If "Master/Slave" terminology in computing sounds bad now, why not change it to "Dom/Sub"?
If "Master/Slave" terminology in computing sounds bad now, why not change it to "Dom/Sub"?
It sounds way less offensive to those who decry the original terminology's problematic roots but still keeps its meaning intact.
You're viewing a single thread.
For IDE drives, Master/Slave is both correct and describes properly the functionality.
Only one device can talk on an IDE channel at a time (one IDE ribbon cable is one channel). The Slave Drive requires the Master drive to be able to connect to the controller. If there is only one drive, it must be designated the Master drive.
We don't share multiple devices on a single channel anymore - SATA, PCI-E, these techs have only one device per channel (or only a certain number of channels dedicated per device).
The old Master/Slave system was a hack to get double the IDE devices connected per controller channel.
We don’t share multiple devices on a single channel anymore - SATA, PCI-E, these techs have only one device per channel (or only a certain number of channels dedicated per device).
Right... This desire to change verbiage on a dead technological concept is kind of stupid. I taught in an R1 institution. This topic DOES come up. And nearly verbatim it was "While it's unfortunate naming convention existed... we simply don't operate that way anymore." It was never a problem.
All this other shit people keep bringing up like git branching... "Master" was a shit name for the main branch, and "slave" doesn't appear at all. So that's not even relevant and I simply don't understand this aversion to words just for the sake of manufactured "hurt".
Others are saying HA... Except that Master/Slave doesn't describe any HA I've worked with. Nor the terms were used in any of them.
I think I could argue that master/slave doesn't describe that functionality. Does the master drive control the slave drive?
I think a better example is the SPI bus, which has a one-controller-many-peripherals bus topology with two data lines often referred to as MOSI and MISO: Master Out Slave In and Master In Slave Out. (in addition to a clock line and one or more chip enable lines) In this case the controller does literally control the peripherals, which aren't allowed to put data on the bus unless commanded to. Newer documentation is using the terms COPI and CIPO, for Controller Out Peripheral In and Controller In Peripheral Out. Personally I prefer MOSI and MISO because there's a definite way to pronounce them; how do you pronounce "CIPO?" See-poh? But it's something for someone somewhere to be uppity about so sure let's expand the glossary.
The activity light on the Master drive will light up in sync with the Slave drive when accessing data on just the Slave drive. At the end of the IDE lifespan, there was a movement to put the names as Primary and Secondary instead. It doesn't really describe the relation to how the hardware works, though.
I mean that does actually work better. If there's only one, it has to be the Primary. If there's two, one of them has to be the Primary and the other one is the Secondary. That makes more sense to me. Because neither drive is sending the commands here, the IDE controller either built into the motherboard or on an expansion card is sending the commands, that's the "master."