WUJUVlDW D11 15 l JWVU DI $W f U S S OO S jJI1 S f OO jJ l1 i WI1 lBrn 'ilrnl1rn JUN-AUG 1985 86-36 NSA'S INITIATIVE ON SECURE VOICE U o oooooooooo 1 TELEPHONE SECURITY 1918 U ooooooooooooooooooooooooo oooo oo o o oo ooo 4 BULLETIN BOARD U ooo ooooo o ooooo ooooooooooooooo oooooo oo oooooooooooo oo 4 SHOPWORK IV U oooooooooo oooo ooooooo oooo 5 AN APPLICATION OF PINSETTER CU 7 VALEDICTORY OF A TRAFFIC ANALYST U oooooooooooo Joseph Starr ooo oooooooooooo 9 BACK UP YOUR DATA FILES U ooooooooooo o oooooo Norman P Smith oooooo o 10 ESCHEW OBFUSCATORY SCRIVENERY PLEASE U ooooooo1 11 NSA-CROSTIC NO 62 CU ooooooooooooooooooo o ooo D H W o ooooooo oooooo oo oo 12 '1 I I I 'fIllS BOetfMI3N eOPl AINS eOBIWIORB MllTERUI GbASSIFIEB BY NSA eSffif 123 2 eclassified and A roved for Release b NSA on 10-16-2012 ursuant to E o 13526 MDR Case # 54 77 aCID 4019704 P L 86-36 Published by Pl Techniques and Standards EDITORIAL VOL XII Nos 6-8 June-August 1985 PUBLISHER 'nalysts in NSA have an opportunity to share their views and xperiences with others by writing for one of the several fperiodicals published in the Agency LI j One of them is CRYPTOLOG BOARD OF EDITORS Editor ---JK963-YfQs LI ' ' L I Collection Computer Security Computer Systems Cryptanalysis Cryptolinguistics Index Information Science Intelligence'Research anguage f k963i 7E 90$' $14t J 9 alno 3 J 9 --4740 96 41 4 1 I 9'$3-5330 963-1145 963-3095 For assistance advice or just plain aid and comfort you may call on any of the editors it need not be the one on your subject 1 As for anonymity we follow the standard etiquette of most publications ' We will honor the author's request to remain anonymous we will even provide a nom de guerre for those who might want one but the identity of the author must be made known to the Editor Anonymous contributions that is lacking identification of the author may as well be written in invisible ink we throw them out unread I I Another point of etiquette that authors must observe is about multiple submissions If you send your article to more than one publication you must so state I 1 1 963-5151 Linguistics 1 96 1 3896 Mathematics 1 963-5655 David H Williams 91iB-11 03 Puzzles 963-4423 Science and Technology Special Research Vera R Filby 968-8014 Traffic Analysis Robert J Hanyok 963-3888 Illustrator Distribution It is specifically intended for informal exchanges among analysts on subjects of interest to them The editorial blue pencil is lightly applied and only in the inte est of clarity it's up to you to sustain the readers' interest and toputyour idea acrosspersuasively 1 963-3057 1 963-3369 Now for a few tips to ensure early consideration The subject should be related in some way to our business Unclassified term papers seldom fill the bill To submit articles or letters by mail send to Editor CRYPTOLOG Pl If you used a word processor please include the mag card floppy or diskette along with your hard copy with a notation as to what equipment operating system and software you used 'via PLATFORM mail send to cryptolg at bar 1cOS bar-one-c-zero- fi ve note no '0' in 'log' Always include your full name organization and se cure phone number For Subscriptions or Change of Add ess send name and organization to I t14 P L 86-36 Contents of CRYPTOLOG should not be reproduced or further disseminated outside the National Security Agency without the permission of the Publisher Inquiries regarding reproduction and dissemination should be directed to the Editor The article should be shorter rather than longer readers' survey showed that long articles are not read A The text and illustrations should be of reproducible quality Every paragraph and illustration should be appropriately classified Hard copies of articles prepared on word processors should be accompanied by the mag card or floppy We send it to the Data Conversion Center where it is transformed to an 860 floppy and then we convert that to an 8010 floppy Be sure to state what equipment and what software you used and the operating system in the case of a pc If you submit a contribution via electrOnic mail notify us by s ome other medium that you are doing so There's slippage every now and then If you're going off on a field assignment or extended training or if you are retiring designate someone who can act as your literary executor Label every submission with your full name organization and secure phone fOR OFFICIAL SE 8l4LY v 4019704 NSA'S INITIATIVE ON SECURE VOICE U P L 86-36 _____t U The Soviet Union and other hostile governments can