United States. Congress. Senate. Committee on the.

The Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) online

. (page 136 of 140)
Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 136 of 140)
Font size
QR-code for this ebook


.7 Honeywell would have avoided involvement with the Bryant units on the
800 and 400 .systems.

15.40.70 The loss to Honeywell of not having suitable random access equip-
ment, available bv the shared technology, during the first half of the 1960's was
substantial and was further complicated by the fact that this was one area where
Hcnevwell was at the mercy of its competitors. IBM and SR.

15.40.71 IBM and SR capitalized on Honeywell's weaknesses by repeatedly
tnkina- pro.spective sales away from Honeywell during a period of potentially sub-
stantial Honeywell growth. .

15 40.72 With a strong random access unit from the shared technology m its
equipment offering, Honeywell's total would have accelerated during the period
from 1959 to 19f;5. , .

15.40.73 The effect of such equipment offering would have been cumulative
since, as Honeywell enjoyed a wave of customer acceptance, it would have re-
ceived further orders from those who had shown previous hesitancy.

15.40.74 With the addition of suitable random acce.ss equipment from the
shared technology, it can be seen retrospectively that Honeywell's sales for the
H-800 series alone would have tripled or quadrupled, and these sale's would have
had substantial impact on sales of the H-400 series and later of the H-200 series.

15.40.75 Hone^'^vell had never had a drum storage unit of its own design and
manufacture and instead of building on a shared technology base, had been forced
to go to other EDP manufacturers for drum storage units whenever a customer
demanded .such a unit with a Honeywell system.

1.5.40.76 Often Honeywell was not able to provide a drum storage unit nt all
when a potential customer demanded one and hence it lost the sale or lea.se.

15.40.77 Honeywell could have offered a drum storage system on. or earlier
than the H-SOO and would have with access to the technology .shared between SR
and IBM in 1957.

15.40.78 Honeywell hnd no desisn or production know-how in the area of tab-
card handling equipment in the late 19.50's and could not afford the gamlile of
undertaking to design and produce its own tab-card equipment with no base to
start from and no guarantee of success.

1.5.40.79 If Honeywell had had access to the technologv shared between IBM
and SR in 19.57, it could have begun to manufacture tab-card equipment by the
end of 1959 with confidence in success and provided its H-800 customers with
card equipment of its own manufacture : e-if^h year of deferred access cost money
and delayed the date of fruition for Honeywell.

15.40.80 Beginning such activity in 19.57 would have produced more competi-
tive tab-card products earlier with Honeywell features designed to complement
its own system designs.

15.40.81 Without access to the shared technology. Honeywell was not able
to deliver a card-reader of its own design and manufacture until May of 1965
after substantial expenditures and difficulties.

1.5.40.82 Similarly to the card-reader. Ploneywell was not able, without access
to the shared technology, to deliver a card-))unch of its own design and manu-
facture until mid-1966; access to the shared technology could have accelerated
this delivery.

15.40.83 If Honeywell had developed its own readers and punches for delivery
with its first H-800 in 1960, based on shared technology, it would have b^en able



5886

to market an "all-Honeywell system" then with single (not split) system responsi-
bility, and thus would have improved sales, market status and customer base
earlier.

15.40.84 With adequate and competitive reader-punch equipment based on
shared technology, Honeywell could have retained the profits paid to SR's co-
conspirator IBM from which units were obtained at list prices (either leased or
purchased) ; and the money spent in modifying its own equipment to suit the
IBM design, such expenditures being non-recoverable since the leases and pur-
chases from IBM were on a full retail price basis, and interface costs could not
be passed on to the EDP customer: earlier access to the shared technology would
have shortened the period of dependence on IBM.

15.40.85 Assuming access to the shared technology, Honeywell would have
acquired over the years after 1957. or after obtaining such access, a competency
in engineering, manufacturing and field maintenance which would have become
increasingly valuable because of the growing numbers of Honeywell systems and
card handling units installed.

15.40.Sr» With access to the shared technology, Honeywell's revenues would
have been increased in the following ways: by selling or leasing an increased
number of systems, by making an increased profit on each system sold or leased,
and by improving long-range opportunity, starting with the date of access.

15.40.87 From 1955 to 1965, the printer was a weak link in the Honeywell EDP
system offering and had a negative impact on the company's image as an EDP
equipment supplier for the whole decade.

