Saturday 29 December 2012

Review of Oracle APEX Best Practices (or the 200 day shower)



Over the past few days I have had two books on my bedside table: on my Google Nexus I have been reading David Eagleman’s novel Sum: FortyTales from the Afterlives, and whenever I put that down, I’ve been burying my nose in the dead-tree (i.e. non-ebook) version of Oracle APEX Best Practices by Learco Brizzi, Iloon Ellen-Wolff and Alex Nuijten. 

This article is about Oracle APEX Best Practices, of course, but let’s talk about the short story collection, Sum for a bit. One of the stories in the book hypothesises that in the afterlife we all relive our lives – with one difference: all our actions are grouped by an ‘order by’ statement. So instead of waking and sleeping, waking and sleeping, we spend 30 unbroken years asleep and then 40 continuous years awake. We spend 7 straight months making love, and then 6 full days clipping our toenails. We shower for 200 days. It’s a good book; read it.

Oracle APEX Best Practices is a good book too. I want to make that clear.  Two of the authors, Learco Brizzi and Iloon Ellen-Wolff are new names to me, but I have visited Alex Nuijten’s blog many times in the past and have never been let down. So I googled the other two authors, and what I read convinced me of their bona fides.

Not that the book doesn’t speak for itself. At 200-odd pages it is a good size; it is written in clear, precise English; and it covers Apex, from top to tail – from installation, through security and debugging, to deployment. It is mostly aimed at Apex beginners, but there is enough in-depth explanation in it to make it invaluable to intermediate Apex-ers too. It is a very good book.

But… 

In the afterlife in Sum, you spend 14 minutes experiencing pure joy, and then 27 long hours of intense pain. I have spent 2 full paragraphs praising Oracle APEX Best Practices; it is time to highlight my quibbles.

The first annoyance is this: to keep the page count down, the authors often ended sections with the url of some Oracle documentation or a blogpost containing more information. This is perfectly fine in an ebook, but what am I supposed to do with this in a paper book? I cannot type a 200-character url into my browser's address bar. Physical books cost more money than ebooks; they should not provide a worse experience.
My kingdom for a bit.ly!

I do also have some questions about the title. Best practices? This is a very good Apex book, no doubt, but I wouldn’t call it a best practices book. I expect best practices books to prescribe the way I develop my application and not just explain the options.

Finally, the authors occasionally struggle to stay true to their topic: Apex. In Chapter 2: Leveraging the Database, they spend 30 pages explaining how to use analytic and aggregate functions in Oracle databases. I was very grateful for the lesson – but what the hell is that doing an Apex book? What next - a fried rice recipe?

<\criticism >

Do not be put off by my quibbles; this is a good book. It was, unfortunately, released just on the eve of Apex 4.2 and therefore only covers up to 4.1; however, I do not think that this is a problem. Best practices are not version-dependent.

In Chapter 4, the authors discuss application security, advising on how to guard against SQL injections and URL tampering, and explaining Access Control Lists and dealing with password complexity rules using APEX_INSTANCE_ADMIN. And Chapter 5 is a 30 page essay on debugging and troubleshooting. Both these chapters are excellent and, along with the chapter on printing (BI Publisher, Apache FOP, Cocoon and JasperReports), are worth the cover price on their own (insofar as any software development book is worth its crazily-inflated price).

However, more than the long pieces on complex subjects, what I found useful in this book were the little reminders of things that I know and overlook: set up your User Interface Defaults before diving in and building your application, for example.

And even though I have said that the section on analytic functions, aggregate functions and index-organized tables should not be in an Apex book, I must concede that it is packed with useful (non-Apex) information.
According to Sum: Forty Tales from the Afterlives, we spend 1 whole year of our lives reading books. I do not regret the hours that I have spent on this one. 

