Presenting (with) Bruce

When I present at a meeting, I like to use plenty of live demos. I can get away with this, because I type very, very well. (I did get into IT via the secretarial pool, after all.) Still, even I get a little nervous about whether I can keep up typing speed and accuracy with the pressure of a whole audience staring at me.

Fortunately, This blog post by Andy Todd (You do read his blog, don't you?) pointed me to Bruce, a wonderful presentation tool freshly created by Richard Jones.

Bruce lets you wrap up decent-looking slides, graphics, and Python interpreter sessions all into a single presentation that scrolls smoothly along at your command. (Simply being spared the fuss of flipping between Presentation slides and a command-prompt window is a blessing.) For demos, you can pre-enter the text you want to type, then Bruce sends it to the interpreter one character at a time. You type 'dfjkasfjdska;dfsafd', but 'def plusOne(x): return x+1' appears on the screen, and the audience thinks you're a keyboard goddess.

My one disappointment in Bruce - as I understand it (I just found it, you know) - is that the wonderful interactive session can only be used with a Python session. I'll need to dig into Bruce and PyGame to verify that, and to see if it can be extended elsewhere. I dearly want to use it when I'm demonstrating SQL.

Until Bruce, incidentally, my best bet was TPP, the Text Presentation Program. I've seen presentations done with it, and it looks OK, except it's written in Ruby and has dependencies (ncurses) that need C compiling - not very practical where I am, unfortunately.

Search engines are not Zen

I decided that I should make myself a list of questions I've never been able to resolve, or that were very difficult to research, because they don't "search" well. That is, in everything from Metalink's Search bar to Google, it's hard to come up with relevant results.

For instance,

- I'm trying to use RMAN DUPLICATE, and it does indeed produce the data files on the auxiliary database, but it doesn't create a controlfile. The documentation clearly says it should. Searching on terms like RMAN DUPLICATE controlfile not generated only brings up more pages saying that it should.

- All sorts of questions like such-and-such service DOESN'T start, program WON'T run, logfile ISN'T created. Negation - "not" - is clearly a failing of a typical search. Searching for green eggs no ham will find you 10,000 pages on green eggs AND ham, with the one page on "My green eggs came without ham!" buried anonymously as result #4,312.

I'm not sure what I'll ultimately want to do with this list. But I began to wonder, more broadly, if someone out there had already started assembling issues of unsearchable questions. Ironically, Can't Find On Google is very easy to find on Google.

Dayton - Cincinnati Code Camp

Hooray! - to the Dayton .NET Developers Group for putting together the Code Camp on Jan. 21, 2006 - a whole day of community-provided seminars on a variety of programming topics. For free! I'm looking forward to it; in fact, I'm planning to present there (on Python, naturally).

If you're anywhere near southwest Ohio, come and enjoy. Better yet, come and share! Whatever cool programming tricks you have up your sleeve - come and show the rest of us!

LibraryLookup published

My article on LibraryLookup has been published in the October 2005 issue of Technology First, a tech newsletter for the Dayton area. (I'm on p. 8-9). (The print article's content ties in with this blog post)

The sixty-second version: LibraryLookup lets you query your library's catalog directly from a page of a book vendor like Amazon, looking up that particular book for you automatically with just one click. You can get it installed in your browser in, oh, about five seconds. It's really cool, not just for helping you find books, but also as an example of clever and sneaky use of JavaScript.

Why do almost all contracts and job postings specify exactly the software that they want you to use (or to have used)?

When they want to incorporate you tightly into a large project that has already gotten underway with some specific software, OK. But I think such over-specifying is also the custom in cases where that is not at all true. It's very annoying to somebody like me, who bizzarely imagines she might be able to select the best tools for her own work.

When I take my car to the mechanic, I don't say, "You must do all the work using a Sears Craftsman #3 box wrench." But maybe, if I were a writer of contracts, I would.

Announcing Dayton-Oracle mailing list

Q. How do Dayton-area users of Oracle stay in touch with one another and find out what's going on in their community?
A. They don't.

Q. Well, that's lame. What can we do about that?
A. You can sign up for the Dayton-Oracle mailing list I just set up at FreeLists.

Please spread the word!

cx_Oracle and Ubuntu

Naturally, after getting Oracle working, my next move was to get it Python-powered by installing cx_Oracle. Installation seemed to go OK, but when I tried to use it...

>>> import cx_Oracle
Traceback (most recent call last):
  File "", line 1, in ?
ImportError: libclntsh.so.10.1: cannot open shared object file: No such file or directory

which seemed odd, because

catherine@ubuntu:~$ locate libclntsh.so.10.1
/app/oracle/oracle/product/10.2.0/db_1/lib/libclntsh.so.10.1

I extrapolated a solution from an old mailing-list post by Karl Putland: I appended /app/oracle/oracle/product/10.2.0/db_1/lib to /etc/ld.so.conf and ran ldconfig.

