[fosscomm] Multiple standards [was DEADLINE [...]]
prabirp at gmail.com
Sat Jul 11 19:15:54 PDT 2009
2009/7/12 Pranesh Prakash <pranesh at cis-india.org>:
> On Sun, Jul 12, 2009 at 00:36, Sunil Abraham <sunil at cis-india.org> wrote:
>> On Sat, 2009-07-11 at 22:08 +0530, Pranesh Prakash wrote:
>>> On Sat, Jul 11, 2009 at 21:38, narendra sisodiya
>>> <narendra.sisodiya at gmail.com> wrote:
>>> It isn't about which standards are chosen. India hasn't gone down the
>>> standard eGIF path where the standards themselves are specified. It
>>> has instead chosen to go in for a policy first and then do the
>>> selection of standards separately.
>> I think it is a parallel process.
Yes, Pranesh is quite right, things are moving in parallel. Thanks
Pranesh also for enlightening me on the patent status of GIF. That
brings up immediately that in this case PNG emerged as a parallel
standard for people who wanted to move away from patent protected GIF.
It was not about a different domain. Now they both exist, what should
be our stand and should we now accept both standards? One position
could be since both have some advantages in specific areas, accept
both. The other have only one standard and improve it to take care of
areas where it might have some deficiencies or live with it. I would
be inclined to the second for two reasons. One is that historically,
the second standard arose because of patent protection of the first
and therefore need to support standards that were created with open
objectives in mind. The second reason is as I have said, once I accept
teh concept of multiple standards, could be the thin end of the wedge.
> Well, given the varying levels of domain specification (sometimes as
> broad as "mobile access" and "graphics", and sometimes as specific as
> "Data Transformation for conversion from XML schema format to another
> format"), it is inevitable that multiple standards will have to
> sometimes be specified. For instance, under graphics, there would be
> one standard for scalar graphics and a different one for vector, thus
> leading to a situation where multiple standards may, legitimately,
I am a little skeptical of the domain issue. If I am clever, I can
always create a domain in a way that I create a domain for my
"standard" rather than evolve a standard for a defined domain. Though
I do grant your point that there are overlapping domains and therefore
there could be reasons for having multiple standards, which would
cover some of these common parts of the domain.
To give an example that supports your position (and not mine), in
network communication protocols for example, we have CSMA/CD (IEEE
803.2), token passing ring (IEEE 803.5) and token passing bus (IEEE
803.4). They could be defined as the same domain or different domains.
So you have opened a contentious issue -- can we really avoid multiple
standards, and if not, will not some clever people use this to get
parallel standards. What this means is that even if we win the current
battle, the war will certainly continue!
> network mailing list
> network at lists.fosscom.in
More information about the network