intercept and exploit with ease and impunity many U S telecommunications in the continental United States and around the world Classified and sensitive communications over unsecured telephone systems alone represent such a lucrative and easy source of intelligence that NSA has undertaken a bold new approach to counter this threat U Each of NSA's past efforts to develop and distribute COMSEC equipments to the field has taken many years to complete To reduce this time significantly NSA decided to tap the knowledge and resources of major commercial manufacturers of telecommunications equipment in a program called The Future Secure Voice System FSVS whereby NSA provides the cryptologic expertise and contractors provide technical manufacturing and marketing know-how The goal is to design and produce a secure telephone the size of a conventional office deskset which will plug into a modular jack cost less than $2 000 and be simple to operate o 1 4 c EXTENT OF THE THREAT P L 1 89 _ 1 nsecured 145 which directs that all U S Government systems carrying classified information be secured including telephone U NSDD-145 requires that NSA as the new National Manager for Telecommunications Systems Security provide a secure communications capability for as many as 500 000 users beginning in 1987 The FSVS program has been conceived to help implement this directive can conversatIons telephone be easily exploited by hostile intelligence-gathering activities to reveal governmental military and industrial secrets and technology That is because communications over commercial telephone systems especially long distance are likely to be transmitted via microwave or satellite which are particularly easy to intercept all it takes is equipment costing about $25 000 small enough to fill the back of a station wagon I PRESENT SECURE TELEPHONE SERVICE The I J L- f t Moreover mission-essential classified information is sometimes discussed over unsecure telephones ' simply because secure communications are unavailable Unclassifi ed but sensitive technological data also pass over the telephone system daily As a recent National COMSEC Committee Biennial Report points out the industrial base which produces the nation's most advanced military hardware and develops it most sensitive technol gy is almost o mplete y Without a secure communlcatlOn s capabillty CritIcal secrets are lost before the senSItive systems and technologies they represent can be fielded by military or other government users U The unsecured telephone has been determined to be the most lucrative source of classified and sensitive information available to unauthorized recipients To counter this vulnerability President Reagan recently signed National Security Decision Directive NSDD Jun-Aug 1985 only DoD-wide secure telephone system available today is AUTOSEVOCOM which serves about 2 500 locations around the world This system however secures only a small fraction of the total telephone conversations of the DoD and much of the equipment is outdated and provides poor voice quality Under the recently implemented AUTOSEVOCOM Life Cycle Extension Program ALCEP older HY-2 KG-13 equipment is being replaced by newer VINSON devices which will update the system and improve voice quality The total AUTOSEVOCOM community will however remain small SA's Secure Telephone Units STU-II and STUlIM are also being procured by military departments and the civil sector Over the next five years the total number of STU-II and STU-lIM units should reach approximately 14 000 units which is only a small portion of the half-million or so of the secure telephones needed Because of their relatively high cost $12-25 000 and their dependence on the Bellfield Key Distribution Centers KDCs each of which can service only a limited nU J1ber Qf subscribers these units cannot be distributed to the field in large CRYPTOLOG page 1 HANQI E vn COWIWP ClIllUllEn ONLi SfCeltE ' aCID 4019704 quantities KDCs act as centralized points for electronically distributing key a function necessary in communication security strategy is necessary to achieve NSA's goal of initial delivery in early 1987 COMSEC DOCTRINE STU-III FAMILY OF TERMINALS U With a half million or more potential customers to satisfy NSA realized there will be a variety of user requirements A family of telephone terminals rather than a single type will therefore be developed to satisfy the diversity This family of equipments has been designated STU-III or Low Cost Terminal LCT U Military planners have stated a requirement