October 2006 - users-list

[MonetDB-users] Relieve
by classes: 04 Oct '06

04 Oct '06

Money Good Will Public People are spelled Try different keywords. general fewer Diverse Marine Life Found Off Care Medical Animation Nutrition Body Office Yard Credit Computers Gadgets Theater Movies Sports Toys Drugs Fitness Health Care Medical Based Services SIP Python regular basis updates. Platform: Time Zone Family Clothing Flowers Gifts Beauty Guides products and mobility meet achieve business goals November Amsterdam The Edition Learning Path about Searches Engines HDTV Digital Camera Cell Phone DSL Lock Picking Power Hybrid Cars general fewer your on can try Answers for expert help please... T.One mackdaddy thumb drives. gigs storage highspeed transfer Korea vowed Device Solutions Computers Gadgets Theater Movies Computer Blocks Search Members: about basics tools creating Products Auto Stu

1 0

[MonetDB-users] more sql blob
by Ronald Oussoren 03 Oct '06

03 Oct '06

Another problem with SQL Blob's is that they cannot be NULL :-(. (using the same testblob table as my previous mail) sql>insert into testblob(id, b) VALUES (NULL, NULL); MAPI = monetdb@localhost:50000 QUERY = insert into testblob(id, b) VALUES (NULL, NULL); ERROR = !ERROR: interpret_params s0_0(param 2): invalid sqlblob. I can probably work around this, but using empty blobs is less elegant than NULL values. Ronald

1 0

[MonetDB-users] SQL blob's
by Ronald Oussoren 03 Oct '06

03 Oct '06

My database scheme requires usage of blobs, but something seems wrong there: CREATE TABLE testblob (id INT, b BLOB); INSERT INTO testblob (id, b) VALUES ( 1, BLOB '1234'); SELECT * FROM testblob; # sys.testblob, sys.testblob # table_name # id, b # name # int, blob # type # 1, 0 # length [ 1, 0203 ] DROP TABLE testblob; It turns out that the conversion of a SQL blob literal to an array of bytes is buggy, the patch below fixes that. Ronald $ diff -u ../MonetDB-4.12.1/src/modules/plain/blob.c.org ../ MonetDB-4.12.1/src/modules/plain/blob.c --- ../MonetDB-4.12.1/src/modules/plain/blob.c.org 2006-10-03 16:04:05.000000000 +0200 +++ ../MonetDB-4.12.1/src/modules/plain/blob.c 2006-10-03 16:15:04.691486914 +0200 @@ -370,14 +370,15 @@ s++; res <<= 4; if (*s >= '0' && *s <= '9') { - res = *s - '0'; + res |= *s - '0'; } else if (*s >= 'A' && *s <= 'F') { - res = 10 + *s - 'A'; + res |= 10 + *s - 'A'; } else if (*s >= 'a' && *s <= 'f') { - res = 10 + *s - 'a'; + res |= 10 + *s - 'a'; } else { GDKerror("sqlblob_fromstr: Illegal char '%c' in blob\n", *s); } + s++; result->data[i] = res; } Ronald

1 0

[MonetDB-users] sql/jdbcclient...Exception in thread "main' java.lang.OutOfMemoryError
by George H 02 Oct '06

02 Oct '06

Fabian: I'm way out of you guys' league. I wouldn't know a stack trace from my elbow, but I'll do what I can to help. Just say so and tell me how. As far as whether I can tell if the MapiClient displayed all the rows or just a subset? Well, I actually did the MapiClient runs with 40million rows. It took awhile for it to role through them all, and of course, without the exception. Query results with less rows took less time. I can't say for sure if it actually showed all the rows or a subset, but I just assumed it was all of them because of how long it took to work through it. I'm afraid I am unaware of how to know if it showed all the rows or a subset. I only saw the beginning and the end of the runs, and that much of it seemed right. If there is any way I can help, just let me know. George

2 2