gilbertd at sunflower.bio.indiana.edu (Don Gilbert) wrote:
>>> The "odd" and "srsbuild" functions both work properly -- one
> can build a full set of srs indices on a mac. But one can also
> transfer indices from another system and have them work right (as
> I did from Sun to Mac). "wgetz" does properly spit out HTML versions
> of the data.
>great! did you compile with a 32bit compiler?
The indices are platform independent so once built they
can be used on any operating system
> I don't yet see a good way of having "wgetz" functionalilty running t
> his on a Mac -- that would involve rewriting an HTML browser to
> work with "wgetz". As far as I know there isn't yet any standard
> way of doing HTML forms between two programs on a Mac or PC,
I don't understand that ..all you would need is a httpd on the mac that
talks to wgetz?
> P.S., SRS doesn't seem to do quite as good a job as WAIS in finding all
> matches for some simple queries of SwissProtein, and probably also
> GenBank/other data. I hope to do some more comparison testing soon, and
> to see if the problem areas can be tracked down. It may be only a
> matter of writing those odd sdl definitions better (e.g., many matches
> were lost because the comment field wasn't indexed in the swissprot.sdl
> I used at first, and the "AllText" index I used didn't really search
> all the other indices).
can you send me the queries you have done? there should be really
no difference ...and if there is it can be all configured.
Alltext is a virtual data-field, ie, a group of indices that will
be all searched - those index groups can be added and modified -
the Alltext field is defined in srsgeneral.sdl - other data-fields
that want to be part of Alltext must point to the Alltext definition
..and yes, i should index the comment data-field in swissprot