for the STU-III to interoperate directly with existing STUn unitso This is being satisfied with the STU-III Command Control C2 version designated KY-77 The C2 terminal will meet environmental requirements for mobile radio telephones and is planned for limited military and civilian government applications It will be slightly more expensive than the other members of the STU-III family and is being developed by RCA on a separate contract The KY-77 should be available early in 1987 U Of the non-military users of STU-III some handle classified information and others handle unclassified but national security-related information Classified user requirements will be satisfied by a Type I LCT while unclassified users will be limited to Type II terminals An example of a Type I application is the securing of conversations between NSA and contractors working on classified NSA projects Typical Type II applications include the protection of proprietary information such as financial organizational marketing and technical data U Another version of the LCT being conceived is a cellular radio telephone model This unit compatible with other LCTs a nd regular unsecured telephones will satisfy secure mobile requirements U Members of the STU-III family will look and act much like conventional telephonedesksets In fact an LCT will interoperate easily with a regular telephone for unsecured communications Secure operations between STU-Ills will require only a few extra steps by the users The terminals themselves will include modern telephone features like repertory dialing and will offer a light-emitting diode LED or liquid c ystal display LCD The display will indicate the identity and security clearance level of the distant end terminal and will help guide users through STU-III operating procedures PROGRAM STRATEGY U In the Spring of 1984 NSA asked five major telecommunications contractors AT T ITT GTE RCA and Motorola to define a concept fa the STUIII LCT and the attendant system necessary to key and support it This concept definition phase lasted six months U NSA's program strategy was to select three of the vendors to p oduce telephones and one contractor to develpp a system responsible for such things as key management The selection process is now complete and the system contract has been awarded to GTE while AT T RCA and Motorola have been contracted to produce the phones This parallel acquisition Jun-Aug 1985 HANDLE 'IA U Physical handling requirements for such a large quantity of COMSEC devices are a major concern COMSEC doctrine is being rethought in order to lessen constraints and make COMSEC more desirable to the user community Older security doctrine dictated the handling of cryptographic equipment according to equipment type STU-III equipment however will be governed by new doctrine which bases the amount of required protection on the specific application of the device For instance a STU-III located within the NSA compound will require less stringent' control than a unit housed overseas where foreign nationals may have easy access to it U In December 1984 NSA commissioned A D Little Company to complete a market survey to determine the size of the potential market for STU-Ills and to establish a priority list of user preferences and demands According to the company's preliminary report the number one requirement expressed by potential users in both Govern ment and industry is that the units be unclassified when unattended This condition presents a problem however since these terminals are to be high-grade cryptographic devices which normally require special handling A device called a crypto-ignition key CIK will therefore be implemented to help solve that problem U The security of U S COMSEC equipment depends on protecting the key which is instrumental in determining the pattern of ones and zeros called the key stream produced by a key generator or cryptologic The CIK is a removable device small enough to fit in a shirt pocket that contains a portion of the key the other portion resides in the STU -1II A keyed STU-III is therefore sanitized by unplugging and removing the CIK and the sanitized phone becomes unclassified and can be left unattended Controls on the CIK will depend on specific applications but in some cases simply locking the device in a desk drawer or taking it home will be sufficient mustBecause STU-Ills will be widely distributed it be assumed that U S adversaries will obtain some of the units and attempt reverse engineering in order to recover the cryptologic