#apertus IRC Channel Logs

can someone tell me what the Execute() function in the CMV12000Adapter expected to do?

But I am not sure about this because I guess the iteration should be done in the IDaemonModule itself, if required.

morning folks!

Good Morning!

hi siddhantsahu, first i would suggest to work on your challenge code to be competitive, as i see a lot of problems there

also monolithic approach is not impressive

it's not only formatting, the code is not really following OOP and is monolithic in many methods

Please give me some more context why it is monolithic.

your methods are too big, make too many things and nested too deeply

nested in terms of unnecessary nesting of if/else and loops

haha ok will work on it

it's up to you, i just have to grade

asap, i wanted to check on your abilities, but current state of the code is not reflecting our expectations, so i'm doubting at the moment that you are ready for the gsoc task

get over at 6, will work on after that

Also if I finish studies early today will work on it late in the night

hey BAndiT1983. do you think my c++ challenge is ok or should I modify it? I tried to implement my own solution for debayering but I know its very lengthy. I also understand I have to work according to requirements of the organisation so my code should also be satisfactory to them.

hi rohan_, if you are referring to the previous discussion today, then it is merely about seeing what the student can achieve, as he wanted to make his own task, but as it is rather advanced his code is not up to it

Nevertheless, as asked so would to improve it

No No my professor don't like to write code.

adding it on april 14 is a bit late, so i have to check it with other mentors, as i haven't seen the history before

At first*

still can't understand why everybody is focussing on the 12bit raw data and not on the proper IPO pipeline with correct conversion of data to either 8-bit or 16-bit, like we have discussed over and over

and you think you are most qualified for GSoC then?

umm will try to understand

Is it a better option?

let's calculate PI by using the LSBs as random bits and squaring the area of the circle

I mean for 12bits to 8bits conversion, instead of removing the LSB, I was thinking for diving the each 12 bits pixel value at 4095 and storing it as floating point numbers with range 0 - 1. Now we can scale it back up by multiplying each floating point pixel value numbers with 255

Feel free to criticize if I am wrong

Throwing away the LSB and doing what you are describing results in exactly the same values

well depens on how you convert from the scaled floating point values back to integer

but you can do the "correct" rounding without using floating point numbers

but i wouldn't go that far for challenge task

either omitting the last 4 bits or using plain 12bit value is okay, as it is more about the code quality then the image itself, although the image should also be correct

*than the image

Hi

I wanted to know which IRC client you use to so that you don't miss messages when offline. Matrix servers are down, so my messages aren't being sent :(.

the main "problem" is that to get all messages you need to be always connected

yes because matrix had a backend server that was always connected

you can get something like that by using quassel or BIP but you will have to run the server yourself

there is irccould, but it is paid

That would be great

did you get my pm's ?

well then give me your email address so i can give you your username and password

that username will only be used to login to the quasselcore

ok thank you so much :)

if I can have the nick without underscore , I have no problem

Thanks vup2, It is working now :)