UlGBPiJVVlIlUlGBPiJl1 15 l JlBVV GBPiJl iI5Ul W V W W fW W THE FUTURE OF CRYPTANALYSIS ooo oooooo ooooo Wi11iam Lutwiniak o 1 TIE N S A Nt ti I UNCLE-A SAM WANTSA YOU ooo oo o oooooooooooo ooo C o A A o N E W S Wo E o S o o o o THE JOYS OF UNIX ooo ooooooooooooooooooooooo I 1 THE EDITOR'S PAGE ooooooooooooooooooo ooooooooooooo o o oo NSA-CROSTIC NO 14 o oo o o oooooooooooooo Sardonyx ooo oo oo 'fIllS BOetiMBN'f eON'ftINS eOBBW6RB flfU'fBRI AL 86-36 13 14 15 19 20 e ioo 9IRNSA 'SIISS8 N8A 'SSSM 118 eaapt P L 0Bi 619S B9 116 11 S k II 1l I Oec ily II N il ioo II lie 9PiIi_te Declassified and Approved for Release by NSA on '10-'1 2- 20'1 2 pursuant to E O '135 26 vl DR Case # 54778 DOCID 4009811 TOP 8ECBM' Published Monthly by PI Techniques and Standards for the Personnel of Operations VOL V MAY 1978 No 5 PUBLISHER WILLIAM LUTWINIAK BOARD OF EDITORS Editor in Chief Arthur J Salemme 5642s Collection o 'L I IIC8955s P L Cryptanalys is 1 1 49025 Language I ItS i 36s Machine Support --- I i 303S 1L - Mathematics Reed Dawson 3957s Special Research Vera R Filby 7l19s Traffic Ana1ysis Production Manager ---' 4477s Harry Goff 4998s For individual subscriptions send name and organizational designator to CRYPTOLOG PI TOP 8HCRH'f o 86-36 DOCID 4009811 'fOP SECRE'f lJMRRA THE FUTUFIE OF CRYPTANALYSIS William Lutwiniak Chief P1 he second most frequent question I'm asked the first concerns promotions is Does the cryptanalyst have a future Predicting is a risky bus-iness Did you happen to catch in all the school closures and cancellations announced on tne radio on that snowy Friday last week the postponement of the meeting of the Clairvoyants Society T I Complete text of the keynote address deliveroed by M ' LutlJiniak on 24 Januaroy 1978 to thE thiI'd annualsemina1'-1JJOrokBhop senes Croyptanalysis Contemporoary Issues The senes is offeroed as a aourse CA-305 by the Croyptanalysis Division of the National Croyptologia Sahool Other paperos proesented during the 1978 senes UJiU appear in future issues of CRYPTOLOG May 78 CRYPTOLOG Page I I'OP SBCBBf f JMBRA _ _ _ _ _ _ _ _ _ 0------ DOCID 4009811 TOP 88CR8T UMBRA EO 1 4 e P L 86-36 May 78 CRYPTOLOG Page 2 TOP S8CR8T UMBRA o DOCID 4009811 1 4 P L e 86-36 TOP SECRET HMBIM Do youwant to attend next JPY-ear S- s emi nar The seminar-workshop Cryptanalysis Conteml porary Issues is offered every winter by the Cryptanalysis Division of the National Cryptologic School It is designed to bring SIGINT and COMSEC cryptanalysts cryptologic mathemati ians and other interested persons up to date n the status and trends in the current practice of cryptanalysis Speakers from various parts of the Agency make presentations covering cryptanalysis and related fields - Usually the seminar lasts three days It consists of the keynote address delivered in the Friedman Auditorium and 15 smaller presentations Each presentation is given twice To receive credit participants must attend the keynote address and five presentations The seminar which changes every year attracts participants from throughout the Agency If you are interested in attending the 1979 seminar look for the course announcement in Vecember 1978 If you haven't seen it by Christmas time check with your training coordinator Distribution may have slipped up J E 0 U May 78 CRYPTOLOG Page 3 TQP SECRET f JMBRA DOCID 4009811 n T E l J 86-36 CONFIBtJNTI- al CALLSIGNS AND WARC 7 P13 L --_ --- - T ----- he Federal Communications Commission FCC has proposed a significant change in radio callsigns for WARC 79 World Administrative Radio Conference Geneva 1979 namelY that callsigns be unique and capable of fully automatia monitoring FCC also proposes that such unique callsigns be applied to every transmitter or transmission which aouZd propagate italics added beyond the boundaries of the country to which they belong The FCC also wants to drop from the radio regulations the exemption for military transmitters This scheme if it is presented as the U S position at WARC 79 could have a marked effect on SIGINT and COMSEC of many countries and hence deserves notice The unique callsign scheme was presented by the FCC in its Fifth Notice of Inquiry NOI for WARC 79 pUblished in FedePaZ Register 31 May 1977 The specific language appears in Appendix 3 as proposed changes to the Radio ReguZationa of the International Teleconununication Un i on lTD and as Resolution G-C Resolution G-C Relating to Automatic Identification states in part The General World Administrative Radio Conference Geneva 1979 considering a the state of the art in respect to identification b the need for unique identification c the possibility of inadvertent operator error d the ever increasing number of active transmitters not only within existing administrations but also noting the ongoing assignments of new call sign and selective calling systems e the economics of presently available equipment which is directly applicable to automatic identification f the possibilities for use of automatic faster message transmission service in conjunction with automatic identification equipment g the increased ease of resolving cases of harmful interference and of ensuring compliance with the provisions of the Convention and the Radio Regulations resolves 1 that administrations shall implement automatic identification as provided in Article 19 MOD modification -_ see below at the eartiest possible time and 2 that automatic means of identification should be a opted by aZZ administrations Italics added The FCC then proposes substantial changes to Radio ReguZations Article 19 Identifica- cur- tion of Stations Ii It proposesreplacing iKe re nt 1976 No 735 with MOD 735 as follows 1 ut13 5 Transmissions without identification or with false identifications are prohibited MOD 13 5 Transmis s ions and transmi tting stations shall be uniqueZy identified Administrations shall make every effort at the earliest possible time to introduce and use automatic identification On frequencies assigned for international use means recommended by CCIR International Radio Consultative Committee of the lTD shall be utilized Italics added The FCC then proposes the suppression of the current Nos 736 and 737A of the Radio Regutattons In order to remove exceptions to the odified No 735 No 736 currently provid an exemption for survival craft and emergency radio beacons the FCC would eliminate this exemption in favor of cal1signs that satisfy MOD 735 No 737A currently provides an exemption for some space stations e g satellites which exemption would also be eliminated No 737A specifies various acceptable kinds of callsigns station identifications or selective call numbers -- which the FCC would accept No 738 would remain unchanged It specifies regular identification signals at least hourly and ends with 'the statement that ' dministrationsareurged to ensure that whereVer practicable superimposed identification methods be employed in accordance with CCIR Recommendations Italics added Nos 739 740 and 741 would be sup pressed No 739 states that the iclentifying signal shall be transmitted by methods which do not require the use of special terminal equipment for reception Italics added Nos 740 and 741 also conflict with the automatic monitoring and uniqueness requirements The FCC gives as its reason for suppression of Nos 736 and 737A To provide universal unique and automated identification Its reason for suppression of Nos 739 740 and 741 is Consequential to above proposals The FCC then proposes that No 742 which allows each ITU member to establish its own measures for identifying its stations used for national defence be suppressed The FCC gives as its reason for this proposed suppression Unneces sary to incorporate Convention May 78 CRYPTOLOG Page 4 CONFIBBNl'fAL IbllN9 E '1IA 6eUHIT Slw HSr elf DOCID 4009811 eONFI9BN'I'IAI provisions in the Radio Regulations that is the general authority for national defense stations is given in Article 38 of the Convention However the deletion of No 742 is clearly intended to encourage and facilitate the unique identification of military transmitters and 'transmissions The treatment of No 742 at WARC 79 could produce interesting alignments of countries I ence Hence most stations are not required to use international callsigns and the burden falls on the victim to prove the harm and identify the station exerting the harmful interference The modified language the FCC proposes is quite different namely Each station whose signal aould propagate internationally shall uniquely identify itself such as by a call sign formed pursuant to No 747 Identification shall preferably be by automated means using the applicable Recommendations of the CCIR See Resolution G-C Italics added This is clearly a complet ly different criterion for