Connectivity Providers Database

From: shem@minster.york.ac.uk To: kenfri-net@Athena.MIT.EDU CC: pmwanik@cadev6.intel.com Date: 23 Apr 1992 00:26:10 GMT Friends, You probably remember that mid last month we suspended kenfri-net discussions partly because of my impending trip to Nairobi -- in the hope that certain things could become clearer during the trip. Part of my/our terms of reference were to address issues like: >1. Who are the potential nodes? 1.1 Who has what? -- equipment -- existing network/s -- personnel 1.2 What they _intend_ to acquire (along similar categories as above) 1.3 Who needs what? 1.4 Is there a willingness to co-operate, or what is their level of interest in and/or committment to such a project? 1.5 Who is/are the contact person/s? --- I must say from the offset that I was not able to visit all centres that appeared in the proposal for various reasons. However, I managed to visit most of the major players, and has a sound basis on which to make some recommendations. The intention is therefore not to individually answer the queries above. That shall be done in the full report. However, they were some of the points considered in arriving at my conclusions. The revised AFRINET proposal should be going to Digital in about 4 weeks time. By that time, we should have exhaustively discussed our ideas, and kenfri-net (as an ad-hoc committee) should be ready to disband. As far as KCI is concerned, most of its deliberations will be subsumed by the Networking Comm. And since some of these discussions will be of interest to that Comm anyway, am cc.ing this message to Peter Mwaniki to share with the rest of that committee. Some quickies: 1. After the trip, I have re-read the original AFRINET proposal, and concluded that it was no doubt a great piece. (I urge you to refresh yourselves with it; if anybody doesn't have a copy, please ask Gachora -- gadj@athena.mit.edu). However, a few things may require revisiting, in the face of new info, and/or more recent developments. 2. I had lengthy and fruitful discussions with Dr Alex Tindimubona of African Academy of Sciences -- the co-architect of the original proposal. We came to at least two conclusions: i. The Academy must play a more pro-active role, and ii. Intra-continental communication should develop hand in hand with North- South communication. Initially, it was proposed that exisiting institutions with established infrastructures (eg UoN ICS, University of Zimb, etc) should constitute phase-1 nodes. AAS was simply to play a "catalystic" role. Uni of Zimb, for example, should have served as the "main" Zimb node, UoN the "main" Kenyan node (communicating directly with London or US), etc. We felt that this should be reversed (for reasons that will be clear in my full report). What we preferred was that: i. Only one major African node directly handles North-South traffic. However, depending on technical requirements and pragmatics, a second "major" node could be explored (eg. one major node in Nairobi, another at Rhodes!). ii. That AAS be this major node, and that it feeds "national" nodes directly. My personal opinion was that national computer societies are the more ideal to serve as national nodes. However, in the shorter term (or where they do not exist), universities or other research institutions could "hold-brief". I further felt that even the main backbone of the continental traffic should be "handled" by a continental _computer_ body, in the longer term. In other words, I would (in the long run) favour an association along the line of the recently formed African Regional Computer Confederation (ARCC) replacing or helping AAS in the above role. Although kenfri-net may limit itself to the Kenyan situation, am sure many of the ideas we'll discuss will be applicable in other situations. 3. We felt that in the shorter term, not too much may be needed in terms of equipment/resources, except: i. a suitable machine each for AAS and KCI (and any other selected "national" nodes). ii. modems for nodes and "points" which do not already have them (if Moi University is connected to the national node, am calling it a "point". It may, however, also serve as a node, eg. by feeding Maseno campus). iii. phone bills (which can be horrendous) and suitable personnel/training, and other day-to-day administrative costs. For many of the institutions that I propose to be considered for the first phase, we may not have to start from scratch as far as training is concerned, afterall! Those sites (ie nodes + points) which do not have VaxNotes yet, and have suitable machines may require such software. (On appropriateness of platforms, and other technical details, I urge you to read recent notes on BESTSD's AFRINET/KENFRINET accounts. I hope you are all now VaxNotes experts!). However, even basic email facilities will still make substantial contribution toward realization of AFRINET ideals. 