Oddly enough I think that actually adds to the problem. Because there are so many Unix-like OSs and each is slightly to significantly different, the solution to a problem on one may or may not work on another.
I liken it to SQL, having worked with many different sql dialects, I can never recall what functions are supported by one versus another. However, I don’t run into that problem with Mongo, cause it’s so different.
In my experience, if you have an issue on a Unix-like OS it’s almost always possible to diagnose and fix the problem. There is almost always a log file with extensive information. Unices are built from collections of simple tools so it’s often easy to find what exactly is going on and to isolate the problem for an easy fix.
Windows on the other hand is a black box. If something doesn’t work, you’re screwed. Error messages are often cryptic to the point of being little more than a 432 digit number. Most subsystems are monolithic monstrosities and isolating an issue is a PITA. Troubleshooting usually isn’t much more than randomly changing things and see what happens, or googling the error and hope someone else found a fix. Even if you manage to fix it, it’s often unclear what the actual issue was. It kind of reminds me of the magic/more magic switch.
Oddly enough I think that actually adds to the problem. Because there are so many Unix-like OSs and each is slightly to significantly different, the solution to a problem on one may or may not work on another.
I liken it to SQL, having worked with many different sql dialects, I can never recall what functions are supported by one versus another. However, I don’t run into that problem with Mongo, cause it’s so different.
In my experience, if you have an issue on a Unix-like OS it’s almost always possible to diagnose and fix the problem. There is almost always a log file with extensive information. Unices are built from collections of simple tools so it’s often easy to find what exactly is going on and to isolate the problem for an easy fix.
Windows on the other hand is a black box. If something doesn’t work, you’re screwed. Error messages are often cryptic to the point of being little more than a 432 digit number. Most subsystems are monolithic monstrosities and isolating an issue is a PITA. Troubleshooting usually isn’t much more than randomly changing things and see what happens, or googling the error and hope someone else found a fix. Even if you manage to fix it, it’s often unclear what the actual issue was. It kind of reminds me of the magic/more magic switch.