UK-based website

The Alternative TPF Homepage
Serving the online TPF community since 1995 : Now receiving >12,000 hits per week

Where To Next ?
TPF Scoop

   TPF Jobs
  Online Store
TPF Talent
TPF Help
TPF Site Search
TPF Net Search
TPF Link Central
What is TPF ?
TPF TravelGuide
TPF vs. Unix
TPF Futures
  Where are they now?
TPF Salary Survey



Any browser will do....

Animated GIFs by:

SCOOP Archives for older news here 


As IBM put it so succinctly on their site here

Lots to say about this one in the months to come I'm sure...until then catch up on your reading at the TPF site with their complete collection of TPFUG Presentations



z/Transaction Processing Facility Enterprise Edition (z/TPF) V1.1 (5748-T15) is a high-performance operating system specifically designed to provide high availability for demanding high volume, real-time transaction processing for mission critical e-business applications. z/TPF runs on and exploits IBM's zSeries servers, an infrastructure offering for transaction processing with high quality of service demands. As the next generation of on-demand transaction processing systems, z/TPF has been developed on the architecture of TPF 4.1. TPF 4.1 is widely used in Travel, Banking/Finance, and Public Sector industry segments and is well suited to business computing environments requiring:

* Management of extreme transaction volumes - z/TPF can process tens of thousands of transactions per second from hundreds of thousands of end users. Additionally, A z/TPF loosely coupled complex can scale to as many as 32 zSeries servers.
* High reliability/availability - For mission critical applications, downtime can effect losses of millions of dollars an hour, not only from revenue loss, but also from bad publicity and dissatisfied customers.
* Fast response time - When response time is critical to help lower call center costs or enhance customer satisfaction, z/TPF is designed to provide fast and consistent response across predictable and unpredictable peaks.
* Low Cost - z/TPF offers a low cost per transaction for high-volume real-time transactions. This capability alone represents a competitive advantage for delivery of business services in the marketplace.
* Efficiency - z/TPF supports IBM's 64-bit z/Architecture with the use of 64-bit real addresses and 64-bit virtual addresses enabling large-scale memory spaces permitting large applications and large memory tables.
* Open development environment - z/TPF also uses the GNU tool chain. z/TPF can share applications, tooling, and development infrastructure with the most common open system available, Linux.
* Flexible cost structure by supporting IBM's Workload License Charges. WLC is a monthly license pricing metric designed to support today's on demand business requirements. It allows customers to grow hardware capacity without necessarily increasing IBM software charges and manage IBM software cost by managing workload utilization.
*The z/Transaction Processing Facility Database Facility (z/TPFDF) (5748-F15), provides database management functions to z/TPF:
* Can help increase programmer productivity by providing centralized database handling routines so that application programmers only need to understand the logical relationships of data, not the physical characteristics.
* Helps simplify the application programmer's job by allowing high speed access to persistent data on z/TPF while providing a simple application interface.
*z/TPFDF also provides the database administrator with the tools to manage the database without necessarily affecting application programs.



It is IBM's intention to provide High Level Assembler (HLASM) capability on Linux for zSeries prior to, or at the time of, the general availability of z/TPF Enterprise Edition V1.1.

It is IBM's intention to provide z/TPF support of the Sub Capacity Reporting Tool (SCRT) within nine months of the general availability of z/TPF Enterprise Edition V1.1. SCRT is intended to allow z/TPF to report utilization for the purpose of sub capacity pricing (WLC).



IBM Grabs Amex TPF System

"26 February 2002--American Express has announced that IBM will take over a
large part of its technology infrastructure in a $4 billion, seven-year deal. 
The agreement will transfer American Express data center processing, hosting
of the company's Web site, network servers, data storage, and help desk
support to IBM. American Express processes more than 1 billion transactions
The move, which is effective 1 March, came after American Express posted a
56% falloff in fourth-quarter net income due to the global travel slump and
a U.S. recession that hurt its charge card, travel, and money management
units. It is expected that approximately 2,000 American Express technology employees will be able to transfer to comparable positions at IBM. 

Full details at:

                                                                           (31st March 2002)

Another Try For An Online TPF "Community" 

Another brave soul is trying to establish an online community for TPFers to exchange views and information. The link is below, it is part of the Yahoo clubs/groups section.  We're really hoping that one of these tries really works and we do get a vibrant, active online area where lots of TPFers really go and exchange messages.  Frequent watchers of this site and the net in general might remember some earlier attempts to get something going.  Perhaps the most disappointing was the TPF newsgroup that took a long time and significant effort to get underway but then died.
                                                                            (31st March 2002)

Father of ACPDB/TPFDB/TPFDF Hans Eisele Retires

A significant figure in the history of TPF has retired, Hans Eisele, the father of ACPDB, the TPF database system that has evolved to become TPFDF.  As he retired he circulated a note to friends and colleagues. We have taken the liberty of reproducing this below, and we hope not to offend Mr Eisele by doing so. 

"Dear All
After more than forty years of dedicated work for Swissair Schweizerische
Luftverkehrs AG and its newer style companies it is now high time to say
good bye.