To counter this potential threat NSA has decided on a special chip coating as a way of preventing the reverse engineering of cryptographic chips within the terminals NSA engineers are confident that this protection will effectively thwart the recovery of information from the circuitry or memory of the chip sets KEYING SCHEME U A major effort is required to provide keying material to COMSEC equipment in the field n most cases NSA produces the key in some hard copy form like key tape key card or key list and then distributes the key among many users Some newer equipments like the STU-II employ remote keying which reduces NSA's burden significantly These remote keying scheme s however generally depend on a KDC or other central point to distribute key electronically to remote users There are however physical limitations on the number of users the KDCs CRYPTOLOG page 2 Cg 4IN' e iANNELs ONLY - SH RET - ero 4019704 can support Therefore if a half million or more LCTs must be supported a new keying method is necessary e new method is calledFIREFL Y II This is a breakthrough in key distribution technology in that it will eliminate tpe need for a KDC by allowing each STU-III to generate and securely exchange key with other STU-Ills The calling and receiving terminals themselves generate random components which are then securely exchanged combined and used by both terminals on Ii session or per-call basis Session keys enhance the security of the system by eliminating back-traffic vulnerability and the entire process of generating and exchanging random components takes less than 12 seconds allow the vendors to have a direct effect on the size of their market Clearly LCT vendors could benefit dramatically from the production mar eting and sale of secure phones Even if NSA's initial estimate of' 500 000 units is inaccurate by 50% it still means a market of 250 000 phones priced at around $2 000 each for an estimated total market of $500 million over the next five -years If the private sector market pans out and over a million units are sold LCT vendors stand to profit handsomely U Parallel production by three vendors should increase the number of units produced which will in turn bring down unit cost The benefits of low cost and high quantity to NSA's COMSEC effort is that many telephone conversations containing classified or sensitive data will be secured and unaccessible to' Soviet and other hostile intelligence activities r v dded benefits of FIREFLY II include positive th tication of distant end terminals and very long cryptoperic ds A cryptoperiod is the length of time a particular key may be used before it is replaced Although each STU-III generates and exchanges key each time it is used secure y there are also keys which reside within each terminal which must be replaced periodically everyone to three years This replacement will be as simple as placing a call to a key management center and receiving new key securely over the telephone line SUMMARY U The huge amount of classified and sensitive information the U S is giving away through unsecilred telephone communications must be stopped and stopped soon Older methods of developing and implementing COMSEC are too time-consuming to be effective quickly enough NSA's FSVS program has therefore been established as a new approach to meeting this COMSEC challenge APPLICATIONS OF THE STU-Ills NSA'S Deputy Director for Telecommunications has expressed support for the eventual replacement of some NSA black telephones with STU-lIIs The FSVS ' initiative in fact seeks to install 500 000 secure phones in the Department of Defense Civil Government sectors and defense contractors That number grows considerably when other potential LCT customers are included Financial institutions academics doing Government research and any company performing high-tech research and development that could be transferred to an enemy's military or security systems must be considered prime LCT candidates U Only a preliminary report of A D Little's market survey has been completed to date but NSA believes that private sector interest in an inexpensive secure telephone coulej eventually increase the market 1'0 1 1 million Independent market surveys performed by some of the bidding contractors indicate an even larger potential market An RCA-financed study established the market size as over 1 6 million while other contractor studies indicate a potential private sector market demand by 1990 as high as 2 6 million The higher numbers reflect the private sector's