That was the only trick - it's working fine now!

Oracle on Ubuntu

I love my new Ubuntu Linux (version 5.10). I love how easy it was to install, how it picked right up on my wireless card.

Oracle doesn't support Ubuntu, but I installed it anyway and it's doing fine. The process was tricky, though. Here's what I learned.

1. RTFM. The Installation Guide is very important, more so than when installing Oracle on Windows. It has to be the primary document you work from, and follow step-by-step. Any other advice on Oracle/Ubuntu that you may Google up is secondary. Where Red Hat and SuSE directions differ, follow the Red Hat ones.
2. This document is useful primarily for reassuring you that the error messages you get are OK.
3. Make sure the Universe is in your Synaptic repository and install these packages: gcc, make, libxp, libaio, lesstif2, lesstif2-dev, rpm, and libdb1.
4. You can't find all the packages that the install documents call for in the Ubuntu Universe. That seems to be OK; just find the ones you can.
5. The magic word is runInstaller -ignoreSysPrereqs.
6. One little misstep and you may get something broken. It took me a couple attempts. Try, try again, following that Installation Guide even more carefully.

Oracle - Python wish list

For my upcoming Oracle - Python talks at OOUG and (maybe) COLLABORATE 06, I want to write a nice Python script in realtime, before the very eyes of the audience, like a cooking show hostess would. I'm having trouble deciding on an objective for the script, though. The ideal script would make Oracle people sit up and scream, "I want that! That power must be mine!" It should also

• Be very Oracle relevant
• Be something that can't be done with SQL or PL/SQL alone
• But still use some SQL (to demonstrate cx_Oracle)
• Not duplicate the function of any existing Oracle utility or Oracle Enterprise Manager (this one worries me since I don't even know everything OEM can do)
• Be writeable within an hour (shouldn't be a problem)

I'm brainstorming for ideas, but I don't know if any of them are really "it".

• Compare parameters in INIT.ORA to live database (this is what I used for my paper)
• Condense or groom a TKPROF output file
• Populate a database table from a website (but what website? What's cool enough?)
• Find and compare TNSNAMES.ORA files scattered across filesystem. Find TNSNAMES entries that are "dead" (unpingable).
• Format alert.ora for better readability; color-code messages?
• Measure log switch frequency from alert.ora
• Find obsolete (unused) configuration files like init.ora, ORACLE_HOMEs, etc.
• Clean up PATH (mine gets cluttered with droppings from various installs/uninstalls)

Any comments on these? Anything you'd like to put on the wish list? Let your imagination run wild - what wishes would you make to a magic Oracle genie? That's the kind of "gotta have it" feeling I'd like to produce.

Ohio LinuxFest

I had a great time at Ohio LinuxFest. Highlights, in my book, were

• Jorge Castro's Ubuntu talk. He pointed us to a lot of great stuff to try out, and his enthusiasm was infectuous.
• Rich Bowen's mod_rewrite talk. I've never encountered Apache documentation that I was truly happy with, but Rich was wonderfully clear. I want his book.
• Ram Rao's virtualization talk. Xen sounds like it has jumped straight to #1 in my List of Reasons My Management Would Let Me Use Linux For The Database Server If They Had A Clue. Moving entire virtual machines easily between production and development servers, with almost no performance impact... wow, that would be incredible.
• The reception. Hey, the only thing better than dancing is dancing with geeks. Too bad I had to leave early, but somebody foolishly located Dayton too far from Columbus.

I heard that the attendance had more than doubled from the last year, and indeed, many of the people I talked to were quite new to Linux. That's the one improvement I'd suggest for the event. There were good talks for newbies, but not enough of them, nor did the schedule provide clear enough guidance to help newbies steer themselves that way.

COLLABORATE 06, featuring me, maybe

I just submitted a proposal for a technical session ("Python and Oracle: Coding that Cooks") at COLLABORATE 06, the IOUG's conference next April. Wish me luck! I'm not sure how hard it is to get a slot; I do know that the speaker quality there has been excellent for the past two years.

introducing sqlWrap.py

A Python script, for your use and suggestions.
Sorry, I know this is awfully long for a blog post - I have to figure out a better place to put it. I wonder if it would be appropriate for the Cheese Shop?

"""
sqlWrap
Adds convenience methods to DB-API2 connection objects.

Sep. 22, 2005 by Catherine Devlin (catherine.devlin@gmail.com, http://catherinedevlin.blogspot.com/)

This script is NOT intended as an object/relational mapper.
Rather, it helps experienced SQL users form their SQL more quickly.

