Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : IFMX - Investment Discussion -- Ignore unavailable to you. Want to Upgrade?

To: Jeff Meredith who wrote (10088)	3/30/1998 1:23:00 PM
From: Mark Finger	Read Replies (2) \| Respond to of 14631

>>(Oracle's) data cartridges (meaning those that run in the backend >>server) do not exist. I'm suggesting that when they do, the >>technical differences between them and datablades will be >>insignificant. The differences may not look to be very much on paper, but there should be a lot more difference in practice. From what I have seen, I see two major areas of difference: 1. Datablades are better integrated with the RDBMS core. By this I am referring to the more complete hooks that the object and core have to each other. For example, how much information about CPU costs for comparisons is made available? How many different index types (R-trees, custom, ...) are available to the object? Can the server based objects work with the "thread manager" of the engine core to improve performance? Can the optimizer get enough information about different objects to determine the best order to do joins? 2. Oracle still is insisting that the CORBA interface will be the only one supported. This is inherently slower than the what will be done for most commercial Datablades. If you are doing a complex search through a document or large picture, this will not make much difference. However, if accessing an index or getting simple information from the object's basic definition is involved, the response difference can be a major factor (1-2 orders of magnitude). The key here is that both may look good on paper, and for the complex new data types (documents, GIF and other graphics files, ...) that Oracle now supports in non-extensible manner, but as VAR's start supporting many new simpler data types (spatial, color, complex numbers, ...), the two above differences will really show up in actual implementations. The difference would be between a successful project and one that may work in the lab on 100 data elements, but fail to scale properly. Coupled with IFMX current lead in actual sites supported, this could be a huge difference. Actually, IBM and Sybase are the two most likely to challenge IFMX over the next year or two in this category. Mark