On 12/10/10 1:55 AM, Robert Dodier wrote:
> On 12/9/10, Raymond Toy <toy.raymond at gmail.com> wrote:
>
>> This is interesting. I had thought about reinstating the old cl-info
>> code that groveled the maxima.info file to get the tags and
>> everything. And the character offsets there would have worked just
>> fine if we read in the whole info file (only takes a 10 ms on a slow 750
>> MHz machine with cmucl) and extracted the data using the character offset.
> The old regex search on maxima.info was slower in GCL than
> other Lisps, as I recall. It was really unworkably slow -- I seem
> to recall waiting a long time (like 20 seconds) to get help text.
That's not my memory of the situation. I know I tested this stuff with
clisp, cmucl, and gcl at the time, and on the slowest machine at that
time (a 300 MHz sparc?) took no more than a second or two. Perhaps
your machine was significantly slower than that? Also, I stole the
cl-info stuff from gcl, which used that for it's describe function and
that was always quite fast.
Anyway, it doesn't matter now.
What I see now is that build-index takes some 15 sec to run (even when
compiled) on a 750 MHz sparc. If this is at build time, that's fine.
Not acceptable if it's at run time.
Ray