harmful interference does not have to manifestly occur Instead the criterion is merely the technical aapability for detectable propagation across a national border into the international ocean areas or into international space particularly the equatorial geostationary satellite orbit which is becoming crowded The effect of MOD 743 would be to require that a much larger population of tYansmitters allover the world send unique identifications in a manner that could be automatically moni Ha t'rrtful Interference V8 Capability to tored Tens of millions of transmitters Pl'opagate Inte r nationaUy particularly mobile stations would be affected by this In Section II Allocation of International Considered in a U S context virtually all Series and Assignment of Call Signs of Article 19 the FCC further proposes a signifi- mobile and CB transmitters which aould propagate cant regulatory change which would markedly in- across national borders would be affected Earth satellite stations and even radio relay crease the number of stations to be assigned stations near borders would be affected The unique international callsigns No 743 now modified language of 743 also requires that reads each station shall uniquely identify itself All stations open to the international while the existing language only requires that public correspondence service all amateur it shall have a callsign This imposes a restations and other stations which are quirement for automatic identification whenever capable of causing hal'rnful interference a transmitter comes on the air rather than at beyond the boundaries of the country to the operator's discretion and the use of impro- which they belong shall have call signs vised or changing callsigns as in CB would not from the international series allocated satisfy the new regulatory language The FCC to each country as given in the Table of has been trying for some years to get automatic Allocation of Call Sign Series in No 747 identification features into mobile and CB Italics added radio e g Docket 2437 and this has been fought by the radio industry But if WARC 79 Here the deciding criterion of other stations adopts such a yision then Regulation is that of harmfUl interference which is 47 USC 303 R will make that new Radio Regulation defined in Annex 2 of the 1973 ITe TIAS 8572 applicable to the United States and FCC-type as approval can exclude all new transmitters which any emission radiation or induction do not satisfy this requirement which endanger8 the functioning of a radioConsidered in a foreign context the adoption navigation service or of other safety serof such regulations not only will affect internal vices or 8eriou8ly degrade8 obstructs or radio operations and monitoring but will also repeatedly interrupts a radio communication produce a great increase in radio negotiations service operating in accordance with the between countries and in the reports to the ITU Radio Regulations Italics added IFRB and CCIR as countries attempt to reduce This is a stringent criterion which requires the interference their stations cause or suffer the occurrence of sustained severe interferAutomatic monitoring with computer analysis MOD 143 The clearly stated aim of all this would be to allow automatic monitoring of all radio stations by causing them to transmit unique callsigns or identification numbers in an automatic and standardized way The adoption of this regulation would affect millions of stations -fixed and mobile CB radio-amateur and earth satellite -- which currently are not required 47CFR 25 206 to identify themselves in the United States and in 150 other countries It would completely change the economics and p actical aspects of radio monitoring interference notifications and regulatory enforcement in developed countries where there are numerous radio transmitters It would also have a marked effect in the ITU allocations and statistical studies of the HF spectrum Although millions of transmitters which already use callsigns would be affected the FCC proposal does not stop there May 78 CRYPTOLOG Page 5 eONFI9BN'I'IlaL DOCID 4009811 EO 1 4 c P L 86-36 CONFIDBNTIAI and accounting will lead to much more compiete' models of pTopagation Merever it will provide a great deal of dafa about radio traffic activity rare one The Soviet delegate to CCIR Sviridenko favors such central planning management and engineering of the radio spectrum using computer propagation models One other proposed FCC modification -- to largely developed by the United States At Article 16 Reports of InfTingemen reads present the priority rights of current spec If an administration has informatrum users especially in HF and space systems tion of an infringement of the Conare defended by the industrial nations because vention or Radio Regulations comradical changes in allocations and radio links mitted by a station over whiah it may would produce unpredictable effects With global automatic monitoring made economical exercise authority it shall ascertain the facts fix the responsibility and practical much more intensive use of the spectrum could be undertaken and propagation and take the necessary action effects predicted Interference or any in Italics as in text fringements could be quickly identified and The FCC states that this change is To permit corrections demanded under Article 16 All action against possibly unauthorized stations committing infringements This closes a legal this would produce a great deal of change in radio usage and data about radio traffic over loophole by which a country could ignore viola- xhe next 25 years particularly in the contions by a transmitter in its territory by gested regions of the spectrum claiming it had not authorized the station Aggregate Eifeat of FCC PrOPosal 8 The aggregate effect of the modifications to Articles 16 and 19 and of Resolution G-C would be to establish a framework for global automatic monitoring and much tighter enforcement of radio laws and regulations If these proposals or something like them are adopted at WARC 79 the global radio spectrum will be a much more tightly managed resource and international engineering of radio systems will become a standard phenomenon rather than a HISTORICAL NOTE ON MILITARY CALLSIGNS J A Meyer ince the first international radio t reaty in 1906 nations have always reserved completeJree domfor th eir miJitary and naval stations opposing international regulation except with regard to distress messages and inter ference This is expressed in ATticle 38 of the 1973 International Telecommunications Convention TIAS 8572 The United States has always ma n- taineatnISsame reservation Section 303 0 or the Communications Act of 1934 states that the FCC shall have authority to designate call letters of an stations italics added Section 305 a of the same Act provides an exemption for government-owned stations as follows Radio stations belonging to and operated by the United States shall not be subject to the provisions of sections 301 and 303 of this Act But then the exemption is reduced for aansigns by Section 305 c which states An stations owned and operated by the UnJted States exceEt 1Il0bil e_stati ns S 'of the Army of the United States and aH other stations on land and sea shall have special call letters designated b the Federal Communications Commission Note that the word except pertains only to mobile stations of the Army of the United States Therefore the removal ofNo 742 would reinforce the authority of the FCC to assign callsigns to U S military stations under 47 USC 303 c Even the basic concept of freedom for national defense stations as declared in Article 38 of the ITC was challenged by the USSR fri 1932 when that government was first invited to the IrC in Madrid The Russians proposed that military stations be regulated by the same rules as nonmili ary stations and although that proposal was defeafed 'ine c hafIenge might be resurrected at WARC 79 where the USSR ould expect greater political support than in 1932 If No 742 is deleted at WARC 79 as the FCC proposes a considerable effoTt by U S COMSEC to get U S military stations to use changing callsigns would be undermined Hence the FCC proposal would affect U S security unless 47 USC 305 c is amended to compensate for this CPOl l May 78 CRYPTOLOG Page 6 CONfIBHN't'IAL - - ---'-- '---- IWIBbE YIA 8SMIIft' etb'dftffihS SlitN DOCID EO 1 4 c F o L 86-36 SECRET The FCC proposal repreS nts a consolidation of ideas which have been emerging and finding application -- especially in aeronautical mobile and maritime mobile radio communications -- for some years Selective calling systems and automatic monitoring equipment as the FCC notes are already in service and have proved their value These facts will be noted at WARC 