United States v. Computer Associates International, Inc. and Legent Corporation, Civ. No. 1:95CV01398 (TPJ) (D. D.C.); Response of the United States to Public Comments Concerning the Proposed Final Judgment

DEPARTMENT OF JUSTICE

Antitrust Division

United States v. Computer Associates International, Inc. and

Legent Corporation, Civ. No. 1:95CV01398 (TPJ) (D. D.C.); Response of

the United States to Public Comments Concerning the Proposed Final

Judgment

Pursuant to section 2(d) of the Antitrust Procedures and Penalties

Act, 15 U.S.C. Sec. 16(d), the United States publishes below the

written comments received on the proposed Final Judgment in United

States v. Computer Associates International, Inc. and Legent

Corporation, Civil Action No. 1:95CV01398 (TPJ), United States District

Court for the District of Columbia, together with its response thereto.

Copies of the written comments and the response are available for

inspection and copying in Suite 200 of the Antitrust Division, United

States Department of Justice, 325 Seventh Street, N.W., Washington,

D.C. 20530 (telephone 202/514-2481) and for inspection at the Office of

the Clerk of the United States District Court for the District of

Columbia, Third Street & Constitution Avenue, NW., Washington, D.C.

20001.

Constance K. Robinson,

Director of Operations.

Response of the United States to Public Comments

Pursuant to the Antitrust Procedures and Penalties Act (``APPA'' or

``TUNNEY Act''), 15 U.S.C. Sec. 16(b)-(h), the United States is filing

this Response to public comments it has received relating to the

proposed Final Judgment in this civil antitrust proceeding. The United

States has carefully reviewed the public comments on the proposed Final

Judgment and continues to believe that entry of the proposed Final

Judgment will be in the public interest. After the comments and this

Response have been published in the Federal Register, under 15 U.S.C.

Sec. 16(d), the United States will move the Court to enter the proposed

Final Judgment.

This action began on July 28, 1995, when the United States filed a

Complaint charging that the acquisition of Legent Corporation

(``Legent'') by Computer Associates International, Inc. (``CA'') would

violate Section 7 of the Clayton Act, 15 U.S.C. Sec. 18. The Complaint

alleges that the acquisition would eliminate significant competition

between CA and Legent in five markets for systems management software

used with mainframe computers that work with the VSE operating system:

VSE tape management software; VSE disk management software; VSE

security software; VSE job scheduling software; and VSE automated

operations software. In addition, the Complaint alleges that the

transaction would substantially lessen competition in the market for

``cross-platform'' systems management software, used in computer

installations where a mainframe computer is linked together with other

types of computer ``platforms'' (such as midrange computers or networks

of workstations or personal computers).

Simultaneously with filing the Complaint, the United States filed a

[[Page 6026]]

proposed Final Judgment and a Stipulation signed by the defendants

consenting to the entry of the proposed Final Judgment, after

compliance with the requirements of the APPA.

Pursuant to the APPA, the United States filed a Competitive Impact

Statement (``CIS'') on August 18, 1995. The defendants filed a

Submission Pursuant to 15 U.S.C. Sec. 16(g) of the APPA, on August 11,

1995. A summary of the terms of the proposed Final Judgment and CIS,

and directions for the submission of written comments relating to the

proposal, were published in The Washington Post for 7 days from

September 3, 1995 through September 9, 1995. The proposed Final

Judgment and CIS were published in the Federal Register on September 8,

1995. 60 Fed. Reg. 46861-46870 (1995). The 60 day period for public

comments began on September 8, 1995 and expired on November 7, 1995.

The United States has received three comments, which are attached as

Exhibits 1-3.

I. Background

The proposed Final Judgment is the culmination of an intensive two-

month investigation of the proposed acquisition of Legent by CA. The

Government interviewed 55 customers and 14 competitors, who would have

been affected by the proposed acquisition in various product lines. In

addition, the Government issued 49 Civil Investigative Demands

(``CIDs'') and reviewed over 950 boxes of documents in connection with

this investigation.

At the conclusion of its investigation, the Government determined

that the proposed acquisition violated the Clayton Act. The Government

challenged the proposed acquisition and negotiated a proposed Final

Judgment with the defendants that adequately resolves its competitive

concerns.

II. The Legal Standard Governing the Court's Public Interest

Determination

When the United States proposes an antitrust consent decree, the

Tunney Act requires the Court to determine whether ``the entry of such

judgment is in the public interest.'' 15 U.S.C. Sec. 16(e) (1988). As

the D.C. Circuit explained, however, the purpose of a Tunney Act

proceeding ``is not to determine whether the resulting array of rights

and liabilities `is one that will best serve society,' but only to

confirm that the resulting settlement is `within the reaches of the

public interest.' '' U.S. v. Microsoft Corp., 56 F.3d 1448, 1460 (D.C.

Cir. 1995) (emphasis in original); accord, United States v. Western

Elec. Co., 993 F.2d 1572, 1576 (D.C. Cir.), cert. denied, 114 S. Ct.

487 (1993); see also United States v. Bechtel, 648 F.2d 660, 666 (9th

Cir.), cert. denied, 454 U.S. 1083 (1981); United States v. Gillette

Co., 406 F. Supp. 713, 716 (D. Mass 1975).\1\ Hence, a court should not

reject a decree ``unless `it has exceptional confidence that adverse

antitrust consequences will result--perhaps akin to the confidence that

would justify a court in overturning the predictive judgments of an

administrative agency.' '' Microsoft, 56 F.3d at 1460 (quoting Western

Elec., 993 F.3d at 1577).

\1\ The Western Elec. decision involved a consensual

modification of an antitrust decree. The Court of Appeals assumed

that the Tunney Act standards were applicable in that context.