(Full disclosure: I received a complimentary copy of this book from the publishers, Packt Publishing, in return for an unbiased review)

Friday 24 August 2012

Authors & Awards: Usain Bolt with a pen

This post has absolutely nothing to do with Oracle, and even less to do with Apex. Sorry - you can stop reading here and I won't hold it against you.
However, I recently turned my attention to my other love, literature, and did a spot of datajournalism for another blog.  I'm reprinting my piece here.

I'll be back waltzing and wrestling with Oracle Apex very soon. Promise.

Africa's Leading Literary Powerhouses



Summer 2012 and the eyes of the world turn to the city of London, England. People across the globe sit in beer parlours, shebeens, pubs, washing unhealthy snacks down with tankards of beer and cheering as the healthiest specimens of our nations run, jump and swim faster, higher, stronger.
But why stop at sportsmen? Why not pitch our countries’ plumbers against the world’s, our street-corner hookers, our brain surgeons? Why don’t Liberians sneer at Sierra Leoneans: “The barefoot kids hawking peanuts in your Kroo Bay slums are nothing compared to the former child soldiers weaving through traffic selling groundnuts in the misery of our West Point”? And why not stand our writing ‘athletes’ up against each other in a sort of literary Olympic Games and see which nation ends up on the podium?

And the last was just what we did. The rules: We sourced our data by analysing the winners of major international literary prizes, filleting out all African winners and noting their country of origin. We limited our scope – and therefore our resultset – to awards for English language literature, with a deliberate bias towards prose fiction. Where a writer has dual nationality, as in the case of Zimbabwean-British author Doris Lessing, we favoured the African nation; with Mauritian-South African novelist, Lindsay Collen, we plumped for Mauritius, as this is the country she identifies with.

Continue reading...

Wednesday 4 July 2012

Introduction to Oracle Flashback Technology (Part 3: Flashback Database)

In this series of articles, we are discussing the collection of tools that make up the Oracle Flashback Technology suite. They are varied in their purpose and in their implementation, but together they act very much like a morning-after pill for your database, a means by which developers and administrators can protect their database from immediate past errors and indiscretions. In the first part of this series, I talked about the flashback tools that every developer should have in his arsenal, and in the second we rummaged through the pile looking for things that might be of interest to administrators.  However, I have saved the biggie for last: Flashback Database.

What is Flashback Database?
 For those times when minor surgery is not enough, Flashback Database can apply a defibrillator to the heart of your database, shout “Clear!” and proceed to shock it back to a happier time. This is because Flashback Database is a mechanism that allows administrators rewind an entire database to a past time or SCN.  Flashback Database has most of the advantages of a point in time recovery, with  only a fraction of the hassle and longwindedness. Flashback Database, unlike some of the lesser members of the Flashback family, is a physical-level recovery mechanism; it uses its flashback logs to access past versions of data blocks, finessing the final result with information from the archived redo log where necessary.

Setting Up Flashback Database
Flashback Database must be enabled before you can use it.  This simple query will tell you whether it is or not:

SELECT FLASHBACK_ON FROM V$DATABASE;

The result will either be YES or NO.  If Flashback Database is not enabled and you wish to enable it, you will need to ensure that the database is in ARCHIVELOG mode first.

There are two other prerequisites for enabling flashback database.  You will need to instruct the database on where to store the flashback logs that it will generate and you will need to tell it how much space it can use up doing so. These are expressed in two instance parameters – DB_RECOVERY_FILE_DEST and DB_RECOVERY_FILE_DEST_SIZE.  Remember to set the size before the location.

ALTER SYSTEM SET DB_RECOVERY_FILE_DEST_SIZE = 9G SCOPE=BOTH;

ALTER SYSTEM SET DB_RECOVERY_FILE_DEST = ‘/oracle/flash_recovery_area’;

When Flashback is up and running, your control over the Flashback Recovery Area, which is what the location you specify in the DB_RECOVERY_FILE_DEST parameter is called, will be limited; the database will create and age out the flashback logs as necessary. One way in which you can attempt to manage the Flashback Recovery Area is to specify a retention target, telling the database how long you would like it to retain its logs before they are overwritten by newer ones. Whatever retention target you specify, however, is not set in stone; in practice, the length of time that logs are retained will depend on the intersection of the retention target and the space made available by the DB_RECOVERY_FILE_DEST_SIZE parameter. If the database runs out of space, it will kill off old logs, irrespective of the retention target.