79 The FCC proposal is a logical generalization of existing practices and a recognition that automatic monitoring --and the automat ic identification of practically aZZ transmitters -- is an essential condition for radio planning and manL -------------------------------- agem t for the next 20 years and beYond Effect on Agency's 8sion The FCC in issuing these proposals on 31 May 1977 invited comments At that time the Office of Telecommunications Policy COTP a White House staff group established in 1970 was the official organ for coordinating and presenting the comments of the government to the FCC The deadline for the comments on this fifth NOI has passed and the next NOI is expected in early 1978 The OTP is being disestablished and it is not clear where the co-ordination function for NOI responses will eventually land The Department of Defense as the largest user of international telecommunications in the world has a special status and in 1973 was a member of the U S delegation to the Plenipotentiary Conference of the ITU Other government departments respond directly to the NOls and the Department of Defense could also reply directly partiCUlarly where national security considerations apply The FCC proposal for unique fixed automatically recognizable callsigns having been presented as the public position of the U S govern ment in the FederaZ Register' will be widely read around the world and ma be introduced and supported by other countries at WARC'79 whether or not the United States presents it r---------------------------------------------- J 6 668 T A IMPLICATIONS OF F C C PROPOSAL 1 ------------ lpt4 L -- ----J EO 1 4 c May 78 CRYPTOBOlL SBEURRB'f 3li ES 668 P L 86-36 DOCID 4009811 SBCRE'f sse Solution to NSA-crostic No 13 P L By CRYPTOLOG April 1978 Ounklr II P William Filby ULTRA Was Secret Weapon That Helped Defeat Nazis CRYPTOLOG December 1975 U l'nhappily it was not unusual for holders of the German decrypts to have to forgo using them for fear of compromising the cypher break One such occasion was the bombing of poor Coventry enemy plans were known beforehand but to aefend the city would have aroused German suspicions The upper right-hand corner of the covername allocation is OVAL ALBUM ROWBOAT COMPRESS PES ADAGE LACEWING SEEDLI NG TONI C GERUND APEMAN LOUVER TYPHOON FIGHTER SCHOOLBOY COUNTRY PLODDER BARRETTE MESA Once you recover the method of generation can you deduce the source of the covernames U U CLA-NCS in June FORE IGN FILM SERIES Thursday 8 June 0930 Both films will be shown in the Friedman Auditorium All are welcome PRESENTS The Crypto-Linguistics Association and the National Cryptologic School will present in May Heroes of Shipka in Russian with English subtitles The Shop on Main Street in Czech with English subtitles Announcements with details about the films will be mailed to CLA members and will be posted throughout the Agency Look for them We'll see you in the Auditorium B Y O P C U Friday 5 May 0930 May 78 CRYPTOLOG Page 8 SECRE'f 1h'dIBbE VIA serlI IT SlWHlEb5 ellbY 86-36 DOCID UNCLASSIFIED P L 86-36 rPRD lECT U TE NSIL The DDO Data Dictionary Directory I J Ip13 hat js a data dictionary directory Just as an ordinary dictionary contains information about words a data dicti l1al'Lc Iltains infc rmation about data It does not contain the actual data that forms the data files but contains pertinent information about that data its attributes and relationships There are many definitions in the commercial world but what the dictionary contains is information in two forms o the what information -- the data diationary description of the data elements and o the where information -- the data direatory the location and use of the data elements and their relationships to other data elements records files procedures etc Why Combine the Two By combining the dictionary and the directory into a single data dictionary directory we have the ability to provide W I o coordination and control in systems development o assistance in search for relevant data during design o a means of identifying and reducing data redundancy o an increased data-transfer capability between systems o data standardization o administrative support o documentation support o data-definition support Thus the system provides a versatile tool for managing the ODD data resource BT'ief History of Data Diationaries Data dictionaries are not new either in the commercial world or within the government Their existence is related to and probably can be entirely attributed to the Data Base Management Systems DBMS and many are an integral part of a DBMS The National Bureau of Standards has published an extensive report on seven of the commercial dictionaries and eleven government-agency systems SLightLy edited veT'sion of a taLk given in 197'1 at two meetings of CISI's SpeaiaL Interest Group on Information Proaessing Systems SIG IPS The National Bureau of Standards also sponsored a Federal Information Processing Standards FIPS a k group -- TG-17 -- specifically to address guidelines for establIshing data dictionary directory systems That task group published its report in late 1977 NSA also has several data dictionary efforts in various stages of development They are o TEDS in which uses a data dictionary on the M-204 computer system that has been operational for some time o an effort in A to utilize the TEDS experience in developing a dictionary for STEPSTONE on the M-204 o HOLLYHOCK a project to support L M N and E which will use a data dictionary developed on the M-204 by T33 o INLAND a project to maintain continuity on R tasks which uses a data dictionary developed in R o an IBM data dictionary that is being used on a system developed for a field site by T and o the most recent addition the Cullinane Data Dictionary which was purchased by T for use with the IDMS Data Base Management System and which is currently being evaluated DDO's Need for a Data DiationaPy Diredtory Why does DDO need a data dictionary directory If you have ever attempted to solve a problem that required you to find all the possible files manual or machine that might contain information about your particular subject you know one reason why we need a data dictionary For those of you who have never made that attempt I have an example of such a situation About 2 years ago I was asked to help locate all the data files containing geographic coordinates grids and or other means of identifying a point on a map The requester also wanted to know if there was any associated software to process the geographic information for selecting records by area I found no means to locate either the files or the software without surveying each organization How then do managers answer questions from auditors customers and the Director if there is no means to assure that they have all the information Other questions that a data dictionary directory system might answer are May 78 CRYPTOLOG Page 9 UNCLASSIFIED DOCID 4009811 UNCLASSIFIED o How many programs in the COBOL language are there on each machine o How many A files are resident on the IBM 370 168 system o Is there a standard for aircraft type unless I have an inventory of what is available on those computers Just having an inventory isn't enough either if you don't know how to identify or access a file You also need to know more about the file if you plan to use it Take the following hypothetical example For some research reason you would like to know the number of current NSA male employees with blond hair born in New York City who were hired by the Agency between 1956 and 1970 You want to be sure you have looked in all files which could possibly have information to answer o What computers are on the NSA Network your query But you do not want to search files PLATFORM which wouldn't possibly have the proper inforAnother illustration of a dictionaryrs use mation Withlidatadictionary directorYdata comes from I I'upresentation base containing pertinent information about all On the G project GEISHA in which he disthe available files you could narr0l'lyour cussed their current operations and the search to only those fil sl'lhichcontain inforproblems that G has encountered He explained mation about Agency employees city of birth that one of the problems is the existence of date ofhire afid color of hair The system many individual processes carrying out similar l'Iouldthen provide from the contents of the functions Elimination of duplication and data dictionary directory data base all the inrelated ills requires a coordinat cl efGBPort to formation available either on how to extract create a single GsystclI ' ithe information yourself or who to contact to get what you are looking for This example _Tuh IJi nkofDDOas being similar to what Mr isn't a typical 000 