5 ms·

History of computing in Scandinavian Airlines System [in Danish] [pdf]

40 points by unixhero 11y ago

yarper 11y ago
Is there a version in English?
- unixhero 11y ago
  I have not been able to find an English translation. Google Translate does a great job at this text and I think its worth the work for people interested in early global networked messaging and computing with Mainframes, Unix, pre-TCPIP, post-TCPIP etc... After googling after some exotic hardware mentioned in the text, I found a translated section from this text in "History of Nordic Computing" [0]. I am helping my girlfriend with an internal project in SAS, and I found this text when I was searching for documentation on an ancient Mainframe "QQLIF file" which is still in use today! This text does indeed contain a description of it: "All messages in and out of systems were logged on logtapes, so we started using them to extract traffic in the network. IBM mainframe computer was we used for our data processing and SAS (Statistical Analysis System) was the "language" we programmed in. At an earlier point, we had been losing track over which terminals we had and where they stood. Therefore, we had already built up a widely used file called QQLIF and contained a variety of information about each terminal. It was both a real time online "direct access" file at Unisys, and a batch copy on IBM. Mainframes known terminal's trust, and by means of QQLIF file could be translated into PLID. Because of strict PLID'ens building could we add line drop, controls, and terminal for individual messages." [0] https://books.google.dk/books?id=c7MlCQAAQBAJ&pg=PA276&lpg=PA276&dq=lorenz+agentset&source=bl&ots=kH7t_NKsLg&sig=v3PTowEl_6lY8X-nxafQ9fOA5Ac&hl=en&sa=X&ved=0CCAQ6AEwAGoVChMIq8jBkeDpxwIVxZ1yCh26zgoK#v=onepage&q=lorenz%20agentset&f=false https://books.google.dk/books?id=c7MlCQAAQBAJ&pg=PA276&lpg=P...
  - fulafel 11y ago
    Google Translate told me "The page you requested was too large to translate" - got a working GT link?
    - unixhero 11y ago
      I am afraid you have to translate the section you wish to read by copying and pasting into Google Translate. A minor hassle imo.
  - PMan74 11y ago
    > an ancient Mainframe "QQLIF file" which is still in use today! I work in the travel industry, I was recently looking at a system's documentation with included revision tracking going back to the 1980s. System does the same thing it did in the 80s (presumably with completely different hardware/software)
JacobiX 11y ago
It would be interesting also to see what they are using now.
- yarper 11y ago
  Small airlines use a collection of software between their business, if they do their own maintenance they use a tool like AMOS (http://www.swiss-as.com/main.do http://www.swiss-as.com/main.do) which is produced by a subsidiary of Lufthansa, then another tool like sabre (http://www.sabreairlinesolutions.com/home/software_solutions/enterprise_operations/ http://www.sabreairlinesolutions.com/home/software_solutions... ) to manage crew, flights planning etc. Bigger airlines will probably use SAP. The hardware is going cloud based/vm but due to a ton of legacy and regulation concerns (i.e. data protection, rules regarding data access in the event of accidents etc) it's very slow going and most companies will still host their own servers, with maybe a company-wide shared oracle/db instance. The market is strangely competitive, though it has become quite stale in my opinion. RAMCO (http://www.ramco.com/aviation-suite/campaign/aviation/ http://www.ramco.com/aviation-suite/campaign/aviation/) are the closest thing to a "disruptive" force, but their software is widely regarded as flashy with little meat to it (and a lot of glitches). Source: developer for one of the MRO software providers (not listed here)
  - atdrummond 11y ago
    Would you have an email where you could be reached? Would like to ask you a few more questions about the industry.
    - yarper 11y ago
      Sure, I'll drop you a email
- justliving 11y ago
  More than 100 airlines worldwide are using Amadeus (amadeus.com) today. There are different components to it such as Amadeus altea inventory (for managing their flight inventory, prices etc), altea reservation and altea DCS (for check-in and boarding). Additionally there are products to manage their websites (such as the e-commerce solution of Amadeus), revenue management systems etc. The main competitors are sabre (as mentioned) but also HP, Accenture, SITA and others are providing solutions in the same field. It's a pretty mature industry and btw one of the first ones to move to the cloud.