Methods: insert, update, delete, select, genericSelect
These methods accept arguments for the WHERE, SET, etc. clauses
that should generally be provided as dictionaries; i.e.
whereClause = {'col1':'val1','col2':'val2%'} implies
  "WHERE col1 = 'val1' AND col2 LIKE 'val2%'",
setClause = {'col1':'val1'} implies 
  "SET col1 = 'val1'"
You may also pass object instances for whereClause and setClause,
with instance attributes corresponding to column names... but this 
has barely been tested!
They also automatically make use of bind variables, which have
performance and security benefits over hard-coding values in SQL.
The different ways of handling bind variables in various DB-API 
adapters are masked from the user.

Currently supports: Oracle (cx_Oracle), sqlite (pysqlite)

Sample usage:
# setup - unchanged from cx_Oracle
conn = OraConnection('scott/tiger@orcl') 
conn.cursor().execute('CREATE TABLE myTable (column1 varchar2(10), column2 varchar2(10), column3 varchar2(10))')
# now try out the sqlWrap convenience methods
conn.insert('myTable', setClause={'column1':'value1','column2':'value2'})
conn.insert('myTable', setClause={'column1':'value1a','column2':'value2a','column3':'value3a'})  
for row in conn.select(source='myTable'): 
    print row 
for row in conn.select(source='myTable', whereClause={'column1':'value1'}, resultProcessor=conn.dictionaryize):
    print row
conn.update('myTable', setClause={'column1':'value1','column2':'value2'}, whereClause={'column3':'value3'})
# as always, must explicitly commit  
conn.commit()
"""      

Get sqlWrap.py here

If you're in Ohio...

notes to self

Do not smash your head against a problem for consecutive hours. Instead:

• Ask for help. It doesn't matter whether you ever send the request; in the process of putting the problem into language, you're more than likely to solve it. (Yes, I know: "Pair programming!", you say. But I'm all alone here.)
• Stand up and walk away. Do not touch anything resembling a computer for at least ten minutes.

Will someone please remind me of these next time I get stuck?

Python for Oracle Geeks: unplugged!

I've been invited to present at the Oct. 27 Ohio Oracle User Group meeting in Columbus. Hooray! After I've practiced at a local group, I'll feel ready to start trying to get to wider areas, maybe even the national IOUG conference. I will bring tanker-trucks full of Python Kool-Aid!

We have a lot of catching up to do - Oracle's OTN has been publishing a LOT on PHP, and Oracle and Zend are providing a handy-looking integrated package. I salute the PHP folks, but I don't want that to be the only dynamic language active in Oracle-land.

sqlite: a flyswatter to kill flies

I'm now using sqlite to support an Oracle production database. I love it!

When I first went to download sqlite, I went away frustrated. I found an executable described as "A command-line program for accessing and modifying SQLite databases", and thought, "OK, so that's my SQL*Plus equivalent, but where's the actual server? The part that keeps the database process running?"

Because I have been an Oracle-only person so long, I didn't understand that sqlite doesn't need anything like that. It's simply this:

1. a single executable program that creates and modifies a database file
2. A database file

And that's it. Really! You can't imagine how dizzying and refreshing that is to someone who's spent hours before a certification exam memorizing Oracle architecture diagrams.

Anyway, there are other simple database engines, of course, but sqlite has gotten a lot of attention recently (like an Open Source Award) for its efficiency and its support across many languages. There's an excellent and honest rundown of its powers and limitations here.

So, anyway - why sqlite to support Oracle? Well, one of my Oracle instances has a bunch of logic applied to it by a nightly batch job. For every data record, a series of decisions are made, and our users ask questions like, "For record #12945, why did it decide X instead of Y last Tuesday?" And I have to be able to answer, "Well, the seventh of the nine tests conducted on that record determined that, since column 'product' was 'lutefisk' and 'quantity_kg' was 22, blah blah blah..." So all those decisions need to be logged every night.

That generates a quantity of data far outweighing the application data itself. It can be discarded after a week or two, but all that inserting and deleting was causing out-of-control generation of archive log files and making a mess of my disk space allocation. By moving that data out into a sqlite database, it becomes a single simple file that can be moved or deleted as easily as any other file.

So, thanks to sqlite, I'm living happily ever after. Hooray!

more on XML

The problem with my XML-generating views is that they break down for rows whose XML is longer than 4000 characters.

Meanwhile, the troubles I cited with XMLELEMENT really are limited to SQL*Plus. I was deterred at first, because I want to experiment with unfamiliar features through SQL*Plus first, but when I swallowed that reluctance and remembered that there are other ways to experiment with ad-hoc SQL - TOAD's SQL window, for instance - I was OK.

But then, I wanted to create views that would store the particular combinations of XMLELEMENTs I wanted for various circumstances. Unfortunately, I found that a SELECT query that runs fine can't be used to generate a working view when the result is longer than - you guessed it - 4000 characters.