---------------------------------------------------------------------------

Tunney Act review is confined to the terms of the proposed decree

and their adequacy as remedies for the violations alleged in the

Complaint. Microsoft, 56 F.3d at 1459. The Tunney Act does not

contemplate evaluating the wisdom or adequacy of the Government

Complaint or considering what relief might be appropriate for

violations that the United States has not alleged. Id. Nor does it

contemplate inquiring into the Government's exercise of prosecutorial

discretion in deciding whether to make certain allegations. To the

extent that comments raise issues not charged in the Complaint, those

comments are irrelevant to the court's review. Id. at 1460. The Court's

inquiry here is whether the relief sought in the markets of concern in

the Complaint has been tailored to maintain the level of competition

that existed in those markets prior to the acquisition.

It is not the function of the Tunney proceeding ``to make [a] de

novo determination of facts and issues'' but rather ``to determine

whether the Government's explanations were reasonable under the

circumstances'' for ``[t]he balancing of competing social and political

interests affected by a proposed antitrust decree must be left, in the

first instance, to the discretion of the Attorney General.'' Western

Elec., 993 F.2d at 1577 (internal quotations omitted). Courts have

consistently refused to consider ``contentions going to the merits of

the underlying claims and defenses.'' Bechtel, 648 F.2d at 666.

In addition, no third party has a right to demand that the

Government's proposed decree be rejected or modified simply because a

different decree would better serve its private interests. For, as this

Circuit has emphasized, unless the ``decree will result in positive

injury to third parties,'' a district court ``should not reject an

otherwise adequate remedy simply because a third party claims it could

be better served.'' Microsoft, 56 F.3d at 1461 n.9.\2\ The United

States--not a third party--represents the public interest in Government

antitrust cases. See e.g., Bechtel, 648 F.2d at 660, 666; United States

v. Associated Milk Products, 534 F.2d 113, 117 (8th Cir.), cert.

denied, 429 U.S. 940 (1976).

\2\ Cf. United States v. Associated Milk Producers, Inc., 534

F.2d 113, 116 n.3 (8th Cir.), cert. denied, 429 U.S. 940 (1976)

(``The cases unanimously hold that a private litigant's desire for

[the] prima facie effect [of a litigated government judgment] is not

an interest entitling a private litigant to intervene in a

government antitrust case.'').

---------------------------------------------------------------------------

III. Entry of the Proposed Final Judgment is in the Public Interest

Entry of the proposed Final Judgment in this case is clearly within

the reaches of the public interest under the standards articulated in

Microsoft and other decided cases. The proposed Final Judgment resolves

the competitive concerns that led to the filing of this case as to each

of the five VSE systems management product markets and the cross-

platform systems management software market identified in the

Complaint.

IV. Response to Public Comments

We received only three comments, one from a customer, one from a

competitor, and one from a former Legent employee.

A. Comment of Pete Clark (Exhibit 1)

Pete Clark, a VSE customer, submitted a comment expressing concerns

as to: (1) Whether certain Legent products apart from the five named in

the proposed Final Judgment (the ``Subject Software Products,'' as

defined in paragraph II.H. of the proposed Final Judgment, hereafter

referred to as the ``subject products'') should also be included within

the scope of relief; (2) the adequacy of CA licensing, rather than

completely divesting, the subject products as an effective remedy to

the competitive harm posed by CA's acquisition of Legent; and (3) the

adequacy of provisions of the proposed Final Judgment aimed at helping

a licensee recruit and hire former Legent personnel responsible for

development of the subject products.

1. Product Coverage

Mr. Clark believes that six additional Legent products should also

be covered by the proposed Final Judgment because of their close

relationship in functionality to two of the subject products--FAQS/PCS,

for VSE automated job scheduling, and FAQS/ASO, for VSE automated

operations. Mr. Clark appears not to regard the six

[[Page 6027]]

additional products as constituting markets of competitive concern

apart from the markets alleged in the Complaint and addressed in the

proposed Final Judgment, in which case, his criticism would not be

cognizable. Microsoft, 56 F.3d at 1459. Rather, he asserts that being

able to market the six products is important to the competitive

viability of the eventual licensee of FAQS/PCS and FAQS/ASO in the

markets for job scheduling software and automated operations software

respectively.

In defining relevant markets and evaluating competitive

capabilities of firms in the markets, the Government considered the

possible effects of CA's acquisition of Legent with reference to many

products and combinations of products marketed by either of the

parties, including Mr. Clark's six candidates for coverage by the

proposed Final Judgment. Our investigation did not, however, support

Mr. Clark's view that a vendor's success or effectiveness in marketing

FAQS/PCS or FAQS/ASO depends on its ability also to market any of the

six additional products.

To whatever extent that it might be useful for users of FAQS/PCS or

FAQS/ASO to also have access to any of Mr. Clark's six products, those

products are likely to continue to be available in the marketplace.

Having acquired Legent, CA now supplies the six products as well as

FAQS/PCS and FAQS/ASO. If Mr. Clark is correct about the existence of

valuable functional inter-relationships among these products, CA should

have the same incentives to continue marketing all of them as Legent

had before CA's acquisition of it, and customers will have the same

access to them.

In addition, a licensee of CA under the proposed Final Judgment

may, to the extent it deems necessary, seek licenses from CA as to any

of the six products. Where appropriate, such additional licenses may be

facilitated by application of paragraph II.H.2. of the proposed Final

Judgment, which defines ``subject software product'' to include ``all

optional modules, add-ons, enhancements and software customization sold

or distributed to customers for use with the Subject Software

Product.''

The overriding objective of the proposed Final Judgment is to

ensure that the contemplated licenses will result in the establishment

of a viable and effective new competitor in the markets where

competition would otherwise be reduced substantially by CA's

acquisition of Legent. Pursuant to paragraphs IV.A.8. and IV.C.2. of

the proposed Final Judgment, the Government has the responsibility to

determine, in its sole discretion, whether this objective is satisfied.