15.40.88 Print quality was considered carefully when vendor-selections were
being made by several potentially large Government and commercial accounts
for which business Honeywell bid in 1955-1965.

15.40.89 Honeywell approached SR in 1956 to attempt to buy or lease for re-
lease the UNI VAC High Speed (600 lines a minute) printer and SR refused to
deal on any terms ; this was included in the technology shared between IBM
and SR ; and, incidentally, the patents on such printers were barred by public
use.

15.40.90 Honeywell then decided to offer Anelex printer at an advertised speed
of 900 lines per minute.

15.40.91 The Anelex arrangement was a costly and unsatisfactory interlude in
Honeywell's search for a solution to its printer problems.

15.40.92 Tlie Anelex printer occupied an inordinate amount of time on the part
of Honeywell's systems engineers and designers, and control over the design
proved to be of only limited advantage.

15.40.93 By March, 1959, the decision was made to manufacture printers for
the 800 system at Honeywell which began a tedious learning process lasting sev-
■eral years for the Honeywell developmental group.

]5!40.94 Print quality remained only marginally acceptable through the Honey-
well 822. 422, and 206 printers delivered by Honeywell during the period of
1960-1965 and improvements did not come until the Honeywell 222 printer was
fii'st delivered in 1965.

15.40.95 Honeywell poiired large amounts of money into printer engineering
and field maintenance, but its loss was far greater in terms of systems sales or
leases lost to competition.

15.40.96 Had the ITNIVAC high-speed printer design information and manti-
facturing drawings shared with IBM been made available to Honeywell in
1957 or later, steps would have been taken to produce the UNIVAC type printer
in-housp immediately.

15.40.97 The first Honeywell printer would have been designed and produced
in-housp Itv the end of 1959 [or later if access to the shared technology had been
later! for 'dplivpry with the H-800 system which would have been a better prod-
uct, both in cost and performance, than the printer Honeywell actually delivered
wit^li the system.

15.40.98 Assuming access to the shared technology, the print quality of the
H-800 or the 400 and the early 200 series printers would have been at least equal
to that of SR's ITNIVAC printer device and the impact upon sales of the H[-800,
H-400 and H-200 would have been substantial during the years 1958-1965 when
such systems were offered.

15.40.90 In 1957. Honeywell had little know-how on which to build an EDP
svstem manufacturing facility.
" 1.5.40.100 The only electromechanical EDP units manufactured by Honeywell
prior to 1964 were niagnetic tape units and printers with tape units receiving by
far the greatest amount of attention.



5887

15.40.101 Had the IBM/SR shared technology been available to Honeywell in
1957 or later, Honeywell engineers would have been able to learn in such areas
as optimal assembly techniques, preferred componentry, preferred materials, cur-
rent manufacturing methodologies and use of special tools and jigs ; knowledge
in these areas was the product of extensive prior cycles of trial, error, and re-
vision by IBM and SR product engineering and manufacturing personnel by the
time of the 1956 Agreement and was shared thereunder; early access by Honey-
well would have meant earlier accomplishment by Honeywell.

15.40.102 Had Honeywell had access to the technology exchanged between
IBM and SR in 1957 or later, there would have been improvements in Honey-
well's production engineering activities, in documentation and implementation
of design changes and in manufacturing performance and cost reduction before
they were accomplished without access.

15.40.103 Access to core storage stack technology shared between SR and IBM
would have increased the ability of Honeywell to choose, work with and evaluate
core storage suppliers earlier, which would have resulted in reduced system
costs and increased system performance earlier.

15.40.104 Information from service manuals exchanged between IBM and SR
in 1957 or later would have benefited Honeywell in the following ways as soon
as obtained and thereafter :

.1 In those cases in which IBM equipment would still have been used with
Honeywell systems, selection of specific units and cost estimates would have
been on a more knowledgeable and effective basis ;

.2 Honeywell service manuals, after exposure to IBM and SR manuals, would
have been improved ;

.3 Adjustment and calibration information necessary to achieve compatibility
with IBM and SR terminal and peripheral units respectively would have been
more readily available ; and

.4 Maintenance tools and test equipment respectively used by IBM and SR
would have been available to Honeywell design engineers and would have mini-
mized one of Honeywell's greatest technical problems, namely, disparity between
Honeywell maintenance aids and those of IBM and SR.

