK with a lovely shaded map. The newsletter refines this to 100% NI,Scotland and Wales, which is a distinct improvement, as Wales in poorly served in the map in the brochure. 99% England. Also mentions 60,000 prestel users, 45%/55% home/business, 5,500 in ABTA Travel Agents. A nice tariff guide too, which includes costs and of gateway links. Robirrelevant (talk) 09:20, 2 February 2010 (UTC)

Nice work, complete with sources too - and as you say an explanation of the naming scheme, although I'm not sure I can identify any names so far with adventurers! From the small count of computers, I believe that this refers to the time following the rationalisation of Prestel down to just three IRC's, with two machines per site. From your previous sources it seems that the two computers per centre was later increased to three. I'm slightly surprised at the reference to Prestel Gateway computers as X25 switches themselves, rather than being connected to X25 switches. I'll have to work out the implications of this. The source also indicates that during 1986 the machines themselves had been upgraded from the original GEC 4082 configuration up to 4190. Anyway, I was planning to tackle GEC Prestel machines next (honest) so this gives me the spur I needed. Inspeximus (talk) 17:45, 2 February 2010 (UTC)

Quick bit of research at http://www.rhaworth.myby.co.uk/phreak/tenp_01.htm confirms that many of the names we had already sourced for Prestel machines are to be found in the list of Post Office exchange names. Note though that all of them are London exchange names, although some of the so-named Prestel computers were sited in Birmingham. I suggest that this naming convention might have been employed to get around any potential issues of "copyright" on the names?

The list specifically confirms the following computer names as exchange names in London:


 * Derwent
 * Dickens
 * Dryden
 * Duke
 * Enterprise
 * Hogarth
 * Juniper!!
 * Keats
 * Kipling

but not
 * Austen
 * Bronte
 * Burns
 * Constable
 * Eliot

Inspeximus (talk) 18:21, 2 February 2010 (UTC)

Prestel travel brochure, tariff guide, and associated leaflets, are now scanned and available at at http://www.viewdata.org.uk/index.php?cat=15_Prestel&page=30_Brochures The Tariff guide gives examples on how much it cost to connect your own system up via Gateway, too (a LOT!). Brochure says there are 62 gateway systems already connected.. Robirrelevant (talk) 13:47, 7 February 2010 (UTC)

From a mailing list I've just joined... The Prestel units were as follows:-

Other units were named:- Not sure where these ones were but there must have been units at Colindale SSC & Kingston SSC
 * Franklin               Based at Baynard House
 * Dryden                                "              "
 * Kipling                                  "              "
 * Duke                    Based at Clerkenwell Exch
 * Byron                   Based at Fleet Building
 * Kelvin                   Based at Ilford SSC
 * Derwent             Based at Croydon SSC
 * Enterprise                          "              "
 * Vigilant                Based at Eltham SSC
 * Atlas                     Based at Ealing SSC
 * Juniper                Based at Wood Green SSC
 * Austen
 * Dickens
 * Keats
 * Bronte
 * Eliot

yay!! Thanks Simon! Robirrelevant (talk) 08:59, 9 February 2010 (UTC)

The plot thickens - I can confirm all the new names: Atlas, Byron, Franklin, Kelvin, and Vigilant were names of London directory area telephone exchanges. I've read somewhere that only four of the six London SSC exchanges had Prestel machines, but not certain. I have anecdotal evidence of additional machines named Scott (as well as Burns) in Edinburgh, Gainsborough (as well as Constable) in East Anglia and both Megan and Dylan in Cardiff. Hoping to make another contact shortly who may have documentary evidence. Inspeximus (talk) 22:59, 9 February 2010 (UTC)

From an (Micronet new users') online introduction to Prestel that dates, I am told, from right at the end - "FALCON is the billing computer: it receives information from the other computers concerning usage, which is used to construct your quarterly bill"

It also says there are 8 IRCs, "Austen, Bronte, Eliot, Dryden, Dickens, Enterprise, Keats and Kipling). As more modern IRCs are introduced older IRCs are being retired" - I note that Derwent is missing from that list - plus Duke and Pandora.

