Wednesday, October 03, 2007

It's all about the data...

I've said many times - go ahead, erase all of my .exe files, but don't you dare touch my data.  Take word away from me, leave me my .doc files.  I'll be able to find something that can process the data eventually.

It is all about the data, applications are sugar on top of them to make them look prettier, but are not the end goal. 

Funny though, in most organizations, the thought process runs the opposite direction - for many people it is all about the application.  That is totally backwards.  Applications come, applications go.  They are secondary.  The instant an application is production - we are looking to replace it already.  Be it with version 2.0 of the application, or be it with another application entirely.

It is all about the data.

Even non-technical people understand that :)  I've always enjoyed Seth Godin's blog entries - short, and sweet and to the point.  That one, from a marketing guy, was good to hear.

POST A COMMENT

42 Comments:

Anonymous Anonymous said....

It's not the telephone, it's what people say on the telephone. But if the telephone doesn't work, you won't hear much.

Substitute any medium for telephone, and you understand why people focus on applications rather than data. They are the gateway to the data. They are expected to be ubiquitous and work seamlessly. When they don't work, people scream.

As to one click saving links... why would anyone want all those ads?

Wed Oct 03, 09:57:00 AM EDT  

Blogger Arun Mathur said....

Yep. Here's a typical scenario:

1) An application(s) gets re-architected using the latest and greatest technology as of two minutes ago, to be obsolete five minutes later.
2) Application launches. Database administrator kept in loop to create tables.
3) Users complain. They experience 2 hours to pull up a report. However, cool design pattern!
4) Architecture team asks DBA for help, specifically, what single database parameter in the init.ora file needs modifying.
5) DBA quietly during the night rewrites the database portion while he *gasp* ensures the procedures scale well before launch, and *double gasp* instruments.
6) Users happy again. Everyone else but the DBA reports that the DBA just needed to add an index, and the problem is solved.

Until the next latest technology comes out, tune in..same Bat time, same Bat channel.....

Good post Tom!

Regards,
Arun

Wed Oct 03, 09:58:00 AM EDT  

Anonymous Robert said....

Some decider up in Oracle's hierarchy should consider making that the new corporate motto.
BTW, what is it these days ?

Wed Oct 03, 10:39:00 AM EDT  

Blogger Jayasankar Peethambaran said....

>>> Take word away from me, leave me my .doc files. I'll be able to find something that can process the data eventually.

Well, practically, don't we still need Word to access the data inside?

Theoretically, someone might be able to reverse engineer an application to be able to present the word data, but would it present all the information and work which had gone into making of that document?

All the comments, the tables and all... In the end you will have to have anothe application as complex as word.

Is not the format in which the application stores the data as important?

Seth did his gradn in Comp.Sci I think. Geek at heart maybe :-)

U've cut down a lot on ur blogging.

Rgds,
Jay

Wed Oct 03, 11:22:00 AM EDT  

Blogger Thomas Kyte said....

... Jayasankar Peethambaran said ...

Yes, you would need something that understood the .doc format - and there are many.

My point was - leave me my data, take my applications but please just leave me my data, I'll be able to find *something* that can deal with it, perhaps better than the application that originally produced it.

Wed Oct 03, 11:56:00 AM EDT  

Anonymous dogbert said....

Jayasankar Peethambaran said ...
U've cut down a lot on ur blogging.


you can read this post while your boken keyboard is being repaired.

Wed Oct 03, 12:04:00 PM EDT  

Anonymous Toon said....

AND:

It's all about data integrity...

(plus, having a single point where it gets enforced)

Wed Oct 03, 01:10:00 PM EDT  

Blogger Dominic said....

Absolutely -- how many people still backup their entire hard drive? Do you really need to back up Word, Excel, etc? Heck, before I started keeping photos and music storage of my data was an order of magnitude smaller than the size of the s/w I had. That's why I like apps that let me keep config and personalization params in files outside of the .exe location (or registry for that matter) -- so I can back them up.