I enjoyed over so many years high motivation to work hard, do a good job,
enjoy the work being driven by the enthusiasm for the Air Transport
Business and the pride for how the little country Switzerland could
steadily build up connectivity across continents and thanks to using
computers early to compensate for the high Swiss labor cost. What I have
learned at school about the explorers who conquered the seas, whilst the
Swiss were sitting in their narrow and cold valleys, having little to eat,
starving a lot, having only human beings as capital to send out as soldiers
or to work on farms e.g. in Eastern Europe, we could now run our business
and produce goods and reach the world. I sincerely hope we will not fall
back to those times but will reconnect to the world. It has been so great
to work with so many good people over the years. I hope that after the
crazy ways of the nineties, it becomes again more fashionable to work
primarily for a sound sustainable wealth of the company and less for
short-term balance sheet massaging and personal ego.

I never expected to see our Swissair to close down just about the same time
I would complete my work contract by this December 2001! It is very sad to
end ones life job such a way. And that Atraxis, which stands for Attractive
Swissair Information Systems, has come to an end too, just shows how much
the Swissair Airline Business was driving so many businesses. I hope that
all of us who can help to let some 26 long haul plane connect to the key
places of the world and some other 26 competitively sized airplaines would
ensure vital links to main European centers, will do whatever they can to
make it happen. I know we all had to suffer a lot after the Grounding Event
which immediately struck us with the consequence of loosing a global
Airline Network which can NOT be operated on a stop and go basis.

And I am happy to see that our core product RES and DCS which bases on
IBM's hardware and software platform 390/TPF will continue to exist with
EDS well beyond Swissair and Atraxis. What a good job IBM did with the 360
system architecture and the forerunner of TPF called ACP Airline Control
Program. Everyone would have bet in 1967 that Swissair would last longer
than the 360 type architecture. It has given a life time job to many of us.
Even so computers base on very simple switches which can either be on or
off i.e. 0 or 1, it is human brain that combines zero's and one's into
meaningful things. It is the same with business, one can do it right or
wrong and if too many things are too quickly done wrong, it really goes

I wish you all the best for your future, enjoy a peaceful and relaxing
Christmas period and start a Happy New Year 2002 which may bring all back
what we missed so much in 2001 (with a little bit of luck).

Thank you all for helping me to have so many years of pleasure and joy
working for what used to be a great company.

All the Best.
Hans Eisele"

We'd just like to add our wish that Hans Eisele enjoys a very well-earned retirement after a most distinguished career in TPF.

                                                                   (31st March 2002) 

IBM Ends Up Running Galileo...

It was announced Monday 3rd December that IBM and Cendant have agreed an outsourcing deal worth around $1.4 Billion for Big Blue to run data systems including the Galileo GDS on behalf of Cendant.

The official IBM Press Release is here.

According to the release 500 Cendant employees will become IBM employees as a result of the deal.  We're not clear whether that means that all Galileo staff will move over, does anyone know ?

It isn't immediately obvious whether this deal is good for TPF and TPFers or not.  On the one hand we have a large GDS in the hands of people that understand the running of TPF systems and have vast resources.  On the other we have a company that has been sitting on a valuable asset (namely TPF) while everyone else in the technical sphere has been wondering how to meet the exploding demand for high-availability, high performance computer systems that can handle millions of transactions a day for low cost and offering almost infinite scaleability to meet growing demands.

By my reckoning this could mean that Galileo might become the centrepiece of an innovative, net-centric system that pioneers the appropriate integration of newer architectures with TPF to form a world-beating combination of speed and rich functionality, or... Galileo might be run into the ground while the last drops of income are squeezed from the outsourcing contract to be eventually replaced by a Frankenstein-like architecture of whatever passing fad grips the industry next.

I know what I'd rather see happen (or at least be attempted guys !).

But what do you think ?  Why not let us know and share your ideas with those that might actually have some influence here....? 

                                                               (4th December 2001)

 EDS Assimilates Atraxis: Are You Next ?

The relentless rationalisation of the travel (and by association TPF) industry continues with the purchase of Atraxis by EDS.  We now have two of the largest players in the global TPF/Travel arena, EDS and Cendant, who are relative 'newcomers' to the market.

Nobody can doubt their newly acquired industry muscle or business acumen but few would have predicted 15 years ago that the principal TPF systems in the world would be owned independently of the major airlines.  It looked more likely that groups of airlines might form and centralise their systems but in reality they have got out of the IT business almost entirely.

Robert Crandall was famously quoted, more than 15 years ago, when he was CEO of AA, that given the choice he would rather keep SABRE and sell the airline.

Rationalisation is often good for the industry in the short term, and for the end consumer, but rarely for the people that work in that industry.  For them (that's 'us' !?) these are troubling times.  As airlines retreat from IT and hand more and more responsibility for TPF over to fewer and fewer entities this inevitably means a shrinking job-market.  Couple that with the shaky financial aftermath of September 11th and there is no doubt 2002 will be a difficult year to be in the TPF business.

From the EDS/Atraxis deal there are further redundancies expected this month.  No doubt the accountants at EDS are preparing for a bountiful holiday season, with bonuses all round as the 3000LB Gorilla of the TPF world enjoys another healthy meal.

For many others who have helped create the wealth that fewer and fewer TPFers are able to share in, the prospects for the Christmas holiday season are not looking quite so good.   There will be a lot of belt-tightening, resume-writing and reading of 'Help Wanted' ads replacing the office party.
                                                                             (22nd November 2001)  

Mr Richard Eastman (Well known TPF Guru) Certainly Generated Some Comment (see TPF Scoop

We have had more e-mails following the report of the comments made by Richard Eastman about TPF's inability to cope with proposed new anti-terrorist cross referencing of passenger data that it has been hard to pick some for publication (some of the language used has also made it difficult...).  But this is not unrepresentative of the opinions expressed:

"The problem with that, said Eastman, is that the government and the airline
industry don't recognize the differences in each other's systems. They just
think they can write some code and make it work, he said.