problem but it enables I Ireported about G processes Many me to add that the system would not allow you individual processes Similar functions information about files if you do not have Duplication The fact is that there is a need-to-know or proper credentials You would system for A a system for B a system for G have at minimum identification of all files a system for V and a system for W I do not which could help you get your answer propose that we design a single system for 000 as a whole but a data dictionary containing So the solution to our data-management descriptions of all the systems In common terms problem is to to have a DOD Data Dictionary would eliminate many problems or at the very Directory With that dictionary directory least would help us to recognize problems where we would no longer have the situation in which they exist the same data used in two data bases would be described differently Let's take a look at ourselves in 000 and diagnose the situation The 000 organization Yes the 000 Data DictionarylDirectory can contains several Groups with similar tunctions be the solution but only if the contents and the such as TA and A This results in similar data base are current and accurate and if they computer processes because the analysts' needs represent all facets of the 000 data resource are similar Therefore it follows that data Descriptions of our data elements data fields bases to support these functions will be nearly records files and data bases are required to identical in structure Well similar or build the Dictionary Directory datil base nearly identical but not necessarily recogThe following are' two examples of descriptions nized as being such because 000 is organized of data elements by al ea and since each is supported by different computer experts the design of similar DATA ELEMENTS data bases would be different Add to that the different terms used by analysts in each area and discipline along with the use of Name Social security identification acronyms and abbreviations and it would be Abbreviation SSN impossible to recognize the similiarities between data bases unless you carried out a Synonym s Social security account number thorough study Therefore we need to agree Definition A unique indication of an upon common names and definitions for common individual and his Social fields of information Security account Another situation is the result of the size Date approved 710701 of the Agency and the number of files required to support its functions PLATFORM a project NDSC identi- NI-0003 to link computers will make access from one fication computer to another a reality I contend that Etc having access to one or more machines doesn't give me any capabilities I don't already have o What files contain aircraft types associated with the Bulgarian or Hungarian air forces o How many B FORTRAN programs are there for the IBM 370 158 system May 78 CRYPTDLOG Page 10 UNCLASSIFIED P L 86-36 DOCIO 4009811 UNCLASSIFIED Name Sex Abbreviation Synonym s Definition The division of human beings into groups based on physiological characteristics Date approoved 710225 NDSC identi- 00056 fiaation Eta Motors etc if you sit down at the terminal and ask for STOCK INFORMATION you will get information you want plus information you don't want HOGS ISH SHEEP CAITLE GEN r rrRS 14678 2 3478 87 62 12345 o 98362 o 4 9231 93 46 Conversely if you sit down at the terminal and request information on CARS manufactured in the United States and get the response REQUESTED DATA NOT FOUND - CARS 0 it could be because File I contains information on AUTOMOBILES etc Foreign and File 2 conThe following is an example of a reaord tains information on VEHICLES etc U S wh1ch can consist of one or more data elements manufactured So you can see that a lot of hard thinking goes into getting the terminology right instead of just dumping all the information into the Name Personnel record data base and causing retrieval problems later Abbreviation Another consideration is that the contents of PERS RCD the data base for the Dictionary Directory must Synonym s have amplification information concerning every Definition The record of a spec iff level that is to be described see Fig 1 NSA employee DATA LEVELS FieZd 1 Name of data element Social security number Length 9 Configuroation Numeric update authoroity M3 Eta FieZd 2 Name of data eZement Sex Length 1 Fig 1 Configuroation Alphabetic update authoroi ty M3 In order to have reference terms with which we Eta can relate we shall refer to specific data levels as entities Entities will be distinguished from one another by attroibutes -descriptors that identify or characterize entiWhat information then should we collect ties and help to establish relationships for inclusion in the Dictionary Directory within the entity or among entities The data base What are the data elements What are-their defintiions Where are the data ele ' following are some of the attributes ments used The document Data Standards for Identification SIGINT Activities promulgated as Annex A o Naae of USSID 414 is the only centrally docuo Abbreviation mented source of data elements with definio Synonym s tions A Computer Record Format File docuo Reference s ments some computer jobs with field names ExDescription cept for the published standar Data Elements o Narrative these field names -- which are arbitrarily as- Purpose - Scope signed -- are subject to the problems referred o Physical to earlier different terminology and ideas - Sequence about similar fields Therefore we have - Size different fields represented in several riles o Organizat ions - Responsible organization s under the same name and the same field repre- User organization s sented in several fields under different names o Dates To use everyday examples assume that one file - Implementation - Change contains information on STOCK in the sense - Other of livestock with data pertaining to sheep Etc cattle hogs etc and another file contains information on STOCK in the sense of Fig 2 illustrates some of the entities and shares with data pertaining to IBM General their hierarchical relationships The data May 78 CRYPTOLOG Page UNCLASSIFIED 11 OOClO 4009811 UNCLASSIFIED what information is made up of the 'attributes entities include data base file record describing the data elements and the where data element etc which describe storage information is made up of the attributes information All entities above the data levels describing the records file etc The human are the functional or management entities and you might say provide the reasons why each data aspects I refer to are those descriptions and definitions which are input or used by individuelement exists By that I mean that there is a als in a form they can understand And finalfunctional reason -- or should be -- for each ly the figure shows the machine needs for data element and that requirement is dictated the computer to maintain the system and the by one of the entities above the data levels relationships among the entities The arrows Attributes form the actual contents of the Data going in both directions indicate that the Dictionary Directory data base We therefore software uses information from the user as well would have attributes for every entity used by as providing information to him DDO to accomplish its data processing RELATIONSHIPS Project UTENSIL I've covered the description of a data dictionary and a few of the uses that DDO could make of such a system Now I should give you some background on Project UTENSIL A task 'force was established by 000 in July 1976 as a result of an A memorandum suggesting the crea- tion of a Data Dictionary Directory for 000 The task force forwarded a statement of require ments to C now T in February 1977 The proje t is currently in the fact-gathering and problem-specification stage The potential of a DDO Data Dictionary Directory as a tool is limi ted only by the degree of commi tment tO the concept that data is a resource to be managed like people money or any other resource Initially the 000 system at minimum will be aimed at providing a central computerize9 resource of information about 000 data It will provide a capability to show data relation DATA ships to all levels contained in the data base LlYIL with retrieval capabilities As with most online systems there will be input and update Fig 2 capabilities to maintain a current file The design of the system must be such that Fig 3 is a representation of the dictionary di- new capabili ies can be added_ without affecting rectory data base -- not how it is actually conthose incorporated in the original des'ign Therestructed but how we visualize it functionally The fore I believe we have to make a careful evalu ation of the full potential