Saturday, June 17, 2006

Saw this and was intrigued...

Saw this and was intrigued. What do you think? Good idea or bad (would something similar about Oracle make sense?)

Me, Iā€™m on the fence ā€“ serious things take serious time, but ā€“ is everything serious?


Blogger Daniel Fink said....

Can a very serious topic be comprehensively covered in one minute? Not even close.

However, is one minute enough to impart some knowledge? Perhaps a topic that provokes additional thought or provides a nudge in a certain direction? Absolutely.

Think about how many issues can be described in a single sentence? Not that this sentence exhaustively covers the topic, but it distills the essence.

"The most optimized statement is the one that is never run." Anjo Kolk

"Work first to reduce the biggest response time component of a business' most important user action." Cary Millsap

"I think therefore I am." Descarte

Sat Jun 17, 10:23:00 PM EDT  

Anonymous Partha said....

"but ā€“ is everything serious?"

If it is technical, then it better to be on the safe side and serious about it. We have too many people giving too many 'easy' answers which never suits all requirements, so, any 1-minute tip especially if it comes from an 'expert', will be treated literally by someone and might be taken out of context.

If this one-minute tip is for a very restricted audience with whom we are comfortable with (like an internal audience) then it "just might" be something that people can think about.

But most of the time in technical, especially in Oracle, the answer will be "it depends".

Word verification: tklean - tom kyte suggesting a lean solution ?

Sun Jun 18, 12:02:00 AM EDT  

Blogger Roderick said....

The plusses and minuses probably run along the same lines as the discussion around the OraQA site a while back. I'm still not into audio podcasts myself since I have a short commute and find it much easier to skim over written material. Video might work for me if it's entertaining.

Sun Jun 18, 01:52:00 AM EDT  

Anonymous Anonymous said....

Daniel Fink said....
Can a very serious topic be comprehensively covered in one minute? Not even close.

Have you picked one to listen to ?
"One Minute" is just a catchy term here - they're longer than that.
Basically the idea is "Short Podcast"

Sun Jun 18, 02:10:00 AM EDT  

Blogger Doug Burns said....

It probably depends on what the tip is about. I don't think it's necessarily the case that every quick tip is a bad one, but it's a definite danger.

In the past, I questioned whether OraQA was a good idea or not, but more because there are so many different sources out there.

I think this type of thing would work best if it highlights the other, deeper sources of information on the subject. In a similar way that a 45 minute conference presentation might not get into the same depth as a course, but give you something to think about and then point you in the right direction.

Sun Jun 18, 07:43:00 AM EDT  

Blogger Niall said....

I rather think that a one-minute tip might be better than a 5 or ten minute tip. One of the presentations that made the biggest impression on me at last years UKOUG was the 60 tips in 60 minutes. Enough to give you an idea of something to look at, not enough to give you the idea that you could just go away and apply.

Sun Jun 18, 11:27:00 AM EDT  

Anonymous Mark A. Williams said....

Hmm... One minute tips might just take on an aura of "rules of thumb". Something to which you are (rightly) fairly adverse.

- Mark

Sun Jun 18, 11:49:00 AM EDT  

Blogger Bill S. said....

If this is used as a means to get people to read the fine manual, sure. AS a replacement for reading the fine manual? Nope. I'm all for spurring on additional learning. Like the others have said, though, this smacks too much of "easy answer" for newbies. Need a good balance. :-D

word verification: rlnsrm
Required Learning Now So Read Manual

Sun Jun 18, 12:28:00 PM EDT  

Anonymous mikito harakiri said....

Just have read on other forum -- related to earlier topic:

"Many high school students (who have to study war and piece in 4 volumes) feel sorry that it was Pushkin who was shot on a duel, and not Tolstoy".

Sun Jun 18, 01:52:00 PM EDT  

Blogger David Aldridge said....

