schnuller wrote:I thought the question was what software you already did write, not what software you are planning to write.
And I thought that arguments, especially technical arguments are speaking for themselves - it's You who have a problem here, so You are trying to qestion my competence, because You can't understand the arguments nor the portions of code which I've posted here. So I could ask : are You just trolling here?, but I'm not going to say this
(well, I've said this, sorry)
Anyway, since I don't have google+ account nor Facepalm account, I can't give You a link to my profile, and for sure I will not show You my CV (You wouldn't understand most of it, cause there are mostly technical terms).
However, I may tell You that I'm working embedded systems/software in the area of factory automation (servo controllers, PLCs, etc).
In case if by "embedded system" You can only imagine toys like RaspberryPi, (what is rather common picture among laics), then I'll surprise You: "a toy" like S7-319F PLC is build on 3 independant CPUs: TriCore, PMC E9 and ERTEC400 (dedicated ethernet controller with integrated ARM core for handling Profibus and ProfiNet stacks).
You won't find this on Google, but You may try...
Oh, and I wrote a reply to a quiz here:
http://forums.debian.net/viewtopic.php? ... iz#p510648
schnuller wrote:You can stop searching for that man: There are people who see it like that.
And that is exactly why i said in another thread that not all problems related to systemd relate to the quality of it's code (You called it nonsense, btw).
I don't think so - most of negative opinions I've seen were lacking techical background, so they could be easily questioned as being not objective.
................
Normally this should be end of this post, but I do uderstand that You believe in some magic or some super powers owned by developers who are working for big-name companies. Well I'm going to show You something:
Some time ago, when I was writting wxEDID (a tool for modyfying EDID data under GNU/Linux systems) I've discovered a really funny thing:
Video Electronics Standards Association (a big name, isn't it) have failed to understand the structure of Manufacturer_ID. In both official documents and on a Wiki page, this 2-byte structure is presented as a
obfuscated set of bits. Even
Linux kernel has wrong implementation of this struct (written by INTEL and REDHAT):
line 1316:
http://lxr.free-electrons.com/source/dr ... drm_edid.c
Wiki: (bytes 8-9)
https://en.wikipedia.org/wiki/Extended_ ... ation_data
What is really funny here, is that this struct is encoded as Big-endian, while the rest of EDID struct is encoded as Little-endian.
Nobody for those years which have past from the day 0, when the standard was settled have noticed this.
And the struct looks like follows:
Code: Select all
struct {
uint letter3 :5;
uint letter2 :5;
uint letter1 :5;
uint reserved :1;
} mfc_id_t;
So, instead of shifting, or-ing and mixing bits, it's just sufficient to just swap 2 bytes of manufacturer id and read those 3 letters directly.
How could this happen? Big Name companies are hiring geniuses right?
Who would be that stupid to mix data fields with different endiannes in one structure? In a specification for a standard?
This is how it's possible:
In big companies NOBODY GIVES A crap - their employees just can't wait for a launch time.
(btw, this field was forced by ... gues who?
Microshit.)
Regards.