"They don't know what they're doing yet," he said.

The above quote from Mr. Eastman shows how truly ignorant he really is when it comes to TPF, or, even, the broader area of OLTP in general. We TPF professionals have been aware for very many years of how different TPF was and is from other systems. Why else would most TPF systems rely on an MVS system (or VM) for other, conventional, data-processing needs. Being a certified network engineer from Novell and a CNA from Cisco, and, having worked both inside and outside of TPF for over 23 years, I challange any OS either mainframe or distributed to achieve the transaction speeds (and data integrity) of TPF. The proponents of Himaleya, NT, Win 2000, SQL Server and Oracle, to name a few DB's and OS's cannot seem to grasp what thousands of transactions per second means. Mr. Eastman says we don't know what we're doing yet. Has he any inkling that TPF was written and developed by the Airlines for the Airlines and is yet to be replaced by any system equally as robust. We TPF programmers were coding our own hashing algorithims in assembler when 99% of the presumably IT literates didn't even know what hashing was, let alone IBM assembler. It never ceases to amaze me how people with such a limited grasp of a subject can be touted as authorities on that subject. The case of Mr. Eastman and his followers must truly be an example of "the blind leading the blind"."
                                                                         (17th November 2001)      


AMTRAK Unable To Escape Sept. 11th Effect

We recently received this from a reader at Amtrak (once thought to be one of the safest and surest TPF shops.... 

"It looks like this quarter will be an interesting for us at the Trak.
The Amtrak reform council wants Amtrak to report to the knackers.
Amtrak disagrees.
Bombardier wants the money it feels it is owed.
Amtrak disagrees

If I were to meet the person who said "May you live in interesting
times.", I would feel obligated to throttle the life out of the SOB. " 

Does anyone have further info on this ?           (15th November 2001) 

Airline Systems Might Be Called Up In Terror War 

According to this article:,4125,NAV47_STO64966,00.html

there may be moves to link airline systems and GDSs to some sort of Government (or International) database to cross-check passenger names against terrorist suspects or known aliases.

This sounds a good idea but the article is notable for some curious views expressed by so-called experts.

"Some of the DOT's recommendations call for the airlines' IT systems to
perform event-based functions, like red-flagging a suspected terrorist's
name in a reservation system. While functions like that are relatively simple procedures in a relational database, the databases used by existing airline reservation systems can't perform them. And adding the capability to screen passenger names could cause bottlenecks that would bring the airlines' networks almost to a halt, according to industry experts. "

At first sight one can only imagine that the experts consulted were experts in some other field.  

"Secretary of Transportation Norman Mineta's Rapid Response Team on Airport Security issued a report Oct. 1 that recommended, among other measures, a nationwide program of voluntary prescreening of passengers. 

The problem is that mainframe-based reservation systems were "never designed to respond to high-speed automated relational queries driven by hundreds of millions of users," said Richard Eastman, president of The Eastman Group Inc. in Newport Beach, Calif., in a soon-to-be published report. " 

Of course, since this is a 'soon to be published' report, we can't actually see whether it is worth the paper it is to be printed on (no doubt it will be printed by The Eastman Group, so lots of independent editorial integrity there).

""Whenever you make even the smallest changes in Sabre or any TPF environment, it takes an unbelievably long time to do it - time and resources," said Tom Cook, former president of Sabre Technology Solutions. That company became Sabre Holdings in 1999, when Fort Worth-based AMR Corp., the parent company of American Airlines, divested itself of its interest in Sabre. Cook now heads Replane Inc. in Chicago. 

"The code is old and structured," Cook said. "Making changes in the TPF environment is orders of magnitude more difficult than it is in a modern environment." "

Presumably here Mr Cook (who's he ?) means that when changing mission-critical systems that process more transactions than any other commercial systems and involve the largest private computer networks on the planet, this sort of change is approached with care and thorough testing before use.   The widespread use of the internet has brought home to many average consumers just how difficult it can be to supply a consistent, fast and robust service 7x24x365.  I would hazard a guess that Mr Cook has limited technical knowledge of TPF and has based his comments on a (senior) manager's perspective that extends to budgets, time-to-market and company image, while worrying little about the less glamorous business of creating, deploying and maintaining robust software that can survive some of  the most strenuous computing environments in the commercial world.

"Even the company that developed the reservation systems in the late 1950s and early 1960s acknowledges that security queries are going to cause performance problems. 

"That could be true," said Michael Hulley, vice president of the Global Travel and Transportation Industry unit at IBM, referring to the possibility of performance problems. "The reservation system is meant for doing mostly one thing: acting at a very secure high speed with high availability. It is not looking for triggers," like a terrorist's name in the passenger list for a flight, he said. 

To screen passenger names, the airlines could add another processing unit to run checks without sending the name through TPF, said Hulley. The reservation system could send passenger names to the old system, as well as to the new unit, working around the old TPF's inability to do queries. Programs that scan passenger profiles, not just names, could run in the second processor, Hulley said. "

This sounds more like the voice of sanity.  Almost regardless of the requirement for these extra checks they could be handled, if necessary, by an external system (provided that external system, presumably running one of those 'easy to maintain/develop', 'modern, non-mainframe based' Compaq !? systems could handle the load....

"The problem with that, said Eastman, is that the government and the airline industry don't recognize the differences in each other's systems. They just think they can write some code and make it work, he said. 