The Government will be monitoring the license negotiation process and

the scope of the proposed licenses carefully in exercising this

responsibility. Moreover, the proposed Final Judgment, at paragraph

IV.C.6., gives the Government the right to seek additional relief

should a Court-appointed trustee's efforts to license the subject

products fail to produce, to the satisfaction of the Government, an

effective new competitor in any of the relevant markets. The Court is

then authorized to enter additional orders ``as it shall deem

appropriate in order to carry out the purpose of the trust * * *.'' Id.

2. Adequacy of Licensing Remedy

Mr. Clark's general assertion that complete and total divestiture

is the only means of effectively addressing the competitive concerns

posed by CA's acquisition of Legent is unfounded. While Mr. Clark notes

specific issues pertinent to the fashioning of appropriate relief in

this case, all of his points had been fully anticipated and considered

by the Government, and all have been addressed in the proposed Final

Judgment with measures aimed at ensuring the establishment of an

effective competitor for each of the subject products.

For example, Mr. Clark correctly points out the importance of

ensuring that any new marketer of the subject products acquires not

merely the right to sell the product but also capabilities to provide

competitive levels of customer support and to engage in sufficient

levels of product research and development necessary for long-term

competitive viability. With respect to these points, various provisions

of the proposed Final Judgment require CA to provide a licensee with

all the software codes, specifications, development tools, and other

information or know-how needed to compete effectively in terms of

product support and development. Paragraph II.H. of the proposed Final

Judgment. In addition, the proposed Final Judgment provides the

licensee with the opportunity and assistance of CA to recruit and hire

former Legent product development and technical support personnel

retained by CA after acquiring Legent. Paragraph IV. B. 4-5. of the

proposed Final Judgment.

In any event, as noted above, paragraph VI.C.6 of the proposed

Final Judgment permits the Government to seek additional relief

consistent with the purpose of the proposed Final Judgment, if that

proves to be necessary. In such case, the Court is authorized to enter

additional orders as appropriate, ``which shall, if necessary, include

disposing of any or all assets of the Subject Software Product

businesses, including Customer contracts and/or software assets * *

*.'' Id.

3. Access to Developers

Mr. Clark raised concerns that provisions of the proposed Final

Judgment requiring CA to assist licensee recruitment of former Legent

personnel are overly restrictive in applying only to individuals whose

job duties related to development or technical support of the subject

products as of the date on which the proposed Final Judgment was filed.

Mr. Clark suggested that prior to filing of the proposed Final Judgment

many Legent employees with relevant product development expertise were

transferred to other assignments to avoid subjecting them to the

provisions of the proposed Final Judgment governing licensee

recruitment.

The proposed Final Judgment, at paragraph VI, prohibits CA from

taking any action that would thwart the disposition of the Subject

Software Products or undermine the Judgment's objectives. Thus, the

proposed Final Judgment already addresses Mr. Clark's concern.

In any event, the Government investigated Mr. Clark's concerns,

particularly in light of his suggestion that the parties may have

engaged in conduct to frustrate a significant term of the proposed

Final Judgment. Our investigation did not, however, substantiate Mr.

Clark's concerns, and we are presently satisfied that expanding the

scope of CA's obligations to assist in licensee recruitment efforts is

not necessary. Moreover, nothing prevents any former Legent employees

interested in working for a licensee--including employees not covered

by the Judgment's recruitment terms--from seeking out the licensee and

pursuing employment discussions without CA's assistance.

B. Comment of Syncsort, Inc. (Exhibit 2)

Syncsort, Inc. (``Syncsort'') submitted a comment expressing

concerns that the proposed Final Judgment does not address a VSE

systems management software product known as sort software, which is

commonly used in connection with two of the subject products, disk and

tape management software. Syncsort markets a sort software product that

it sells in competition with a CA product. Legent does not have a sort

software product, so CA's acquisition of Legent does not reduce current

competitive choices for VSE sort products. However, Legent has

[[Page 6028]]

in the past cooperated with Syncsort by providing it with software

interface information to help Syncsort develop a sort product that

works well with Legent's disk and tape management products.

Syncsort believes that Legent's new owner, CA, being a competitor

in sort software, will not have the incentives that Legent once had to

cooperate with Syncsort; instead, CA may have incentives to try to

disadvantage Syncsort by withholding information on future Legent

interface developments and by making new versions of Legent's disk and

tape management products increasingly less compatible with Syncsort's

sort product. To address these concerns, Syncsort suggests that the

proposed Final Judgment be modified to require CA and its licensee to

maintain the levels of cooperation and interface information sharing

that previously existed between Syncsort and Legent.

The issues raised by Syncsort are adequately addressed by the

proposed Final Judgment. As noted before, the central purpose of the

proposed Final Judgment is to enable another firm to step in Legent's

place as a viable and effective competitor in the markets for the

subject products. The accomplishment of this objective should alleviate

Syncsort's concerns by establishing and maintaining an independent

developer and marketer of tape and disk management software with which

Syncsort could work to develop compatible sort software. There is

little reason to suppose that Legent's competitive replacement would

have any less incentives to cooperate with Syncsort on software

interfaces than Legent had. To the extent that this interface

cooperation confers significant marketplace advantages to the new

supplier of the subject products, competitive pressures may compel CA

itself to engage in such cooperation.

C. Comment of Brian W. Gore (Exhibit 3)

Brian W. Gore, a former employee of Legent, stated concerns similar

to those of Pete Clark relating to the scope of the products that are

the subject of the proposed Final Judgment. Although Mr. Gore

identified different additional products for coverage than those named

by Mr. Clark, his reasons in support of adding the products are similar

to the views expressed by Mr. Clark. For the reasons previously stated

in response to Mr. Clark's comments, the Government does not believe it

appropriate or necessary to provide relief focusing on any of the

products identified by Mr. Gore.