The retention target has a default of 1 day and is set in minutes, thusly:

ALTER SYSTEM SET DB_FLASHBACK_RETENTION_TARGET = 2880; /* 2 Days */

This done, you are now ready to flick the switch on Flashback Database. To do so you will need to shut down cleanly and mount the database, and while in this mode run the following:

ALTER DATABASE FLASHBACK ON;

Continue reading...

Tuesday 3 July 2012

Introduction to Oracle Flashback Technology (Part 2: Database Administrators)


In the first in this series of articles on the tools in the Oracle Flashback Technology armoury, I discussed what is available to application developers, the various means by which they can rewind the clock and correct recently-past data errors.  However, if this was a world in which the worst thing that could happen was a small mix-up with data, what a happy world it would be. However, this is not Disneyland, and data errors are not the big, bad wolf at the door; it is errors that raze whole database objects that database administrators fear. But fortunately, Oracle Flashback can come to our rescue in those situations too.

First a quick recap: Oracle Flashback Technology is actually a clutch of different Oracle tools that enable developers and administrators reach into the immediate past of their database to recover from errors, without having to resort to point-in-time recovery options. For this reason, they are relatively fast and incur a relatively low logistical overhead (since the database does not need to suffer any downtime to implement them).

FLASHBACK TABLE:
Flashback table enables you revert a table to its state at a specified timestamp or SCN. Any rows of data created after the parameterized timestamp disappear and any deleted rows zombie back to existence; in addition, associated attributes such as triggers, indexes and constraints remain unviolated. And all this magic occurs without even the slightest blip of interference affecting the rest of the database; it stays up and available.

Flashback table is able to sidestep the unpleasantness of restoring from backups by using information in the undo tablespace to restore the table.

Imagine this scenario: some catastrophe has befallen the EMP table and the developers are unable to unpick the data. All eyes turn to you: Save us, they beseech, wave your wand and erase the past hour. This is where you would find Flashback Table useful:

FLASHBACK TABLE EMP TO SYSTIMESTAMP – INTERVAL '60' MINUTE;


And it’s done. Day saved. Simple.



To use flashback table, however, there are a few prerequisites that need to be in place:
  • You must have the FLASHBACK privilege on the table. Or you must have the FLASHBACK ANY TABLE system privilege.
  • You must have the SELECT, INSERT, DELETE and ALTER privileges on the table.
  • Row movement must be enabled on the table.
  • Finally - and crucially - the information in the undo tablespace must extend back far enough to cover your timestamp or SCN.
Continue reading...

Monday 2 July 2012

Introduction to Flashback Technology (Part 1: Application Developers)

Imagine, if you will, that someone invented time travel. What would you do with it? Nip back into the past and fix that error you made? And maybe you’d want to do something about that silly, ill-advised haircut you had when you were 17. Well, the good news is that, in a manner of speaking, time travel has been invented. At least as far as Oracle databases go.

And the bad news?  There’s nothing it can do about your teenage haircut.

Oracle Flashback Technology is a suite of features that enables Oracle application developers and database administrators to wind the clock back and examine their database objects and data in a previous state without having to resort to backup media. Simply put, Oracle Flashback Technology is time travel for databases.

In this series of articles I intend to expound on the benefits of Oracle Flashback to application developers and to database administrators. This first installment will concentrate on the tools that will be of interest to application developers.

Oracle Flashback Technology is actually a collection of features that enables the developer to examine the database’s recent past in a number of ways:



Oracle Flashback Query:
Flashback query is a neat trick. It allows you run a query against your data as it was at past time. Imagine you accidentally delete a number of rows and commit your changes. With flashback query you can easily reach into the past and retrieve those lost rows.  Using the SELECT... AS OF clause you can examine the ghost of data past by referencing its timestamp.



