@h[Future Terminal Concentrators] The ERCC is proposing to install @b[Camtec PADs] instead of new PDP 11 concentrators and start a conversion process which will mean that our terminal handling protocol ITP will be replaced by the international standard X29 (or the variant TS29) over a period of two to three years. This will bring us into line with other British Universities where the @b[Camtec] @b[PAD] is widely used. There are a number of reasons for making this change, connectability of new hosts, cost, emerging standards for micros and the ability to connect a @b[Camtec PAD] to a number of communications media. @s[1. New Hosts] As some users know only too well; when a new host is connected to the network with standard X25 software, users on TCPs have to communicate through a protocol converter @b[(XGate)] in order to communicate with the host. This procedure is wasteful in resources, slow and complicated. @b[XGate] is only just capable of supporting the current load and will be completely overloaded when new developments such as the Library, the Administration's IBM and the proposed Data General machine are attached and are heavily used. The only solutions are to insist that @b[every] host that connects to our network talks @b[ITP] (extremely expensive and no longer practicable) or that the network uses @b[X29] as the terminal protocol. @s[Cost] The cost of our concentrators has been increasing as a result of the rise in the dollar. We started to produce a second generation TCP which it was hoped would be cheaper as well as better. Rising prices have beaten us on that; meanwhile the @b[Camtec PAD] has got faster and more flexible. @s[Emerging standards for Micros] The current method @b[(Xtalk)] of connecting micros asynchronously to the network is not entirely satisfactory. There is now an Academic Community standard proposed that will allow micros to connect asynchronously to the network as if they were connected synchronously. This will mean that they will have easy File Transfer access to both our local hosts and to hosts on JANET etc. @b[Camtec] have been given a contract to implement the software in the PAD, but there is no chance of our being able to fit similar software into our current PDP 11s. @s[4. Different Communication Media] @b[Camtec] will be producing an ethernet interface later this year and by the middle of next year we should be able to use @b[Camtec PADs] on our ethernets. To attach our PDP 11s to ethernet would cost #3000 per TCP, plus considerable reprogramming effort. @s[Disadvantages] There are, however, a number of disadvantages for EMAS users using PADs as against TCPs. They are:- @b[1. Prompts] ---------- The X29 protocol does not allow for prompts to be suppressed when a user types ahead. Although EMAS will still be able to suppress prompts under some circumstances, a number of unwanted prompts will appear. A user who does not type ahead will not see any difference. @b[2. Screen Editing] ---------------- The X29 protocol will not support screen editing as used currently in Screed. It may be possible to change VECCE so that it will work on a PAD. The recommendation to users who wish a screen editor like Screed will be to buy a BBC SSMP @b[(Simple Screen Management Protocol)] will be produced that should allow fast, efficient and more powerful screen editing than is currently available. @b[3. Interrupts] ------------- The @b["Int:"] message will no longer be generated on pressing the escape key. It will be necessary to press @b["ctrl P"] then @b["B"] on a PAD terminal, or @b["Break"] followed by @b["b"]. @b[4. Xtalk] -------- The current version of Xtalk will not work on a PAD. From investigations that we have carried out, it should be possible to extend Xtalk so that it can be handled both on TCPs and the PAD. That is the present intention. @b[5. Tabs] ------- The Tab handling on a PAD is not as flexible as on a TCP in that only a simple tab stop 'distance' can be set. @b[6. Dial-up] ---------- More work is needed to get dial-up support working on a PAD although this will not actually be necessary until TCPs are phased out in, at least, 15 months time. We are discussing the problems with @b[Camtec]. @s[Advantages] In addition to the advantages listed at the beginning, there are a number of advantages that the average user will see from being on a PAD. 1. Every line can be run at speeds up to 9600 bit/s. 2. Each terminal should get a better response as there will be fewer users sharing a switch port (16 max as against 48 max) and the X29 protocol means fewer network packets per host interaction. 3. Access to remote hosts (on JANET) or local hosts using X29 will be much easier, faster and more reliable than at present. 4. The PAD is more flexible in terms of the number of printers, plotters etc that can be connected (up to 15 on a single PAD). 5. Calls can be made directly from one port to another - useful for micro to micro communication for micro use. B. A. C. Gilmore 4th June, 1984.