One minute sounds just about right. "Bind variable in one minute" ought to be enough to raise the major points and caveats so that a person is stimulated enough to start investigating further. I bet the the importance of good statistics could be similarly encapsulated, or the major benefits of partitioning.

Personally I don't think that even a one hour talk on a topic is enough to get a person fully familiar with any topic -- until they have sat down and types in an example themselves they cannot know more than 25% about a topic. The key thing is to stimulate them enough to spend their own time typing, experimenting, and researching, and a minute ought to be long enough for that.

Sun Jun 18, 02:41:00 PM EDT  

Blogger Noons said....

Not sure I agree with the one-minute thing. There is plenty of that in the game of golf and it has done immense damage to folks who want to play it seriously. I'm siding on with the "too close to rules of thumb for comfort" view.

Sun Jun 18, 08:45:00 PM EDT  

Blogger Rachel said....

depends on how much is covered in that one minute. If it's a VERY specific "how to do this" sort of thing, then sure, it's great. I wouldn't want a one minute tip on how to recover a database though!

And if the one minute tip is enough to get someone interested in learning more, then it's definitely worthwhile

Sun Jun 18, 10:19:00 PM EDT  

Anonymous Anonymous said....

Finding no way to ask you a question I am seeking help from you which not a correct place to do so?Is there any way to restrict the DBA from dropping the triggers created by the users?If yes, please let me know how to do it.


Mon Jun 19, 08:39:00 AM EDT  

Blogger Jeff Hunter said....

I think if they were used appropriately for building blocks, they might be OK. However, even with the simplest decision sometimes there are more things to consider than the basics.

Mon Jun 19, 09:13:00 AM EDT  

Anonymous Anonymous said....

A little knowledge is a dangerous thing.

Mon Jun 19, 11:08:00 AM EDT  

Blogger Thomas Kyte said....

restrict the DBA

sure, don't use DBA, use MY_DBA after you create it and grant to it whatever you want (or don't want!)

No one says all DBA's should be created "equal"

Mon Jun 19, 11:31:00 AM EDT  

Anonymous Anonymous said....

I really like the 3-minute-ish video blogs called 'whiteboards' at 'zdnet' (

Although perhaps not especially suited for detailed "how-to's", there great for getting a heads-up on big areas -like say- Flashback.

Mon Jun 19, 11:59:00 AM EDT  

Anonymous Mark from NY said....

I think it could work. Obviously you couldn't have a quick tip on "Why is my query slow?" But there are some quick tips that could really work in my opinion. Examples:
- Do not use SYS for regular DBA duties, and explain why
- Your mantra about using SQL if possible, then considering PL/SQL, then Java. Back it up briefly.
- Simple tips about SQL Plus, such as how to format columns (I often hear complaints about SQL Plus formatting, but no one reads the manual). Could always recommend Raptor here.
- Explain why forcing the optimizer to use an index on a column in your where clause is not a good thing, and may actually be slower than a full table scan (well, this one might be iffy)
- Why you should nearly always use ARCHIVELOG mode

My two cents,

Mon Jun 19, 01:02:00 PM EDT  

Blogger Joel Garry said....

I don't see the difference between this and an elevator speech.

Given the percentage of audio comprehension and retention, it may be a bad idea. You may be able to recite the Oscar Meyer song, but how many times have you heard it? Are you going to listen to a 3 minute podcast more than once?

Don't blink.

word: xadpafy
I couldn't make that up.

Mon Jun 19, 01:21:00 PM EDT  

Blogger Alberto Dell'Era said....

Instead of a 1-minute podcast, why not a 1-page pill, such as

"instead of

for r in (select x from source)
insert into target (x) values (s.x);
when dup_val_on_index then
end loop

you can write

insert into target (x)
select x from source
log errors reject limit;

and you'll get not null, etc violations as well.

Read this link to know more."

A written pill is much more effective (vastly greater bandwith) than an audio podcast IMHO.