I'm having a little trouble figuring out where exactly the 4000 character problem kicks in - at first I thought CLOBS were simply converted back to VARCHAR2's whenever I used the concatenation operator |, but that doesn't explain why views fail when the equivalent bare SELECT statements don't.

Anyway, the end of my adventure was that I simply wrote Python scripts to get what I want. I'm beginning to get some nice conveniences built into my personal wrapper for cx_Oracle, and maybe I'll eventually float it around for others' use. There are plenty of tools like SQLObject out there, of course, but as far as I know, they're all focused on OO programmers who hate to handle SQL directly or to think of their data as "rows in a table" rather than "instances of an object". I, on the other hand, think very naturally in SQL statements and relational concepts, and just use some functions to streamline my use of SQL.

Finally, there's one cute little possibility for producing XML that I haven't explored. SQL*Plus can generate an HTML table as output really easily. That could be coupled with a XSLT to produce XML.

querying XML without fighting with XDB

My coworker and I want to query some nice, simple XML from my Oracle table. For example, from the table
CREATE TABLE pet (name VARCHAR2(22), species VARCHAR2(22), weight_kg NUMBER);
I want to query up results like

  <pet>
    <name>Martin Luther</name>
    <species>cat</species>
    <weight_kg>6.2</weight_kg>
  </pet>
  <pet>
    <name>Jordache</name>
    <species>horse</species>
    <weight_kg>450</weight_kg>
  </pet>

My first thought was to use an XDB's XMLType View of my table. Okay... that looks... sort of intimidating. But I want to be able to query from it with the same old SQL I'm used to; I want to grab the XML for all rows
WHERE species = 'cat' OR (weight_kg > 100 AND name LIKE 'J%')
Now it looks really intimidating. I don't want to go back to school for a Master's of Science in XPath right now, and neither does my coworker.

Then I thought, "I know! I'll just use XMLForest, like this!"
SELECT XMLForest(name, species, weight_kg) FROM pet

On my platform (Windows 2003 / SQL*Plus 10.1.0.2.0 / Oracle 10.1.0.4.0 ), this produces
"Oracle SQL*Plus has encountered a problem and needs to close."
Attempted workarounds - like
EXEC SELECT XMLForest(name, species, weight_kg) INTO :petxml FROM pet
give me results like
"PLS-00801: Internal error [*** ASSERT at file pdw4.c, line 782; Cannot coerce between type 43 and type 30; _anon__2C9F5D70__AB[1, 7]]"
Okay, I give up - this XDB stuff is not mature enough for me today.

My handmade solution is fairly simple to use.
SELECT xml FROM pet_xmlvw WHERE species = 'cat' OR (weight_kg > 100 AND name LIKE 'J%')

Setting up views like pet_xmlvw is a headache and a half, but my ugly code will do it for you. (Can you hear PL/SQL crying? Can you hear it crying, "Somebody write Cheetah for me!"?)

CREATE OR REPLACE FUNCTION sf_columns_xml(table_name IN VARCHAR2)
RETURN CLOB
IS
  result CLOB := '';
  column_name VARCHAR2(30);
BEGIN
  FOR c IN ( SELECT utc.column_name
             FROM   user_tab_columns utc
             WHERE  utc.table_name = UPPER(sf_columns_xml.table_name) )
  LOOP
    column_name := lower(c.column_name);
    if length(result) > 0 
    then
      result := result || chr(10);
    end if;
    result := result || 
              '    <' || 
              column_name || '> '' || ' || column_name || ' || ''</' || 
              column_name || '>';
  END LOOP;
  RETURN result;
END;
/

CREATE OR REPLACE FUNCTION sf_table_xml(table_name IN VARCHAR2)
RETURN CLOB
IS
BEGIN
  RETURN 
    'CREATE OR REPLACE VIEW ' || lower(substr(table_name,1,24)) || '_xmlvw AS
SELECT ''
  <' || lower(table_name) || '>
' || to_char(sf_columns_xml(table_name)) ||
'
  </' || lower(table_name) || '>'' xml,
  ' || table_name || '.*
FROM   '  || table_name;
END;
/

Now running EXEC execute immediate TO_CHAR(sf_table_xml('PET')) will generate the view pet_xmlvw, whose column xml contains what I want.

To generate such views for all the tables in my schema, I run

spool buildXMLViews.sql
 select 'exec execute immediate to_char(sf_table_xml(''' || table_name || '''));'
 from user_tables
 where dropped = 'NO';
 select 'exec execute immediate to_char(sf_table_xml(''' || view_name || '''));'
 from user_views;
 spool off
 @buildXMLViews 

(Hey, I didn't use q'| |')! Well, you may not all be on Oracle 10g.)

Sometimes, it's a sausage factory in here. Don't look at how it gets done, just pass the ketchup.

Geek event aggregator

Click me!

I've written up an aggregator script (Python, of course) that browses an assortment of event announcement webpages and parses out the name/city/state/date information. Then I made an HTML DB application to serve it up according to your state or province (sorry, non-North-Americans).