"They don't know what they're doing yet," he said. "

Is it just me or do others of you see the irony (humour) in this last 
statement ?  I'm sure when the two parties do recognise any need for technical guidance Mr Eastman will be hovering somewhere nearby to offer his own brand of (expensive) advice.   As it stands Mr Eastman clearly has little respect for those of us intimately involved in these antiquated TPF systems.  Obviously we wouldn't have the first clue how to interface our out-dated monolithic systems with any other architectures.

There is another article here:,4167,STO64967_KEY241,00.html

that mentions in a fraction more detail where these people believe the problems lie. Most TPF installations around the world have recognised this already anyway and have been starting to work towards some sort of more centralised and widely accessible repository for passenger data.  The problem was, is and is likely to remain one of performance.  In the article above there is reference made to the fact that an airline agent's set has to 'wait' to be 'polled' to gain access to the central system.  I might be a little behind the times here but aren't response times across most TPF networks still far superior to other online architectures ?  

Have I missed something ?

I really think we need some debate on this issue.  If you'd like to get involved and express an opinion we'd be grateful to receive it and hope to publish them to foster a better understanding of the problems and possibilities.  So send us your thoughts. 


EDS Joins List Of TPF Users 
To Shed Jobs

EDS has completed 2 rounds of layoffs. One a general layoff in the
Information Solutions jobline, of which TPFers are a very small cog. Some
people were moved to other positions (like project managers and business
analysts). Of that round of layoffs, 11 were from the LAX shop (Southern
California Solution Centre), the main TPF site. Only 2 of those (that we know
of) were actually programmers.

The next round of layoffs occurred one week later, and staff were told it was
directly related to airline development commitments due to September 11. 18
people were laid off from the LAX shop, 7 that we know of from the Miami
shop, From LAX, that included 6 TPF programmers and 4 MVS programmers. The rest were project managers and business analysts. From Miami, they were
mostly programmers, some who had recently completed a TPF class. 

EDS are currently going through their annual staff reviews, in which everyone is
placed into one of the dreaded "quintiles". People rated a 1 or 2 get raises
and bonuses and more opportunities. People rated a 3 or 4 are considered
nuts and bolts type programmers. People rated a 5 are encouraged to improve
or leave. The assumption was that people laid off were rated a 5, but we
don't know for sure. So now everyone will be very nervous to see what their
quintile is for the next year.

More could come depending on Continental's business needs over the next few
months. None of the Indian programmers are known to be affected. There is some controversy over whether EDS will save any money in the long run, because so much support is necessary for the Indian workforce. One EDS employee was quoted a saying "Basically, we have to hold their hands and do all of the research for them. It comes down to 'code this line in this subroutine' type support".

So far we don't have too many examples of remote (aka "inexpensive") development or support projects working terribly well in the TPF world.  If anyone reading this has any other stories of past or future projects of this type please let us know.  Many of the same issues would need to be faced if large scale remote or teleworking was considered by TPF shops as well so it might be interesting to hear of experiences people or companies have had dealing with this difficult but potentially exciting working arrangement.            (24th October 2001)

Belt-Tightening Continues: TPF Job Losses Continue To Grow 

United TPF staff & contractors are reported to have dropped from 110 to 55 in 2 months. Galileo's new owners Cendant are in the process of reducing staff but so far are only resorting to 'natural wastage' methods.

At Atraxis the decision was taken to reduce the workforce by 25% which Atraxis management hope may be accomplished "with the minimum possible hardship to the employees concerned" .  The statement also added the following ominous byline: "The Atraxis Executive Committee will continue to monitor developments on the customer front, however, and must reserve the right to take further action if necessary at a later date" 

A process of prioritization of projects and job functions is now taking place within Atraxis to identify the surplus people and positions.  This exercise will complete before the end of October and redundancy notices will be issued at that time.

Now would appear to be a troubling time to be in the job marketplace, whether a TPF veteran or even those with the most up-to-date skills.  I'm sure most TPFers will be battening down the hatches and hoping the shadow of redundancy does not fall across their desk....                                (7th October 2001)

What Will September 11th 2001 Mean 
To The TPF World ?

Rumours have been circulating ever since the shocking events of a few weeks ago regarding cut-backs in IT expenditure and the financial plight of many of the airlines.  British Airways have announced job cuts and their new IT partner Amadeus have posted reduced earnings estimates already as a direct result of people's reluctance to fly.

In the US, as everyone must have read, the major carriers are shedding jobs in even larger quantities than in Europe.  There has also been the collapse of Ansett, who were rumoured to have been a future customer for Amadeus.

The latest financial casualty seems to be Swissair, where desperate finsncial rescue plans by Swiss financial institutions like Credit Suisse may still not be enough to prevent the loss of an airline whose TPFers have been key figures in some notable TPF developments over the years e.g. TPFDB, FDBR etc.

If anyone has any news of how the current chaos in the airline and financial worlds is affecting TPFers please let us know.  This is a situation we will have to follow closely over the coming months.     (2nd October 2001)  

Marriott Lets Contractors Go

This from a reader in the US: "Marriott Hotels has axed all its contractors (9 in applications, 2 that I know of in systems - about 30% of staff overall) once current projects are complete - anything from 'effective immediately' to 'end of October'. They have committed to keeping all employees through the current round of cuts, and they're talking about looking for opportunities to acquire weakened hotel chains after the current downturn 

Some TPFers' Views Of The SABRE / Compaq (HP) Announcement

"What's always missing, too, is the fact that these announcements come, the stock prices are affected, but then when they fail nobody ever knows.