Also a note within a description of the transision to VASSCOM (the new network alluded to already) "The real advantage of VASSCOM has appeared since all the IRCs were moved to London, and that is load sharing:   "

I'll get the originals to this up on viewdata.org.uk soon...

Robirrelevant (talk) 11:18, 24 February 2010 (UTC)

This brings back yet more memories - still fairly hazy though: FALCON was the name of one generation of the Prestel billing computer system - which did just that: i.e. issue paper bills to users & IP's. Both IP income and user page charges were actually recorded by the Prestel IRC servers, as were messaging charges when these appeared. These were sent to FALCON via magnetic tape I think. It did not do accounts receivable processing or debt collection either. It was based on a UNISYS mainframe machine sited in a building in Apsely, variously known as Network House, Dialcom house etc. I believe that it superseded an earlier system running on a Burroughs machine elsewhere in London around the time that Burroughs took over Sperry to form UNISYS. It was written in a kind of high level 4GL possibly called something like MAPPER. The system was later again moved out of Apsely into a standard BT computer centre.

Secondly, what we have with all these machine names must be the final phase of the Prestel infrastructure during public ownership - after all the regional IRC's had closed. I'm still trying to get contacts of people who were around at the time who should know how this network developed and then contracted over the years.

Finally, the name VASSCOM (VASCOM?) presumably refers to a packet switched X25 network for Prestel and other BT services - see the interesting article at Packet Switch Stream. Inspeximus (talk) 15:39, 25 February 2010 (UTC)

That sounds about the right level of technology for Billing - I've had (and got) hand-corrected bills, which I presume means there was little input they could make to the system, and the reminders I have seem to be stapled to copies of the main bills.. (although that might have been me!)

VASSCOM was an acronym for Value Added Systems and Services (the BT grouping) COMmunications, or something like that anyway. I read a reference recently, will have to dig it out. Yep, it was a switched network shared with Gold et al.

The Micronet intro that I referred to above is now up at http://www.viewdata.org.uk/index.php?cat=15_Prestel&page=17_Micronet-nbsp~Virgin-nbsp~Users Robirrelevant (talk) 20:36, 25 February 2010 (UTC)

Just a quick confirmation of Juniper- Screenshot of welcome and goodbye pages in "Viewdata in action", Rex Winsbury, 1981. ISBN 0-07-084548-4. Page 1 and page 237 ! 90.198.191.134 (talk) —Preceding undated comment added 14:36, 21 February 2012 (UTC).

And Vigilant features in screens within Viewdata and the Information Society, James Martin, 1982. ISBN 0-13-941906-3. 46.18.104.147 (talk) 15:02, 25 February 2012 (UTC)

From a period handwritten list posted on Facebook - https://www.facebook.com/photo.php?fbid=10213083056338851&set=gm.1433816050046075&type=3&theater

Robirrelevant (talk) 11:10, 18 June 2017 (UTC)
 * Arkwright - Manchester
 * Atlas - Ealing
 * Burns - Edinburgh
 * Byron - Fleet
 * Chippendale - Leeds
 * Cbonstable - Chelmsford
 * Derwent - Croydon
 * Dickens - Birmingham
 * Duke - Clerkenwell
 * Dylan - Cardiff
 * Enterprise - Croydon
 * Gainsborough - Chelmsford
 * Juniper - Wood Green
 * Keats - Birmingham
 * Megan - Cardiff
 * Priestly - Leeds
 * Scott - Edinburgh
 * Vigilant - Eltham
 * Wordsworth - Manchester

Locations
From an old welcome pack: "Your Prestel Centre is: Birmingham Prestel Centre, Berkley House, 245 Broad Street, Birmingham. B1 2HQ. Looks like it's a retail place though, based on present usage! Robirrelevant (talk) 00:44, 24 January 2010 (UTC)