No, I haven't written a brilliant AI to figure this out. It's just a bunch of regular expressions.

To do:

1. Mine from more event sources
2. Refactor the source code so it doesn't embarrass me and I can post it
3. Debug multiple hits for some events (InOUG especially)
4. Add non-North-American region support
5. Automate the mining (currently I kick it off and upload to HTML DB by hand)
6. get somebody to host the app in some more prominent site
7. Provide access to the pure XML as generated by the aggregator, so others don't have to go through the heck of parsing I do

Python for Oracle Geeks

IOUG posted my article, Python for Oracle Geeks, and referenced it in their 7/20/2005 "5 MINUTE BRIEFING: Oracle" e-mail. Ah, the fame, the adoring crowds, the paparazzi...

Tom Kyte

Tom Kyte was at OOUG in Columbus. He's even better in person than the is at AskTom.

Here's one little gem he gave that will change your life, if you're an oracle person. Or at least hurry you to upgrade to 10g, where it can be used.

The magic is q'| |'.

INSERT INTO my_tbl(str) VALUES 
  (q'|I don't have to worry when I write a string like 
    'INSERT 'r2d2' INTO droids'|');
INSERT INTO my_tbl(str) VALUES 
  (q'|It's almost as good as Perl!|');

SELECT * FROM my_tbl;

STR
-----------------------------------------------------
I don't have to worry when I write a string like
    'INSERT 'r2d2' INTO droids'

It's almost as good as Perl!

All those single-quote single-quote, single-quote single-quote single-quote monstrosites we used to make to get quote marks into SQL strings are history. Let the people rejoice!

LibraryLookup

For Mozilla, Firefox, or Netscape, drag-and-drop any of these links to your toolbar to create a LibraryLookup bookmarklet. For Internet Explorer, use Add to Favorites... and create it in Links.

Public libraries

• Dayton Library
• Greene Co. Library
• Clark Co. Library
• Troy Library
• Middletown Library

College and university libraries

• SCC Library (Sinclair Community College)
• UD Library (University of Dayton)
• WSU Library (Wright State)
• CSCC Library (Clark State Community College)
• Wittenburg Library
• Wilberforce Library

Now, when you're browsing any site that describes a book by its ISBN (amazon.com, for example), clicking your new toolbar button or link will automagically look that book up in your library's catalog. Yay!

Hooray for Jon Udell, creator of LibraryLookup! (And a little bit of hooray for me, who slaved over a hot keyboard to make these public library bookmarklets.)

TODO: These bookmarks are written to extract the ISBNs from the URL, not the webpage text. Some sites, like O'Reilly, don't use ISBNs in their books' URLs - but still list them in the webpage text. A more sophisticated bookmarklet might be able to find the ISBN there, too. I'd like to write bookmarklets that could find and use those ISBNs, too.

the easy tools

• If only I had known about tooltips months ago!
  A tooltip is what you get when you hover over a word like these ones. I went off looking for how to generate them, expecting to get myself hip-deep in JavaScript, and found that this is all you have to do:
  <span title="I love tooltips.">these ones<span>
  
• If I had known about CherryPy months ago!
  With all due respect to other Python-based web application platforms like Zope, CherryPy is so much easier to get into, and soooooo satisfying. And it really does run much faster than Zope, too. Yum!

Flaming about Curl and open source

Once upon a time, I read some interesting buzz about the Curl programming language. It sounded different, and it came from MIT (yeah, I know, alum chauvenism), so I had an inclination to check it out.

I'm a hopeless addict of the MicroCenter bargain rack, so when I spotted the Curl Programming Bible there for $5, I couldn't resist.

Wow, am I disappointed. It's no wonder Curl is apparently dying; it was doomed from the start.

One problem is the syntax. I didn't get very deep into it - just far enough to read

Notice that the entire herald is enclosed in curly braces ({ and }). All programming expressions in the Curl language are enclosed in these braces. They are used so often that the language itself was named after them!

Well, tee-hee-hee! Isn't that just precious? Maybe they should have named the language "Repetitive Motion Injury"! Okay, so I'm a Python geek; requiring gratuitous punctuation is an excellent way to send me reeling away in disgust..

But the real problem is the business model. The Curl plug-in is built so that, every time a user hits a Curl site, it notifies the Curl corporation to charge that site's owner. I don't care how small the charge per hit is, that's crazy. I'm going to get approval, in advance, to pay an unknown, usage-based fee, just to put up a site? For a tool that is (within our organization) totally unproven?