4. To start off this discussion, I am cross-posting parts of our recent discussions with Mike L. and/or Beryl (I hope none of them is holding any copyrights!). Some of them may be technical (hold your breathes -- but thank God this group has KCI's best, as recently demonstrated with Banyans!). For example, the X.25's and X.28's (Mike what are the x28's?): recently in Nairobi, I realised many people/organizations were declining from using them 'cos of their "running" costs. People tend to prefer ordinary phone lines (see our exchanges with Mike/Beryl elsewhere). The last day before coming back, I spoke to the Kenpac boss -- and he was quite positive. But let me not jump the gun. 5. Before leaving for Nairobi, Ayisi asked me to obtain a copy of the Kenyan current network topology -- and I did precisely that. Unfortunately, documents and "actual" topologies will always be out of phase -- but I hope to receive a more recent topology from Doug Rigby of ELCI or from Oscar Mann of Afrinet project (a different Afrinet!!). For those of you curious about this other "Afrinet", am separately posting a copy of its proposal -- I only received an electronic version of it in a diskette from Oscar today. Essentially, there are two nodes: ELCI and ACTS (for African Centre of Tech Studies). Most of the sites appearing below are/will be fed by one or the other of these two. (Then there are the "independents" like KEMFRI/RECOSIX, but since Peter Reyners is on this mailing list, we can "hear it all from the horse's mouth" -- Peter, how sad it was we couldn't/didn't meet!). Again, I'll also try to post some info which tells who feeds who. 6. I visited most of these sites, and there are various reasons for my proposal that they be part of the original AFRINET: (Order insignificant) UoN KEMFRI ACTS ICIPE/AAS ELCI IDRC KENGO All-African Churches Congress KCI HealthNet WHO/HLM AMREF KEMRI Kenya National Library Kenya Posts & Telecoms UNEP Moi University Oscar Mann Ministry of Health Ministry of Technical Training & Applied Tech Regional AIDS Research Centre (CEDHA, Arusha) A few comments: i. These are obviously skewed towards Kenya, but there are a few regional institutions (eg CEDHA, AACC, etc), and international agencies. The latter have a great funding potential, and like the others, are very much committed to email. Met a couple of good guys at IDRC, UNEP, and WHO/HLM (Health Learning Materials). ii. You may realise there is a bit of bias towards health institutions; this is one area emphasized in the original AFRINET proposal; besides, am simply trying to put "our hands where our mouths are". However, I have sampled a diverse "team" nonetheless -- NGO's, state corporations, ministries, unis, etc. iii. The University of Nairobi (UoN), in particular, had loads of depts (including EE, Survey, Economics, Physics, Economics, Civil Eng, Commerce, Building Econ, etc), Institutes (ICS, IDS, Housing Research & Devpt, etc), and the Library which were particularly keen on this project. UoN's main campus services other off-shore campuses (eg Kikuyu), and other universities (eg Kenyatta), and (technically) would not be very difficult to network. It (UoN) may, however, require a mini-project for itself! iv. Oscar Mann, Doug Rigby and ACTS folks will be useful, least for providing local expertise. As I understand it, these folks currently keep existing systems running, so we have a lot to learn from them. v. There is no way we can leave out KPTC (or rather, Kenpac) from all this; in fact it is them to spearhead this course. (Who knows, they may even infuence the relevant policies, and make x2? communications real!). vi. In all, there are currently about 30+ emailers in/around Nairobi, including NGOs, Unis, private sector, individuals, etc. Most of them are members of Kengele, a group co-rdinated by Mazingira (with the help of ELCI, ACTS, Oscar and others). Doug has kindly agreed to liaise with that group (and our Networking Comm), to create kcik-net -- as a Kenyan-based sub kci-net. ^ Mike Jensen might want to tell us more about his experiences with these groups! BTW, are there any spanner-thirsty kciers who could hold screws and bolts if they were to be called upon to do so? There shall be a great need/demand for fundis; existing engineers alone will not be able to cope. Any guru who wants to make a 2-4 week holiday to Kenya, tie up with Peter M. -- with of course, your CV. 7. The following could be approached for some co-sponsorship: UNESCO & other UN bodies (eg UNDP) AAS/ICIPE Kenpac Ministries (of Health, Sci & Tech, and Technical Training) Commonwealth Secretariat IDRC, FORD Foundation, etc The Private Sector (too important to ignore!) Most of the existing networks are Fido-based. We cannot afford to ignore APC in all this, so Mike J. and Randy, I leave the floor! Shem