Dmitry Yemanov wrote:

> Leyne, Sean wrote:
>
>> The 32K page size is not an easy fix; it requires the same clean-up as a
>> 64K page size.
>>
>
> It depends on what changes you make ;-) The 32K page size requires a
> bunch of SSHORT->USHORT replacements, while the 64k page size requires
> to change everything into ULONG. The latter implies an ODS change as well.
>
>
>> In fact, about a year ago, I checked-out the code and did a search for
>> "page_size/pagesize" and found that there were less then two dozen
>> references.
>>
>
> Actually, there are more than a hundred code lines to change. And this
> is an optimistic estimate.
>
>
>> The change to a 64K page size can easily be reached (~1 man day of work)
>>
>
> I'm afraid it's somewhat harder to reach.
>
>

Rather than increasing the page size, has Firebird considered truncated
index keys, where only the first hundred bytes (or whatever) are used in
the key? The two phase index retrieval handles these almost for free.
It's not a universal bullet, but it does get around the semantic
restrictions of relatively small page sizes.