15.40.105 Honeywell could not have "reverse-engineered" the IBM and SR
electronic and electromechanical devices as to which IBM and SR shared in-
formation in the technological merger of 1956 and which would have been help-
ful to it in the period 1957 to date for the following reasons :

.1 It is impossible on small parts to determine what type of metallic material
was used, whether steel, brass, bronze, or some alloy ;

.2 It is impossible to tell from examination of parts which parts were made
by the i^eripheral device manufacturer and which parts were purchased from
vendors and, if some were purchased, it is impossible to tell the sources, pur-
chase specifications and inspection criteria ;

.3 It is impossible to tell by examination of the device what kinds of tooling,
or machine tools, fixtures, jigs and dies were used by the manufacturer of the
device to prodiice the desired shajies and tolerances ;

.4 It is impossible to decipher tolerances and adjustments of intricate electro-
mechanical and electronic parts since the particular machine which is torn down
may be in the middle, low or high position within the permissible tolerance range
with respect to each part ; and competitive tolerance ranges are not known and
are impossible to decipher by inspection ;

.5 It is impossible to determine by inspection the order of manufacture and
assembly and the procedures and finishes used in the manufacture and assembly
processes ; and

.6 Most important, it is impossible to reproduce any device without making
some unintended or intended modifications or adjustments and as soon as any
such modification or adjustment is made a part may be moved outside the per-
missible tolerance range since there is no way of knowing which adjustments
ai-e incidental and which are fundamental.

15.40.106 Had Honeywell been given access to the shared technology, a sub-
stantial forward momentum would have been delivered to Honeywell at an
earlier time when the leverage was much greater: hence, potential beneficial
effects from such an exchange to IBM and SR were greater in 1956-57 than
they would be today and made it that much more difficult for those who had no
access to the shared technology to compete profitably in the year after the
exchange.



5888

15.40.107 The following is the actual schedule for production and delivery of
the Honeywell H-800, H^OO and H-200 systems :

.1 The iirst H-800 was delivered on December 31, 1960, but production problems
were not resolved until July of 1061 :

.2 The H-800 used IBM card readers and punches, had no drum storage unit^
had no acceptable disc random access storage unit, and had a significantlj^
inferior Honeywell-built printer ;

.3 Tlie tirst H-400 system was delivered in late December, 1961, but production
problems were not resolved until about July, 1062 ;

.4 Tlie H-400 had acceptance problems in the marketplace and suffered fmm
the same peripheral equipment disadvantages as H-8C0 ;

.5 The first H-200 system was delivered in June of 1964 but did not go into
volume production until December of 1964 or January of 1965 :

.6 The H-200 suffered from split system responsil)ility in its early years and
lacked quality peripheral devices until Honeywell was able to deliver its own
reader, its own punch, its improved printer, and its own disc random access
storage unit ; and

.7 Honeywell has never had its own drum storage equipment, did not come out
with a printer which was satisfactory to its customers until February of 1065.
did not produce its own card reader until May of 1965. did not produce its own
card punch until mid-1966, and did not produce its own disc random access
storage unit until 1960.

15.40.108 Had Honeywell had access in 1057 to the technology shared between
IBM and SR. Honeywell could have designed and manufactured a set of its
own peripherals for inclusion with the first H-800 system delivery in 1060 : later
access would have saved mone.v but not alleviated the problem of the 1060s.

15.40.100 That set of Honeywell peripherals would have included a card
reader, a card punch, a drum storage device, a disc random access storage
device, and a high speed printer with print quality and speed equal to that of
SR's high speed printer.

15.40.110 Had Honeywell had access in 1056-1957 to the information exchanaed,
the first H-800 system could have been in volume production by July 1. 1060.

15.40.111 Delivery of the H-SOO with its own set of Honeywell-built peripherals
at an earlier date would have enhanced tlie image of the system in the market-
place and improved the reliability and performance of the complete package.

15.40.112 Had Honeywell had access to the shared technology in 1956-57. first
delivery and volume production of the H-400 would have been stepped up to
July 1. 1961, resulting in a time saving of six mfinths in first delivery and a full
year in volume production over the actual schedule, again with additional com-
petitive advantage of greater dimension.

15.40.113 Assuming Honeywell had had access to the shared technology in
1056-57, the H^OO which had acceptance ])roblems in the marektplace, would
have had its image enhanced and its relialnlity and performance improved because
of the availability of a complete set of high quality Honeywell-built peripherals.

