roode@orc.olivetti.COM (David Roode) (06/19/88)
A variable which significantly affects reliability of the slide-lock is the weight of the transceiver cable. Some manufacturers have much heavier cable than others. Some (DEC) make an extra-light cable for use when the run is not too far. Also, right angle connectors are available from some transceiver cable manufacturers (DEC again). I second the comments for strain relief. Even the alternative use of screws and nut posts is subject to mechanical damage due to the vulnerable way these jacks are often located, and strain relief via an offset bar or other means would be a good idea regardless of the use of slide locks. The slide locks are excellent when used to couple two transceiver cables together. In many cases the built in thin-net transceiver makes operation without external transceivers over thin-net more desirable than using the 15-pin transceiver port anyway. For the price of 10 transceivers, a multiport thin repeater can be purchased. This will allow operation of 8 independent thin segments with a length of 180 meters each, fanning out from the multiport repeater. On each segment, up to 31 thin transceivers may be located. Holding each segment to 5 or so stations yields a compartmentalized network which is easy to debug, and robust in the event of a disturbance to the (now thin rather than thick) coax.
darrelj@samsun.UUCP (Darrel VanBuer) (06/21/88)
Slide locks are not perfect, but when correctly implemented do not cause much trouble. Sun's violation of the 802.3 spec (and the more implicit specs in the orginal Ethernet document -- where you have to read the connector manufacturers' catalogs to get all the rules for correct adjustment) is a major headache. By putting a chassis between the connector flange and the slide, they steal about 1.5 mm of pin insertion out of a total of 3 mm pin insertion, leaving the connector quite wobbly. At one time, we dealt with this by modifying the drop cable to remove two washers which the spec says SHOULD be between the lock posts and the connector flange, resulting in a cable end which complements the Sun error, and gets a good fit. Of course this cable now fails to properly fit a conforming computer or extension cable :-(. When properly fit, a slide lock is quite strong. With poor QC or design, they are abysmal.