of a data dictionary DllKTORY directory for DDC'to assure that we design the WHIRl initial system to be flexible enough to'allow for any possibility The fact that we already have several dictionaries available in the Agency proves that dictionaries are feasible but at the same time it demonstrates that given the iack of central management each dictionary has its own merits and shortcomings What we need is clear concise direction from management concerning its goals with regard to managing data Then we can clearly plan not just the short-term goals of a dictionary but also the data dictionary directory system as a tool for the long term The data dictionary directory can be just a glossary of terms used by only a few technicians or with good planning it can develop into a tool to be used at all echelons The uses of a data dictionary directory system and its data base will belimitedonlyby the ingenuity of its users Fig 3 May 78 CRYPTOLOG Page 12 UNCLASSIFIED -nUCID 4009811 CONFIBBN IAL EO 1 4 c P L 86-36 for something more to do besides the people there are very nice I found a course EG 421 entitled Effective Writing and startecl atching television To my delight the heart of the 6-hQllr course explained how to edit and unscramble obbled ook That was recisely what I needed The impact of taking the course was that although I hadn't been able to do much translating since failing my first attempt at Part II of the PQE I passed on the second try EG 421 was the only visible influence that could have made the difference So what does EG 421 have in common with hiring P L 86-36 linguists Well the prerequisite for the course was that anyone wanting the course was supposed to have job duties requiring extensive writing It hadn't dawned on me that I was doing a lot of writing after all my work was to reduce material written by someone else into English I had not been doing any composition per se just very time someone defines a bigamist as translating No wonder I had failed the PQE I what an Italian calls a dense fog I am finally passed it when I began to think of myself reminded of the article by Robert E as a technical writer whose material is usually Gould in the December 1975 CRYPTOLOG Linguists dictated by but sometimes only inspired by the from the Melting Pot The author's main point Portuguese it represents I rarely had trouble was that the Agency's dream of recruiting sucunderstanding Portuguese my problem was writing cessful translators from ethnic neighborhoods English The people who edit my translations was being frustrated because the aspirants were i sist I still have lots of problems not working out Using examples taken from My point is that WTiting English is the major English as spoken by Italian immigrants he portion of a translator's job the foreign lanclaimed that anglicisms had polluted the guage is secondary There are lots of good applicants I foreign language so badly that they translators in this Agency whose command of their did not have a chance of succeeding What he job-related f017eign language is far from native asked those potential translators from ethnic or even that of a college graduate This means neighborhoods in effect was not What'sa that i f an applicant with a childhood foreignmatter You no spicka da English but rather language background doesn't do well it is not What'sa matter You no spicka you owna because his neighborhood was polluted with anlanguage glicisms it is because his English was polluted That idea struck me funny at the time because or in some other way deficient like mine In I could think of no real reason why a person from other words That'sa matter He no write-a an ethnic neighborhood should not be able to do da Englishl well in the Agency It certainly was not because Having diagnosed the problem it's now time the applicant would not know the special vocabu- to prescribe the remedy If the Agency ever lary found in Agency work The Portuguese I had needs to recruit translators again 1 specify learned in Brazil as a missionar was ver different translator a person who writes English -- very differentfrom a transcriber a person who and yet I somehow passed the pro lClency writes some other language -- it was a sad mistest Surely a child growing up in an ethnic take to confuse the two and mislabel them both neighborhood would learn the patteI'nS and many of linguists it should stop restricting the the idioms of the foreign language With a search to language majors The ideal recruit mastery of those patterns and idiom$ the new really might be a Journalism or English major hire with an exotic surname could be given a with a Language minor and some course work in glossary of special terms and be expected to do International Economics and Political Science well no Apparently not But why My degree is in Economics Because of that A year after reading tne article I discovered no one ever asked me if I wanted to go to the the reason why the Agency shouldn't be able to find a rich source of good translators in ethnic D C area as a GS-7 and use my Portuguese No one asked my roommate who spoke German to come neighborhoods Fot technical reasons I was to NSA either he was a chemistry major I knew left for a while without much to translate and lots of people who never took college courses in I finally decided to go to the Learning Center f L-- ' UNCLE-A SAM VVANTSA YOUI 051 E May 78 CRYPTOLOG Page 13 EO 1 4 c P L 86-36 CONFIBBN IAL DOCID 4009811 CONFIBBN'fIAI the foreign languages they spoke fluently Indonesian Japanese or Korean in addition to French German Italian or Spanish because they felt they could never make a living with their language and went into other fields and because they were not language majors they were not interviewed by NSA People like this should be talked to by Agency recruiters who are searching for translators Perhaps recruiters could find people like this with ads in campus papers saying something like If you got an 'A' in Freshman English and know a foreign language Uncle Sam wants you An additional thought on recruiting linguists high ambition in an applicant should be viewed as a criterion for nonselection This view contrasts sharply with the one expressed by Daniel G Buckley in another Agency publication Can A Linguist Development Program for High School Graduates Work at NSA CIyptoZogia SpeatPUm Winter 1977 The problem with highly am i ious or highly mo ivated people whether college degree holders or Agency-trained high school graduates is that they expect and NEWS OF THE CAA CQl1lll1uni 'fi nl An lgjil AII' i'fi n ' E 8 Week in Marah CAA had a busy week in March On Wednesday I March l Ispoke in the Auditorium drawing upon her years of experience in the White House Situation Room On Thursday morning 2 March the CAA ODerational Briefing series featuredr - Busy emotionally require high productivity to be rewarded and government service is not set up to give rewards for productivity The result is that highly motivated linguists begin to look outside the 'Agency for advancement when they see little opportunity for promotion inside If the object is to recruit career linguists no more than a moderate amount of ambition or motivation should be allowed in a recruit A word of warning is in order Due to the accelerated decline in the number of college graduates who can actually write well in English the Agency may find itself in a position with translators similar to the one it is in now with programmers and engineers as the general demand for good writers increases in the economy translators will gain more outward mobility not for their ability to understand'foreign languages but for their ability to write plain English Then the Agency may have a language problem that will make today's situation pale in comparison entry came from I IA64S tho will receive a DOOK O flfls c oice Mo e ab9ut the winning entry n xtmonth U band Daii BOcKley might assist and encourage professionalization among ci vilians and militaryo MU9 abOut hOw tlie CAR I Then on Friday 3 March the CM 's Special Interest Group on Cryptologic His ory had a session on Oral History featuring Dave Goodman former NCS History Fellow and Art Zobelein The presentation included anintroducti l1to the principles and teChniques of 0I aJ history and a description of the oral history program here at NSA e-eee P L 86-36 Winner in Logo ConteST It was grueling work but the board knows its oats After much animated discussion about the wide variety represented in the entries the CM Board finally made up its mind Out of the sixtyplus ent ries in the CAA Lo o Contest the winning 86- 3 6 - CAA E'reunutwns Our Progr8ll Co_ittee chaired by' thas been busy too The co_ittee has lined up theJoilowlng speakers 10 May Admir l loun 21 June csubject