15.40.114 Had Honeywell had access to the shared technology, in 1056-57. the
H-200 would have been first delivered and in volume production by late 1063,
instead of 1064, re.sulting in a time saving of six months in first delivery and a
full year in volume production ; this six month to one year advantage would liave
greatly increased sales of the H-200 .system since SR's co-conspirator IBM had
problems in meeting promised delivery schedules of its own 360 system as late
as 1065 and 1966 which promises were forcedly announced to meet the H-200
system.

15.40.115 The experience of having designed and built a complete set of Honey-
well peripherals on the shared technology base for the H-800 and the H— 100
systems would have meant that Honeywell's H-200 system would have contained
an attractive set of Honeywell-built peripheral devices which would have en-
hanced the H-200's market image for reliability and performance for customers.

15.40.116 As a result, Honeywell would liave become a much more efficient
organization earlier with a totally competitive EDP system product to offer in
the marketplace and thus would have had substantially increased efficiencies,
sales, revenues, and profits throughout the period after access to the shared
technology.



5889

15.40.117 Had Honeywell had access to the shared technology, it would also
have been able to penetrate, user markets earlier from which it was excluded by
its lafk of peripheral e(|uipment capabilities.

15.40.118 These effects are cumulative.

15.40.119 The most important loss to Honeywell was the loss of the opportunity
it could have had to advance its image and customer base position in the field
and its consequent reduction in revenue and profit during the years following the
teclMK'lot'icjil merger of SR and IBM.

15.40.120 The anticompetitive effect of the sharing of technology between SR
and I !).\l \v,;s ro p!;K-e tioiieywcJl (and others) in a •'re-invention" cycle in com-
petition with both SR and IBM; SR and IBM conspiratoi-ially obtained an
advanced and combined base of know-how in 1957 against whicli Honeywell and
the EDP industry had to compete ; it enabled SR and IBM to stay two or three
years ahead of Honeywell and the Industry which were "re-inventing" the merged
technology whereas SR and IBM were able to concentrate on more advanced
products or fresh approaches and to keep tliem in reserve until competition was
about to or did announce or introduce similar items to the market.

15.40.121 This never-ending disadvantage continued with Honeywell and the
industry still having to "re-invent" what the co-conspirators shared in technology
and which provided a base on which they could l)uild.

15.40.122 Finally, the technological merger dampened the need and diminished
the reason for SR and IBM to expend as large sums, relatively speaking, for
current R & D, thus reducing the total penetration of EDP equipment and systems
and keeping total market size smaller and more risky to enter than otherwise
would have been the case.

15.40.123 Honeywell, throughout the period 1957 to 1969, made R&D expendi-
tures for design of EDP equipment, some of which would have been unnecessary
at some point in time had Honeywell had access to the technology shared between
SR and IBM in 1956 or thereafter.

15.40.124 Honeywell EDP Division's financial history for the period 1955 to
1969 demonstrates the heavy expenses and losses which Honeywell incurred in
tryinc to turn the profit corner in the EDP industry dominated by co-conspirators
SR and IBM.

15.40.125 Comparing the relationship between R&D productivity statistics for
Honeywell against those with access to the merged technology produced injury
and probable damage to Honeywell.

15.40.126 Honeywelfs wasted or marginally required EDP R&D expenses
produced injury and probable damage to Honeywell.

15.40.127 Honeywell's lost EDP business due to defendants' pattern of conduct,
including their control of and shared technology, produced injury and probable
damage to Honeywell.

15.40.128 Defendant's pattern of conduct, including the technological merger of
IBM and SR, imposed and perpetuated tlie effects of an entry barrier on the EDP
industry and artificially depressed the size and rate of growth of the industry and
its market, and depressed Honeywell's share therein and produced injuiy and
probalile damage to Honeywell.

15.40.129 Hindsight demonstrates that certain of the injury and indicated dam-
age aforesaid were probably provable with reasonable certainty prior to May 26,
1963: certain thereof were not susceptible of proof at all or to such a certainty
until a later time.

16. SR-IBM Settlement of Interferences

16.1 Plaintiff claims that the settlement of the interferences violated Sections
1 and 2 in that SR and IBM each knew of the invalidity of the BNIAC and SSEC
applications and other patents and applications.