SELECT *
FROM emp AS OF TIMESTAMP TO_TIMESTAMP('2012-04-21 17:00:00','YYYY-MM-DD HH24:MI:SS');


And if you wanted to view those rows that you accidentally deleted 5 minutes ago?



SELECT *
FROM emp AS OF TIMESTAMP (SYSTIMESTAMP - INTERVAL '6' MINUTE)
MINUS
SELECT *
FROM emp;


Continue reading...

Thursday 16 February 2012

Introduction to Materialized Views: Query Rewrite


In the fairy tale, The Shoemaker and The Elves, a poor cobbler works to the best of his ability; and somehow, when he isn’t looking, magical beings would sneak in and polish and perfect his handiwork.

It seems the people at Oracle are up on their fairy tales. Perfectly-tuned queries are the Holy Grail to developers and DBAs; we are constantly fiddling, forever optimising, shaving milliseconds from the total run time. Well, since 8i, functionality has been in place that can take our DML statements and magically improve them. 

Oracle do not admit that this involves any elves. But I have my suspicions.

How It Works:
Here’s how it works: part of the reason queries involving joins between data-heavy tables can take such a long time is that the engine needs to compute aggregates and disentangle the joins between the tables before it can run the query. However, materialized views, by their very nature, have all that information precomputed. Therefore, if you can find the right materialized view and match it to the right query, you can cut running times dramatically.

This trick, called Query Rewrite, only works with SELECT statements. However, those statements may be hidden away in a CREATE TABLE … AS SELECT statement, or an INSERT INTO … SELECT statement, or they may be squirrelled away in any type of subquery or sub-clause. 

Additionally, Query Rewrite must be enabled at the initialization parameter level:


Alter session set query_rewrite_enabled = TRUE;

Continue reading...

Tuesday 7 February 2012

Alter session set triggers disabled (or The personal anti-ninja assassin insurance policy)

If your database was a city, your tables would be buildings, your constraints would be the streets and motorways that connect them, your records would be the citizens that populate them, your views the nightclubs and restaurants where like-minded records go to hang out.

And triggers? Well, they'll definitely be ninjas. Think about it: easily overlooked, moving with stealth, 'firing' on command, shrouded in mystery.

Triggers, of course, are immeasurably useful. You don't need me to tell you that. However, there are times when you wish you could flip a switch and turn them off: imagine you are inserting or updating a truckload of data and cannot afford the overhead, or you are tracking a simple statement and do not want it to take any unexpected detours. 

Of course, you can knock up a quick anonymous pl/sql block to loop through all your triggers (or at least all the relevant ones) and dynamically disable them.

/*
** This block will disable all database triggers. Do NOT do this.
*/
begin
  for i in (select trigger_name
            from user_triggers) loop

            execute immediate 'alter trigger '||i.trigger_name||' disable';
  end loop;
end;

"Ladies and gentlemen, this is your pilot speaking. We will soon be flying across the Himalayas mountains. I will fly this section with my eyes closed. Because I can. Please, sit back and enjoy the complimentary peanuts and your in-flight movie, Alive."

Just because you can do a thing, does not mean that you should. You wrote those triggers for a reason; so unless you can guarantee that no other users will be using your database, switching them all off for any length of time might be a tad unwise. What you need in those circumstances is the ability to disable triggers for only your session. It'll be like your personal anti-ninja assassin insurance policy.

Unfortunately, ALTER SESSION SET TRIGGERS DISABLED; is not a valid Oracle command. So until it is, we are going to need to create a substitute for it.

The easiest way to do this would be to create a session variable, a boolean that you can switch on or off, and then edit your triggers to only fire dependent on the state of this variable.