interest in acquiring secure telephones to protect against loss of proprietary information U The STU-IH family of secure tel phone equipment is now in contract with initial distribution to the field planned for April 1987 It will provide advantages in size acquisition and life cycle cost security 'and performance over current secure voice equipme'nt and meet a broad variety of needs ' U Various STU-Ill models will provide compatibility with conventional office telephone systems military strategic command and control requirements and conventional and cellular mobile radio telephone systems Most significant is the very scale of the program with quantities ultimately expected to exceed half a million units U All aspects of NSA's Future Secure Voice initiative add up to an strong program to protect tne voice communications of the U S If this effort is successful and the contractors and NsA believe it will be we can truly button up U S voice communications by the end of this decade and at the same time deal the Soviets a severe blow 0 U NSA will permit direct controlled sales of LCTs by the contractors to authorized purchasers This will 17 Jun-Aug 1985 HANDLE CRYPTOLOG page 3 im0liELS ONLY - ero 4019704 DO YOU WRITE AND EDIT IN YOUR JOB TELEPHQNESECURITY 1918 Ex tr a t from Confidential ordi rstolst Lt Chaplain John McOowe l Alexander Lacy October 12 1 t 18 1 laving reported at this office this date or transportation in compliance with order Headquarters Port of Embarkation Hoboken N J or from Navy Department you will report at Officer's Gangway YesSel No 56 Pier N9 59 o North River Foot of West 18th Street New York City at o4p uOct16 1918 3 You are tlirect ed not to visit pier to whichy ou have been assigned before date of embarkation except for the purpose of delivering your baggage You will then go onlyto'the Baggage Room and under no circurrtstances wan jer about piers or go near vessel 8 No information EO 1 4 c P L 86-36 GONFl9ENTIAL will begiven by telephone U If youdOasignifiGantamount of writing and or editing in your job and do not holdJl C SC p s a o editor writer please make yourself knownt0L-j c JT54 SAB 2 Door 3 972-2355s CALL FOR CA TERMS i'et TllE SIGINT Terminology Group is compiling an on-line database of cryptanalysis terms and definitions Initial inpllt has been drawn from the following sources MilitaryCryptanalytics II LambrosDo Callimahos October 1959 Basic Cryptoiogic CloSe sary PI 1971' NSA TechnicaL Journal' CRYPTOLOG Suggestions for additional SOUrces are solicited Please forward your submissions t oc J P13D 968-8162s Feee The database now contains about 1600 terms and definitions When the total reaches 2 000-2 500 it will be reviewed by a panel of experts appointed by the chiefs of key components 9 THIS INFORMATION ISSTRICTLY CONFIDENTIAL covrtesy SUGGEST-A-BOOK ofl 15052 grandson of Chaplain Lacy BULLETIN BOARD FOR DATA BASE ADMINISTRATORS fOIJO The Computer Record Format CRF form H3173C has been revised It is noW on the new standard paper size 8hll to be consistent with the Computer Services Work Request form H3173 which it accompanies The CRF is a vehicle for documenting files it is the official repository for information about the individual file s and record formats subordinated to a particular file or data base which are going to reside on T-controlled re sources such as CARILLON CARONA HOLDER METEOR PLATFORM WINDMILL U The NSA Library is soliciting suggestions for books and periodicals to be purchased for the Library For each book provide the following information and send it to T51 Author Title ISBN Price Publisher Source of Information and your name organization and telephone number It would be helpful if you also include a review of the book or an ad for it GLOSSAR Y OF MACHINE TERMS peuej Contributions are solicited for the Supplementary Working Aid for Machine Terminology SWAMT for the updated edition planned to be published early in 1986 the first since 1981 SWAPT is an informal glossary rather than a standard one so the less formal style allows for the frequent updates which such a rapidly changing technology demands The' next edition will be printed in both upper and lower case as it is now being processed on an IBM PC to be printed on a laser printer Contributions suggestions and comments FOYO may be sent to jP13D FAN X III or you may call her on 968-8162s FOR PINSETTER USERS ON THE XT OR ASTW FOIJO A new PINSETTER applications manual How FOYO The National Data Standards Center PJ3D tll get Started with the fBM PCiXT or ASTW