tempodox 11y ago
I can only curse myself for not learning enough Danish by now. Another interesting source of practical applications of computing machinery is the book “Computer” by Herb Grosch: http://www.columbia.edu/cu/computinghistory/computer.html http://www.columbia.edu/cu/computinghistory/computer.html He also lets on that remote electronic flight booking was already possible in the early 1960s. Also have a look at Frank da Cruz's pages about computing history at NYC's Columbia University: http://www.columbia.edu/~fdc/ http://www.columbia.edu/~fdc/
danso 11y ago
If you enjoy reading about the history of airlines, I highly recommend Thomas Petzinger's "Hard Landing"...a good chunk of the book is devoted to the airlines' brutal fight over computer systems...reading it is a great look into how something so seemingly mundane as arranging ticket reservations was so revolutionary in changing the business...and how so fundamentally even primitive computers impact our lives: https://books.google.com/books?id=GF5GZaTd8qsC&lpg=PA62&ots=EKsOUN6uRE&dq=hard%20landing%20thomas%20petzinger%20computer%20george&pg=PA59#v=onepage&q&f=false https://books.google.com/books?id=GF5GZaTd8qsC&lpg=PA62&ots=... Here's an excerpt regarding the state of the industry before computers: > An airline seat is like fresh food—a grapefruit, say—in that it spoils after so much time on the shelf. Every empty seat taking off on every flight is a spoiled grapefruit and exactly as valueless. Both required time, effort, and money to create, and both came to a wasteful, meaningless end. And on an exceedingly large number of flights, the sale of one last seat, according to the First Rule of Airline Economics, could easily decide whether the plane flew the entire distance in the red or the black. As the builder of American Airlines, C. R. Smith had agonized continually over the problem of matching passengers with seats before the latter perished upon departure. The trick was to balance reservations against inventory, a far greater task than it might seem. At first, in the early 1930s, a single ledger had passed from hand to hand as agents in a central office recorded reservations and erased those that were canceled. > With the advent of the DC-3, as the fleets and schedules of the airlines swelled and as the seating capacity of each plane increased, a single book became impractical, so multiple books were maintained. These had to be reconciled frequently—a massively time-consuming process—and even so, each book was inherently out-of-date at all times. Books gave way to chalk and slate boards and eventually to electric light boards. More and more agents were crammed into the reservations offices, with the line of sight to the blackboard limited by the space between structural support columns—22 feet in most buildings of the time. Some agents peered through opera glasses. Teletypes clattered. Clerks known as “card boys” scurried from desk to desk, eventually to be replaced by mechanical conveyors. One blackboard gave rise to multiple blackboards, which quickly encountered the same problems that multiple ledgers had caused. > A science called “reservations theory” was born to cope with the exponentially worsening challenge, but to little avail; as with so much else in the airline business, there was no economy of scale in handling reservations. The more the airlines grew, the less efficient they became. The need to coordinate reservations among multiple cities—indeed, among multiple airlines—added still another dimension of complexity. For most of their history the airlines had been scheduled like railroads: numerous stops along a single line, with passengers boarding and disembarking at each stop, sometimes to board a second flight on a different airline flying in a roughly perpendicular direction. Confirming the itinerary for a single passenger traveling on two connecting airlines required no fewer than 20 separate communications, making the airlines, even in the 1940s, among the world’s biggest users of leased telephone lines. But what if the entire process could be automated? Petzinger Jr., Thomas (2010-12-01). Hard Landing: The Epic Contest for Power and Profits That Plunged the Airlines into Chaos (Kindle Locations 1270-1275). Crown Publishing Group. Kindle Edition.

