[Ilugc] is FOSS methodology structured ?

Ashok Gautham thescriptdevil at gmail.com
Wed Dec 17 20:55:16 IST 2008


On Wed, Dec 17, 2008 at 6:45 PM, Selvakumar Rajeswaran
<selva_rsr at yahoo.com>wrote:
>
>             So now, the Q'n is does FOSS adhere to formal Engineering
> practices ?

Case Study: The Hungarian notation
It is considered good etiquette by Microsoft programmers(They claim it
eliminatesbugs caused by typing). But the Linux coding standard takes a
completely different approach to this:

[quote]

Encoding the type of a function into the name (so-called Hungarian
notation) is brain damaged - the compiler knows the types anyway and can
check those, and it only confuses the programmer.  No wonder MicroSoft
makes buggy programs.

[/quote]
Also consider the confused plight of duck type language programmers who
use your API.

http://lxr.linux.no/linux/Documentation/CodingStyle
(It most certainly confuses me. F is Flag and not Float for example!)
Also, if the functions are actually well-written, they wouldn't exceed a
single screen. So I can figure out the type anyway.

            If not then for what other one or more reasons we should embrace
> FOSS ?
>             Iam raising this Q'n as a responsible Engineer.

Cathedral and the Bazaar explains this better. (ESR prefers calling
himself a scruffy hacker :P. I love his anecdote as to how he handled
Craig Mundie)

            Also I can tell that if FOSS donot adhere to any formal
> Engineering practices, then Iam sure that most of
>             the activities done would yield themselves redundant and that
> would spell an eddy-current in the dynamics             of economics. How
> can such a scenario be usefull for a layman user/consumer in every-day life
> ?


I see excellent roadmaps in FOSS projects. And much unlike Microsoft's
Vista's roadmap, which was expected to release in 2003 and came out in 2006,
most are almost on track.
Also the longest step in any SE process(maintenance) is best applied to FOSS
since if the lead dev quits, we have either a fork or a set of developers
taking up
the same project.(audacious from xmms)
Testing is comprehensive too.(Alphas, betas, RCs). This is too slightly
different
from the Cathedral style since we have nothing to hide from the people here.
So
we can involve them more in testing. Testers are always welcome at any
project.
The bug fixes/triage of most software is good.

There are two more [OT] reasons why we do not need to "spend 10 months in
planning and 2 days in execution"
1) The UNIX philosophy: KISS(Keep it Simple, Stupid). Every app does one
thing
and does it well. So we focus on the target alone
2) The development starts with a small subset and keeps adding more and more
(something like the spiral). Since most tools dont charge for updates, this
is
welcomed by the users.

---
Ashok `ScriptDevil` Gautham


More information about the ilugc mailing list