I think that open source really is the way to introduce a language these days. Very few languages are selected in advance by people with money-spending power. (Unless, of course, they are pushed by a megacorporation with the advertising budget to send flocks of salespeople to razzle managers and take them to lunch. But that's another discussion.) Rather, a developer or sysadmin or whoever finds a language she likes, starts using it informally for her own this and that, eventually builds something that her manager sees and likes, THEN gets permission to start working it into formal projects - the kind that have budgets. Then we'll start paying money for an IDE and training and maybe consulting and so on. At least, that's the way I've experienced it.

The sad thing is that the fate of Curl reminds me very much of the Water language - another MIT baby that interested me enough to sell me a book, but that is also more-or-less stalled. Again, I think the choice not to go open-source is a big part of that failure. (It probably doesn't help that Water is built on top of, and demands a little familiarity with, Java - so the ideal target audience is one that already has a preference for open-source.) Geeks just don't seize something and energetically play with it when they're told (1) This is ours, not yours and (2) Pay at the door.

I wonder if some foolish person in MIT's Office of Making Money Off Our Babies simply doesn't realize this, and is twisting arms to impose doomed business practices on MIT's babies. Ironic, for the institution that spawned the Free Software Foundation to go that way.

If anybody wants a barely-used Curl book, just ask.

HTML DB

A couple months ago, I built a D&D-suitable "random goods" generating website with HTML DB as an experiment. Parts of it are very frustrating, but overall it's an awfully powerful product. My example non-masterpiece is here.

making default column values work in PL/SQL's %ROWTYPE inserts

I'm going to look/post around the web for the answer to this puzzle...

I like PL/SQL's capability to insert a row all at once, using %ROWTYPE.

Example:

CREATE TABLE TEST
 (a NUMBER,
  b NUMBER DEFAULT 0 NOT NULL);

 declare
   t test%ROWTYPE;
 begin
   t.a := 22;
   t.b := 0;
   insert into test
   values t;
 end;
 /
 
PL/SQL procedure successfully completed.

However, I'm trying to use it to do inserts on a table with many NOT NULL, DEFAULT 'N' columns, and I get ORA-01400 CANNOT INSERT NULL errors.

declare
  t test%ROWTYPE;
begin
  t.a := 22;
  insert into test
  values t;
end;
/

ERROR at line 1:
ORA-01400: cannot insert NULL into ("EQDBW"."TEST"."B")
ORA-06512: at line 5

I don't want to clutter my PL/SQL by manually setting all these columns to their default values; after all, reducing clutter was the whole idea behind using %ROWTYPE in the first place. Besides, if I hard-code the default assignment into my PL/SQL, then it will become out-of-date if I change the column defaults in the table specification later.

If life were perfect, the automatic behavior would be to use the default values for these columns when nothing is specified in the %ROWTYPE variable. I'd settle for having some manual way to specify that as the behavior I want, but I can't find it.

Ironically, there is a way to do this for a record type explicitly defined in PL/SQL - "To set all the fields in a record to default values, assign to it an uninitialized record of the same type". But that doesn't work for %ROWTYPE records (I tried).

reference:
PL/SQL User's Guide and Reference
5 Using PL/SQL Collections and Records
http://download-west.oracle.com/docs/cd/B14117_01/appdev.101/b10807/05_colls.htm#sthref655

Win32::GuiTest

I have given laborious and painful birth to an automated GUI tester. I'm feeling a bit of postpartem depression, because it's nowhere near as clean as I'd like. I chose some complex ways of coding simply because the obvious ways produced errors or strange results that I didn't understand. It was also hard to get consistent results... things would stop and start working without apparent cause.

Bleh. I don't like programming voodoo.

This article says, "the first rule of GUI testing is: Don’t!", and it's really true. It would have been far better to get my developer to design his GUI app for testability in the first place, or at least get him to embed handy hooks in the code itself. Working totally from the outside, trying to write code that looks at and uses a GUI on the screen exactly as you do, is

But still, there are times when you do have to work as a complete outsider to the code. It is nice that there's a way, however painful. And here are some things I learned. Who knows if they will have any applicability to anybody else's Win32::GuiTest project? I'm sure it depends on a million quirks of the system you're testing.

1. Most things I saw on my screen could not be probed by Win32::GuiTest, as far as I could tell. About all I could see were the titles of windows and window elements, and not even all of those.
2. If you SendKeys, don't forget to bring your window to the foreground with SetForegroundWindow first. For me, the same GUI automatically seized the foreground on one machine, but on another machine, it stayed back, and the wrong window intercepted my keystrokes.
3. I wanted to know how long an hourglass wait lasted while a new window came up. I found no good solution. FindWindowLike succeeded too soon - it found the window immediately, long before the user could see it. What I settled for was to count the number of dependent items that GetChildWindows would find for the new window - and keep counting until the number was up to its maximum value. Then I knew the window was ready for business. Kludgey? Oh, yes!
4. Insert lots of sleep()s. They help reduce errors enormously.
5. I had trouble stringing several tests together into a single session. The first test in a single session would work fine, but inconsistent problems cropped up with subsequent tests. Eventually, I remembered that it's a machine and it doesn't get bored, so I made it close the application, get all the way out, and go all the way back in for each test.