16.2 IBM repeatedly took the position that the ENIAC application was invalid
and did file another petition witli the Patent Office in 1959.

16.2.1 See 15.24.2, 15.24.13, 15.24.21 and 15.24.30.

16.2.2 The following lists the Patent Office interferences which were settled by
IBM and SR pursuant to the 1956 Agreement, the subject, the application num-
ber, the filing date, the applications and the SR "EM" file number for each SR
application involved in the interferences settled :



5890



Application
Interference No. Subject of application No. Date filed SR applicants EM No.

86,576 ENIAC 757,158 June 26, 1946 Eckert, Mauchly EIVl-6.

86,997 Information storage 98,178 June 10, 1949 Eckert. EM-8 Lukoff.

system.

86,998 do 98,178 do do Do.

87,203... .. - do 98,178 do do Do.

87,230 ENIAC 757,158 June 26, 1946 Eckert, Maucfily EM-6.

87,231. . . BINAC 179,782 Aug. 16, 1950 Eckert, et. al EM-22,

87,388 ENIAC 757,158 June 26, 1946 Eckert, Mauchly EM-6.

87,779 ENIAC 757,158 do ...do.. EM-6.

87,872 ENIAC 757,158 do.. do.... EM-6.

87,957 Magnetic reading 253,189 Oct. 25, 1951 Rubens ERA-30.

device.

88,118 ENIAC 757,158 June 26, 1946 Eckert, Mauchly EM-6.

16.2.3 In the August 21, 1956 Agreement, IBM reserved the right to proceed
with a challenge to the ENIAC patent application on the public use ground.

16.2.4 In 1958, IBM became concerned because it feared that, for antitrust
purposes, it might be argued that IBM's 1956 attempt to institute a public use
proceeding against the ENIAC patent application was insuflScient since the peti-
tion had been denied on a procedural matter and not on the merits.

16.2.5 In 1958, IBM was aware that there could be antitrust ramifications if
it did not proceed further and that it could charge in a renewed petition that
Eckert and Mauchly committed a fraud on the Patent Office by not disclosing pub-
lic use information to the Office at the time they executed their oaths in 1947.

16.2.6 IBM, with SR's knowledge, never raised the question of the commission
of a fraud on the Patent Office by Eckert and Mauchly in its petitions for insti-
tution of public use proceedings.

16.2.7 On February 20, 1959, IBM renewed its previously filed petition for the
institution of a public use proceeding against the ENIAC application and did not
bring out all of the public use information it had nor the information concerning
inventive derivations by Mauchly from Atanasoff which was then in IBM's
possession.

16.2.8 Furthermore, both parties knew that the Patent Office had a long estab-
lished policy of refusing to process public use petitions.

16.2.9 IBM had more to lose than to gain by establishing invalidity of the
ENIAC patent application once the 1956 cross-license had been executed because
IBM knew that once it had agreed to pay $10,000,000, its competitors would be
financially burdened by the ENIAC patent application only if it issued ; further-
more, IBM knew that any of the claims of its SSEC application common to
ENIAC claims would also be invalid and other SSEC claims reading on ENIAC
would be similarly invalid.

16.2.10 When IBM renewed its public use petition in 1959, a patent attorney
representing SR wrote a letter to Army Ordnance asking that Ordnance not give
IBM access to evidence which might establish the fact that ENIAC had been
in public use before the critical date ; however, no effort was made by IBM to
subpoena the records.

16.2.11 The foregoing acts and non-acts are consistent with conspiracy.
16.3 In 1960, the Assistant Commissioner declined to order a hearing on the

claim of public use and IBM took no further legal action.

16.3.1 On May 26, 1960, the Commissioner of Patents by Assistant Commissioner
Crocker advised IBM and the ENIAC applicants that the IBM petition to in-
stitute public use proceedings had been denied on the ground the affidavits and
exhibits submitted on behalf of IBM failed to make a prima facie showing that
the subject matter of the ENIAC application was in public use or on sale more
than one year prior to its filing ; thus, the question of public use of the ENIAC
was never considered on the merits by the Patent Office.

16.3.2 After the Patent Office denied IBM's petition for institution of a public
use proceeding in 1959, IBM made an internal report to state that it had pursued
the public use proceeding in good faith even though any direct benefit to IBM
to be obtained from the public use proceeding was eliminated and even though



Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 136 of 140)