comp.lang.lisp Google Group

Re: lightweight database

2010-03-19T07:59:50Z

??>> As I understand you can get something like O(N*M) complexity if one
??>> table is not indexed and all other tables are indexed.
??>> That is, you do a full scan on a table with N elements, doing M
??>> operations on each element. Well, you also have a (log N) somewhere
??>> there, but who cares?

Lisp internships and jobs at HPC Platform, Boston and Paris

2010-03-19T03:15:08Z

It's been a while since I last posted here as I have been really busy
with HPC Platform so I'm happy to announce that:
There are some positions for Common Lisp internships at HPC Platform in
Boston and Paris. Some full time positions will be available in Boston
in the next months.
HPC Platform makes the fastest hardware (FPGA) based processing systems

Re: to RG - Lisp lunacy and Perl psychosis

2010-03-19T02:51:45Z

+---------------
+---------------
Mostly for checking whether a (possibly-abbreviated) fixed substring
exists at some specific but variable location in a string [especially
with the :START2/:END2 options], *without* having to do a SUBSEQ first
to extract the portion to be tested [avoiding unnecessary consing].

Re: Please comment on this code from a newbee

2010-03-18T23:13:18Z

Here's my 20-minutes hack: [link]
Ariel

Cheap Wholesale NIke Air Jordan (paypal payment) (www.globlepurchase.com )

2010-03-18T22:58:20Z

Cheap Wholesale UGG Shoes (paypal payment)
Cheap Wholesale NIke Air Jordan (paypal payment)
Cheap Wholesale Ugg Shoes (paypal payment)
Cheap Wholesale Gucci Shoes (paypal payment)
([link] )
Cheap Wholesale GUCCI Boots
Cheap Wholesale Lacoste Shoes

Re: Please comment on this code from a newbee

2010-03-18T22:46:56Z

Or perhaps better (list* a b).

Re: Anyone heard of Bee Lisp?

2010-03-18T22:31:06Z

Are you aware that McCarthy work is 50 years old?
That is, it is almost as old as the Big Bang, in computer history...
I, for one, implemented the AIM-8 LISP in Common Lisp
[link]
but it is more of a joke or rather an archeological reconstruction and

Re: Please comment on this code from a newbee

2010-03-18T22:12:40Z

In article
<2fec99a1-773d-4a45-b2ae-a0f9f efce...@q15g2000yqj.googlegrou ps.com>,
Quick remarks:
* the code looks much better than the previous one
* the macro is not needed. Use a function instead.
* (let ((a (foo))) (if a a ...)) is simply (or (foo) ...)
* `(,a) is simpler (list a)
* (concatenate 'list ...) is simpler (append ...)

Re: Anyone heard of Bee Lisp?

2010-03-18T22:01:28Z

Personally, the only thing that I find frightening in [link] is
that they do not mention that it's a Common Lisp (or scheme or ISO
Lisp). All the rest looks rather nice (apart from a few hints such as:
The majority of Lisp systems are provided with a considerable
quantity of libraries often frightening users by its large size and

Re: Anyone heard of Bee Lisp?

2010-03-18T21:47:45Z

If it was a repackaging of GPL code, the sources should be provided, so
there would be no trojan.
People wanting to sell trojans would use another license, such as MIT or
BSD....
Once again, the goodness of GPL proves over the evilness of BSD! :-)
PS: No need to mention that people distributing trojan woundn't mind

Re: lightweight database

2010-03-18T21:06:15Z

Yes, you're right, I waved my hands there and you deserve an explanation. My
minute and a half assessment was hasty.
The way I was picturing the algorithm that performs joins across the entire
database is as a merge sort performed n times.
m, number of records in database
n, number of joins
s, average number of records in a selection set

Re: lightweight database

2010-03-18T20:59:41Z

The lookup of the indexed element would typically be an
O(log(m)) operation, unless the lookup used a perfect hash (maybe even
an identitity hash :-) Thus, the full scan for the single join should be
a O(m*log(m)) - as I stated earlier.
I'm waiting for vanekl's analysis - best case, I learn something

Re: Please comment on this code from a newbee

2010-03-18T20:50:03Z

Hello again,
I'm done with the next task: "Watersheds" --
[link]
I've tried to be functional, and to write functions but not macros;
and not to use global parameters.
Some exceptions are needed (IMHO):
1. One macro to generate two functions and to avoid copy-n-pasting of

Re: lightweight database

2010-03-18T20:45:08Z

RW> I'd like to know the details of your analysis - I cannot see
RW> that a join can be a linear operation.
As I understand you can get something like O(N*M) complexity if one table is
not indexed and all other tables are indexed.
That is, you do a full scan on a table with N elements, doing M operations

Re: lightweight database

2010-03-18T20:03:33Z

^ is the exponential operator, right. What do you think the cost
of a self join on a table of m records is?
My guess would be
O(m^2) if the table is not indexed on the (single) join column, and
O(m*log(m)) if the table is indexed.
This does not take into account actual data retrieval, which would carry