The first bit is easy. A variable in the specification section of a database package makes a perfect session-specific boolean.

create or replace package trg_pkg as
   skip_trigger  boolean := FALSE;
end trg_pkg;

Now all we need do is add a line as the first bit of executable code in our triggers checking the state of our trg_pkg.skip_trigger variable. If it is TRUE, we exit the trigger without executing any code; if not, we fire the trigger as normal. Locate the triggers you may want to switch off and add the following to them:

 
...
BEGIN
  if trg_pkg.skip_trigger then return; end if; -- do not execute the trigger.
...

And that's it. Whenever you do not want triggers to fire in your session, all you need do is flip your trg_pkg.skip_trigger switch. The triggers won't fire for whatever commands you carry out, but will work as normal for everyone else. (For security reasons, you will probably want to control which users are granted execute permissions on your trg_pkg package.)

One last thing, if you want to add the clause to all the triggers in your schema - as I have needed to do recently - you might find the following code handy.

/*
** This script will rebuild all triggers adding a clause which
** prevents the trigger from firing if trg_pkg.skip_trigger is TRUE.
** This code presumes that no trigger is longer than 32767 characters long. 

** 04/02/2012  David Njoku * Created this script.
*/

declare 
 vSql  varchar2(32767);
 vBackup  varchar2(32767);
 vString  user_source.text%type;
 
 vCLAUSE  CONSTANT varchar2(500)  := ' if trg_pkg.skip_trigger then return; end if; -- Do not execute this trigger if trg_pkg.skip_trigger is set to true. ';
 
 
begin 
 -- Find all the triggers that relate to this schema's tables.
 for i in (select at.trigger_name, at.owner
    from all_triggers at, all_objects ao
    where at.table_owner = USER
    and at.status = 'ENABLED'
    and at.trigger_name = ao.object_name
    and at.owner = ao.owner
    and ao.object_type = 'TRIGGER'
    and ao.status = 'VALID'
    and not exists (select 1
      from recyclebin
      where object_name = at.trigger_name)) loop 
    
    -- Now that we've got the trigger name and owner, we need to grab a hold of the existing code, in case we need to revert.
    vBackup := ' CREATE OR REPLACE ';
    for j in (select replace(text,chr(10),' ') text 
       from all_source
       where owner = i.owner
       and name = i.trigger_name
       and type = 'TRIGGER'
       order by line) loop 
       
       vBackup := vBackup||chr(10)||j.text;
    end loop;
    
    /*
    ** We now need to start building our new version of the trigger.  The plan is to add our if clause immediately after the BEGIN 
    ** Some triggers will contain more than 1 BEGIN. Since we cannot tell which is the correct one, we'll need to add our clause after
    ** all of them.
    */

    vSql := ' CREATE OR REPLACE ';
    for j in (select replace(text,chr(10),' ') text 
       from all_source
       where owner = i.owner
       and name = i.trigger_name
       and type = 'TRIGGER'
       order by line) loop 
       
       -- do we need to add our clause here?
       if replace(upper(j.text),' ','') like '%BEGIN' then 
        vSql := vSql||chr(10)||j.text||chr(10)||vCLAUSE;
       else 
        vSql := vSql||chr(10)||j.text;
       end if;
    end loop; 
    
    -- Now we've got our trigger we should try building it.  
    begin 
     execute immediate vSql;
    exception 
     when others then 
     dbms_output.put_line(i.trigger_name||' '||sqlerrm);
    end; 
    
    -- If the trigger was not built successfully for whatever reason we should revert to the orginal.
    for j in (select 1
      from all_triggers
      where trigger_name = i.trigger_name
      and owner = i.owner
      and status = 'DISABLED') loop 
      
      begin   
       execute immediate vBackup;
      exception 
       when others then 
       dbms_output.put_line('BACKUP '||i.trigger_name||' '||sqlerrm);
      end;
    end loop; 
 end loop;
end;
/

Wednesday 18 January 2012

Introduction to Regular Expressions in Oracle

Regular expressions – RegEx or RegExp to their friends – were fashionably late in coming to the Oracle party. The seeds of what we know today as regular expressions were formulated in the period immediately following the Second World War in fields as diverse as formal language theory and neurophysiology; but it wasn’t until the PL/SQL Web toolkit arrived for 8i and 9i that they popped their heads over the Oracle parapet. By this time they were already rock stars in the worlds of Unix and Perl.