Interstellar SIGINT September JI INI If youh Yeallyi4 asforotn ipresentations let Earl knOw UJ I Meet On Thursday afternoon the CM Board held its monthly meeting by the way we're back to monthly-changing rooms again and we spent a large part of our time talking with Sergeant 'P L P L 86-36 EO 1 4 c Unda The newest CAA Board Member is I aCQH SEC Analyst in S1 She earned an AB lJ1 Psychology at ettysburg COllege and CRo o to work here at NSA in 1966 P L Her cryptologic experience has included assignments on the staff of the Nat nal Cryptologic School where she was involved with CY-OO and CY-300 in B Groun where she was associated with the I and in S wl'n i e 1 she became a COMSEC Intern P L 4 c 86-36 She is certified in Traffic Analysis and COMSEC tip Ask her about collies CAA BOIU'd President David Gaddy President-elect frlnk Pgrrjoo Secretary I TreasuJ ' erTim Murphy Boa d members I I I 3247 5879 8025 3791 4935 5991 3573 3369 Inside U Four Easy Steps for Joining the eM 1 Get a membership card from any of the members of the CAA Board 2 Fill it out 3 Attach $1 00 Is that all it costs Yes Fred that's all it costs 4 Mail card and money to Tim Murphy B09 U May 78 CRYPTOLOG Page 14 CONFIBENI'IAL 86-36 lWlShE VIA EURellHff EURIWmEhS em y P L 86-36 DOCID - 4009811 UNCLASSIFIED I l J T The Joys of UNIX I Ip13 R53 F or the last 6 months or so I have been experiencing a new and in my jaded and cynical mind totally unexpected turn of events in working with computers I have actually been enjoying the process of using a computer system the R53 MYCROFT PDP 11 70 under the UNIX operating system After several years of being turned off by discouraging encounters with various computer systems some involving on-line terminals I find UNIX and In particular the RAND editor restoring much of my lost faith in the value and promise of computers I have found myself choosing to do at the terminal more and more things that I would once have done with pencil and paper with a typewriter or with cumbersome POGOL procedures on the IBM 370 Even more important I have 'chosen to undertake many things that I would hardly have considered before the convenience and accessibility of UNIX makes new things or old things done in newer ways seem pleasant challenging and possible rather than prohibitively painful burdensome and remote I have been so struck by the dramatic contrast between my former feelings of disgust and discouragement and my present optimistic and positive feelings about UNIX and by extension about computer technology in general and the tasks I perform with its aid that I have spent some time in considering which specific features give UNIX its remarkable value for me as a user I felt that some informal comments on this topic -- ways to make a computer system more supportive and hospitable to its day-to-day users -- might be of general interest Our Agency is becoming more and more committed to on-line interactive systems Larger and larger numbers of users will soon be attempting to accomplish an increasingly broad and heterogeneous set of tasks on an ever-growing spider web of intersecting networks The question of how to design and maintain a friendly user interface linking a wide range of users to a wide range of computing facilities is becoming increasingly crucial For one set of users with varied needs UNIX appears to have provided one good answer that bothered or pleased me When I first began programming on ATLAS I I felt that programming was a supremely enjoyable and challenging activity Debugging was done on the computer in octal commands were numbers as were all addresses in memory and addresses were fixed Programmers were also operators and learned as wlch as possible about the hardware of the computer since we had to demonstrate and prove each hardware error to the maintenance men before they would fix it Input and output were on punched paper tape printed out on a teletypewriter corrected with sticky tape and a hand punch Programmers could understand and get at everything and we could carry out every step of coding debugging and running our programs at our own pace and on our own terms using simple equipment directly accessible to us Given a chance to vote on whether we wanted an assembler language for ATLAS I we voted it down The manual consisting of two mimeographed 8 x 11 inch sheets was a miracle of clarity and succinctness which I have never since seen equaled it simply listed exactly what each command did with each bit in each register In any case we all knew most of it by heart Who needed an assembler As successive new computers came and went things necessarily got more and more complicated Assemblers compilers and subroutine libraries came along and operating systems began to assume increasing importance Addresses became relative or relocatable so we had to add a base address to everything before we could read our octal dumps and magnetic tape replaced paper tape for input and output Programmers were banished from the machine area to the outside of a counter and a new hierarchy of operators and systems specialists reigned supreme backstage Computers rose on the horizon and fell away to make way for still newer ones -- 704 705 709 7090 DCS to name those which which I myself became most familiar Still until the advent of IBM's third generation -- the 360s and 370s -- and time-sharing the changes involved primarily a slow accretion of added features which programmers could asFirst Joy similate a step at a time I myself still felt I would like to set the stage with a bit of that I understood DCS hardware and software alhistory -- a rapid glance back over my own exmost as well as I did for the earlier machines perience with computers as an applications pro- and I still felt that it was worthwhile trying grammer since 1951 I am sure many readers to do so reading and studying maintenance and will recognize some of the stages of computer software reference manuals to learn as much as usage at NSA that I mention in passing and will possible perhaps also recall similar reactions to them Joy Abating though many will not perhaps have been bothered or pleased by the same developments With the coming of the 360s there seemed to May 78 CRYPTOLOG Page 15 UNCLASSIFIED P L 86-36 DOClD 4009811 UNCLASSIFIED be an abrupt discontinuity Suddenly the manuals a programmer jleeded to study if he really aspired to understand the system had stretched out to fill a rack 10 feet long The workings of the complex agglomeration of hardware components and peripheral devices in evervarying configurations were all buried unJer endless layers of software comprehensible only to the IBM men and a very small number of o hers who chose to specialize full-time in these arcane matters For programmers whose interest and knowledge were centered around an application rather than programming as such the effects of this sudden increase in complexity coupled with a loss of understanding and control could not fail to be discouraging Against this changing background my own experience was for the most part one of growth to accommodate the added complexity Up until and including DCS again in the IBM computer series that I know best most of the changes seemed to bring improved capabilities without too much of a sacrifice for the user in terms of convenience and control over what was happening After the 360s arrived however my experience began to be one of progressive deterioration in my ability to get what I wanted out of the system the hardware the software and no less important the fleshware -- the people behind the computer installation and their ways of dealing with me as a user I became in fact less and less of a programmer at all and more and more simply a procedurewriter who tacked together canned routines or previously debugged POGaL steps to do dull things in a dull way Since it invariably took me 2 or 3 days just to catch up with all the control-card and rule changes transgression of which invariably resulted in a canceled run accompanied by little or no helpful information correct JCL errors achieve the necessary two -valid POGOL listings a compile and a go listing and cope with all the other things that usually went wrong with FILE cards to keep me from reading my input tape and getting an output tape it hardly ever seemed worthwhi Ie to try anything the least bi t conceptually challenging I was just glad to get a job done any way I could available and POGOL in particular was always a convenient useful tool for accomplishing the data processing functions I needed Unfortunately so many people were trying to do so many things with the system some of them apparently mutually incompatible at times that some of us were unable to get much out of it Thus for me the time during which