40 points by unixhero 11y ago

yarper 11y ago

Is there a version in English?

unixhero 11y ago
I have not been able to find an English translation. Google Translate does a great job at this text and I think its worth the work for people interested in early global networked messaging and computing with Mainframes, Unix, pre-TCPIP, post-TCPIP etc... After googling after some exotic hardware mentioned in the text, I found a translated section from this text in "History of Nordic Computing" [0]. I am helping my girlfriend with an internal project in SAS, and I found this text when I was searching for documentation on an ancient Mainframe "QQLIF file" which is still in use today! This text does indeed contain a description of it: "All messages in and out of systems were logged on logtapes, so we started using them to extract traffic in the network. IBM mainframe computer was we used for our data processing and SAS (Statistical Analysis System) was the "language" we programmed in. At an earlier point, we had been losing track over which terminals we had and where they stood. Therefore, we had already built up a widely used file called QQLIF and contained a variety of information about each terminal. It was both a real time online "direct access" file at Unisys, and a batch copy on IBM. Mainframes known terminal's trust, and by means of QQLIF file could be translated into PLID. Because of strict PLID'ens building could we add line drop, controls, and terminal for individual messages." [0] https://books.google.dk/books?id=c7MlCQAAQBAJ&pg=PA276&lpg=PA276&dq=lorenz+agentset&source=bl&ots=kH7t_NKsLg&sig=v3PTowEl_6lY8X-nxafQ9fOA5Ac&hl=en&sa=X&ved=0CCAQ6AEwAGoVChMIq8jBkeDpxwIVxZ1yCh26zgoK#v=onepage&q=lorenz%20agentset&f=false https://books.google.dk/books?id=c7MlCQAAQBAJ&pg=PA276&lpg=P...
- fulafel 11y ago
  Google Translate told me "The page you requested was too large to translate" - got a working GT link?
  - unixhero 11y ago
    I am afraid you have to translate the section you wish to read by copying and pasting into Google Translate. A minor hassle imo.
- PMan74 11y ago
  > an ancient Mainframe "QQLIF file" which is still in use today! I work in the travel industry, I was recently looking at a system's documentation with included revision tracking going back to the 1980s. System does the same thing it did in the 80s (presumably with completely different hardware/software)

JacobiX 11y ago

It would be interesting also to see what they are using now.

yarper 11y ago
Small airlines use a collection of software between their business, if they do their own maintenance they use a tool like AMOS (http://www.swiss-as.com/main.do http://www.swiss-as.com/main.do) which is produced by a subsidiary of Lufthansa, then another tool like sabre (http://www.sabreairlinesolutions.com/home/software_solutions/enterprise_operations/ http://www.sabreairlinesolutions.com/home/software_solutions... ) to manage crew, flights planning etc. Bigger airlines will probably use SAP. The hardware is going cloud based/vm but due to a ton of legacy and regulation concerns (i.e. data protection, rules regarding data access in the event of accidents etc) it's very slow going and most companies will still host their own servers, with maybe a company-wide shared oracle/db instance. The market is strangely competitive, though it has become quite stale in my opinion. RAMCO (http://www.ramco.com/aviation-suite/campaign/aviation/ http://www.ramco.com/aviation-suite/campaign/aviation/) are the closest thing to a "disruptive" force, but their software is widely regarded as flashy with little meat to it (and a lot of glitches). Source: developer for one of the MRO software providers (not listed here)
- atdrummond 11y ago
  Would you have an email where you could be reached? Would like to ask you a few more questions about the industry.
  - yarper 11y ago
    Sure, I'll drop you a email
justliving 11y ago
More than 100 airlines worldwide are using Amadeus (amadeus.com) today. There are different components to it such as Amadeus altea inventory (for managing their flight inventory, prices etc), altea reservation and altea DCS (for check-in and boarding). Additionally there are products to manage their websites (such as the e-commerce solution of Amadeus), revenue management systems etc. The main competitors are sabre (as mentioned) but also HP, Accenture, SITA and others are providing solutions in the same field. It's a pretty mature industry and btw one of the first ones to move to the cloud.

