| swordsmanz | aeoril experement | 00:38 |
|---|---|---|
| aeoril | swordsmanz I am sorry, I do not understand | 00:39 |
| aeoril | swordsmanz expirement? Try different books/resources myself? | 00:40 |
| swordsmanz | aeoril no experenment with the kernal | 00:40 |
| swordsmanz | im asumeing that you know how to compile the linux/bsd/mach kernals ? | 00:40 |
| aeoril | swordsmanz no | 00:40 |
| swordsmanz | well you know how to compile other programmes? | 00:41 |
| aeoril | swordsmanz but that I think should just be a matter of following recipes. I am thinking it might be best to learn some of the theory behind OSes before I dive into the practical realm | 00:41 |
| aeoril | swordsmanz yes, I can compile other programs | 00:42 |
| swordsmanz | well compileing a kernal is no different | 00:42 |
| aeoril | swordsmanz I'll give you an example of why I was thinking of pursuing theory and taking a more academic approach at first | 00:42 |
| swordsmanz | and no its best to just learn, the hardware will have manuals for the specific proc types and stuff, and its a matter of translateing that into code that will run well when compiled | 00:43 |
| aeoril | Some years ago, I worked on a problem in a low level system that involved needing to send data across a bus. I could not use a semaphore because there was no way to coordinate with something like that - it was to disparate systems existing across a bus. I did my own FIFO routines to accomplish this communication. I tested it it tortuously, and thought the theory behnd my FIFO routines | 00:45 |
| aeoril | was solid, but years later, on reading Tanenbaum's book I learned that the way I did it was not right - it was not a lock-tight, solid implementation and could have sporadic errors, however unlikely | 00:45 |
| aeoril | So, I found in his book the perfect algorithm that I needed for this specific problem. I could never have come up with that algorithm on my own (most likely) and that particular algoritm if I remember correctly had come along somewhat late in the grand scheme of things even for the experts in OS development that had worked without this algorithm for years before | 00:47 |
| swordsmanz | yeah but there is always a better way to do something | 00:48 |
| aeoril | So, I wanted to learn some of these things to get some of the theory internalized so I could know better how to do OS and kernel development before diving into the practical side of htings | 00:48 |
| aeoril | swordsmanz so, you would think looking at kernel code, etc. might be more useful? | 00:48 |
| aeoril | modern kernel code | 00:49 |
| aeoril | experimenting with it, playing with it, reading it, understanding it ... | 00:49 |
| swordsmanz | i think that unless your comfertable makeing small tweaks and compileing your not gonna find it simple to do more complex stuff | 00:49 |
| aeoril | My idea was (1) theory (2) small tweaks and experimentation (3) go from there to more complex things - baby steps. But I wanted to get some of the theory and stuff I would normally get in university done first, or maybe alongside, the baby steps with the actual linux kernel, modules, etc. | 00:50 |
| swordsmanz | hmm in my experiance theory is opnly really good for referance | 00:51 |
| aeoril | swordsmanz I am sorry, my wife is calling me to eat - I will be back in a few. Thanks for the help | 00:52 |
| swordsmanz | kk | 00:52 |
| aeoril | swordsmanz back! | 01:09 |
| swordsmanz | YAY | 01:09 |
| swordsmanz | its proberbly not a good idea to listen to my thughts on this anyway as im just getting my head around the basics of how the ernal is structured, even tho i have been compileing my own for a while now | 01:10 |
| aeoril | swordsmanz I was thinking there is great value in what you are telling me. I think it would be best to work through some of the basic theory behind operating systems, but as I do so learn to compile, alter and test the kernel and find examples of the theory as I learn in the kernel and experiment with changing it, breaking it, fixing it, etc. to best learn not just the theory but how | 01:10 |
| aeoril | it relates to real implementations in the Ubuntu/Linux kernel | 01:10 |
| swordsmanz | aeoril yeah that sounds sensible | 01:11 |
| aeoril | This will also give me a less "dry" path and allow me to more quickly assimilate into the community to start doing minor helpful things, like bug triage, etc. and get my toes wet | 01:12 |
| aeoril | First my little toe, then my other little toe, then go from there ... :) | 01:12 |
| aeoril | swordsmanz I found some other resources that recommended Tanenbaum's book so I went ahead and bought it. I am now reading it. Wish me luck! | 01:13 |
| swordsmanz | im only really here becouse there is a powermanagment bug on amd apu's that is really realy iritateing, and i want to see if i can have a go at fixing it becouse its bugging me :S | 01:14 |
| swordsmanz | GOOD LUCK | 01:14 |
| aeoril | swordsmanz ty :) | 01:35 |
| === FreezingAlt is now known as FreezingCold | ||
| === Guest47523 is now known as maxb | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!