is now recommends that data base administrators use this available at the PCIC and P14 It provides helpful form to document all files created under their cog- hints to new users of the PC XT or ASTW whether nizance including those entered experimentally or experienced or not in using the TSS UNIX version of provisionally on micros or terminal subsystems PINSETTER Then when the programmer is sud-denly transferred you'll have some idea about what those strange files 1 YOUP For further jnformation 9 r a Sjst n Dce call are The failure to document small private files J which go on to become big shared files is becoming a serious problem in the Agency U The new forms should be available in Supply by the time you read this Jun-Aug 1985 o CRYPTOLOG o page 4 GONFIBENTIAL J aCID TOP SRCIU T YI 1 IHL 4019704 SHOPWORK IV U P L P L 86-36 EO 1 4 o c 86-36 - Ag 1985 CRYPTOLOG page 5 JU TOP i iIoWIoU T UMBRA 4019704 'fOP SECnC'p UMBRA Jun-Aug 1985 o CRYPTOLOG o page 6 'POP SSCIU 'F Ul IBIIA P L 86-36 1 4 c EO 4019704 AN- APPLICATION OF PINSETTER U P L I o I - ' - 1 II II I '_ _ _1 i - _ - - f ' II I - I----- l I 86-36 I' - i Jun-Aug 1985 o CRYPTOLOG o page 7 SECRET SPOKE P L 86-36 1 4 c EO ----- DOCID 4019704 tWfU8'P SPOICE Jun-Aug 1985 o CRYPTOLOG o page 8 SSCIUH ePOKl' P L 86-36 EO 1 4 c SEClU' T 401V atEDICTORY OF A TRAFFIC ANALYST U '--- U On August 1 1985 I retired from the National Security Agency after almost 40 years of cryptologic work which included at one time or another most of the things NSA does in the fields of collection analysis and SIGINT reporting I enjoyed every minute of it I honestly don't believe I could have chosen a more enjoyable or rewarding career If I had been born rich IwouId have done traffic analysis for nothing if I had been independently wealthy I would have paid the Agency to let me do it Hang with me for a few minutes of serious discussion about machine processing problems and I'll share with you a few observations comments and recommendations that may contribute to your enjoyment of the profession and perhaps help you to avoid some of the pitfalls that beset the paths I wandered in U The most important question confronting the profession today is why from a traffic analyst's viewpoint data processing systems have not yet produced the Utopia we were led or stampeded to expect back in the early 1960's et work with This is a fairly easy exercise intellectually but it is much more difficult to structure your thought processes in order to design a machine processing system that will give you essentially the same versatility Unfortunately we have some managers and analysts who have produced processing systems and tried to make the target fit the processing scheme and I include in this category most of thOSe procedures which were designed to replace a hand log kept by the analysts with a similar log prepared by the machine The results were predictable What we have now are logs prepared by machine that in many instances are inferior to those the analyst prepared by hand And without the benefit that accrued from working the traffic by hand today's analyst has even less understanding of the data and may in fact be at a total loss to explain observed anomalies We can reverse this trend We need to build on the good decisions we have made with regard to automating data processing for traffic analysis and we must be willing to discard the bad decisions What 1m In the January-March 1980 CRYPTOLOG r-I we really need to do now is to get on with developing orrectly sorts working analysts il1t an expert system that will process automatically all incoming traffic compar'e what is observed to a categorIes loggers case analysts and reSearch dictionary of stated norms and send appropriate alert analysts He then predicts that automate d' processing m s s to nits nd re orters on trouble s ot techniques will permit us to eliminate the loggers and assesses that the research analyst is tJ 1e one with the bright future because he is concerne 1 with the why of analysis driven by his desire to eXplain But every analyst should be saying why and George's prediction that the logger will become obsolete just hasn't happened although it should have Indeed we now have a new generation of loggers who have even less understanding of traffic analysis than those who preceded them ' I D After George disc usses how analysts are going to have to adapt in order to be productive in a modern machine oriented world he puts his finger squarely on the major prOblems area manageable machine