On Tuesday 03 August 2010 16:00:01 Vickram Crishna wrote: > ["India is so big that you risk being satisfied with your internal > market. Don't. The world needs your device and leadership. Your > tablet is not an "answer" or "competitor" to OLPC's XO laptop," > Negroponte wrote.] > > Without getting into Mr Negroponte's philosophy behind OLPC, about > which there is a huge risk of getting into pointless arguments with > its faithful supporters, it is significant that he suggests having > the tablet definition diluted with the needs of other users (the > "world needs your device and leadership"). Vivek Wadhwa addresses > this same point in a blogpost at TechCrunch (here > http://techcrunch.com/2010/08/01/opportunities-in-the-patent-free-z >one/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+Tech >crunch+%28TechCrunch%29 ). > > There is a significant problem, which I think is based on a false > sense of self-centerdness, that one size fits all. Far from it, > even a tablet meant for some of India's people won't serve the > needs of some others. With several hundred million potential users, > does it even matter? I don't think so.
I could not agree more. Observing usage patterns and failure modes, points to the need for a self rolled kit, rather than an engraved in stone device. This will enable components to be moulded into local needs rather than "USE THIS AND PRAY IT WORKS" handouts. Useage need not be confined to education, there are many many more uses. > > A walk around any consumer mall will show that the simplest of > product categories are crowded with choices. Why should a complex > device be any different? Actually it wasnt any different. In the beginning (1979) there were pcs of every flavour (in the far east and US), but eventually they converged into several separated streams, with the IBM designed INTEL PC architecture cornering the desktop. As usage and economic patterns change , we will see several new categories and spill overs from the other streams onto the desktop. The current lowpower low cost contender Acorn Risc Machines (ARM) was actually a mind blowingly fast Acorn Desktop PC. The Acorn was 10 times faster running an interpreted BASIC graphics program v/s an IBM pc running a C graphics program. C ran even faster on the Acorn. And surprisingly another CAD and Graphics workstation champ of the past is also in the lowcost lowpower space - Silicon Graphics MIPS (it is present in most of your adsl and wireless routers, also in the LEMOTE netbook). Intel as usual is there with yet another kludge of the original 8086 kludge with the ATOM series. If IBM had not kept the hardware design open, which allowed hundreds of addon cards and DELL not cloned the bios, which allowed hundreds of taiwanese clones, and a small 5 man company not made the Creative sound card, and... Intel would have been a fossil. > Is a PC any better for having such complex > software that purports to meet every differentiated need, all in > one gadget? I don't think so. If anything, it fails to be good > enough on so many counts that the head spins. There was this guy from New Delhi, who had got an import licence (because his dad was a very highly placed bureaucrat) and collaboration with Acorn in 1986. At the time HCL, Wipro and many others were pushing clone PCs and assorted servers (HCL-Apollo, Wipro-Sun, DEC-VAX etc). The rest of the ecosystem, spurred on by the Taiwanese independent mobos and hardware addons, independent software etc) favoured the others over the Acorn, so the rest grew into giants and the Acorn died here. But in Japan, where superlative graphics was an absolute must, to manage the Japanese character set, many alternative hardware architectures continued to do well (Sord- Motorola 68000, Apple, NEC, Toshiba ) with the IBM PC at the bottom of the heap. So one architecture fits all (even for the desktop) is more a result of circumstances and machinations, rather than any virtue. Technology is at just the right point to make an independent push. -- Rgds JTD _______________________________________________ network mailing list [email protected] http://lists.fosscom.in/listinfo.cgi/network-fosscom.in