Mr. Gore also raised concerns similar to Mr. Clark's comments with

respect to the primary requirement of the proposed Final Judgment that

CA license with subject products rather than completely divest them.

Again, the Government's previously stated response to Mr. Clark's

comments is equally responsive to Mr. Gore's.

Lastly, Mr. Gore indicated that the proposed Final Judgment does

not contain sufficient provision for actions against CA for violations

of the proposed Final Judgment. Here, Mr. Gore's concerns appear

largely to be based upon CA's terminations, previously brought to the

Government's attention, of several former Legent employees associated

with the subject products. The Government has thoroughly investigated

these terminations and has concluded that they did not pose violations

of any provisions of the proposed Final Judgment.

V. Conclusion

The Court should enter the proposed Final Judgment upon the

Government's compliance with the APPA. The issue in this proceeding is

whether the settlement is ``within the reaches of the public

interest.'' Microsoft, 56 F.2d at 1460. Because the proposed decree is

within the scope of the public interest, the Court should enter it

after the Government's responses to the public comments are published

in the Federal Register and the Government certifies compliance with

the APPA and moves for entry of judgment.

Dated: February 1, 1996.

Respectfully submitted,

John F. Greaney, Weeun Wang, Minaksi Bhatt,

Attorneys, U.S. Department of Justice, Antitrust Division, 555 4th

Street, N.W., Room 9901, Washington, D.C. 20001, Tel: 202/307-6200,

Fax: 202/616-8544.

From: Pete Clark, Technical Support Manager, Olan Mills, Inc., P.O.

Box 23456, Chattanooga, TN 37422

To: Judge Thomas Penfield Jackson, United States District Court for

the District of Columbia, Washington, DC 20549

Weeum Wang, United States Department of Justice, Washington, DC

20549

Paku Kahn, Tennessee State Attorney General's Office, Nashville, TN

Christine Rosso, Illinois State Attorney General's Office, Chicago,

IL 60601

Subject: Case # 1:95CV01398--Computer Associates/Legent Acquisition

The information following is a result of having read the

Department of Justice Complaint, of having been gainfully employed

in the VSE systems software arena for the last 30+ years, of having

been a customer of both Legent and Computer Associates, and of

having been immediately involved with this industry, its vendors,

and its customers since the industry began.

Introduction

While it is somewhat presumptuous of myself to lay claim to

being an expert in the field of VSE system software. It is perhaps

more accurate to indicate that many users, many vendors (including

Computer Associates and Legent) and many trade press persons have

certainly labeled myself as ``the expert in the VSE systems software

arena''.

I certainly have spent the last 30+ years in efforts to become

proficient in the VSE systems software. In my 30+ years of

employment, I have been involved in almost every position in a VSE

data center. Operations, programming, system programming, education,

systems design, system analysis and management are just a few of the

areas. In addition to the preceding areas, I have taught various

VSE-related college level courses, written many articles that have

been published in national and international periodicals, have

conducted many seminars for VSE user groups and VSE software vendors

around the world and have done numerous private software/hardware

consultations for both VSE vendors and users I have throughout the

years written several modifications to the VSE operating system and/

or vendor products that received wide spread adoption among users

and these modifications have historically been incorporated into the

facilities they were written for by the respective vendors.

The purpose of the preceding paragraph is simply to convince the

court that I have sufficient knowledge of the VSE systems area to

make valid, accurate observations that have merit.

I have several concerns with the Department of Justice Final

Judgment, Civil Action Number 95 1398. These concerns all relate to

maintaining a healthy competitive VSE system software market.

Product Issues

The DOJ Final Judgment specifically addresses five products. My

concern is that there are several other products, that inter-relate

closely with the five products, that are not addressed. These

products are FAQS/CALL, PREVAIL/PCS, PREVAIL/XPE, EXPLORE/VSE,

EXPLORE/CICS and EXPLORE/VTAM. These six products are closely

associated with one or more of the five products that are to be

available for licensing.

Excluding these six products from the licensing agreement

significantly devalues the original five products value to a vendor

and to the ultimate customer. Not including these six products in

the licensing program seriously impacts the probability of creating

a successful competitive arena. There are defined interfaces and

functional relationships between the five licensable products and

the six excluded products that are critical to attracting and

maintaining customers.

Separate licensing of the five products without some or all of

the other six products results in a significant function loss for

many of the customers. This loss of function

[[Page 6029]]

dramatically affects the competitiveness of the VSE systems software

market, requiring customers to remain with Computer Associates to

prevent function loss, even if they prefer another product licensee.

To explain: FAQS/ASO and FAQS/PCS are closely allied with

PREVAIL/PCS, FAQS/CALL, and PREVAIL/XPE Manager. WHY? Because all

revolve closely around operator console automation and job

scheduling. Having access to only FAQS/ASO and FAQS/PCS via the

licensed vendor means I cannot institute cross platform scheduling.

I cannot automatically notify persons of problems via computer and

telephone interfaces of issues or problems. I cannot manage my

complete multiple platform systems from a single control station. I

basically have a very one dimensional automation and scheduling

capability. THIS IS NOT ACCEPTABLE IN TODAY'S BUSINESS ENVIRONMENT.

The functions discussed with automation and scheduling are critical

to my business capability and strategy and to many other VSE

customers.