Mon Jun 19, 02:26:00 PM EDT  

Blogger Noons said....

"Read this link to know more."

That reminded me of the constant media line in Starship Troopers. You know, when the news flashes come in, then 10 seconds later the "do you want to know more?" button comes up.

Might be a way of approaching this whole thing?

Mon Jun 19, 09:00:00 PM EDT  

Blogger Gleisson Henrique said....

I wonder what Howard J. Rogers , aka Dizwell, would say about it. I guess I already know ! Haven't we covered the issue of short and quick advise about oracle on this entry . I know one is q&a and the other is podcast, but I see quality of the content being the problem and we have to agree that there is not a quick and short way of doing serious things. I didn't want to, but I will have to use a quote "serious things take serious time"

Tue Jun 20, 12:00:00 PM EDT  

Anonymous Anonymous said....

I believe that ,regardless of a one-hour presentation or a few chapters on the same topic, it still takes that one-minute thought that occurs somewhere, somehow and actually helps us understand the whole thing..Therefore the urge to read/understand it all over again which David Aldrige rightly says "experimenting, and researching"

Tue Jun 20, 12:08:00 PM EDT  

Blogger jk said....

Bad Idea, Too much to sort through to find a one minute nugget of information from possibly questionable source. I would rather check well-known blogs or podcasts of interest, like TK's, HJR, J. Lewis, etc. Good Example was Cary Millsaps web conference filling up within hours of being advertised.

Tue Jun 20, 01:19:00 PM EDT  

Blogger Doug Burns said....

Okay, here's an example of a (less than) one minute tip.

"The problem with the INDEX(tab1 ind1) hint that will direct the cost-based optimiser to use the ind1 index on tab1 is that, should the index be dropped and renamed for some reason, the hint won't work any more. In 10g, there is an improved version of this hint

INDEX(tab1 tab1(ind_col1, ind_col2))

By specifying the index columns rather than the index name, Oracle can pick up any index with these columns as the leading portion of the index key"

It doesn't explain the depths of the optimiser, or hints, or execution plans or doing the minimum amount of work to retrieve the required result (all enormous topics), but it does tell me something useful in a compact format that I might have missed otherwise.

So (and I really mean this) I'm open to hearing what's 'wrong' with this?

Possibly the tip works better the more you know about Oracle, and the danger of hot tips is in trying to reduce complexity where it shouldn't be reduced, but then why not cater to every level of experience?

The first time I heard that nugget of information (and it was in much less than a minute) I thought 'mmmm, that's useful'. One minute tips are *not* a good way of learning about Oracle from scratch, but I think they can be useful in their own right.

Tue Jun 20, 05:43:00 PM EDT  

Blogger Howard J. Rogers said....

I think it must be all the additives in the food these days. How else to explain most people's complete inability to concentrate?

Or maybe, most people are just lazy and looking for easy answers... instead of doing what they ought to be doing, which is to read deeply about a subject and actually acquire some **understanding** about it.

If you understand a subject, you have no need for 1 minute "tips". Conversely, only those who can't be bothered to acquire understanding have a need for the 1 minute tip.

Life is already broken down into the 7-second soundbite. Let's not encourage the idea that databases can be managed in the same way, please!

Tue Jun 20, 05:55:00 PM EDT  

Blogger Doug Burns said....

Howard said ... (Things I tend to agree with, but ...)

"If you understand a subject, you have no need for 1 minute "tips". "

I don't agree with that. There is so much Oracle-related information to keep up with that I often appreciate someone saying - 'Do you know you can do x in 10g?'. It's much what Niall mentioned earlier here about last year's UKOUG presentation. I didn't learn how to use the features mentioned, but I knew there were some things that I should go home and learn about.

Otherwise I might have been completely unaware of a good solution to a problem whilst studying a less efficient solution.

I definitely think caution is required here, though.

Tue Jun 20, 06:06:00 PM EDT  

Anonymous Arup Nanda said....

