Found a problem when not all nodes had the TDMA interface, and solved it.

[erlender]

Removed actual node id from slot allocation, to allow some nodes to not have the TDMA interface present and still allow the TDMA cycle to work. Introduced a simple iterator named slotUserId instead. It's the MAC pointer that is needed for the TDMA cycle to be successfully run, not the node id.

[knud]

Hey. While I get the issue, without revisiting the assumptions in the simulation, I am not sure that this is a general solution, so am reluctant to merge. That is, the slotUserId is just another kind of constraint that another user of the sim may not like and have to undo. Think I'll leave this unmerged...

[erlender]

Hi, I understand your reluctance, and it could be a better way to fix this. It just became a showstopper for me, and I spent several hours trying to identify the problem, so I thought I would propose the fix for all. The slotUserId is just a continuous counter to ensure that the time slot allocation table correlates with the created TDMA interfaces. Erlend From: Steven Knudsen [mailto:notifications@github.com] Sent: 22. juni 2018 18:10 To: knud/simple-wireless-tdma Cc: Larsen, Erlend; Author Subject: Re: [knud/simple-wireless-tdma] Found a problem when not all nodes had the TDMA interface, and solved it. (#2) Hey. While I get the issue, without revisiting the assumptions in the simulation, I am not sure that this is a general solution, so am reluctant to merge. That is, the slotUserId is just another kind of constraint that another user of the sim may not like and have to undo. Think I'll leave this unmerged... — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#2 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFCkfy4c1NDQh3PF1vNAxlevP5FyFH0Sks5t_Ra7gaJpZM4Uzuk1>.