The FAQS/ASO and FAQS/PCS relationship with the EXPLORE group of

products (VSE/VTAM/CICS) are somewhat less dramatic but are

definitely important. With the integrated EXPLORE products I can

gather performance information and monitor critical performance

thresholds and take action automatically via FAQS/PCS and FAQS/ASO

to limit degradation, improve performance and thruput, and enable

automatic notification of problem areas. Again a significant set of

functions that would not be available without a consistent set of

product interfaces, typically via a single vendor.

If licensing is appropriate for the 5 products identified in the

Judgment then it is also especially appropriate for PREVAIL/XPE,

PREVAIL/PCS and FAQS/CALL and definitely warrants serious

consideration for EXPLORE/VSE, EXPLORE/CICS and EXPLORE/VTAM. The

eleven products complete a cohesive functional product suite that

can be truly competitive with Computer Associates existing product

suite.

Having five products from the licensee and the other six

products from Computer Associates presents a daunting challenge. I

have personally had experience in this environment before, trying to

interface Computer Associates products closely with other vendor

products. Because of co-operation issues product problems and

interface errors, after 2 years we closed that project and committed

to not ever utilize that approach again. It simply is not a workable

alternative.

We currently hold permanent licenses for four of the five

licensed products and all six of the additional products mentioned

in this document and in addition six other Legent products that were

purchased by Computer Associates that are not discussed in this

document.

Product Licensing

Is licensing an acceptable way to ensure competitiveness in this

market place?

NO. I do not think so. This is system software, a significant

competitive part of system software is ingenuity, unique solutions,

complementary product interactions, proprietary system interfaces,

product support, product enhancements, developer capability, and a

close vendor/customer working relationship.

Most of these issues are not adequately addressed with this

Judgment and all are very critical to maintaining a competitive

environment. This Judgment does not address these issues in a manner

that ensures and maintains a competitive market place.

This Judgment segregates and separates products preventing

complementary product integration and negatively affecting

competition and customer ability to effectively build a product

suite that utilizes cross product synergy to maximize capabilities.

By instituting licensing rather than divestiture Computer

Associates is the benefactor of having complete and total access to

both their existing product line and complete and total access to

all of Legents product line. A significant advantage Legent had over

Computer Associates in the market place was incorporated into the

software it had developed.

The licensee only has access to the licensed products and is

definitely placed into the market at a distinct disadvantage. As if

startup was not already enough of a challenge the licensee must deal

with a competitor with ``inside product knowledge''. This scenario

ensures that the licensee is NOT competing on equal footing within

the market place.

Complete and total divestiture is the only way to ensure a truly

competitive market

Access to Developers

While the Judgment makes provisions for the licensee to be able

to potentially obtain developers with knowledge of the product set,

it severely restricts who the licensee may considered. Perhaps it

was not known that many of the developers, who had expertise in the

area, were ``transferred'' to other assignments prior to this

Judgment. This had the effect of making them ineligible for

consideration by the licensee and severely limits the talent pool.

Almost without exception the original developer was not associated

with the licensed product on the day of Judgment signing.

This part of the Judgment must be modified to include persons

involved with the product in any substantial way within one year

prior to the initial Legent/Computer Associates acquisition

agreement.

Conclusion

Three modifications must be made to the original Judgment to

make it a viable competitive environment:

1. Add the following products PREVAIL/XPE, PREVAIL/PCS, FAQS/

CALL, EXPLORE/VSE, EXPLORE/CICS and EXPLORE/VTAM into the Judgment.

2. Alter the Judgment to require divestiture instead of

licensing of all 11 products.

3. Alter access to personnel to include anyone who has performed

substantive work on any of the products in the past year, dating from

5/25/95.

Many VSE customers including myself believe that without these

three modifications the Judgment has very little if any chance of

being successful. Who will be impacted if these three issues are not

addressed? Every Legent customer.

State's Attorney Generals

I respectively request that the State's Attorney General's of

states with customers affected by this Judgment intervene to ensure

that a fair, competitive market in VSE system software products is

maintained and that active harm is not done to customers information

systems installations by allowing this acquisition to proceed.

Thanks

Pete Clark,

Technical Support Manager, Olan Mills, Inc.

November 6, 1995.

VIA FEDERAL EXPRESS

John F. Greaney, Esq., Chief, Computers & Finance Section, Antitrust

Division, United States Department of Justice, Suite 9901, 555 4th

Street, N.W., Washington, D.C. 20001

Re: United States v. Computer Associates International, Inc. and

Legent Corporation (95 CV 1398) (United States District Court for

the District of Columbia)

Dear Mr. Greaney: On behalf of our client, Syncsort, Inc.

(``Syncsort'') we submit these comments to bring to your attention

certain facts about competition in the market for VSE sort software

and the impact of the proposed consent decree on that market which

we believe require a minor, but nonetheless important, modification

to the Final Judgment.

Syncsort is a company which, among other things, specializes in

developing sophisticated, high performance sort software for main-

frame computer environments, including the VSE system environment

which is the subject of the proposed decree. A summary of the

technical specifications of Syncsort's current VSE sort product,

SyncSort VSE Release 2.3, is enclosed as Attachment A. Sorting

software permits efficient operation of main-frame computers,

effectively speeding their operation and increasing their practical

capacity through use of sort algorithms in virtual memory.

Competition in price and improvement of sorts benefits VSE computer

users by reducing computer time and enabling them to use their

computer resources with maximum efficiency, reducing overall

computer costs.

Syncsort's sort product must interface with the systems

management software which is the subject of the proposed decree, and

particularly the disk/tape manager programs. In the VSE environment,

this has meant attempting to interface either with the Dynam/D and

Dynamt/T program of defendant Computer Associates International,

Inc. (``CA'') or the EPIC/VSE program of defendant Legent

Corporation (``Legent'').

CA markets its own sort product which competes with Syncsort's

and therefore has an incentive not to cooperate with Syncsort. In

fact, CA's systems management software is structured so that

Syncsort's product does not have ``PreOpen'' access to file

[[Page 6030]]

information although CA's own sort product does have such access.

