Wikipedia on Cullinet and my comments on same
Wikipedia’s current article on Cullinet is long, detail-laden, and slanted. The difficulties are not of the sort to be fixed with my usual pinpoint Wikipedia edits. So I’ll just reproduce it here, commenting as I go. As for copyright — this particular post is as GPLed as it needs to be to comply with Wikipedia’s copyleft rules. All other rights remain reserved.
The company was originally started by John Cullinane and Larry English in 1968 as Cullinane Corporation. Their idea was to sell pre-packaged software to mainframe users, which was at that time a new concept in an era when enterprises only used internally developed applications or the software that came bundled with the hardware.
Actually, Applied Data Research got there first.
Rather than write its own products, Cullinane approached IT departments of major enterprises, particularly banks, to identify internally developed applications that he felt had potential to be productized and licensed to others. However, it proved difficult to sell these applications because most weren’t generalized and supportable systems. As a result, the company had to create its own utility packages. The first was a tape based source code management system, TMS, that competed with Pansophic’s (PanDA) and UCC’s products (UCC-1) in the space. TMS had the handicap of being “tape” and not “disk” based so it was never successful. The first breakthrough product was a report writer named Culprit, actually developed in-house by Gil Curtice and Anna Marie Thron, who had built the PHI payroll system. The product competed with Mark IV from Informatics but was perceived as a late entry in the report writer category.
I don’t know about those details one way or the other, actually.
The company struggled with financial stability until it branded a variation of Culprit, EDP Auditor, which was nothing more than a second name for the same product with a collection of predefined reports, but more importantly, special services aimed at the new discipline of EDP Auditing including the first EDP Auditors User, special support to give auditors independence of data processing which was very important to them. What was remarkable is that many corporations licensed essentially identical products. This led to serendipitous prosperity for Cullinane. As EDP auditors developed knowledge about business systems and computers, they could invariably produce reports faster than slower-moving internal IT departments. As a result, MIS departments would feel compelled to buy the Culprit version for their own use — to compete.
Also not unreasonable, and also before my time.
As the company prospered in the early ’70s it was approached by a consultant to BFGoodrich, Naomi O. Seligman, to consider taking over development of a Honeywell database management system called IDS that had been modified to operate on IBM and IBM compatible (RCA) mainframes. Actually IDS was originally developed by General Electric, and a Bill Curtis had supposedly gotten the rights to convert the system to run on IBM equipment.
First I heard of those individuals’ involvement.
The decision was made in early 1973 — primarily by John Cullinane, Jim Baker and Tom Muerer — to bet the company on the effort. Several executives joined the effort over the next three years, including Andrew Filipowski, Robert Goldman, Jon Nackerud, Ron McKinney, William Casey, Bob Davis, Bill Linn, and Ray Nawara.
Shortly after I became an analyst in 1981, the Cullinane folks bragged to me about their low turnover. John Maguire of Software AG promptly put me on the phone with a Cullinane ex-pat named Grant Osasa, who in turn told me of four VPs who left around the time of Cullinane’s IPO — Flip Filipowski, Jon Nackerud, Tom Muerer, and I think Bill Casey. Flip went on to found DBMS, Inc. with Ray Nawara, which led to a bitter breakup of their partnership. Of course, he bounced back amazingly successfully with Platinum Software. Flip’s version of his departure from Cullinane — not really contradicted by anything John Cullinane told me — is that he rose from being the first salesman to EVP, and was essentially running the company while John was out doing the IPO. When John returned to to more hands-on management, it was time for Flip to leave.
IDMS was to be a great bet for the company as it became the leader among many capable and popular products of the mainframe era. It competed with Cincom‘s Total, Software AG‘s ADABAS, Applied Data Research‘s DATACOM/DB, Computer Corporation of America’s Model 204, MRI (later Intel‘s) System 2000 and IBM’s IMS & DL/1.[2]
Good list.
John Cullinane mentored a series of future entrepreneurs and software industry executives. One of the early executives was Andrew ‘Flip’ Filipowski, who later founded Platinum Technology, Inc.. Another was Robert Goldman who became the CEO of several public software companies including AICorp.
I don’t recall Bob Goldman having actually been CEO of natural language pioneer Artificial Intelligence Corporation, but that doesn’t mean he wasn’t. He definitely was CEO of Trinzic, the company formed by merging AICorp and expert-system shell vendor Aion. He also ran Object Design, which merged into Excelon in a financial play; Excelon was eventually bought by Progress Software.