timeout with threading

I'd gotten the impression that python's threading module was a very frightening thing, and had put off getting to know it. That's too bad, because when I finally looked at it, it provided a very simple way to add timeouts to some of my code. And - unlike os.fork(), etc. - it works on Windows!

    t = threading.Thread(target=self.code)    
    t.start()
    t.join(self.timeout)
    if t.isAlive():
        completed = False
        self.msg = 'Timed out after %s seconds' % (self.timeout)
        if self.killOnTimeout:
            t._Thread__stop()

interactive perl

I'm writing a GUI tester, which involves me deeply in Win32::GuiTest. It's been a mix of joys and frustrations. Two things have saved my mind.

1. I'm hooked on Python's interactivity, especially when exploring a poorly-documented module. I was missing that badly in Perl; Komodo's interactive mode is only a partial substitute. This tip saved me: http://www.devx.com/tips/Tip/17304?trk=DXRSS_WEBDEV
   
2. Getting a module to export its functions was giving me fits. Just adding the function names to EXPORT_OK in the .pm file only gave me "Can't locate
   auto/... (function).al" errors. This is black magic to me, but it fixed it: http://groups.yahoo.com/group/libwww-perl/message/1470
   

IOUG Live! sent me into rapturous joy. More on that later. I can't believe I was seriously considering not going.

Here's one of the problems that the conference set my mind to.

Some of my homemade procedures fill up some tables with huge amounts of somewhat ephemeral logging data. It's only needed for a couple weeks, after which I need to delete the rows. Doing it the conventional way generates cruel amounts of redo. I've wanted a good script to duplicate the effects of the imaginary command
DELETE FROM logtable WHERE blah blah blah NOLOGGING

The basic algorithm I've seen is
CREATE TABLE newlogtable AS
SELECT * FROM logtable
WHERE (criteria for keeper rows)
UNRECOVERABLE;
DROP TABLE logtable;
RENAME newlogtable TO logtable;

The trouble is that it's hard to get all the dependent DDL to move along. Especially since I don't want any names to change - I name all my constraints, down to the NOT NULLS. The script gets more complicated then.

I found out about DBMS_METADATA.GET_DDL and DBMS_METADATA.GET_DEPENDENT_DDL at Joe Trezzo's "New 10g & 9i features", and it looks like that might be a good basis. It's still tricky, though, because I can't create new dependent DDL until the originals have been dropped. It feels like juggling: generate the new DDL, drop the original dependent objects but not the original table, create the new table with its data based on the stored DDL, then finally drop the original table... hmm.

I'm thinking of relying on the recycle bin (more new IOUG knowledge) for some help here. Deleting objects in 10g really only renames them and designates them as being in a "recycle bin"; you can still access them, you just need to consult the right views to learn their new obscure system-assigned names. (They make an awful clutter in your schema viewed through TOAD, btw.)

Could I drop the table (with its objects) first, then build a new table based directly on the one in the recycle bin? That means all the old dependent object names would already be out of the way, and I can do it all in fewer steps.

Something seems scary about relying on the recycle bin that way, though. Like mailing checks against the deposit you expect to make while your envelopes are still in the mail. I guess Oracle makes no guarantees about the continued availability of objects in the recycle bin; they stay until they're manually purged or the tablespace fills up. But, since it is a very bulky table I'm working with, it's not too farfetched that the recycle-bin version could become unavailable while I'm halfway through this process. Hmm. Darn.

O-verkill-racle

Oracle's had a philosophy for a while now that, "We're the database company; so anytime our software has to interact with any data, for any reason, we'll have it create an Oracle database for the purpose."

The trouble is, an Oracle database represents a LOT of overhead. For instance, an empty Oracle database takes ten times the disk space of an empty postgresql database. Worse, every new database is a new administration burden.

We wanted to do a little bit - a very little bit - with web-based Oracle Discoverer. That meant installing Oracle Application Server, and OAS created its very own infrastructure database. Now I've got this "infrastructure" database sitting around, doing next to nothing, soaking up almost as much disk space and memory as a small production database would. It's 9.0.1 - way behind my own databases - and I have to attend to its obsolete 9.0.1 security needs, or the network security scan gripes at me.

And the shame of it is - a use like this doesn't need the things an RDBMS provides. A store of configuration data for an application server doesn't need multi-user access, transactional consistency, etc. etc. - totally unnecessary. A few flat files would do fine.