What makes a question or topic serious? Drop partition operation with "update global indexes" clause - will that pass as a serious or very informative tip? to someone who knows this specific tip, it may seem trivial to mention. How about someone - with years of experience and a lot of grey hairs to prove it - didn't catch it the first iteration of manuals? It may save him (or her) a ton of work - with minimal explanation in the podcast. I just fnished giving a talk to roomful of DBAs and dvelopers on security and mentioned in one my of demos this clause. Not a single one in the room knew about it. And that was a pretty seasoned crowd. The trick is to use topics or question that can be reasonably covered in short paragraphs. I like Doug Burn's example; it's a great candidate. Others - set the audit_trail to "db" when you create a database, even though you never intend to audit any table. Reason: in the future if you temporarily turn on auditing to catch, say, a bad apple truncating tables, you don't need a database bounce. Setting the audit trail to "db" does not start auditing; you have to set auditing options for objects.

Tue Jun 20, 09:41:00 PM EDT  

Blogger Howard J. Rogers said....

There is so much Oracle-related information to keep up with that I often appreciate someone saying - 'Do you know you can do x in 10g?'

Well, maybe Doug and I need to discuss this elsewhere, because I don't want to hijack someone else's blog... but what Doug's talking about there is not a 1 minute tip. He's talking about being pointed in the right direction for further research. There's a difference: a tip claims to encapsulate a "truth" within itself. General guidance to investigate X, Y or Z isn't a "tip" in that league.

Case in point: I didn't know only lower-version RMAN executables can be run against a higher version of the database. (That is, 10gR1 RMAN can backup a 10gR2 database, but 10gR2 RMAN cannot backup a 10gR1 database). I should have known it, because it's not new. But I asked on the Dizwell Forum, and got a one-word answer. Not even a 1-minute answer, mind! Just one word... a word which was a Metalink document number, and thus all the insight on the matter you could ask for was contained within that one word.

But that doesn't make '76337.1' (or whatever) a tip, does it? It makes it a correction, a note pointing out my error, a reference to a store of knowledge. But a little "nugget" of tip-like information. Nah.

And the same goes for someone telling you feature A, B and C should be looked into. I'm all for that. Looking into things is what proper research is all about, after all. But I am completely against the idea that you can 'nuggetise' what feature A is all about, let alone how to employ it, in a one-minute brain fart.

Same problems with Arup's example. if you sit there thinking that setting AUDIT_TRAIL=DB switches on auditing, you don't understand auditing. If you understood it, you wouldn't be surprised that setting audit options has to come into the picture. Which is not to say a reminder of these things isn't useful to those of us prone to forget stuff, but again: you can't miss out the issue of intent here. If your intent is to jog memories, then brain-fart away! All very useful, no doubt. But if your intent is to try and encapsulate a nugget of knowledge, and to impart that knowledge to others for the first time, then stop it immediately, because that's not a good thing to be doing (IMHO, of course).

There's a very fine dividing line between 1 minute nuggets and 'silver bullets', and I've had enough of the latter and those who dispense them to last me a lifetime!

Wed Jun 21, 12:13:00 AM EDT  

Blogger Doug Burns said....

Howard said .. "Well, maybe Doug and I need to discuss this elsewhere, because I don't want to hijack someone else's blog... "

It wasn't so long ago that I worried I was guilty of that but Tom told me not to worry about it.

I think we both understand each others opinions in any case, so I'll shut up now.

Wed Jun 21, 05:15:00 AM EDT  

Blogger David Aldridge said....

Howard said, "...being pointed in the right direction for further research"

I think that's exactly what the purpose of a one-minute tip ought to be. Any subject that can be completely tackled in one minute is going to be too specialized to be of value to many, I think, but as an aide and prompt to further research, and possibly with an appropriate set of supporting links, it ought to be a winner.

Wed Jun 21, 11:06:00 AM EDT  


<< Home