Jon Nackerud was a co-founder of Relational Technology, Inc., formed to commercialize the Ingres database management system. Prior to becoming a public company in 1978 the company’s name was changed to Cullinane Database Systems, Inc. The company changed its name again to Cullinet Software in 1983, partly because John Cullinane wanted to distance his name from the personal connection to the business when he turned the company over to Bob Goldman, and also in a nod to the importance of computer networking. Joe McNay, a board member, was particularly important regarding the company’s IPO, the first ever in the software products industry. Of note is that Greylock purchased some shares from John Cullinane in 1977 less than a year before the company was to go public. It was to be the early foundation on which their Greylock’s software technology investment prowess rested. It was Greylock’s first investment in a software company.
Just to be clear, Joe McNay was at an outfit called Essex. He wasn’t affiliated with Greylock, where the key guy was Henry McCance.
Cullinane’s public offering was of note as it was the first successful offering of a pure software products company ever and the first software company Hambrecht & Quist ever took public. Cullinet was also the first software company to have a billion dollar valuation, and the first to do a Super Bowl ad. Specifically, Cullinane Database Systems, Inc., went public in 1978. On April 27, 1982 the company became the first computer software firm to be listed on the New York Stock Exchange and later, the first to become a component stock of the S&P 500 Index.
There always are definitional debates, but those claims are not unrealistic. On the other hand — the way it was executed, that Super Bowl ad was not exactly anything to be proud of …
However, two quarters after the company went public IBM introduced its 4300 series. Its salesmen told all mutual clients that IDMS didn’t run on the 4300 series and that all IBM software of the future would be built with IMS/DL1. This caused a major problem as every IDMS customer went ballistic and every prospect went on hold. The company only had three months to solve this marketing problem, and technical problem, and remarkably, they did. Technically, it only required the modification of one instruction to get IDMS running on a 4300.
Obviously hyperbolic.
The solution to the company’s revenue problem turned out to be its new Integrated Data Dictionary. By moving very fast, the company used it to put IBM on the defensive and made its numbers, no small accomplishment. It then went from winning one out five competitions to winning four out five and this fueled its growth.
Beginning in 1979, in an attempt to promote less dependence on the database sales alone, Cullinane fully integrated financial and manufacturing applications with IDMS and decision support systems, another first. The company acquired financial applications from McCormick and Dodge, and completely rewrote them using IDMS. They also acquired an MRP system from Rath & Strong and completely rewrote it using IDMS. Thus, Cullinet had a suite of integrated financial and manufacturing systems, the first on-line database driven applications, and was a major competitor in what is now called ERP. The company had become a software power house. Eventually, it acquired a small Boston-based company called Computer Pictures whose graphics-focused decision support system had already been integrated with IDMS and was very successful. This team developed Goldengate, a Lotus Symphony-like PC product.
Lots of inaccuracies there. If memory serves, IDD was earlier than suggested in that passage, and the apps were later. Computer Pictures hadn’t sold much of anything before Cullinane acquired them.
Goldengate was a part of Cullinet’s flawed ICMS (Information Center Management System). The promise of ICMS was the ability to move data between the mainframe and PC desktop. Apple Computer was supposed to do the same for the Apple Lisa, but never delivered. ICMS was unveiled in 1983 as part of a splashy 20+ city closed circuit TV broadcast that focused on IDMS/R and fueled the market for Cullinet for the next two years, but it was obvious that it was getting harder to maintain its unbroken string of quarters with sales and earnings in excess of 50%.
IBM had introduced the term “Information Center.” The idea was pretty much the same as that of today’s data warehouses — keep two copies of the data, one for transactional update and one for analytics. But IDMS’ CODASYL/network/linked-list architecture wasn’t at all well-suited for analytics, so this wasn’t an area of strength.
That said, Steve Jobs did do a cool video for them in connection with the partnership, banging on a washing machine-sized disk drive to show his frustration at the difficulty of getting data out.
Goldengate was a mistake. The company should have developed PC based IDMS development tools, instead. Ironically, it had the technology under development which was later to become the foundation of PowerBuilder at Powersoft. In fairness many failures mark the landscape in that space and era including the infamous Ovation product introduced with great fanfare by Ovation Corporation in a race with Lotus’s Symphony suite attempting to create the early office suites now dominated by Microsoft Corp. Goldengate’s other flaw was that it was built pre-Windows which was expensive for Cullinet because of all the permutations and combinations of PC hardware and memory configurations.
