intelligent that it can read documents with a text header and address list, to transmit and store them in IR, while notifying the addressees through MP that the T- cument is stored in IR. Later the addressees ma-y access IR and retrieve the-file on their own facsimile device TXX2. At the recipients’ option

does mails realy comes on sunday

( or possibly the sender’s To mimic “recorded delivery”) each addressee may automatically inform the originator that the document was received. In addition, we wish to store the facsimile data in a canonical form, so that it may be retrieved from different facsimile devices. There is an important difference between the storage of facsimile and short text messages. Such textual information is usually small (less then 500 bytes), whereas the facsimile data is much larger typically 25 K bytes for an A4 size page with optimum data compression. Hence the textual data can be stored in multiple copies, one for each addressee in his “mailbox”. The facsimile information is stored only once in IR, and the path name to retrieve it is known by the addressees from their text messages. In our implementation of this system we have used ARPANET as the data network. For FAXi, we attached a Micro-Processor to an analogue 4-6 minute facsimile device ksection 3). This was then connected via an asynchronous link to the UCL TIP. However, for reasons of speed and flow-control, this llni was switched over to one of the UCL PDP-9 as described in Section For the MP of Fig. 1, we use one of the Tenex systems on ARPANET,supporting the message systom MSG TRef 2]. For IR the Data-Computer at CCA [Ref. 3) with =- on-line storage capac-ity -of 10**12 T”5Tts, is a natural candidate, and has been used exclusively. As for FAX2 we are using the XGP printer (Ref. 4] at ISI, whichis controlled by a PDT-! Tenex configuration as Tecribed in Section 5. o:i-. i_ 1 1 i II 1 I + I ‘ 1 i i i ‘1

return to sender usps is possible

72 1.3 FA X1 TIP Fig.1 Overview of on Idealised FocSIMile System FAXOO p.x C 3RPNE Fig.? UCL Facsimile Terminal

73 ,-., ” UCL Facsimile Terminal and its Use 3.1 Introduction A system such as the one outlined in Section 2 can be described at many different levels. In this section we will describe the UCL facsimile terminal itself [Section 3.2) and give examples of its use in message transmission. The examples given contain two user dialogues, one for message transmission [Section and the other for message retrieval [Section 3.4.2). The message transmission dialogue is mainly concerned with the composition of a message file [containing textual and facsimile information), its local verification, its submission to the MSG system and the Data-Computer, and confirmation of deliv-e-r. This dialogue also illustrates the binding of linkages between the two portions of the message and the supervision of message transmission which are almost transparent to the user.
ups lost tracking number is not find is that amazing

The retrieval of a message can have either one or two stages. Because the notification comes via the text message system MSG, the user may receive notification of facsimile message-s–uring straightforward text retrieval from a conventional terminal. In this case the retrieval of the facsimile portion must be a separate exercise. Alternatively, as illustrated in the retrieval dialogue, the user may access the MSG system via the facsimile terminal. In this, somewhat–impler case, the notification and retrieval become an integrated operation. 3.2 The UCL Facsimile Terminal Fig. 2 shows the basic components of the facsimile terminal we have developed. Physically it consists of : [(1 FAXD : The facsimile device which is a Plessey 4-6 minute analogue tranceiver [ KD-111 incorporated with a two level Analogue-to-dgi-tal converter.jita [2) up An Intel 8080 mico-processor with 24K [8-bTFword] Random ccess Memory, and peripheral interfaces. [3) FDD. A Fjlopy-Disc driver. [4) KBT: A Keyboard Terminal. [5) ACI An asynchronous communications _ interface.