tempodox 11y ago

I can only curse myself for not learning enough Danish by now. Another interesting source of practical applications of computing machinery is the book “Computer” by Herb Grosch: http://www.columbia.edu/cu/computinghistory/computer.html http://www.columbia.edu/cu/computinghistory/computer.html He also lets on that remote electronic flight booking was already possible in the early 1960s. Also have a look at Frank da Cruz's pages about computing history at NYC's Columbia University: http://www.columbia.edu/~fdc/ http://www.columbia.edu/~fdc/

danso 11y ago

If you enjoy reading about the history of airlines, I highly recommend Thomas Petzinger's "Hard Landing"...a good chunk of the book is devoted to the airlines' brutal fight over computer systems...reading it is a great look into how something so seemingly mundane as arranging ticket reservations was so revolutionary in changing the business...and how so fundamentally even primitive computers impact our lives: https://books.google.com/books?id=GF5GZaTd8qsC&lpg=PA62&ots=EKsOUN6uRE&dq=hard%20landing%20thomas%20petzinger%20computer%20george&pg=PA59#v=onepage&q&f=false https://books.google.com/books?id=GF5GZaTd8qsC&lpg=PA62&ots=... Here's an excerpt regarding the state of the industry before computers: > An airline seat is like fresh food—a grapefruit, say—in that it spoils after so much time on the shelf. Every empty seat taking off on every flight is a spoiled grapefruit and exactly as valueless. Both required time, effort, and money to create, and both came to a wasteful, meaningless end. And on an exceedingly large number of flights, the sale of one last seat, according to the First Rule of Airline Economics, could easily decide whether the plane flew the entire distance in the red or the black. As the builder of American Airlines, C. R. Smith had agonized continually over the problem of matching passengers with seats before the latter perished upon departure. The trick was to balance reservations against inventory, a far greater task than it might seem. At first, in the early 1930s, a single ledger had passed from hand to hand as agents in a central office recorded reservations and erased those that were canceled. > With the advent of the DC-3, as the fleets and schedules of the airlines swelled and as the seating capacity of each plane increased, a single book became impractical, so multiple books were maintained. These had to be reconciled frequently—a massively time-consuming process—and even so, each book was inherently out-of-date at all times. Books gave way to chalk and slate boards and eventually to electric light boards. More and more agents were crammed into the reservations offices, with the line of sight to the blackboard limited by the space between structural support columns—22 feet in most buildings of the time. Some agents peered through opera glasses. Teletypes clattered. Clerks known as “card boys” scurried from desk to desk, eventually to be replaced by mechanical conveyors. One blackboard gave rise to multiple blackboards, which quickly encountered the same problems that multiple ledgers had caused. > A science called “reservations theory” was born to cope with the exponentially worsening challenge, but to little avail; as with so much else in the airline business, there was no economy of scale in handling reservations. The more the airlines grew, the less efficient they became. The need to coordinate reservations among multiple cities—indeed, among multiple airlines—added still another dimension of complexity. For most of their history the airlines had been scheduled like railroads: numerous stops along a single line, with passengers boarding and disembarking at each stop, sometimes to board a second flight on a different airline flying in a roughly perpendicular direction. Confirming the itinerary for a single passenger traveling on two connecting airlines required no fewer than 20 separate communications, making the airlines, even in the 1940s, among the world’s biggest users of leased telephone lines. But what if the entire process could be automated? Petzinger Jr., Thomas (2010-12-01). Hard Landing: The Epic Contest for Power and Profits That Plunged the Airlines into Chaos (Kindle Locations 1270-1275). Crown Publishing Group. Kindle Edition.