Legent, on the other hand, does not offer its own sort product, and

Legent has historically cooperated with Syncsort, permitting the

sort to access crucial information through EPIC/VSE before

a file is open.

Without the modification Syncsort proposes, there is a danger

that the acquisition will disadvantage Syncsort--and ultimately VSE

users--despite the best intentions of the proposed Final Judgment.

Under the proposed Judgment, those VSE users who continue using the

Legent products will now be divided among two companies (CA and the

licensee). One of these companies has a history of not affording

competitive third party sort products PreOpen access to file

information through its disk and tape management software; the other

company has no history either way but faces uncertain prospects for

a long-term role in the market. As a step toward maintaining the

status quo, the decree should provide that the EPIC/VSE

PreOpen interface or its equivalent will be maintained--by both CA

and the licensee--for all Legent/VSE products or VSE products

subsequently derived from the Legent products.

Even with this relief, the competitive equation will change

after the acquisition takes place. Another small step is therefore

in order. Since current Legent users can choose to become CA users

(and since some at least will conclude that this is the least risky

choice), CA is likely to have even more users of its software

management programs than in the past. CA will therefore have more

market power and more opportunity than in the past to engage in

strategic behavior to extend that market power into the sort product

market. To deal with this change in market conditions, the decree

should provide explicitly that neither CA nor the licensee will

discriminate among other sort programs (including their own sort

programs) in the interface and interface information made available

for the sort function.

These are relatively minor modifications to the Final Judgment,

entailing no real costs or burdens on the parties. They are

nevertheless of considerable importance for the future. They serve

much the same purpose as, and are even lesser mandatory in nature

than, the provision in the decree requiring CA to assure competitors

potential access to PIPES for cross-platform customers. (Final

Judgment para. VII.) Suggested language to accomplish these purposes

is set forth on the enclosed attachment B.

The need for provisions such as these is well illustrated by

past history. Legent has cooperated with Syncsort in the development

of EPIC/VSE so that file information is exchanged before a

file to be sorted is opened. The information provided includes the

following nine items:

1. file size

2. tape/disk

3. device type

4. blocksize/CIsize

5. concatenated

6. record length

7. record format

8. file type

9. spanned

The PreOpen availability afforded by EPIC/VSE permits

dynamic device switching by the customer--switching between devices

without the computer user having to change programs or its job

control language (``JCL''). PreOpen availability also permits

dynamic reblocking--changing from one blocksize to another without

the computer user having to change programs or JCL. Finally, the

PreOpen interface improves performance of the sort by allowing the

optimal sorting algorithms to be chosen before the file is open. In

short, the current, PreOpen EPIC/VSE interface permits

Syncsort to design, and VSE customers to use, efficient, state of

the art sorts without sacrificing flexibility; reduces the amount of

computer time needed for a particular operation; and provides a high

performance sort option for main frame users in the VSE environment.

Syncsort's history with CA, which markets its own program in

competition with Syncsort's, has been quite different. CA has

arbitrarily refused to provide PreOpen access to Syncsort of the

type afforded by EPIC/VSE--but nevertheless has provided

such access to its own sort product. File information can now be

obtained by Syncsort's program only much later, after the file is

actually opened. This denial of access means that, for many users,

Syncsort is unable to provide dynamic device switching or dynamic

reblocking, providing less flexibility and degrading the sort's

potential utility for the customer. Moreover, without PreOpen

information about file size, record length and the like, the

Syncsort sort may be precluded from choosing the optimal sort

algorithms.

There is no technological, cost or other acceptable reason for

this difference in access. It has been explained to Syncsort as

dictated entirely by CA's perceived competitive advantage. After the

divestiture CA's ability to exploit this unfair competitive

advantage is likely to be greater, not less, than it is today.

According to the complaint, CA already has 96% of the market for one

of the software management products (disk management, para. 19) with

which the sort must interface; if even as few as one quarter of the

Legent customers switch, CA will control nearly 60% of the other

(tape management para. 18). There is no guarantee, absent the

suggested decree modification, that CA will maintain PreOpen

Access--or any access at all--for third party sorts for any of these

users. If, ultimately, the licensee should fail or be unable to

compete effectively with CA, CA could abandon or change the former

Legent products and Syncsort and VSE sort users would have no

protection at all.*

*Syncsort believes the 25% figure for switching customers is

low; if one half the Legent customers switch, CA would have market

shares of approximately 95% and well over 70% and virtually no

market constraints on its behavior.

---------------------------------------------------------------------------

These circumstances mandate that the Judgment be modified so

that whoever inherits a former Legent customer--the licensee or CA--

will continue to maintain PreOpen access in EPIC/VSE. In

addition, protection is required against the type of discrimination

CA has employed in the past to favor its own sort product so that CA

cannot anticompetitively translate any market power gained through

the acquisition into a forclosure of the competition and VSE choices

that now exist in the sort market.

Support for such terms can be found in the proposed Final

Judgment in United States v. AT&T and McCaw Cellular Communications,

Inc., 59 F.R. 44158, August 26, 1994. There, the Department of

Justice recognized that, after its merger with McCaw, AT&T would

possess both the incentive and the ability to discriminate against

additional third parties. 59 F.R. at 44168. As a means of requiring

AT&T ``to continue to deal with its customers on terms in place

prior to the merger [with McCaw], and on terms not less favorable

than those offered to McCaw,'' (59 FR at 44158), that decree

proposes requiring AT&T to provide on-going support for ``locked-

in'' customers and to arrange an alternative source of supply for

certain products if they are discontinued by AT&T. 59 FR at 44164.

