olivier |
Nov 21, 2009 00:18
Status update:
- still waiting for some stuff.
- while I originally thought I would quickly finish it and dump it on github, I'm fascinated by the direction it is taking, and polishing it and thinking about the "how could people get one?" bit. learning a ton of stuff.
- i have started a far more ambitious thingy in parallel, using the couple of skills I learned on this project (current stage: evaluating several technical solutions, and thinking...).
- originally based on open-source stuff it totally outgrew, not relying on it anymore.
- will do over the week-end an important redesign of 2 parts of the project, in order to make things leaner and even more "minimalistic". The game will be to remove as much stuff as possible while keeping things working. And reducing the "Rainer Buchty ratio" of the thing, which for now is super high.
- 10k lines of google-style C++, 4k of which generated by 1k lines of python, the rest mostly in templates (wtf???).
- dumped new sounds in the usual directory.
- name of the project: shruti-1.
- project does not involve any sample.
- small, minimalistic. Lubitel, not D3s.
- you're going to hate me for wasting my time doing this :)
 
|
JT |
Nov 22, 2009 15:50
olivier wrote: Status update:
- you're going to hate me for wasting my time doing this :)
... HATE??... it is a bad word... keep at it Oliver...
... maybe there are some of us just waiting... huh
|
kosotodeluxe |
Nov 23, 2009 06:35
JT wrote:
... maybe there are some of us just waiting...
Stick of gum anyone?
|
drewzle |
Nov 23, 2009 07:47
*lights a 'cigarette' and loiters*
|
JT |
Nov 24, 2009 14:46
... coffee?... biscuits?...
... trraak trraak trraak...
|
olivier |
Nov 25, 2009 22:55
- long week-end! redesign of the 2 parts that I was not satisfied with. went fine.
- found a bug I could not explain - it never looked like a bug to me until I redesigned the section in which it occurred. Once solved, makes the thing sound even better. looks like I will need to record new samples... woooow!!
- the part that doesn't depend on me is taking more time than I thought, let's see if I can work with someone else to make it happen faster...
- working on "hackability" - making it easier for other people to tweak, modify, replace. One of the "features" I am the most proud about is the readability of the code - making the code descriptive, and getting rid of hacky constructs that even I wouldn't understand in 3 months. For example, I have made a couple of changes this week-end to parts I haven't touched in 2 months, and I was really pleased to find out that I could immediately dive into the code.
I would surely love to work on stuff like that full-time, but hey, I also have a lot to do at last.fm!
|
JT |
Nov 28, 2009 06:38
olivier wrote: -
I would surely love to work on stuff like that full-time, but hey, I also have a lot to do at last.fm!
... the shop quoted Singapore $70 to repair the Zire 72...
... but I think I should wait for the new platform and then put the $70 into that...
... the present Tungsten T2 I have is too weak...
... keep at it Oliver... you can do it... we know...
|
olivier |
Nov 28, 2009 10:35
JT wrote: olivier wrote: -
I would surely love to work on stuff like that full-time, but hey, I also have a lot to do at last.fm!
... the shop quoted Singapore $70 to repair the Zire 72...
... but I think I should wait for the new platform and then put the $70 into that...
... the present Tungsten T2 I have is too weak...
... keep at it Oliver... you can do it... we know...
A new platform? For $70? Running something like Bhajis Loops?
Never heard of that!
|
olivier |
Dec 04, 2009 15:44
tj_mothy and drewzle made the best guesses.
I originally experimented with an arduino, and decided to go with my own board design/dev libraries later, but it's still using the same chip (ATMega328).
|