That’s pretty confused. For example, Powerbuilder was an inherently Windows-based, client/server product. But it is true that the Powerbuilder team started out at Cullinet, before finding a home at Mitchell Kertzman’s company.
In 1983 John Cullinane, after 25 years in the software business, handed over the helm of Cullinet to Bob Goldman while he began to pursue other interests. Things continued to go well but eventually the company ran into trouble and Cullinane brought in a recent acquaintance, David Chapman, as CEO of the company. At the time, Cullinet had some $50,000,000 in cash reserves. David Chapman, a veteran IBM and Data General executive, started an aggressive campaign to acquire technology from other companies. The reason for bringing in Chapman was that the company had got hung up on the open architecture and relational issues. In other words, a company with an unparalleled record of outpositioning competition every two years, for sixteen years, including IBM, allowed itself to get outpositioned by IBM, and others, with the help of E. F. Codd and C.J. Date. This was the company’s fault, not theirs.
In 1986-87, David Chapman attempted to move the company to the more and more powerful minicomputers such as Digital Equipment Corporation’s VAX line of computers. In the process, Cullinet acquired some very questionable VAX companies but one had an outstanding relational DBMS but by then it was too late, the company’s $50 million nest egg had been burned.
In 1988, John Cullinane returned to Cullinet, fired David Chapman, and tried to salvage the company. By repositioning the company’s product line with a new product called Enterprise Generator, he solved the open architecture problem and the company was able to return to profitability by the fourth quarter. This made it possible to negotiate a deal with Charles Wang and Computer Associates.
In 1989, Charles bought the company for $330,000,000 in stock. It was a good deal for investors because the CA shares increased in value ten times.
Actually, Cullinet was already losing share to ADR (and others, but especially ADR) due to pre-relational product architecture issues. (Most of the competitors had inverted-list architectures, and these were more flexible than Cullinet’s network structure.) That said, Cullinet was still holding its own until IBM introduced DB2, and Computerworld ran Codd’s criteria for defining a relational DBMS. At that point, the bottom dropped out of all the independent mainframe DBMS vendors’ markets. In one quarter shortly before it was acquired, Cullinet got exactly two new-name accounts. I was told this by John Landry and Bob Weiler, who ran the company before it was acquired, after their little company Distribution Management Systems (DMS) was acquired by Cullinet.
Edit: I missed this earlier, but “25 years in the software business” by 1983 was obviously hyperbolic. For one thing, there wasn’t a software business in 1958.
Comments
12 Responses to “Wikipedia on Cullinet and my comments on same”
Leave a Reply
I always enjoy reading details about the history of Cullinet Software. I spent the first 5 years of my career working at the headquarters building on Blue Hill Drive in Westwood. One can hardly blame Cullinet or claim it to be a “flaw” that they developed GOLDENGATE before Windows. Let’s remember that it was May 22, 1990 before version 3.0 of Windows shipped – the first version that offered even the slightest real hint of multi-tasking. And Excel (version 2.1d) was the only GUI app that Microsoft themselves had available for their Windows offering. Oh well – no matter – GOLDENGATE was an awesome product. And Cullinet was a great company to work for. I learned a LOT during those years.
I worked for Culcorp from 1979 thru 1989; but out in California. A few notes: 1) both Nakerud and (the other) Casey returned to Culcorp at one point in the 1980s, 2) IDD was an extension of the original “directory”, 3) It would be nice if some mention of IDMS-DC (tp monitor) is included. IDMS-DC had features way beyond CICS when it was introduced, and even today it is hard to pick which is technically superior. The ability to run ordinary application logic WITHIN a combined TP monitor/database engine was revolutionary.
I worked for Cullinet until 1989 in Scandinavia. Today in 2009 I can positively and firmly say that IDMS was the best database ever, if one compares the database features. IDMS was more than a database ie IDMS-DC, IDD and ADS shipped with that making it an almost complete development and runtime environment on IBM/370. IDMS DB/DC ran clean on MVS, DOS/VSE and native VM/CMS. IDMS-DC ran under VM (Even IBM doesn’t have CICS running under VM native). Application code was portable across all the three. ie COBOL, ADS code and data definitions basically needed no change. I have never seen such compatibility anywhere else even in the modern era!