Similarly, the Final Judgment here should be modified to require (i)

that CA and the licensee maintain the EPIC/VSE PreOpen

interface, or its equivalent, and (ii) that neither CA, nor the

licensee, will discriminate among other sort programs in the

interface and interface information made available for the sort

function.

Respectfully submitted,

James B. Kobak, Jr.

cc: Richard Rosen, Esq., Arnold & Porter, 555 12th Street N.W.,

Washington, D.C. 20004

Michael Byowitz, Esq., Wachtell, Lipton, Rosen & Katz, 51 W. 52nd

Street, New York, NY 10019

Attachment A

SyncSort VSE

Technical specifications

Release 2.3

Introduction

SyncSort VSE is a high performance sort/merge/copy utility

designed for IBM VS, VSE, VSE/SP, and VSE/ESA operating systems.

SyncSort provides significant savings in program and supervisor CPU

time, elapsed time, and I/O activity.

Performance

In benchmark tests of SyncSort VSE Release 2.3 against SM2

Release 5, SyncSort reduced total CPU time by 25-30%, elapsed time

by 25-30%, and SIOs by 30-40%.

SyncSort achieves superior performance through optimization for

specific computer make and model, proprietary sorting algorithms,

advanced access methods, and Data Space utilization. SyncSort

dynamically responds to system activity such as real and virtual

storage availability, and paging rates to ensure optimum

performance.

In a VSE/ESA environment, SyncSort VSE exploits Data Space

technology with two unique features,``virtual library'' and

``virtual sortwork''. These capabilities maximize the use of high

speed virtual memory, minimizing resource consumption and reducing

elapsed time.

SyncSort VSE's Dynamic Storage Manager ensures that all sorts

attain optimum

[[Page 6031]]

performance by intelligently managing a Data Space so that numberous

concurrent sorts can exploit virtual sortwork.

Sort/Merge/Copy Processing

EBCDIC or user-defined collating sequences.

Up to 64 control fields, with length up to 4092 bytes.

Fields in fixed length records may be located anywhere in the

record.

All standard field formats, including character,

binary, packed decimal, zoned decimal, fixed point, floating point,

and various signed formats.

High performance MERGE combines up to 9 pre-

presequenced data sets into one output dataset sequenced identically

to the input datasets.

High performance copy function (SORT FIELDS=COPY) can

be used alone or with data editing.

Input/Output

SyncSort supports:

SAM, VSAM, and VSAM-managed SAM formats and devices,

including devices connected via the ESCON architecture.

Fixed-length and variable-length records.

Processing of variable-length records shorter than

control field.

Intermediate Files

Disk.

Automatic secondary sortwork allocation with up to 31

extents.

Automatic space release for DASD output files via disk

space manager.

Resource Management Features

Dynamic Storage Manager. Automatically monitors and

controls memory utilization, and reduces or eliminates physical

sortwork I/O for concurrent sorts. Optimizes the use of a Data Space

by allowing up to 15 concurrent sorts running in different

partitions to use the virtual sortwork area. Maximizes sort

performance while optimizing overall system throughput.

Disk Space Manager Interface. Minimizes DASD resources

used for sorting while preventing ``sortwork capacity exceeded''

abends. Compatible with all disk space managers.

Attachment B

Computer Associates and any licensee or successor in interest to

Legent's interest in the Subject Software Programs (``Legent's

Successor'') shall each maintain and provide, from and after the

effective date of this Final Judgment, at least the same degree of

PreOpen Access to file information through EPIC/VSE (including

without limitation any successor to or substitute for EPIC/VSE, any

upgraded or modified version of EPIC/VSE or any program derived from

the EPIC/VSE program) as that made available to sort programs

through Legent's EPIC/VSE program prior to the acquisition of Legent

by Computer Associates. In addition, and without limiting Computer

Associate's or Legent's Successor's obligations with respect to the

foregoing sentence, neither Computer Associates nor Legent's

Successor shall, from and after the effective date of this Decree,

discriminate among sort programs, including any sort program of its

own, concerning (i) the timing and manner of access to any disk or

tape manager or similar program made available to VSE customers and

(ii) provision of relevent information.

November 7, 1995

U.S. Department of Justice, Antitrust Division, 555 4th Street,

N.W., Room 9903 JCB, Washington, D.C. 20001

Re: Civil Action No. 95 1398; U.S.A. v. Computer Associates, Int'l.

and Legent Corp.

Gentlemen: This a comment concerning the Proposed Final

Judgement for the aforementioned case. As a 20-year veteran of (IBM

mainframe computer) VSE operating system software operations and

support, I find the Proposed Final Judgement to be deficient in the

following four areas:

1. No provisions for other Legent VSE products also using G.S.S.

common code.

Explanation: G.S.S. is a proprietary integrated on-line

transaction processor subsystem used by all (or at least most)

Legent VSE products that contain an on-line component. While some of

those products such as FAQS/ASO, FAQS/PCS and EPIC/VSE are covered

by the Proposed Final Judgement, others such as Mastercat, SAR-

Express/Delivery, FLEE, etc.) are not. This poses a serious dilemma

for any Legent customers running VSE products in both of the

aforementioned categories.

Because while it has already been ascertained from discussions

with D.O.J. lawyers assigned to this case that the G.S.S. code would

be included with any license agreement, there is no requirement that

Computer Associates and the licensee keep their respective copies of

G.S.S. compatible once a licensee has been assigned. Indeed, such a

requirement would not be practical, and at some point (most likely

soon) in the future, the Computer Associates and the licensee's

versions of G.S.S. would become incompatible, requiring any customer

running G.S.S.-based VSE products from both companies to run

separate copies of G.S.S.