I recall but unfortunately don't yet have any source material for the following: in the early days everything was run from Prestel Headquarters at Telephone House in Temple Avenue, while the computers and operations staff were located on the 6th(?) floor of St Alphage House alongside London Wall. The local office functions of Customer Support, Customer Registration, Sales, Marketing and Billing spread out to regional locations across the country, based in Regional Headquarters buildings I think, such as Berkley House, Birmingham. The computers were based at the IRC's in major exchange buildings, e.g. Sector Switching Centres in London I believe, including Croydon and Ealing(?). Baynard House took a central role when both the National Operations Centre and machines such as Duke were moved there. As reality about user numbers sank in the regional offices were concentrated back to just two: Birmingham, which handled all functions for the north and west of the country and Brixton for the south and east. Still later these two centres were merged into one at International House, Brixton, while in the final phase everything returned once more to Telephone House. As far as I know the machines remained generally in pairs but sometimes just singletons located in the remote regional exchange locations. Inspeximus (talk) 10:41, 24 January 2010 (UTC)

"Initially the Prestel viewdata service provided by the British Post Office started with three regional centres at London, Birmingham and Ipswich." p143 Teletext and Viewdata, Steve A Money, Newnes Technical Books, 1979 (1981 reprint) ISBN 0-408-00597-1. Robirrelevant (talk) 18:15, 24 January 2010 (UTC)

In the article "Prestel: The First Year of Public Service" (W.R.Broadhead, July 1981, Post Office Electrical Engineers Journal, Vol. 74, p129) it states ".. continuity of service is provided by installing in each centre, 2 computers currently accessed by different routes. An exception is London where 4 centres have single computers." Thus we have 4 singleton and 14 paired IRC machines, so a total of 4 + ((18-4)/2) = 11 IRC sites to locate. So far, we have positively identified Baynard House in the city of London, and Ryland house, Croydon, and by implication sites associated with the regional centres at Birmingham and now Ipswich. My recollection is that IRC's were generally sited in so-called "Sector Switching Centres" because of the requirement to house the extensive comms equipment, modems etc. The BT Archives online catalogue names three such centres in London, namely Eltham, Croydon, and Ealing and four outside namely, Nottingham, Newcastle, West Bromwich, Edinburgh and Leicester Cardinal. I seem to recall that there were in fact six SSC sites in London, and so potential sites for IRC's but cannot be sure. I also seem to recall nine Regional Centres in the British Post Office around the relevant time. Of course this all constitutes original research and therefore is not suitable for the Wikipedia main article but at least it gives us somewhere to start looking! Inspeximus (talk) 12:24, 25 January 2010 (UTC)

Referances
There's a distinct lack of citations in the article, so I'm making some notes now as I find things which may be of use. Not enough time at the moment to do a proper write up, but if anybody else wants to, feel free!

http://www.independent.co.uk/news/business/prestel-rings-the-changes-as-bt-hangs-up-1367604.html

http://www.independent.co.uk/life-style/steps-one-two-and-three--hey-prestel-1331077.html

Those have some useful dates and figures in it.

Oblique references in this one http://www.independent.co.uk/extras/indybest/gadgets-tech/computers-minitel-gets-on-the-road-frances-successful-telephone-information-and-sales-service-based-on-prestel-is-being-made-available-in-britain-frank-barrett-reports-1468011.html

Probably more on there too. Robirrelevant (talk) 08:21, 26 January 2010 (UTC)

Looks like the "sold by bt, bought by financial express" is a bit of an over-simplification, too: http://www.financialexpress.com.hk/uk/about/michael_holland.htm Robirrelevant (talk) 08:54, 26 January 2010 (UTC)

Proposed Replacements for "How It Worked" section
I offer the following five new sections to replace the majority of the existing "How it Worked" section of the Prestel article. Given that they are a total re-write I though it appropriate to solicit comments here before posting. Note that the section headings have been deliberately corrupted so as not to create extra sections within this talk page. Also please note I'm still struggling to get my head around formal citations & footnotes so forgive me if they do not appear as intended.

