[sf-lug] OLPC info for the frustrated (Kristian Erik Hermansen)

[RBV]

Hi:

If I can be allowed to make a small side observation about the OLPC's apparent, and apparently non-trivial, delivery problems.

At the last OLPC meeting, I was *hugely* impressed with the design and operation of the OLPCs that Jim brought in for us to examine.  The OLPC is not only an exemplar of brilliant design, it's a testament to the ability of non-proprietary technologies to fulfill such designs.

But those with a historical perspective on technology can remember many instances in which superior technology was trumped by superior pre- and post-sales customer support.

For example, IBM became a mainframe giant not because its systems were the most technically compelling (the company generally preferred to sidestep or befog A-versus-B performance comparisons), but because the company offered customer service and support that was far superior to its competitors', albeit in sometimes heavy-handed ways.

Intel learned from IBM, and so commandeered market share from more elegantly designed chip products by offering superlative design and manufacturing support.

Like many, I'd be genuinely unhappy to see malicious proprietary vendors -- including, not incidentally, Intel -- undermine the OLPC idea.  But I'd also say that the critical challenge for OLPC is not one of technologies but rather support.  Given that OLPC seems to be antagonizing those who've voted with their pocketbook for the system and its goals, one can but wonder if anyone at OLPC is prepared to understand and react the importance of that support challenge...

Sorry if this submission seems a bit to the side of the central issue, but I believe it to be of some relevance.  Open source -- especially Linux -- is a lovely thing, but too often considers user and customer needs to be annoying distractions from the "interesting" technological bits...

Cheers,
Riley
SFO