SDT (Sabre of 1992) tried to replace the Freight piece of the FOS box with UNIX. After 2 years and $11million, it failed. Freight is still TPF. And how hard could that be? A few thousand airbills to keep track of, some prices and then send billing information to a billing system???

SDT had a contract with Transavia (circa 1993) to develop a booking system on UNIX boxes. After many millions with no solution in sight, it was canceled.

And I'm sure there were others I don't know about.

Sabre announces in Aug. 2000 that they are laying off and stock prices jump. Lawsuits are filed and they were force to offer everybody their job back That fact somehow didn't make the news."

"Although Compaq has provided a lot of "spin" to this story, they
can't hide the fact that Sabre is going to take 3 or 4 years and $100million
just to take some of their "look" traffic off some of their TPF mainframes (by
re-writing their "availability search" Applications to work on the Tandem

This may make sense for Sabre, as they are suffering from an
enormous number of "look" transactions being driven from the Internet. Moving
seat and fare availability functions to hundreds of cheap Intel-Itanium
chip based servers may be cheaper than simply adding more TPF mainframes. 

However, hidden In the small print of the press release you can read
that Sabre's "booking" Applications will stay on TPF for the "foreseeable
future". This creates a problem in that they will have severed the instant
connection between the shopping system and the back-end inventory system by
relying on what is essentially a front-end memory cache of availability
which may or may not be synchronous with actual inventory. 

Orbitz have tried the same model by having Worldspan's TPF system
send them seat availability updates as they change. The problem is that
when the number of these updates peaks at anything over 650 per second,
Orbitz cannot keep up. This has resulted in many publicised cases of customers
getting flight confirmations from Orbitz only to find that the airline won't accept
them because that particular inventory was already sold out. 

The quoted 75% improvement on the time taken to update fares has
nothing to do with the speed of TPF or the Tandem machines. They hope to
achieve that by redesigning their process of accepting data from ATPCO and
that can be done very easily by any airline reservations system regardless of
the platform used. "                             (28th September 2001)


COMPAQ Hype Sabre Announcement 

It would appear that early reports of TPF's imminent demise vis-a-vis the recent pronouncements of Compaq that they were to 'take over' the Sabre mainframe systems, have been somewhat exaggerated. Once again marketing people have been allowed to write the Press Release without their balance wheels. It would seem that far from replacing Sabre as we know it this project's scope is somewhat less ambitious. As we understand it the project committed is to replace 30% of the function of the FPC (the Fares functional complex), not the PSS RES complex, nor DCS. It is the bargain finder part of fares that will be replaced and it seems that simply doing that will require the 192 Compaq processors quoted in the press releases..

So instead of a project to replace the Sabre system we have a project to replace a small part of a subset of the Sabre system. Apparently the only parts of the system they feel safe attempting to migrate are those not directly involved in income generating transactions and that are essentially 'read-only' in nature. Not quite so ambitious or grand as we were being led to believe from the overblown announcement of a few days ago.

Yet even this stripped down project, estimated to run to $100 million, is likely to prove more of a problem than those involved believe.

The cynical amongst us might also see real capital gained from this overstatement of the Compaq role in Sabre in the light of the announcement today that Hewlett Packard is to take over Compaq.

Something else that may emerge as a real problem in all this is if TPFers currently working on the Sabre system believe that their future is in the balance, with all this grand talk of cost savings, productivity gains (40% wasn't it ?) and the sweeping away of this old mainframe technology by this bright, shiny, new 64-bit Unix variant on machines named after mountains. There are always plenty of eager TPF shops that would be only too happy to persuade some Sabre/EDS TPFers that they would have a more secure future with them.

So join us for a glimpse at what we see as the real context and significance of this brave decision......

Sabre, TPF, IBM & a leap of faith 

If anyone has further details about any aspects of this please let us know.          (4th September 2001)


Has SABRE Finally Lost The Plot ?  

No doubt many of you will already have read about Sabre's apparent decision to migrate off TPF.  In the unlikely event that you haven't heard you can check on the basic facts here 

I'm sure your initial reaction as a TPF professional will have been something like " That's ridiculous " , " That's impossible " , " How can that be ? ".   Those were certainly my reactions at first.  We are talking about the busiest TPF complex in the world being replaced by a collection of 'Open System' (i.e. Unix) servers.  Surely this must be a misprint, check the calendar for April 1st, am I awake, did someone slip something into my last coffee/beer/chilled vodka (delete as applicable)  ???   Perhaps another mischievous piece of spurious hyperbole from Anderson Consulting (or whatever they recently metamorphosed into...)  ?

Well something as momentous as this deserves a slightly closer examination and that's what we plan to do.  There are so many questions that spring to mind surrounding this announcement that it would be hard to know where to start.  So we're going to devote a whole section of the Alternative TPF Homepage to exploring, and hopefully discussing with you, just exactly what this announcement means, to Sabre, to EDS, to Compaq, to the travelling public, to IBM, to TPF and last but not least to TPFers everywhere.

TPF has never been the calmest of environments to work in and this latest event in its unique history has done nothing to change that...

So join us for a glimpse at what we see as the real context and significance of this brave decision......

Sabre, TPF, IBM & a leap of faith 

Of course there is also the very real possibility that we might all be overreacting to this announcement.   We don't actually know for certain that the plan is to completely replace all of the TPF functionality within Sabre.

Sabre has what is probably the largest 'functional complex' of TPF processors dedicated to satisfying Fare requests.   Fare quote has long been understood to be an application that could benefit from a massively parallel approach.  Perhaps this announcement simply involves the migration of the Fares complex to the Himalaya architecture.  