I have in draft but as yet incomplete, two further sections: one covering the Prestel hardware, and the other outlining the organisational structure of the Prestel business. These should facilitate the replacement of the remaining sentences of the existing "How it Worked" section. All being well I hope to be able present these for consideration shortly. Inspeximus (talk) 18:27, 26 January 2010 (UTC)

It's fairly good, though a bit off in some of the technical details. Not sure what the protocol here is, but rather than bog down this talk page with lots of re-writes, I've copied it and put some corrections up at: http://forum.viewdata.org.uk/viewtopic.php?f=3&t=82

I've just dog out a copy of the Prestel Customer Handbook and Startup Directory (still in it's 1986 envelope!) which covers a lot of the how-to-use-the-service type stuff. I'll try and get that scanned and posted ASAP. (I've got several full directories too, which are great for colour screenshots, but are fairly daunting from a scanning point of view!) Robirrelevant (talk) 22:03, 26 January 2010 (UTC)

Thanks for the corrections and additions - it's invaluable to get the IP perspective on this since I have no experience whatsoever as an IP. I've copied your version back here to the Talk page and then made a series of minor amendments, largely to fix the references and to make some copy edits outstanding from my own first draft. I'm assuming that no one will actually make further edits on the Viewdata.org site but will only make suggestions which can be incorporated back into the version here as the "master copy".

You say "Somewhere I read that the links between the sites were via V22 private wires, initially. I can't find the reference at the moment." According to this article ITU-T V-Series Recommendations on Wikipedia V22 was full-duplex communication between two analogue dial-up modems using PSK modulation at 600 baud to carry data at 1200 or 600 bit/s. I suggest therefore that it might have been used for high-speed editing access for IP's, perhaps you could check the relevant guides. Both POEEJ and PTT references are clear about the use of X25 4.8 kbit links for intercomputer links so have removed your comment from the main text until we can be sure where V22 was used. Inspeximus (talk) 16:19, 27 January 2010 (UTC)

Yup, that looks a lot better! If you want an account on forum.viewdata.org.uk just let me know - I've got no worries about publishing original research there! (really must flesh out the write ups on the main site..)  V22 was one of the options available for IP dial-up bulk updates (1200 baud full duplex) but it's certainly possible to run it across leased lines - in my post-Micronet job, I had customers linking sites via similar analogue modems on private wire circuits. I'll have another hunt for that reference though. Robirrelevant (talk) 23:08, 27 January 2010 (UTC)

I've found some more relevant sources so have re-written the "Topology" section with a new "Infrastructure" section which tries to trace the expansion of the Prestel network from just one machine to 18 in all at various named sites. Some details still missing but I've just managed to contact another former colleague so may get more facts and figures from this. Inspeximus (talk) 17:05, 28 January 2010 (UTC)

I've just completed a few copy edits to the Topology section and am now ready to go so if no one has any objections I'm going to replace the whole "How It Works" section of the main article tomorrow with the material below (and see if I can the references to merge successfully! Inspeximus (talk) 18:29, 30 January 2010 (UTC)

Looks fine to me. I've just changed slightly the description of *0# - page 0 itself was never used - each user had a default main index set. For standard users, this would be page 1a, but if a user signed up through, or later joined, certain services, such as Micronet 800, Club 403, Prestel Travel, CitiService, etc, they could be given a different welcome page, and their root frame would be changed so that it took them to 800a, 403a, etc. I don't know the exact mechanism for how this was arranged as I only experienced it as a user.

Excellent - your explanation is so good I've incorporated it at the start of the User's View section - with slight re-wording to make it fit with rest of text. I'm starting to move the whole replacement text now - deleting it from here so we only have one version to work on. Inspeximus (talk) 16:09, 31 January 2010 (UTC)

I've got the IP price list over on viewdata.org.uk if you want to use that as a reference (and they did change over time, prices going UP) and perhaps a brief mention of Private Prestel could be added at some point? Robirrelevant (talk) 09:33, 31 January 2010 (UTC)

Five sections formerly entitled:  Prestel - the Database   Prestel - the Information Provider's View   Prestel - the Users' View   Prestel Infrastructure  <ul> <li>Prestel Messaging</li> </ul>

together with the corresponding Notes and References section removed to main article and deleted from here to avoid confusion of versions. Section titles are slightly renamed on main article page for stylistic consistency. Inspeximus (talk) 17:38, 31 January 2010 (UTC)