This type of arrangement would not be acceptable to most

customers since it needlessly complicates installation, maintenance

and usage of the VSE products, reduces integration and is fraught

with operational problems since G.S.S. was never designed to be used

in such a fashion. Thus all customers with G.S.S.-based VSE products

that are not covered by the Proposed Final Judgement and remain only

available from Computer Associates would be forced to get their

G.S.S.-based VSE products that are covered by the Proposed Final

Judgement from Computer Associates as well to avoid the

complications of incompatible versions of G.S.S. This situation ends

up creating a ``restraint of competition'' condition that would

promulgate the Computer Associates monopoly in VSE products that the

Proposed Final Judgement was originally designed to prevent (or at

least reduce).

(I estimate this situation involves a substantial portion of the VSE

product customer base, possibly even a majority.)

2. No provisions for other Legent VSE products also using the

EPIC DSN catalog.

Explanation: The EPIC DSN catalog is a proprietary database file

used by EPIC-based products on various mainframe platforms to

accomplish disk and tape file management across those platforms. In

this case, while the EPIC/VSE product is covered by the Proposed

Final Judgement, other EPIC-base products, namely EPIC/CMS for the

VM operating system, is not. This poses a serious dilemma for any

Legent customers running EPIC-based products in both aforementioned

categories, (or in this case, platforms).

The arguments for this point are essentially the same as those

outlined in #1 above; however, this case concerns a database file

shared across operating system platforms (VSE and VM) instead of a

subsystem shared within the same operating system (VSE). The end

result however, is the same: restraint of competition. Since there

is no provision in the Proposed Final Judgement to keep the database

file shared by these 2 products compatible nor any mention of the

EPIC/CMS product (meaning that it would not be available from the

licensee), those customers running both the EPIC/VSE and EPIC/CMS

would effectively be forced to obtain them both from Computer

Associates.

(I estimate that this situation affects about 10-20% of the EPIC/VSE

customer base.)

3. No specific provisions for action(s) against Computer

Associates when conditions of the Proposed Final Judgement are

violated.

Explanation: It appears to most of us in the VSE community that

Computer Associate's intent IS to create a monopoly in the VSE

systems software market, and they are quite ruthless and devious

about it. They have already directly violated certain provisions of

the Proposed Final Judgement, and also seem to be deliberately

delaying its execution. Specific retribution for willful disregard

of the provisions of the Proposed Final Judgement need to be clearly

defined and carried out.

For example, under section ``VI. PRESERVATION OF ASSETS'',

Computer Associates is ordered to ``* * * continue to commit

resources, development and support to each Subject Software Product

at a level not materially less than that committed prior to the

announcement of the subject acquisition * * *''. However within 2

weeks after the Proposed Final Judgement was issued, in just the

EPIC/VSE group alone, 8 out of 20 employees were let go, including

developers and technical support personnel. The D.O.J. was notified

immediately, yet to date, nothing known has been done.

More recently, technical support was moved to a different office

to be handled by inexperienced personnel, and EPIC/VSE developers

have been assigned to other products. Computer Associates is

definitely not pursuing a ``hands-off'' approach to the subject

products while the terms of the Proposed Final Judgement are being

carried out, but rather one that appears to be deliberately

sabotaging them.

4. Non-exclusivity of the license proposal.

Explanation: In the VSE tape and disk management arena alone,

Computer Associates started with a product it developed, called

Dynam/T/D/FI. Then it brought up all the other major players: Epat,

System/Manager, and IPIC/VSE, creating a complete monopoly. It

appears that the D.O.J.

[[Page 6032]]

compromised with Computer Associate's lawyers in coming up with the

non-exclusive license idea.

Who ever heard of 2 companies marketing the same product(s) to

foster competition? Do Ford and GM market any of the same products?

No, they market different products. If Computer Associates could be

equated to General Motors, it would already own Ford and all the

Japanese and European automobile manufacturers; and Legent would be

Chrysler. Then the D.O.J. Proposed Final Judgement would be

equivalent to an order requiring GM to jointly market Jeeps with

Hyundai, while maintaining ownership of the engine and vehicle

assembly plants. It's ludicrous, and simply won't work in the real

world.

In conclusion, the only workable solution I see is to require

Computer Associates to divest, i.e. completely sell-off and cease

marketing, all Legent products that are in any way integrated with

the five already covered by the Proposed Final Judgement. And this

must be done quickly, before Legent's entire VSE product line and

customer base are destroyed. And finally, Computer Associates should

be severely fined for all present violations of the Proposed Final

Judgement and forced in complete compliance ASAP.

One final note: although I am a former Legent employee, I am not

``disgruntled''. I worked in the VSE community long before I worked

for Legent, and still desire to see it prosper. A Computer

Associate's monopoly on VSE systems software is in no one's best

interest except theirs. I urge the court to modify the Proposed

Final Judgement to prevent such an occurrence at ALL levels.

Sincerely,

Brian W. Gore,

101 Mira Mesa, Rancho Santa Margarita, CA 92688.

Certificate of Service

The undersigned certifies that he is a paralegal employed by the

Antitrust Division of the United States Department of Justice, and

is a person of such age and discretion to be competent to serve

papers. The undersigned further certifies that on February 1, 1996,

he caused true copies of the Response of the United States to Public

Comments, and this Certificate of Service, to be served upon the

person at the place and address stated below:

Counsel for Computer Associates

Richard L. Rosen, Esq., Arnold & Porter, 555 12th Street, NW.,

Washington, D.C. 20004 (by hand delivery)

Dated: February 1, 1996.

Joshua Holian,

Paralegal, U.S. Department of Justice, Antitrust Division, Computers &

Finance Section, 555 4th Street, NW., Room 9901, Washington, D.C.

20001, (202) 307-6200.

[FR Doc. 96-3393 FIled 2-14-96; 8:45 am]

BILLING CODE 4410-01-M