lauren@vortex.UUCP (Lauren Weinstein) (12/19/84)
[Technical sidenote: the reason that tuning for cabletext is so critical is that the ideal tuning point for receiving the megabit data streams for cabletext is often not the same as for receiving the best picture. One of the very tricky constraints of cabletext is the balancing act regarding the 0 and 1 video levels that are chosen (and must be tracked through all sorts of conditions) to avoid interference with the retrace on the main TV picture -- particularly on older TV's. If the video levels sent for data are not chosen very carefully, you get buzzing and even visual interference on the picture. The end result is that 0 and 1 are much closer together than one would ideally like, and the decoders have to be pretty sophisticated to track the fairly small variation between 0 and 1 (which tends to wobble with time) accurately. The custom IC's that are used for this are usually in a bizarre configuraton that continuously tracks both sync and the current 0/1 levels, and adaptively slices in the middle (on a changing basis) to try get the best data it can. Especially in the light of the damage that many cable companies do to their signals, this tracking gets pretty messy and it must always be done at full video speeds.] --- The current visualization is that people will post to the groups that move on stargate just as they do now to moderated groups. In fact, it is presumed that some of the moderated groups can be fed pretty much "as is" to stargate. Ideally, if the project grew, there would be more mail paths created toward those moderators. The moderators do the actual posting (mostly via direct calls) to stargate. So, to the extent that this technique is used, a given message only has to traverse a single mail path to the moderator and thence to stargate, not be broadcast all over the net by phone. Since any given site RECEIVES *much* more news than it itself generates, the high bandwidth receive, lower bandwidth send nature of this plan would seem to fit well into the traffic patterns we'd expect. --Lauren--
paul@masscomp.UUCP (Paul Cantrell) (12/27/84)
Lauren: I wonder if you would clear up a couple of points about how stargate would work? If we assume that not all people on the network will be connected to stargate for whatever reason, then it makes sense to me that all mail will be sent in duplicate: to stargate and also across the currently existing network. Thus, I really don't see how this would decrease the load on the current backbone machines. As for people who want a more efficient way of receiving news, it is true that they can elect to connect to stargate. However, since this is only a censored (sorry, moderated) subset of the news, I for one would not be satisfied with ONLY stargate; I would still want the other news too... Now if ALL of the news was going to be broadcast, this might be different. I have seen several people guess at why the news is to be 100% moderated. However, I wonder if you would post your reasons for this, straight from the horses mouth, so to say, as I feel it is one of the strongest arguments AGAINST the system. If it is indeed because there are standards for what can be broadcast, isn't something like ROT13 effective for this? If the reasons for the moderation is simply that the network is getting loaded with a tremendous amount of fluff (we all admit that this is true), then it seems to me that more moderated newsgroups are in order, but that this should not be tied directly to the stargate project. I am willing to be convinced otherwise, I just haven't heard your reasons why you think there should be such a strong binding between the distribution method and the control of content on the network. Anyway, I guess that what I am asking for is a clarification of how machines would connect to stargate in both send and receive modes, why restrictions on content are necessary if indeed they are, and most importantly: how all this will affect the rest of the network. Don't get me wrong, I think it is a very interesting experiment, and we are all interested in how it proceeds. I would just like to better understand what the impact is likely to be. PC
lauren@vortex.UUCP (Lauren Weinstein) (12/28/84)
I have discussed the reasons for screening netnews numerous times in the past (legal, technical, and "publishing"-wise) and I am not going to go over it all again. I'm getting very tired of repeating the same stuff over and over in this group, and I'm sure other people are also. As I have also mentioned numerous times, materials for stargate would ultimately be routed quickly (via mail) to moderators for satellite distribution. This will require some carefully designed software and routes, but the end result is that the satellite people will get the stuff by satellite and the non-satellite people will receive materials through conventional means, usually later. To the extent that moderators are screening conventional netnews materials (and not materials mailed directly to them, like mod.* groups are now -- this is the best model, by the way) it will help if the moderators are well connected to primary nodes so they get the news early. However, my own view is that the mod.* model, where moderators are mailed the materials, or where the materials pass to several moderators by mail (not by netnews), is the most accurate picture. Of course, these moderators would want to be well located to avoid mail delays, but many of the people who are currently volunteering for such tasks are already in such postions (topologically-speaking). It is possible that, over time, the satellite service would tend to split off from Usenet proper in some ways. For example, if people could submit articles to the moderators via direct mail (perhaps via an 800 number or some such, perhaps not) the moderators would then be even more in the position of magazine editors, and the service is a big electronic magazine, hopefully of high quality. People can still submit whatever they like, and most stuff (that fit the topic catagories and is of general interest) would get sent -- quickly too. There has never been anything like this before, where random people could submit items, that are usually accepted, for a satellite broadcast magazine running 24 hours/day. While such a service would not be suitable for the quick flame sessions that some people have now on Usenet, such people (and anyone else who wanted to) could of course continue to use Usenet proper. In other words, it is incorrect to view (ultimately) the concept of satellite netnews as a high-tech Usenet. Rather, it is another way for people to communicate. Presumably, even if the sort of scenario discussed above came to pass (and it is only one scenario) the moderators could be watching for interesting Usenet items to send off. Also, presumably, there would be gateways so that the ordinary Usenet sites could receive some or all of the satellite materials (albeit much more slowly than by satellite) if this were deemed desirable. The two services could co-exist in a symbiotic way, with each handling the sort of traffic that would be best suited for its mode of operation and sharing when desirable. This is all speculation of course. But it's a model that I think could result in a very useful information service with which many people would want to participate. --Lauren-- P.S. I'm starting to spend too much time generating messages like this. Please do not be surprised if you find me not responding directly to some of the netnews messages that continue to appear on this topic, with my messages only appearing occasionally for status reports. I have laid out the project as best I can for now, and just don't have the time for doing messages like this every day. Please remember that this is an experiment. I suggest that you sit back and relax a little bit to see what comes forth from the discussions among the various parties involved in the project, who are not under my control in any case. I'll keep the network informed of what occurs with the experiment and the discussions, which are in extremely preliminary stages right now. But I'm going to have to bow out of explaining this stuff over and over again, especially when it's so much in its infancy and I have infinite piles of TECHNICAL issues to deal with. I really can't handle the politics too at this point, and given the experimental status of the project right now, the politics don't really mean very much. If you have specific questions, feel free to send me mail, but I'm going to avoid spending the time (that I don't have!) to do too many more messages like this one. We may be getting a bit ahead of ourselves. Some of the discussion now would be like a biologist, having successfully cloned a worm, worrying about to which college he'll send his first artificial human for post-graduate work... --Lauren--