Prestel office location images
Found this useful image of Baynard House on Geograph site [] which is already licensed for use on Wikipedia etc. []. Geograph also refer to planned bulk upload to WikiMedia Commons, so did not want to upload myself. Edited main article with appropriate image reference, but link does not work. Then found that bulk upload is actually in progress right now: [] so have left link in place pending imminent arrival of image. Inspeximus (talk) 16:53, 1 February 2010 (UTC)

Prestel Computer Hardware & Software
The following is offered as a draft for a further section or two on Prestel Hardware. At the moment references are simply codes, but full sources will be incorporated once the text is transferred to the main article. The stuff about MCACCU's will need deleting from the current article to avoid duplication. I realise that the bits about VAMPIRE and DRACULA may be hard to swallow but they are genuine - and will be properly referenced!

I'm next planning a short section of the role and location of Regional Prestel Centres so should be able to link this in to the monitoring section below later.

Inspeximus (talk) 19:28, 8 February 2010 (UTC)

In absence of further feedback, removed the two draft sections to main article for onward editing & readership by wider audience. Sections are entitled


 * Prestel Computer Hardware


 * Software & Prestel Monitoring Equipment

Inspeximus (talk) 14:56, 17 February 2010 (UTC)

Erm sorry, yes, I had nothing further to add to these, being outside my direct experience, and not having anything, documentation wise, relating to this. Robirrelevant (talk) 20:49, 17 February 2010 (UTC)

Prestel Messaging - new draft
I offer a fresh draft of the Prestel Messaging section, incorporating material from a newly received Prestel Marketing brochure, reference Pestel P545, and dated 9/1984. In this draft the brochure is simply referenced by the dummy link [MKT]. I don't think that any material has been left out from the current section in the main article, but some wording has been amended here slightly and much has been added, including reference to the Telex link service, volumes and charges etc.

Inspeximus (talk) 18:08, 17 February 2010 (UTC)

This I can add to: there was a "new" mailbox system introduced later on that added mailing lists, message summaries, etc. Will add with refs later once I've dug out the appropriate docs. Robirrelevant (talk) 20:50, 17 February 2010 (UTC) Couple of minor edits and a new paragraph added. Robirrelevant (talk) 15:43, 18 February 2010 (UTC)

I've reordered and reworded sentences on Telex Link, charges & volumes to make clear that they refer to 1984, making the new paragraph dealing with enhancements in 1989 into the final paragraph. Inspeximus (talk) 18:09, 18 February 2010 (UTC)

Yup that looks better. Added sign-off warning, plus a bit about ex-dir mbx numbers. Not quite sure how to word that it was last 9 digits of telno, eg my old one of 061-427 1596 was 614271596, given the example already quoted. Ex-dir numbers also used up all of n1999nnnn and moved onto 01111nnnn (I got 011111111 yay!) but I think that is too much detail perhaps.Robirrelevant (talk) 21:02, 18 February 2010 (UTC)

I've had a go at clarifying this detail... Inspeximus (talk) 22:01, 18 February 2010 (UTC)

Looks ok, and I can't think of anything else to add, even after leaving it a day or two and coming back to it! Robirrelevant (talk) 09:08, 22 February 2010 (UTC)

I've removed the draft section to main article as replacement for the section entitled Prestel Messaging leaving the discussion only in place here. Corrected title of this section of Talk Page to Prestel "Messaging" not Prestel "Mailbox".) Inspeximus (talk) 17:59, 22 February 2010 (UTC)

Some new documents
The BBC have turned up some interesting documents for me http://www.whatdotheyknow.com/request/presence_on_prestel_1980s The file "Radio Management Registry" includes some details on a "Closed User Group" service Prestel were proposing - essentially a entirely separate computer where pages *and ports* could be rented for private use in-house by a company. Also, details on the trial for Prestel International - again, a separate computer with terminals proposed in multiple countries. Make for some interesting reading. Robirrelevant (talk) 21:03, 2 March 2010 (UTC)
 * Lots of useful documents popping up on the newly relaunched BT Digital Arcives, e.g. TCC 23/852/3 about the monitoring systems, incl. VAMPIRE ! Robirrelevant (talk) 09:21, 21 July 2013 (UTC)
 * Also quite a few useful articles in the Daily Mail archives! Robirrelevant (talk) 09:21, 21 July 2013 (UTC)