But I’m being rude here; I should introduce you. Regular expressions are a codified means to accomplish flexible pattern-matching in strings of text.

And, if you’re rereading that definition and thinking ‘huh?!’, that’s a feeling you might want to get used to, because, while regular expressions are powerful and definitely useful, they can seem very much like hieroglyphics to the uninitiated. But don’t worry, I’ll hold your hand.

String pattern matching has always been possible in Oracle sql and pl/sql, of course. With the LIKE condition and the simple metacharacters of “%” and “_”, character patterns could be described and integrated into queries. However, the moment you required a little complexity these basic wildcards cowered in fear. An example: my name is David, but friends tend to call me Dave when they have a beer or two in them. To match either spelling of my name, not using regular expressions, you would have to query the table for ‘Dav%’. But then your results might be flooded with Davidsons and Davises and Davinas. And I’m definitely not pretty enough to be called Davina!

Tuesday 10 January 2012

Introduction to Autonomous Transactions (Or If Obama And Merkel Had A Lovechild...)

I don't know if you've been paying attention, but you may have noticed that there's a bit of a recession going on at the moment. Stock prices are crashing around the globe, and I'm pretty sure that the Euro is now worth less than Monopoly money. It's very much like the Great Depression of the 1930s all over again - except that, this time, the bankers don't have the decency to throw themselves out of windows.

The solution, David Cameron, Angela Merkel and Barack Obama assure us, is for everyone to go out and spend more money. That'll kickstart our economies, they assure us. More transactions. A lot more.

So they'll probably thank me for this article that I wrote over at All Things Oracle about autonomous transactions. Have a read; you may thank me too.


Autonomous transactions are a bit like that dark part of the forest, beyond the rickety bridge, where the shadows are shaped like witches and bats swarm soundlessly from the trees. We all know it’s there, but the wise are in no hurry to visit.
Here, however, is a definition: an autonomous transaction is a completely independent transaction that is started by temporarily suspending the parent/calling transaction, which is then resumed after the independent transaction is completed. Or, if you prefer a real-world analogy: an autonomous transaction is a little like getting married, but having an affair on the side.
Morality isn’t the reason autonomous transactions are frowned upon; they possess the potential to get out of hand. Autonomous transactions are completely self-contained transactions that commit independently, and so to start one is to juggle two transactions...



Tuesday 3 January 2012

Should Your Next Project Be Written In Forms or Apex? (Or The Polygamist's Dilemma)


The polygamist pauses, his hand hovered above the doorknob. He looks back down the hallway, blinking in the harsh light of the naked bulb he has just flipped on to do war with the shapeless darkness of the night. A few paces away - no further away than a firm decision - stands another door: silent, inviting. His gaze returns to the door in front of him; also as silent as the entrance to Aladdin's cave. Which door should he open? Which of his beloved wives should he visit?

Okay, I admit it, I know nothing about polygamy. I once caught an episode of Big Love on telly, and I enjoyed reading Lola Shoneyin's The Secret Lives Of Baba Segi's Wives, but that's about it.

I do, however, know a bit about Oracle Forms and Oracle Application Express. Which is kinda fortunate since I'm guessing you're not here for marital advice. A question I do get asked from time to time - most recently on LinkedIn - is this: What technology should I choose for my new project - Oracle Forms or Oracle Application Express?

At the risk of sounding like a guru sat atop a Himalayan rock, there is only one real answer to this question: look within your heart,  the answer floats like an asteroid in the galaxy of your inner space. Or something. Personally, if I was making the decision a year ago I'd have said Apex cos it was new and exciting; today, I'd probably say Forms, cos it's old and I better understand its foibles: the polygamist's dilemma - the new, nubile wife, or the old wife who he fully understands?