I used the IBM 360 and 370 systems was a very discouraging nadir in my interaction with computers Joy Regained It was at this point that I had the good fortune to discover MYCROFT UNIX and the RAND editor Now suddenly I have the best of both worlds -- the illusion of having the computing resources all to myself at the terminal though many others are enjoying the same experience at the same time with all the power and richness of a modern com uter Once again I have a chance to understand some of what is going on under the cover if I wish to make a reasonable effort to do s in the office where I work there are helpful and patient people who understand the system and can aid me when the documentation is not enough I can calIon a wide variety of programming languages and I can also call up generalized functions sort select dedupe translate or convert spelling check for English words and several report generators all in a very simple and flexible manner I can create new files and execute my own programs or generalized functions on them directly and easily The whole system is consistent and unified so that I can quickly learn to use it at my terminal If I have been able to use and enjoy this system I am certain that anyone else could do so at least twice as quickly since I have always had a very hard time learning any new programming language and usually require a long time to feel at all comfortable with it The key feature of the MYCROFT system for me and probably for many other users is the RAND editor With the editor I can write a program jot down rough notes or draft a report placing it in a UNIX file I can then immediately attempt to compile and execute the program find the errors go right back into the editor to make changes rerun the changed To sum up this quick sketch of my own view of progr re-enter the editor and so forth until I have either checked out the program or the trends in NSA computer technology as seen else decide to leave my terminal to search for through the eyes of a day-to-day user I recall food or water or to satisfy some other basic an early period of maximum accessibility comneed In fact I have once gain the same plete control and understanding of the computer by the programmer This was followed by a peri- ideal debugging situation I enjoyed back in the days of ATLAS I When I do leave the terminal od in which hardware software and the human in the middle of this process I can rest asprocedures within which these were embedded besured that my files will usually be safe and came increasingly complex while many features will reappear when I log on again just as I were removed from the programmer's direct conleft them a certainty that I never had with trol the added power and conceptual richness the other on-line system I tried After of the facilities at my disposal more than roughing out a report or some initial jotbalanced these losses For me at least and tings an outline for example I can come to judge by many comments r have heard for bac to the terminal and rapidly reshape and many others as well the coming of the 360s refine the draft or fill in the outline as upset the balance of power greatly to my disadeasily as I could with pencil and paper and vantage There was a lot of comput ng power May 78 CRYPTOLOG Page 16 UNCLASSIFIED DOCID 4009811 UNCLASSIFIED more effectively For example if I press ARG then make a mistake while keying the integer or string parameMany readers may be taking issue with me ter I get a crisp and clear warning and can somewhat as follows Aren't these just the start over again by simply pressing ARG again things anybody can do with a system like LODEOften the user can recover from a mistake by STAR ' TSO or CANDE What's so special I am repeating a few simple key strokes In ny certain that these sy ems ha many advan case the mistake and its correction do not tages and would like to see an informal write- spoil or interfere with the text on the screen up on their good points seen strictly from or the previous correct actions I remember the applications-oriented user's point of view well how pleased I was when having written It should be remembered however that UNIX down to the last line of the screen without operates on a minicomputer in contrast to noticing and then having pressed CARRIAGE the large-scale systems mentioned above UNIX RETURN to get a new line I saw the editor provides a remarkable amount of power coupled obediently roll the window down to display a with an outstanding user interface all within new page for me to write on having apparently a computer systems which costs less than read my m1nd $300 000 -- about as much I am told as the How It Works disc storage alone of the big IBM CDC or In order to gain a more vivid picture of the Burroughs systems editor let us imagine a user -- me -- sitting I would like to describe for the interested down at the terminal to write a report I reader some features of UNIX which I find most have some ideas of what I want to say but helpful First and foremost the RAND editor nothing written down yet I log on a matter is beautifully designed from a human-factors of typing two tiny character strings devoid of point of view No other software tool that I syntax and in response to two simple prompts have ever used or studied can equal it in this Having decided to call my paper report I key in respect Most other editors are line editors re report that is I call for re theRAND requiring that editing be carried out on lines editor to work for me on a file called specified by number as was usual with card report Since this is a new file which I am orieqted file-update procedures These line about to create the editor displays a polite editors also require that some set of commands message telling me to press a USE key to REPLACE FIND DELETE MODIFY be cause the file to be set up Immediately the keyed into the terminal with the necessary editor then provides me with a window strings to specify the sought string the replacement string and so forth These commands labeled with the name report and all ready for me to begin writing also have syntax rules which must be learned and which are easily transgressed TransgressI tab over to a preset margin which I can ing the rules brings upon the user the need to change if I wish and set tabs for indentarewrite the command edit the command to the tion or tables if I need them with a few easy editor and try agaiilL key strokes Then since I am in early stages planning my paper I begin an outline I In contrast to this line-oriented programming- of start out _ language-like type of editor the RAND editor 1 Introduction allows the user to do most things by pressing 2 Essentials of the problem a single key Pressing a special ARG key 3 A summary of past solutions permits the next integer or character string to be fed as a parameter to the function designated At this point I dec1de that I want another by any of the other keys Thus a user can acheading between 2 and 3 I position the complish many complex actions by simply pressing cursor anywhere on line 3 and press the ARG keying in a number or a string of letters OPEN key The editor moves line 3 down one line depending on the action desired and then press- and I am all ready to write in the new line 3 ing the single key that stands for the action then change the oid 3 to a 4 After 1 finish This is all the syntax that has to be learned my rough outline and wish to write in subheadand it is consistent over the whole set of acings I can open up space and squeeze any numtions provided by the editor In fact I can ber of them in with ease Then when I am guess at what will happen with a set of key satisfied with the outline I can write the presses I have not tried before simply by extext in after each heading in the same way No trapolating from the editor's behavior after muss no fuss no scratch paper and the copy I the k presses I already know In many insee on my scope is always clean and well forstances of guessing the resul ts of key sequences I matted without crossouts or strike-overs have never so far been disappointed I cannot Suppose I am writing along on a line and think of any other programming tool I have seen inadvertently continue writing past the right which can be counted on to behave so transparborder of the window A flashing message and ently so logically and so sensibly on some terminals also a beeper demands my attention and I see a right-pointing arrowhead In addition to being designed to behave as warning me or the overflow r moveThe cursor a user expects it to behave the editor even achieves sensible behavior when the user goofs back to where I meant to end the line press May 78 CRYPTOLoG It Page 17 UNCLASSIFIED DOCIO 4009811 UNCLASSIFIED the DELETE