If anyone has further details or can confirm one way or the other what the actual plans are please let us know.          (3rd September 2001)  

Off-shore Off-Shoots Off-Side ?

After all the concerns there have been lately about off-shore TPF shops threatening the jobs of TPFers at some of the larger shops we have heard this past week from at least one concerned Indian technician that perhaps relations between the company he works for and their large US customer might not be so healthy after all.

Out correspondent was very worried that the US TPF user might run-down the amount of work being done in India.  Although they weren't aware of a specific problem they had heard rumours that things were not working out as well as originally hoped.

Whether our Indian TPFer friend's fears are unwarranted only time (or perhaps someone out there who knows ?) will tell.  In the meantime, does anyone have stories about past (or present) projects of this type ?  We'd like to compile some sort of information about how well these attempts at globally distributing the development efforts have really been.  In the Brave New World that is already seeing some of us buying our groceries on the net, perhaps we need to embrace what seems like a threat and see if there are ways we can make remote working improve our own lives.  Wouldn't it be nice to be able to consider living anywhere in the world instead of just those (ever reducing) places that have TPF shops ???

Send those experiences or observations of remote projects to us here and we'll create a dossier on the real effectiveness of them, as seen by you, the people that actually know....                                (27th July 2001)

Amtrak Off-Track ?

What follows is a link to a Washington Post article reported to us by a vigilant TPFer. It describes an impending 15% cut in Amtrak management.

Apparently, our contributor informs us, 'management personnel' are any people not in a union. Managers are people who get to give orders. Non-agreement covered personnel are not in a union, and they can't give orders. Programmers are non-agreement covered personnel, and not managers, but they are in "management". Hmmm, things might not be looking too good suddenly for techies, including TPFers at Amtrak.

The CEO is reported to have sent out an email warning of problems ahead.

Amtrak is currently under a Congress imposed death sentence that is
supposed to be carried out by the end of fiscal year 2002.

Apparently this is not deterring aggressive sales-folk from trying to sell Amtrak a series of solutions for their problems.  Out contributor puts it more colourfully:
"The only people who have nothing to fear are the snake oil vendors who repeatedly show up at our door offering the latest, greatest cure-all.
"Our product is guaranteed to make you profitable or we'll sell you
something half as effective at twice the price."

(27th July 2001)

Atraxis & Lufthansa Systems Group To Merge 

The Swissair Group and Deutsche Lufthansa have announced that they have signed a letter of intent to merge Atraxis Group and Lufhansa Systems Group. Swissair has already sold off its hotel chain and has also agreed to swap Rail Gourmet, and several other small catering businesses, for the airline catering operations of Compass, the UK service group.  

So let me see... Amadeus is 33% owned by Lufthansa. Atraxis will be owned by Lufthansa. These appear to be competitors in many ways. How will Lufthansa handle this situation ?

Does anybody have any more information on this one, or some ideas as to what this might mean for Lufthansa, those affected TPFers and the airline IT business in general ?  If so let us know.

(12th July 2001)

What's The Real Situation At Galileo ?

Following our recent news about the Cendant takeover of Galileo and the dim prospects for TPFers due to the temptations of cheap technical skills from India, we have received a couple of reports that at the same time Galileo had been offering incentives to staff to find TPF skills to work in Denver...

Obviously Indian technicians have been in the news recently in connection with BA and their decision to do away with so many contract staff, but we understand that no TPF people are affected by that move since almost all are now with Amadeus.

So what is/was really going on here ?  Does anyone have any information about the previous attempts to utilise skills in less expensive parts of the world ?  Does anyone have the real scoop on what faces the Galileo TPFers now that Cendant management are in control ?   Let us know.

One person at least wont be sad that Cendant moved in on Galileo, or that he no longer has a job with the new company.  That is outgoing CEO James Bartlett.  He stands to receive a payment in the region of $ 4.5 million for his part in turning Galileo into a takeover target.  Severance benefits to surplus executives are said to run to $ 15 million in total.  Something to bear in mind perhaps next time you sign a TPF contract that says you can be let go with two weeks notice.....                               (12th July 2001)

Galileo Staff May Have To Walk The Plank...

Word has it that some employees in Colorado have been served with an ultimatum by the swashbuckling new owners of Galileo: "Relocate to India for reduced salaries or....." .  This is unconfirmed of course (nobody from Cendant was available for comment) but cheap, offshore technical labour has been a temptation for other TPF shops in the past.   One way or another this plan has not worked so far, but with the increasing sophistication and organization of various offshore enterprises, in places where highly talented people can live well for much less than they can in the West, technicians from Europe and the US might find their biggest future problems consist of preventing the 'virtual corporation' ideas of the internet age from slashing their value in the job market.

So, what do you think about the threat of cheap, skilled, highly motivated but remote staff where you work ?  (22nd June 2001)

Galileo International'Rudderless' Galileo Ship Boarded By Cendant Buccaneers

Below we have the way the purchase of Galileo was announced via the Galileo website  Keen, no doubt, to put the best spin on the deal, little was made of the rapid departure of Galileo CEO James Bartlett, who was set to 'pursue other interests' (one of which is likely to be consulting the Jobs section). Suspicions have been widespread for some time that Galileo was making least use of its position as one of the major GDSs to move into the wider realm of e-commerce and full coverage of leisure revenue streams.  It had been referred to variously as 'rudderless' and 'adrift' with 'no perceptible strategy' by financial and industry commentators.  Perfect prey for the opportunistic Cendant, who are as agile as they are diversified, to bolster their travel and leisure interests.   Rumours persist that they may even pursue Worldspan and if that acquisition were to go ahead they would quickly become the major player in the GDS marketplace.  (Amadeus & SABRE/EDS take note)

Also a statement made by  Cendant soon afterwards included a comment that: it does not anticipate ``a reduction in the employee base that would impact service to customers,'' but a spokesman declined to elaborate.  If they continue and absorb Worldspan it is almost inevitable that some economies could be made, although not until the disparate systems were somehow connected (contracts anyone ???).