systems that wiH function as designed And he goes on to point out our track record for the development of such systems is not impressive U In the December 1981 CRYPTOLOGI I I Ipublished an article on ma9hine processing that everyone who claims to be a traffic analyst should have read Part of what Dale addressed was the problem with keeping-competent programmers on the job and how the traffic analyst could be freed from total dependence on the programmer U Traffic analysis is basically a very simple art if you are able to keep your mind loose so that you can recognize and exploit whatever the target gives you to P L 86-36 U None of this should be difficult Buildingand testing dictionaries will be tedious but in the process of c loing this the analyst will be compelled to learn a great deal more about his cases than he nOw knows and supervisors will be compelled to work with analysts to a much greater extent than is taking place now When an expert type system is in place and functional one of the benefits wilLbe that time will become available for a different and much more meaningfUl kind of on-the-job tl'aining ' 6 eeS We began exploring how to develop an expert system for processing Ijust pefore 1 ret ired I expect initial development to be slow painful and frustrating but r do expect the system to function as designed What we must have to make it work is a sufficiently confident estimate of our analytic judgments and capabilities to tell the machine Jun-Aug 1985 CRYPTOLOG page 9 EO 1 4 c ll UBf JB VIA eS1HU'3 emmmu s GN I5l L 86-36 S CR E'f SFJEURlUJT aCID 4019704 how to do all the dog work for us and then let it do the job There can be no doubt as to our ability to make the right technical and reporting decisions based on the alert messages because each analyst must understand his responsibilities well enough to have described his case net and network norms to his analytic dictionaries U We will see vastly improved data bases as Ii side benefit of this kind of processing If the input data aren't good the expert system is going to make noise it will not permit any analyst to do a poor job on editing input data without sounding off treated them with respect in the past they will often help you get out of trouble U I finished tj1is article just as my NSA career came to an end Reading it over 1 find a line or two that may be mildly objectionable to one person or another It was not my intention to offend anyone by my choice of words or by my comments and if I have given offense I apologize This doesn't sound like me but it does sound nice and I think I'll leave it right there Goodbye good luck God bless you all 0 U I am not persuaded that an expert system can be developed and implemented by evolution probably the best and certainly the quickest way to effect such radical changes in the way traffic analysis is accomplished would be to select the right analysts and programmers charge them with developing an expert system to do traffic analysis on a specific target and set a deadline of perhaps a year hence for the initial job to be finished There should be no restrictions as to how the task is to be accomplished and there should be no arbitrary limitations as to the techniques to be used e g storage of norms for ordinary comparisons should not be made excessively difficult by the intrIcacies of an existing data base BACK UP YOUR U There is no question that the agency will deve-Iop an expert system at some point in the future if not for normal traffic analytic functions then to provide rapid data evaluations for tactical support of military forces I regret that I will not be here to see it DATA FILES U U Here is the advice I promised earlier Norman P Smith H215 U Learn to be persuasive as well as informative in any presentation you make The best ideas you have aren't worth much if you can't sell them You cannot depend on decision makers to steer the right course because it is or appears to you to be the logical thing to do FeUOj Reporting the results of your analysis is part of your job Learn what the intelligence requirements are for the target you are assigned and relate them directly and specifically to how you work your traffic If you don't understand what it is that you are supposed to be getting out of the traffic and don't try' to find out you are still a logger and that's all you're ever going to be U Avoid meetings I am convinced that many people at the Agency really and truly believe that they are making progress as long as they are talking with each other