CHARACTER key and hold it down while the overflow letters are neatly gobbled up and disappear If I decide after I have typed a few words that I want to leave one out in the middle I can press the DELETE CHARACTER key to squeeze out the unwanted word To squeeze in some words I press the INSERT MODE key and type them in Material to the right on the line moves over to accommodate them Now suppose that I wish to incorporate a paragraph from another report already on a UNIX file in my work space into this new report By positioning the cursor along the left margin pressing ARG then keying the name of the old file and then pressing CONTROL and Z I can split the screen into two horizontal windows I can have up to ten windows and call for the old report to be displayed in the second window while I keep the new one in the iirst window I know that the paragraph I want starts with the words It is obvious that so I press ARG then type in these words just as I expect them to appear in the text Then I press SCH the editor finds the phrase and displays it wlth its surrounding text in the secona window I count 12 lines in the paragraph by pressing ARG then 12 then PICK I copy the entire paragraph into a buffer -- the pick buffer I press CONTROL and c to move the cursor back to my new file in the first window position the cursor where the paragraph is to be inserted and press PUT -- the paragraph magically appears and I am ready to go on writing This feature is a delight in writing programs one need only code one version of a routine then PICK it and PUT it over and over again wherever a similar routine is desired changing the details later Once I have set up windows and filled them with files I can switch from file to file in each window and move the cursor from window to window with a few quick keystrokes Nothing Lost When System Crashes lines words and paragraphs jumping in and out appearing and disappearing all untouched by human hands In fact the keystroke file saved by the editor to produce this re-run of my session is just like any other UNIX file so t at I could get_ it into a window and edit it 'wlth the RAND ed1tor to change sequences of key strokes and thus rewrite the history of my own editing session if I wanted to As if this were not enough there is a macro facility in a special form of the RAND editor This facility permits the user to perform sequences of key strokes and then treat them like little programs he can designate an entire sequence by one key for example X then position the cursor wherever he wants to and press X tv execute the entire sequence This would be convenient for example in reformatting a fielded file a macro could be de-fined to accomplish reformatting of a page then executed for each page Summary While the features described above are those have proven most useful to me UNIX provldes many other advantages for more sophisticat d users For the benefit of readers who may be lnterested I will quote a summary of UNIX strengths from a technical report prepared for RS3 by a contractor t at In general the UNIX world view appears ready-made for user-controllable multiprocessor systems The capabilities that make UNIX attractive include process creation forking process intercommunication piping file directories and referencing via a MULTICS-like tree addressing scheme the shell concept command language elegance the equivalent treatment of system and user procedures and the user extensible command language Charles Kellngg Alternative Architectures for Deductively Augmented Data Management System TM-600S 000 00 Draft Sygtems Development Corporation December 1977 Now let us imagine that I have been working I hope that the previous paragraphs have at the terminal for about an hour nd suddenly conveyed some of the ease and pleasure of using the system crashes I have not been saving my file as I went along so I fear that I have lost MYCROFT with UNIX and the RAND editor I find an hour's work -- work that I might have trouble that I can get much more work done with far less trouble and effort using this facility as duplicating from memory The editor has autocontrasted to what I could do with other commatically saved a backup version of my file puting systems or with pencil and paper The i'report as it was before I began th-e latest RAND editor became familiar to users of the editing session in a file called report bak ELROND computer under Programmer's Workbench But that is no help with the changes I have and it is available on the KEPLER facility in made during the session With most editors I R17 It is also a part of STEPSTONE lIon would be in a very annoying fix the RAND PLATFORM and will undoubtedly become a valued editor however saves a record of every keytool of many NSA employees In closing I stroke I have made during the current session would like to emphasize the importance of the By keying in a simple sequence of commands I user-friendly design of the RAND editor and can call this record in and execute it on the urge that software designers use it as a model report bak file so that every motion I made for future systems With tools like UNIX and is duplicated until the cursor stops at just the RAND editor computers can come into their the point where it was when the system went own at last as real aids to human performance down and I am ready to go again It is very amusing to watch the cursor scooting around May 78 CRYPTOLOG Page 18 UNCLASSifIED DOCID 4009811 UNCLASSIFIED The Edit r's Page cD been presented as a Special Teenager's Award to the son of I IAZS Peteuuwrote The puzzle had me going in 'circles ' I was toying with it at home when my l3-year-old boy asked what I was doing I gave him the info and 15 minutes later he gave me the answer a cylindrical wedge If this is one of the first ten I would like for him to have the trophy Although the solution was not among the first ten we are pleased to make this award as- a sign of appreciation for spreading the nam E 36 of CRYPTOLOG - T h e solution to the puzzle that appeared in the February issue of CRYPTOLOG was printed in the March issue Because of print-shop deadlines this is the first opportunity we have to print the names of the winners TIle response to the puzzle was tremendous -- 48 _ Other CRYPTOLOG readers who provided the correct answers were rna-fled in carried in or tele- correct solution but too late to win a trophy phoned in telephone solutions were accepted if it's a cheap old thing anyway are the person could explain what he or she had in mind even without the use of hands Other than tne drawn solutions representing the easy-to-make version the three pieces of cardboard or a cone with a square plane intersecting it vertical ly or a piece of a cylinder with two facets shaved off the editorial office received some look-alike descriptions a vacuum cleaner attachment a washing machine agitator etc We also received some three-dimensional models an actual piece of a wooden dowel wi th facets shaved off it the woodworking equipment to make it must have cost hundreds of dollars a shaved piece of a pencil eraser a cut piece of a pencil a raggedly cut chewed piece of artgum eraser a piece of graph paper sort of squished into was an honest attempt at the right answer And When some of the contest winners finally we got two offers of models that never showed up in the Editorial Office materialized including an appropriately cut that's what I call this mess I would ask Now that you see how easy wedge of salami The following is an alphabetic list of the it is to contribute to CRYPTOLOG why first ten persons to provide the correct solution don I t you contribute an article Almost invariably the answer would be I might someday when I get a round TUIT Okay then here's a round TUIT If anyone -- contest winner or just a casual reader -- is thinking of submitting an article L D-a-v-i-d-H- -W-i-l-l-i-am-s- -P-1-6- -------- ---l to CRYPTOLOG cut out the TUIT fasten it to your article and send everything to CRYPTOLOG PI Flimsy gold-papered cardboa dtTophies have Room 2N039 Offer void where prohibited been sent to the ten winne s Another one has whatH re it is - --- ---------------- --1 May 78 CRYPTOLOG Page 19 P L 86-36 UNCLASSIFIED P L 86-36 ----------------- DOCID - -- - 4009811 UNCLASSIFIED NSA-crostic No 14 by guest NSA-crostician Sardonyx DEFINITIONS L 86-36 The quotation on the next page was taken from the published work of an NSA-er The first letters of the WORDS spell out the author's name and the title of the work WORDS UNCLASSIFIED DOCID 4009811 UNCLASSIFIED SoZution next month May 78 P L 86-36 CRYPTOLOG Page 21 UNCLASSIFIED PI-APR 78-S3-26362 DocrD 4009tsll -- - IIIS BOCt1MEN CON 1 INS COBEWOftB I'f11 'fEftI1 L 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