It wasn't too long ago that Cendant had a cloud hanging over their fiscal dealings when they were forced to correct falsified financial declarations extending over a period of years.              (21st June 2001)


Galileo Sale: Press Release Extract

NEW YORK, and ROSEMONT, Ill., June 18 /PRNewswire/ -- Cendant Corporation (NYSE: CD) and Galileo International, Inc. (NYSE: GLC) today announced that they have signed a definitive agreement for Cendant to acquire all of the outstanding common stock of Galileo at an expected value of $33 per share, or approximately $2.9 billion. Cendant will also assume approximately $600 million of Galileo net debt. For the year ended December 2000, Galileo reported revenues of $1.64 billion and EBITDA of approximately $570 million. The transaction will combine Galileo, one of the leading providers of electronic global distribution services (GDS) for the travel industry, with Cendant, a diversified global provider of business and consumer services, with a significant presence in the travel sector. The transaction, which is expected to close in the fall of 2001, is subject to customary regulatory
approvals and the approval of Galileo's stockholders. United Air Lines, Inc. (UAL), the largest stockholder of Galileo with approximately 18% of the outstanding shares, has entered into an agreement with Cendant to support the transaction and has provided Cendant with a proxy to vote the Galileo shares owned by UAL in favor of the transaction.     (21st June 2001)

Atraxis Seeking 'Institutional' Investor 

According to the Atraxis website:
"The SAirGroup has announced its intent to find an institutional investor for its Information Technology subsidiary, Atraxis. The process was formally initiated yesterday, and provisional results should be available soon.
The Atraxis Executive Committee and SAirGroup Executive Management are confident this is the next step in Atraxis' evolution to worldwide industry leadership. The decision will enable Atraxis to consolidate its global market position, and it ensures continuation of the impressive growth that has characterized the company in recent years. With its comprehensive network of partnerships and its use of the most advanced technologies, an independent Atraxis will provide even more flexibility in meeting the needs of its customers in the airline, airport, cargo/logistics and air transport-related industries throughout the world "

Since the term institutional usually implies non-industry-related, e.g. pension fund, bank etc. we don't appear to be looking at any of the current TPF players expanding further on the back of this news. However if outside interests believe they can streamline business processes then the TPF community might be in for some fresh challenges as a result.  What do you think ?  Let us know       (30th April 2001)

Atraxis To Open Development Facility In UK (?)

Atraxis are investigating potential sites for a UK development site in the Reading & Bracknell areas to the West of London.  Out information is that they are in the latter stages of a study aimed at providing a more stable base for development of TPF applications than Switzerland, with its limits on foreign nationals working there.  

Certainly if Atraxis was to locate a development centre in that area they would hope to recruiit some people directly from both British Airways and the fledgling Amadeus installation set up following the recent TPF outsourcing deal between BA and Amadeus.  

Perhaps this could signal other TPF businesses to consider locating in the UK as it is attractive to Americans and Europeans alike, with, apparently, fewer of the problems of locations such as Switzerland and Germany ?
(12 April 2001) 

IBM Finds First User For Super e-mail Server

IBM has found a user for the newly developed 'Super e-mail Server' operating under TPF.  Unfortunately we haven't been able to find out who it is (yet).  If anyone has any clues please let us know.  We'll keep our ears to the ground of course.    

The software was developed by AMSYS and apparently would allow maintenance of approximately 300 million e-mail addresses, more than enough for everyone in the US to have a separate e-mail address.

It can only be a matter of time before other popular internet applications, such as Usenet News are ported to TPF.  We wonder whether the TPF community will be more interested in using Usenet if they thought the server might be running TPF !?  (12th April 2001) 

Have The Wheels Finally Come Off The SABRE Juggernaut ?

Following hard on the heels of the significant layoffs of technical staff from SABRE that was reported earlier this year it now appears that former outsourcing and professional services rival EDS have purchased nearly $ 700 million worth of SABRE assets and also won a $2.2 Billion contract from the beleaguered behemoth. 

So, does this spell the end of the road for the former 2,000lb Gorilla of the TPF technical market place ?  Executives are naturally putting a positive spin on the situation and explaining that the streamlining is necessary for SABRE to retain(?) its No. 1 position in the Travel Services sector.

Alternative opinions are rife amongst the staff, where a general feeling that their leadership may not have a viable strategy are high on the list of misgivings.  Certainly it will be a while before SABRE can reclaim the standing in the eyes of their peers and potential customers following these two apparent acts of 'streamlining'.

What do you think ?            (15/03/2001) 

Who Wants To Know ?

Now I might be completely paranoid, but I saw this item posted to a usenet news forum and I can't help wondering why someone might want to request manuals for ALL these various major CRSs....anyone have any inside info on this ?