regardless of whether or not anything is actually being done I have attended some meetings where the only concrete thing accomplished was an agreement to meet again at some specified date in the future to discuss the same subject And so help me they all left happy U Don't take sinful pride in your own words If the staff officer who reviews your 'message or report wants to change your words for his don't be offended Staff officers do serve a useful function If you have Enough cannot be said about the importance of local backup of data files especially data being stored in micros The first thing new users should learn when they receive their own ASTW PC or other micro is how to backup' all files and the operating system The command below will allow all ASTW-PC lX users to backup their own directories find pathname -print where pathname is the complete pathway to the directory being backed up Le usr npsmith Users of TSS and other systems depend upon their systems administrator and the operating systems themselves Backup copies should probably be made at least every 12 hours if not every 6 hours and held for 24 to 48 hours The rule of thumb for backing up data is to back up your data for the span of time you find acceptable to spend re-creating your work Another method to accomplish this on the ASTW is to work directly from floppy disks where each floppy contains an entire working directory You then use the MOUNT and UNMOUNT commands to bring up and take down access to the floppy These commands are covered in the ASTW Users' Guide 0 Jun-Aug 1985 o CRYPTOLOG o HAPiBLr YIlt priv dump -i -v CO ilI T page 10 CIIArNEbS Q JbY SFJCRE' P S1 J6R8'f' SPOttS 4019704 ESCHEW OBFUSCATORY SCRIVENERY PLEASE u p L 86-36 I This article is classified jf J ff jfSteE in its entirety Many operators frequently use the TR NOTE during transcription to elaborate on the traffic yet few stop to realize just why and for whom they are making the additional entries When properly employed the comments contained in the TR NOTE can be of great assistance to those who ultimately read the transcripts giving them useful information and insights When abused or misapplied the TR NOTE adds little of any worth wastes the transcriber's and the reader's time and can be potentially embarrassing for the unit as well as for the U S intelligence effort as a whole Notes such as the latter are most often rendered on mids on the last page of lengthy transcripts of poor readability They are personal expressions of frustration and exasperation and do little more thantrivialize your work and detract from its overall impact Please avoid the temptation to indulge 0 Such notes are DrIer InrOrmatIVe proresslOnauy executed and to the point They add insight correct mistakes present information not reflected in the words themselves and save leg-work for the re der Contrast them with the following p L 86- 3 6 EO 1 4 c Jun-Aug 1985 CRYPTOLOG - SEtRET iPOl i Reprinted with permission from the Summer 1985 issue of VOX TOPICS page 11 aCID 4019704 IISA-Crostlc 110 82 s S A Put away the swords Heather said 5 B Ornithogalum thyrsoides 2 wds c Surrey town located above U S Marine D Twins age fast at new museum 2 wds E And so he mediates twice near The tides than wash on old Algier 2 wds F Head of Metro and DDE visit'Doiiglas G Get Helen Kennedy ' H c Does it hurt _ I laugh 2 wds 164 -6- 91' 63 186 223 107 128 I Hilaire Belloc hero usually has yellowiSh hue J Twenty-four hours ago he was a steady rye drinker K Wrapped in the wads of bandage L Not on hand M 'it I s the wbnderful lather we feel leads to such a state of perfection comp N Bonnie - - 2 o wds Wearily I left to see my solicitor P Disregard Q Double checker of a tergiversation 15 R Restive not asleep var Jun-Aug 1985 CRYPTOLOG page 12 FeR 9FFlGh b l JSE aNbY' 4019704 S Worse than a vile sea monster T Indian town for two idiots u Large ice sheet in Antarctica var 2 wds v To make the stout weed tender he boiled it longer than anyone else w I betrayed him when I saw him swipe a cheddar cheese x Smelly seat in church Y Remark directed at Dr I Q 8 wds z 126 165 196 The salamander went home 221 120 Zl' Develop a large state in the mind 74 38 un 134 D44illm 7 J R 9 C 156 T 170-P 171 183 R 184 Jun-Aug 1985 Pl-Nov 85-Yl-87742 f6ft CRYPTOLOG 6FFh IAt tiS NLi page 13 10 R This document is from the holdings of The National Security Archive Suite 701 Gelman Library The George Washington University 2130 H Street NW Washington D C 20037 Phone 202 994-7000 Fax 202 994-7005 nsarchiv@gwu edu