It's like a trucking company that won't use a hand dolly because, well, they're a trucking company. So, of course, they need diesel pumps and regular oil changes and overhauls for the big rigs they use to get the copy paper down to the copy room, and maneuvering the 18-wheelers between the cubicles may be a little troublesome sometimes.

passwords in LSNRCTL

Oracle's listener control (LSNRCTL) can take a password. The way you actually supply the password is very confusing, though.

c: > lsnrctl
LSNRCTL> stop
TNS-01169: The listener has not recognized the password

What do you mean, "Did not recognize the password?" You never asked for a password!

LSNRCTL> set password badguess
LSNRCTL > stop
TNS-01169: The listener has not recognized the password
LSNRCTL> set password goodguess
LSNRCTL > stop
The command completed successfully

In other words... 'set password' doesn't mean 'set password' the way normal human beings mean it - 'establish this as the password'. It means, 'this is what I think the password is, so transmit this whenever I do something that demands a password'.

Humph. There should be a category for "not exactly a bug, but nomenclature confusing enough that it may as well be".

For the first time in my life, I find myself worried about mutability.

I want to take advantage of Python's new sets (who wouldn't?). It turns out that sets can only be made of immutable, hashable objects, and a user-created object type is (by definition) mutable.

I can see where it makes sense. If I say
box1 = Box(contents='eggs')
box2 = Box(contents='eggs')
basket = set([box1, box2])
should basket contain one or two items?
Now suppose I
box2.contents = 'yogurt'
Now does my basket have a box of eggs and a box of yogurt, or just one box of yogurt? It's not clear.

The problem is, there are objects I will create that truly never will change, and it truly would be useful to use this wonderful set functionality with them. So now I'd like to be able to create immutable objects, but Python has no particular way to create them.

The only way I've found so far feels a little bit dangerous - to define __hash__. As far as I can tell, it means "You can know me by this". If I do forget and later change data that is the basis of my hash, my object no longer hashes to what it once did - it has moved and left no forwarding address. There are some clumsy ways to try to block later changes to my object's values, but they are... clumsy. (Redefining the object's __setattr__ and raising exceptions therein. Ick.)

Thus, it feels like I'm delving into eldritch secrets, like my soul is toast if I scuff the pentagram. __str__ and __cmp__ are the only double-underscore methods I'd defined previously. They feel downright warm and fuzzy compared to this.

undead Apache processes

Beware of surviving Apache processes that you thought you had stopped.

I just had a miserable time making some changes to httpd.conf on an Apache (1.3)/Windows installation. After making each change, I'd stop/start Apache and test the page with a browser to see if I'd mucked httpd.conf up fatally. The trouble is that I got a lot of false readings with this test, which confused me horribly. (Make change X; page works. Make change Y; page doesn't work. Undo change Y; page doesn't work. Undo change X; page doesn't work. Scream.)

It turns out that, although I'd executed 'apache stop' properly (I think), there were quite a few apache.exe processes lurking alive in the Task Manager's Processes tab. Apparently they were the source of the trouble, because killing them manually during my stop/starts eliminated my false test results.

Oracle 10g upgrade woes

I really ought to initiate a blog with a non-grumpy post! Oh, well. You should have been here a month ago when everything was sweetness and light.

I'm the early-adopter type. I upgraded to 9i pretty early on, and it went well. A couple weeks ago, I upgraded to Oracle 10g. This time, it hasn't gone very well.

At first, performance was absolutely miserable - the CPU was 100% busy and stayed there. But that was my fault. Shame on me, I had accepted the init.ora parameters written by DBCA. Fixing some of them (cursor_space_for_time, sga_max_size, sga_target, fast_start_mttr_target, cursor_sharing, session_cached_cursors) corrected that. I know that's my responsibility, but I'm really surprised that neglecting it caused such an utter CPU jam.

Then, last night, the database started dumping lots of "ORA-04021: timeout occurred while waiting to lock object" errors into the alert log and refusing to do anything whatsoever. The associated trace files mean nothing: "No current SQL statement being executed." I shouldn't have to be looking at binary stack dumps, dangit. At the time, I was trying to do a full export with EXP; I don't know whether that caused the problem. Supposedly, you can still use EXP with 10g, but I wonder if Oracle has stopped bug-checking it vigorously. Tonight, I'll switch to EXPDP.

There were also many unhandled PL/SQL exceptions in packages like WKSYS.WK_JOB - packages that Oracle supplied and that I don't even know what they're doing. Oracle's ever-increasing sophistication means that it's always doing more and more stuff behind my back; with 10g, that stuff sometimes clogs the performance, and sometimes it contains errors. I am not impressed. Postgresql looks more appealing all the time.

The problem might be that I'm one patch behind - I'm at 10.1.0.2, not 10.1.0.3. The patch procedure is pretty involved, though - a lot more involved than it ought to be, for a top-dollar product, IMHO. It will mean a couple hours of downtime, which means it should be done in the middle of the night or on a weekend. I wouldn't mind that, but as a contractor, I'm not allowed to be in the building at those times without dragging my supervisor in to babysit me.