Prestel Worldwide
Just discovered an article in InfoWorld 1982-02-08 about the launch of Prestel in the US. http://books.google.co.uk/books?id=eD4EAAAAMBAJ&pg=PA1&lpg=PA6&dq=%22Frank+burgess%22+prestel&hl=en

Anybody seen another reference to the "Jefferson" computer? Robirrelevant (talk) 09:41, 25 June 2010 (UTC)

Is it still running?!
The article doesn't say for definite one way or the other - there's reference to the service being "run down" after the rise of the internet, but there's also mention of the prices charged for certain features "at the end of 2009"... long, long after all notable urban areas of the UK had received broadband provision, let alone high-speed dial-up services. It strikes me that, in much the same way as faxes and physical answering machines, it may still have a small niche for dedicated, secure data access services with guaranteed responsiveness and the like... Can someone please clarify whether Prestel or its modern equivalent is now entirely dead, or if it continues to soldier on? 87.112.66.153 (talk) 18:37, 14 May 2013 (UTC)
 * Prestel itself closed in 1995. Most other PTT systems had gone by the end of the 20th century too, although the closely-related Minitel (France) soldiered on until 2012.  Private viewdata systems have mostly been superseded, though I gather that it is still in use in certain niche areas, notably the travel industry, but probably only as a result of inertia rather than it being the best choice technologically.  The days of dedicated terminals on the desks dialing direct have long gone, however, with access almost entirely via terminal emulation on PCs accessing across the Internet.  I think the references to 2009 are to clarify how much the costs were "in today's money" Robirrelevant (talk) 09:18, 21 July 2013 (UTC)
 * The travel industry indeed left Prestel onto two competing networks, one being Midland Network Services or MNS - and later Imminnus that came out of Thomas Cook who used to be owned by Midland Bank, that was later Telewest. The other network was ISTEL - which May have come out of the car industry and then bought by NTL - so both became Virgin Media. I would suspect they finally shut down in about 2010-2012, my memory is hazy. The move was caused by a lot of Travel Agents using quizes and chatrooms, bosses wanted a different system. Although attempts to move to a closed IP based system started around 1997 that didn't work as people saw the internet was free,  I do remember writing an implementation of Prestel Gateway to allow users to the Thomson system to access their purchase of Crystal Ski via the Thomson system (T#) rather than the CRY# mnemonic. Woodnorton (talk) 19:20, 21 August 2020 (UTC)
 * BT sold Prestel to small company (called variously New Prestel Ltd, Prestel On-Line Ltd, Citiproducts Ltd) formed by a couple of the GEC Computers former salesmen. They tried to keep it running for some niche business areas, but it folded after a year or two. The Prestel name was then bought by Demon Internet (UK ISP), but they didn't do anything with it as far as I know. 81.187.1.83 (talk) 22:01, 8 April 2015 (UTC)

Vandalism?
I'm struggling to understand what's going on with edits around 30 March and 7 April that between them removed some 10,000 bytes from the article and left just 75% of the original. The edits are not done randomly but in a structured way but still leave the article in a state that is difficult to understand in most areas. Unless anyone objects here in the next few days I propose just to revert to the version as at 21 March 2014. Inspeximus (talk) 17:43, 19 June 2014 (UTC)


 * No need to wait - those removals are clearly suspicious and unambiguously damage the article. Good spot! Bonusballs (talk) 18:52, 19 June 2014 (UTC)
 * From March 2014 to March 2015, there appear to have been vast revisions of this article by multiple single-purpose accounts. I don't know enough about this subject to assess whether the article is better off now or before. If in doubt, a rollback to may be in warranted. Ibadibam (talk) 23:12, 17 May 2016 (UTC)

Strange language in the 'Public take-up' section
"I suggest therefore that it might have been used for high-speed editing access for IP's, perhaps you could check the relevant guides." Clearly not right for a wikipedia article! Scatterkeir (talk) 08:55, 13 September 2015 (UTC)


 * I agree. The last two sentences look like someone's comment and should be removed.--GwydionM (talk) 19:26, 13 September 2015 (UTC)