Cary, As a DBA, I tend not to rely on names anyway because I believe that documentation etc is out of date and incorrect two seconds after it is completed.
But for my developers, it does help to have some sort of convention when they read explain plans, especially if I also impose the rule that no one can create anything except me. And then there is the unanswerable argument of "corporate policy dictates we have naming standards" :) Rachel --- Cary Millsap <[EMAIL PROTECTED]> wrote: > Rachel, one of the SQL statements in our Clinic that people find the > hardest to optimize is one that has a thing that looks like > "id_number = > 10000" in the where clause. "id_number" is the table's primary key, > yet > the query spends 20 seconds executing a full-table scan. Any guesses? > > It's because "id_number" was actually defined as a varchar2 column. > Oracle's implicit type coercion converts the predicate into > "to_number(id_number) = 10000". Presto: the PK index is useless. > > This and dozens of other unnecessarily pathological problems await > people who try to embed too much information into their names. > > > Cary Millsap > Hotsos Enterprises, Ltd. > http://www.hotsos.com > > Upcoming events: > - NCOAUG Training Day, Aug 16 Chicago > - Miracle Database Forum, Sep 20-22 Middlefart Denmark > - 2003 Hotsos Symposium on OracleR System Performance, Feb 9-12 > Dallas > > > > -----Original Message----- > Carmichael > Sent: Tuesday, July 30, 2002 8:09 PM > To: Multiple recipients of list ORACLE-L > > I can see your point, In the data warehouse we are building here, the > modeler is planning on prefixing tables with the type of table (D_ > for > dimension tables, F_ for fact, etc) > > Hm, you mean we have to go back and revisit the naming standards that > they developed? Can I please suffix the column names with an > indicator > of the datatype? :) > > The biggest problem is that most management wants "naming > standards"... > > > Rachel > > > --- Cary Millsap <[EMAIL PROTECTED]> wrote: > > I just think it's a waste. You can tell by context what kind of > thing > > a > > thing is. For example, consider: "select a.flarg from bloing a > where > > a.croopoo > 7". This can be understood by syntactical context (even > > with > > the nonsense names), without having to rename "bloing" to > > "bloing_table". > > > > Most of the embedding of type names into object names that I've > seen > > has > > been implemented by users who were inexperienced at the time they > > created the standard. They were worried that without embedding the > > type > > name into the object name, they might forget what kind of object it > > was. > > ...Most such naming conventions become onerous over time, long > after > > you > > find out that you can find the type of something in the data > > dictionary, > > but after it's too late to save the thousands of extra characters > of > > typing that'll waste people's lifespans over time. > > > > In my old OFA paper, I made a joke about how we don't embed type > > names > > into object names in daily life, with just a few exceptions (Billy > > the > > Kid, Winnie the Pooh, Atilla the Hun, and the younger family > members > > of > > the old Walton Family TV show are a few examples). If you have both > a > > dog and a child named "Rex," though, it's probably a good idea to > > expect > > them both to come when you call. With SQL, though, I can't think of > a > > case in which it's not easy to tell by syntactic context what kind > of > > thing you're talking about... > > > > > > Cary Millsap > > Hotsos Enterprises, Ltd. > > http://www.hotsos.com > > > > Upcoming events: > > - NCOAUG Training Day, Aug 16 Chicago > > - Miracle Database Forum, Sep 20-22 Middlefart Denmark > > - 2003 Hotsos Symposium on OracleR System Performance, Feb 9-12 > > Dallas > > > > > > > > -----Original Message----- > > Carmichael > > Sent: Tuesday, July 30, 2002 4:49 PM > > To: Multiple recipients of list ORACLE-L > > > > Cary, > > > > you said > > "* Don't embed the object type in the object's name. I used to see > > this > > all the time with tablespaces called XYZ_TS, indexes called > > IND_THING, > > and so on." > > > > what's your logic behind that? > > > > Rachel > > > > > > > > --- Cary Millsap <[EMAIL PROTECTED]> wrote: > > > Here's a start. Not a checklist by any means, just kind of a > micro > > > pet > > > peeves list. > > > > > > * Decide today whether table names will be singular or plural. Do > > you > > > want a THING (singular) table? Or a THINGS (plural) table? > > > > > > * Don't use case-sensitive names. E.g., use THING (without > quotes) > > in > > > your CREATE (DDL) statement, which can be spelled "THING", > "thing", > > > "Thing", or even "tHiNG" in your developers' SQL. But don't make > > > developers type stuff like this... > > > > > > select "Name" from "Thing" where "Id" = y /* won't work > > without > > > "" > > > */ > > > > > > * Don't embed the object type in the object's name. I used to see > > > this > > > all the time with tablespaces called XYZ_TS, indexes called > > > IND_THING, > > > and so on. > > > > > > * Decide today whether you want to abbreviate or not. If you do, > > then > > > construct a formal, standard, consistent list of accepted > > > abbreviations. > > > Don't name one table CUSTOMER_THING and another CUST_HISTORY. > > > > > > > > > Cary Millsap > > > Hotsos Enterprises, Ltd. > > > http://www.hotsos.com > > > > > > Upcoming events: > > > - NCOAUG Training Day, Aug 16 Chicago > > > - Miracle Database Forum, Sep 20-22 Middlefart Denmark > > > - 2003 Hotsos Symposium on OracleR System Performance, Feb 9-12 > > > Dallas > > > > > > > > > > > > -----Original Message----- > > > Chambers > > > Sent: Tuesday, July 30, 2002 3:37 PM > > > To: Multiple recipients of list ORACLE-L > > > > > > All... > > > > > > Will some of you please provide some insight on your table naming > > > conventions? I'm in the very early planning stages of what will > > > likely > > > be a large and complex schema (IT asset inventory). I have a > > chance > > > to > > > start it correctly. TIA > > > > > > Gary Chambers > > > > > > //------------------------------------- > > > // Lucent Technologies GIO/Unix > > > // 4 Robbins Road, Westford, MA 01886 > > > // 978-399-0481 / 888-480-6924 (Pager) > > > // Nothing fancy and nothing Microsoft > > > //------------------------------------- > > > > > > -- > > > Please see the official ORACLE-L FAQ: http://www.orafaq.com > === message truncated === __________________________________________________ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Rachel Carmichael INET: [EMAIL PROTECTED] Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051 San Diego, California -- Public Internet access / Mailing Lists -------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