"From: Tiago Santos (
Subject: Galileo, Amadeus, Sabre, Apollo Worldspan Manuals 
Date: 2000-12-17 17:54:07 PST 

Does anyone have or know how to get Galileo, Amadeus, Sabre, Apollo orWorldspan manuals
Please reply to this e-mail adress

I'm sure it's quite innocent but one can't be too careful these days, there are so many shady customers out there...(ask Microsoft, Amazon, Yahoo etc.)       (15/03/2001)

So Which CRS Is Best ?

For some of the more widely travelled of you out there (in the TPF sense especially) you probably have formed an opinion as to which of the Big 4 CRSs is the best.   You will almost certainly have been subject to 'subjective' opinions of the CRSs themselves.  You may even have seen results of 'surveys' or collections of performance statistics or whatever.  But how should we, as TPFers rate a CRS system ?

Size ?  We're told size doesn't matter in some things but does it matter in a CRS ? Presumably if this was your major criteria there would be only one winner. But then does that mean size in terms of CPCs, DASD, Terminals connected, Number of major customers, Number of Flights Sold what ?

Sophistication ? Many of us will remember the days when TPF systems (or ACP) were very good at doing very simple things. Nowadays they still are, but get called upon increasingly to do more complicated things too. When Fare Quotes were first done online it was considered to be an unsuitable application for TPF because of the high ECB existence times and stretched the then definition of a 'transaction' to the limit.   TPF systems have gradually become more and more sophisyicated and more powerful hardware has allowed longer path-lengths without sacrificing that all important speed. Now it is vital to present the results of the work done by the underlying TPF CRS in the most useful way possible using intelligent client packages and web engines of varying degrees of complexity.  But for TPFers there is another level of Sophistication and that is at the TPF code level.  We've all worked at places and seen code and thought "That's a pretty neat/impressive/slick solution to that problem".  The question we should ask then perhaps is 'Which of the major CRSs has the most elegant code ?'

Stability ? Not much good having a big system with lots of flash code if it falls over on a regular basis or produces many thousands of dumps a day, with no hope of fixing them.  In this category we could include those CRSs that have been innovative in reducing planned downtime too.  This also indicates a level of 'Sophistication' but we'll leave it in this section.

So, we have our own ideas here but we'd certainly like to hear yours, and so would everyone else out there I expect.  So send us your nominations for Best CRS In The World and we'll try to make some sort of sense out of it all.  Try to give reasons why you nominate whichever system you choose and we'll publish as many responses as we can.

Incidentally those people over at the World Travel Awards made their mind up sometime last November, but then they're just customers, we can't expect them to really appreciate a TPF system now can we ?         (11th January 2001)

Interest Keen In Union For TPFers

We received this today, and it seems interest is quite high in a Union for TPFers after we published the activities of the folks at Delta and Worldpsan.

"Thank you for posting information regarding interest in Union for
Worldspan and Delta Technology. It seems that many Sabre employees are
also interested in information. Anyone interested should contact us at:
With TWA possibly moving from Worldspan there is even a greater need
for protection for our jobs.  Many have inquired about the pilot work slowdowns at United and Delta. They have unity, without unity we have no chance. Your work is important and critical to the airlines too. Unions can negotiate anything. What is important to you? Perhaps medical insurance upon retirement (seems to be a major concern in the TPF community). What about telecommuting (we all know that this is possible and would be welcomed by those who can complete assignments at
home). Let us know what you would want to see happen if you are represented by a union. If you are interested in meeting with CWA (Communications Workers of America), let us know that also. Thank you for your continued feedback." 

(11th Jan 2001)


A (Fine) Rumour To Kick Off The New Year:

There was a message on a Yahoo Message Board recently claiming that an article had appeared in a newspaper local to IBM HQ suggesting that Big Blue might be considering purchasing SABRE

Whilst this is likely to be mere speculation at best it is worth considering that SABRE's stock price has edged higher in the last week or so.  Industry speculation seems to be that SABRE would wish to close any deal by year's end... 

It would be interesting to imagine what such a purchase might mean though...It would finally give IBM a proper stake in TPF, rather than regarding it as the black sheep of the OS family and trying to pretend that IMS/CICS is just as good and easier to manage.  What do you think it might mean to the TPF world if it was true ?  We'll publish the best answers for a bit of light relief to welcome in the New Year 
(30th December 2000)    

'Irreverence' or 'Candour' ?

We recently noticed the following review of the Alternative TPF Homepage on another site:

"This home page with an irreverent attitude presents editorials on TPF issues in the news, technical information, job listings, and a salary survey. Check out the TPF Link Central section for what is, in our opinion, the most comprehensive list of TPF resources available on the Web."


Our aim on the Alternative TPF Homepage is to be just that, an alternative.   If you want to read the 'official' line on TPF where better to go than IBM's ever improving TPF resource. If you want to read a press release you can search from here for those sites that are simply acting as distribution channels for larger companies.  We prefer to think that we provide a little of what is missing from other sites, that might be of interest or use to TPFers.  That might be an alternative opinion or information that you just can't get anywhere else. 

Whatever Happened To 
comp.os.tpf ?

After the dogged determination shown by numerous people to establish the first proper internet forum for TPF activity now seems to be pretty non-existent.  So what went wrong ?  Plenty of people visit the growing number of TPF-related web sites, so clearly TPFers have internet access... So why don't we like to post messages ?

If you've got an opinion about this why not let us know and we'll try to get to the bottom of this puzzling mystery.  Perhaps what's missing is the virtual beer to go with the virtual discussion ?  You tell us.  

SCOOP Archives for older news here 


Updated: 14/05/02