(8) S. F dge – TCP Perfornimce Measurements from of all proportion to the results obtained. Simulation. l NDRA Note(, 619, May Each of th; thre lmplvrentations involved in (9) S. Erskine – Dataputc: A Paickt Switchedx Netthese measu cm.nts was notably inefficient. The work for Canada. Int.ernal.ional C)nfer’nce l)rc ( -or-&-.1..dmcnl reasons for this have been on Data Camunicatiorts Networks. lhndon discussed. The weakness of the original TCP win- Way 1977 (to be published). do mchanism was a Mijor factor in this ineffic- (10) Y Jacquemnrt & A. Danthine – Lixiely Coupl(d iency, and the lack of suitable algorithm to con- Interface between Transport Station and X25. trol window size was noted by all groups. Tbis The Third European Network User Workshop, has led us to initiate a set of slnulations (8) IIASA, April to investigate the effects of various flow coa- (11) R. Karp – ICP Experiment Plan. Privat v Cmnitrol strategies on TCP performance. unication, Septenber Finally the crude retrananission conflicts %12) L. Kleinrock, W. Naylor and H. Opderbc,’k – observix! point out the kind of difficulties that A Study of Line Overheads, in the ARPANET, can arie.- in a transsnet situation if there is no CAC), Vol.19 p3, January stratgy Itfor itttrac(‘,ing with lowr level con- (13) L. Kieinroek – Queueing Systems Vol.2. tnrls. IL i,, clfar that. such strateries “m”it be John Wiley v)l vl vd and lhat thli-4- will probiably lit, snaged (14) J. Postel – Survey of Netvork Control Progby thu rgattm.ys, hiut at. present. dis.lusion is rata in the ARPA Ompxuter Netw)rk. Mitre very open on exart ly what is needed. Corn. Report JIM 6722, Junra I (15) Packet ional Projec~t Radio Net Quaterly Development Report. – SRI TR Internat- 2=25 May ACKNOWLZXNF)GI S (16) A. Stokes – A User’s Guide to Babbage. UCL TR Report No.13, Tt-,q4upport ol the Science Rfsearch Council (17) C. Sunshine – Interproess Cbmmunicaton under grant B/RG/67022 and the AlUA IPTD office Protocols for Computer Nttworks. SU-DSL TR under ONR grant N C-0280 is gratefully Report 105, Deoaeber acknowlxiged. Additional sup ort has been (18) R. Tomlinson – &electing Sfquece Nmbers. nr.(‘ived fivm the, Ministry of l)ifenre under INWG Prot(xol Note 2, AugLst agr#vwmnt A7/2047/064 and the NAMT travel fund. (19) S. Treadwe-1 – The bieas ired Charu, trist i’.s Much of the inplamentat ion work described of Traffic in the UCL Node in tihe ARPA here ms done by Frank Deignan. The Stanford Network. UCL TR Report 33, DIcember group: Yogen DIlal, Judy Estrin, Dick Karp and Jim Matthis wi’,vr-,,’sntial collaborators in the.amwikemr’twit Ie., and Bi I I Plumaer w s nainly nrevnhsible for the. fiun inpla rntatton. IaalI Spillin of MW. was with ius for part of the nuw-munwtnt lw’riloi, aeix L,;sis.vl with definition of mw of tlim e’xprizmnts. Finally, Professor Vint Crrf, the head of tth, Stanford group at


what work does ups does on sunday

67 *7I “15 VII FACSIMILE ACTIVITIES 7.1 Overview The UCL Facsimile work is another activity which has reached fruition during the last year. The system of analog Facsimile device, use of Message Systems and the Data Computer have been completed. A paper describing this system is given in Section 7.2. The paper will be presented at EUROCOMP in London in May We have analysed the components of this system, and concluded what technological improvements are required to make such a system economically viable. This work is described in Section This is a paper which will be presented at the International Computer Communications Conference in Kyoto in September A key aspect of the economic viability of Facsimile as a method of inter-person communications is the cost of the communications transmission itself. We have therefore analysed the tariff structure of several of the PTT supplied data and telephone networks. We conclude that their present levels are rather unattractive for Facsimile. This work (KIRSTEIN 1978)-will appear in Computer Networks. Although these aspects of the Facsimile work are completed, the project is entering a new phase. Hardware has been developed to allow the terminal to be attached to X25 networks; the software for this purpose is practically completed. This work, together with the X25 network developments of Chapter 4, should allow the Facsimile terminal to be attached directly to an X25 network – including these versions of EPSS and SATNET. Other future aspects of the work being considered are modifications, hopefully being done by ISI, for incorporating some of the UCL features directly into forms of the TENEX message system and into the PDP 11s driving other XGP printers. We at UCL will also replace our analog Facsimile device by a digital one. We then hope to develop a very useful and replicable selfcontained terminal. I_

68 1 – PAPER *
the asooum thing is that fedex tracking number is not available at all