One can write a lot on why Cullinet failed in 1989. It was wrong poisitioning, market switch to relational db and euphoria on client/server computing. Cullinet itself could have easily saved te company if they had introduced the SQL in 1989, which was needed for marketing reasons. But that time the management at Cullinet were wrongly focussing on VAX products. In fact, IDMS Release 11 was announced in 1987 but shelved in 1988. CA brought out this as Release 12.0 in 1992 with sql support.
IDD – the data dictionary was an underestimated product. Even today there is no such complete dictionary on any platform! Active at development (100% ADS and database source is stored here) and active at run time (for ADS load modules).
One can go on… IDMS and Cullinet were victims of pseudo market perceptions. The company leaders gave up too early. CA made more money from the product than Cullinet. Applications made in the mid eigthties are still running at a variety of sites even now…
see the site for more stories !
Just when most ex-Cullinet people are quiet on the product, one comes across new comments on IDMS in connection with NOSQL databases! This is a new front!! (google on NOSQL and IDMS) …
The failure of relational dbs to properly support java objects (domain model) and the emergence of yet another layer of ‘database type’ coding in ORM products like hibernate to mimic an E-R model(!) have also given rise to a new breed of products – NOSQL databases. A sort of re-inventing the wheel!! And in one of the discussions, someone who learned IDMS only in the late 90s has praised the product!
Also Microsoft’s own documentation on Entity FrameWork uses the terminology E-R Model. Incidentally it is in the Microsoft product, one sees a proper database diagram comparable to the IDMS’s Bachman Diagram. All the well known relational products use an incomplete or non-standard diagram. Either too much info or too little, never standardized!
New databases talk about supporting BLOBs (Binary Large objects). Nothing new if one recall that IDMS dictionary load area already stored load modules in binary form and used them in OLTP live applications. The old product ICMS/IDB also could store unstructured data. Finnish Firm Vegasoft stores pictures (JPG, GIF) etc in the IDD dictionary in their Webserver for IDMS. As for character data, XMLS have been stored in IDMS record fields defined as, for example, PIC X(10000). Compression will also optimize the storing.
That’s all for this year!
[…] writers were around in various forms for decades; consider for example the early 1970s history of Cullinane/Cullinet. By the time I became an analyst in the early 1980s, these were mainframe tools that let you […]
If memory serves, I recall attending training sessions for Cullinet software in the late 1980s. I even recall the names Dan Lynch and Don Casey as instructors of some of those training sessions. 🙂
[…] grow with them. Often these were programs written at enterprises, rather than third-party packages. Most of Cullinet’s product line, including its flagship DBMS IDMS, was came into the company that way. ADR, if memory serves, […]
[…] a famous story of Cullinane Software making payroll only because a customer’s check arrived. Less well known is one of MSA in the […]
[…] Computer Pictures and thus Cullinet used a touch screen over 30 years ago. Great demo, but not so useful as an actual product, due to […]
[…] Aion came along slightly later than the four horsemen, and outlasted them, perhaps because its product was targeted at the IBM mainframe/enterprise market. Aion later merged with natural language query pioneer Artificial Intelligence Corporation to form Trinzic Corporation, which later was absorbed into Platinum Technology. Ironically, Bob Goldman ran Trinzic, while Platinum was of course founded and run by Andrew “Flip” Filipowski, Bob’s predecessor as the #2 guy at Cullinet. […]
“Technically, it only required the modification of one instruction to get IDMS running on a 4300.
Obviously hyperbolic.”
No, exactly right. I was the systems engineer that investigate the problem and determined a parameter change in the sygen was needed, to load ADS modules from the Dictionary Loadlib, using DB IO instead of system services.
Where did Cullinet go wrong? As a software engineer in Corporate Tech Support writing PTFs for IDMS-DC, I attended a developer meeting for IDMS/R and Goldengate with 4GL, ADSO dynamically under the covers. The presenter was working closely with Dave Litwak on the presentation, and it became clear to me that it was still an active project in the roll-out phase, when it really looked to me, to be in the late planning stage. Goldengate was a PC app, and I knew the app technology in the PC world was way too primitive for the Integrated Dictionary Database word. Simple things like data integrity, had no meaning in the PC world. Mr Cullinane was ill advised, in my observation. The intended product was the future alright, but the way to get there with mainframe discipline was just not possible.