You will probably want to base your decision on factors more concrete and quantifiable than imaginary asteroids, so here are some things to consider.

Skills: What skills do you have? What skills are you willing to invest in? What skills are you capable of acquiring? I would imagine that you have a greater wealth of Forms experience. This need not be the deciding factor, but it should definitely count as a tick in that box. I was able to retrain my team of Forms developers in Apex, but it took some time and some expense. And, crucially, it was a challenge they were happy to face: I still have nightmares about The Big Oracle ADF Disaster of 2008. My therapist says I'll get over it one day, but I doubt it.

Platform: A huge part of your choice must depend on the demands of the project itself. For instance, what platform are you hoping to run it on? If you are planning to take advantage of the current boom in mobile computing, then Apex should be your choice. I must admit that I have not yet tried to run a Forms application on a phone or a tablet, but I do not suspect that that is an experiment that would end happily.

If, on the other hand, you expect your application to be run only from PCs and laptops using a myriad of browsers, with or without javascript enabled, then possibly Forms with its browser-agnostic java applet might be worth a look-in.

Cost: If you are a millionaire Saudi Arabian prince who has only got to this page cos you googled "I am looking for the oracle on polygamy", you can skip this bit. The rest of you, pay attention. With its middleware costs,  Forms is undeniably the more expensive option. I am not allowed anywhere near the money at my company, but I am assured that the difference in cost can be considerable.

Audience: Who is your application aimed at? If it is to be published on the open Internet, then you only really have one choice - Apex. I am not denying the work Oracle's Forms team has done to modernise the old girl, but I doubt that it will ever be truly suitable for the world-wide web.

However, if you are building an enterprise application, then there is definitely something to be said for the solidity and gravitas that Forms still exudes. And if your application will be used mostly for data-entry, then Forms provides speed that Apex, even with its clever javascript and ajax hooks, cannot match.

Extendability: The answer to the question of which is the more extendable technology may not be as obvious as it may first seem. Being a web technology, Apex can more easily sup at the banquet of ajax and javascript, can more easily flirt with jQuery and whatever the next big thing will be. But with each iteration of Forms, its door is opened wider, with Java pluggable components and javascript and CSS support. However, I believe that Apex will always outpace Forms in this respect; the story of the tortoise and the hare was only a fairy tale.

Speed: I cannot get my hands on the exact figures now, but I have read that for every tenth of a second that Amazon manages to shave off their website's response time, sales increase by many millions of dollars. Google experienced a similar increase in the stickiness of their site when they introduced Google Instant. If response times are as crucial to your project, then perhaps you should be leaning towards lightweight Apex. But do not dismiss Forms out of hand; from 11g, it is possible to slash load times by pre-starting runtime engines. The tortoise may have bought itself some roller skates. 

Deployability: Our IT team are forever engaged in a battle with Oracle Application Server. One of them - OAS or the IT guys - isn't very good at their job. Since no one at Oracle has ever bought me a drink at the pub, I'll assume it's their fault. Apex is easier to deploy than Forms; indeed there is a growing number of companies that will host your Apex application for you.

The Big Picture:  You have to consider the wider context. In my case, the option of building Forms applications is a receding one, since my company has taken the decision to move towards Apex. We now have a cohort of new, young developers to who Forms is Betamax.

Conclusion: The polygamist makes his decision and quietly opens a door. He enters the room and closes the door behind him. Are you awake? he whispers into the darkness. There is a crack in the curtains, and a shaft of milky moonlight is lying across the bed like an abandoned sword. Yes, a voice whispers from beneath the covers. He walks towards the sound, his pace quickens with excitement.

There is no universally right or wrong answer; it depends on your project and the factors I have named above (and the many I have not thought of - which is why there is a comments section below).

PS: I'd like to thank Steve Cairns for his help with this post. He is one of the country's leading experts on polygamy. Or Oracle technologies. I forget which