Wed Oct 03, 01:22:00 PM EDT  

Blogger Gary Myers said....

"That's why I like apps that let me keep config and personalization params in files outside of the .exe location (or registry for that matter) -- so I can back them up."

Great idea from Windows. You can have all your documents and settings in a 'Documents and Settings' folder. And to make it perfect for backing up....we'll add your Temporary files in there too - a massive chunk of cached ads from internet Explorer. Great idea, spoilt by a single cock-up.

PS. its not about the data...its about the backup.

Wed Oct 03, 04:29:00 PM EDT  

Blogger David said....

... Jayasankar Peethambaran said ...

Of course you still need an application to read the data, but which of the following scenarios do you think would alienate more customers:
1) Microsoft releases a new version of Word that retires a popular feature.

2) Microsoft releases a new version of Word that can't read .doc files from any previous version of Word.

Scenario 1 is unlikely, but scenario 2 will *never* happen.

Wed Oct 03, 05:34:00 PM EDT  

Anonymous Michael Olin said....

Re: Scenario 2 will *never* happen.

My accountant just got a new PC for his office. We transferred a bunch of Lotus 123 worksheets that his (just) retired partner had on his old PC. The new box has the latest and greatest version of MS office and reports that the Lotus files are unreadable. So now we open each file in an OLD version of Excel (which reads Lotus worksheets just fine, thank you), save them as Excel workbooks and the latest and greatest can now open them in "compatibility mode".

Wed Oct 03, 07:10:00 PM EDT  

Anonymous Anonymous said....

Scenario 1 is unlikely, but

They did it with Visio. The database generator is now only available with Enterprise version.

Wed Oct 03, 10:38:00 PM EDT  

Blogger Saager Mhatre said....

I guess Chet expresses similar sentiments here. However, I disagree here.

@Tom Kyte
Besides, I don't think Seth was talking about Data vs Application from a development perspective, but from an end user's point of view. Sure, to a user their data (tax return, inventory records, manuscript) is more important than the application they use to access it (Acrobat, OO Calc, OO Writer; but then, if a spreadsheet is a spreadsheet is a spreadsheet... why prefer OO Calc over MS Excel; if email is email is email, why do people (end users!) swear by their GMail; if music is music is music, why iPod, why not Zune?
As a developer, I take considerable offence to the idea that, "[i]t is all about the data, applications are sugar on top of them to make them look prettier". As a developer I also take offense to be typecast as anti-data and anti-database. Persistence is a cornerstone of any non-trivial application and every developer worth their salt eventually learns to respect it. But every good developer also keeps in mind that design is a set of tradeoffs to achieve the right balance.

@Arun Mathur
Remove that 'anti-developer' gadget from your utility belt. I could 'belt' out a whole bunch of real (as in personally experienced, as opposed to 'typical') scenarios that'd put DBA's to shame.

Thu Oct 04, 12:41:00 AM EDT  

Blogger Thomas Kyte said....

@Saager,


Saager - I too am a developer and I take high offense to developers that think it is all about the application and persistence and objects and classes and user interface.

Seth was saying "the applications - they stink, but the data is getting better and better all of the time - and the data is what is important, these applications, they come and then go - but the data - it persists and is getting bigger and better"

And who's point of view is relevant here - the end users or the developers? The developer works for the end user, not the other way around.

Some people swear by gmail, others swear at it. I have a multitude of choices for UI to access my email at Oracle - I chose versamail and thunderbird. I don't do gmail, I use yahoo - but not their UI, not unless I have to.

and when a better IMAP client comes around, I'll throw out thunderbird and use it.

Email is in fact, just that - email. I access email from many different clients - the email data is relevant, the client - somewhat secondary. It is like a piece of clothing, I use the one that makes me feel good that day.

And music is music, I do not own an ipod at all - so indeed, why ipod?

You have to admit that that vast preponderance of developers are anti-data and anti-database. There is a huge difference between persistence and knowing how to manage data, process transactions, do data integrity, process in sets - not slow by slow.


There are good developers, there are good DBAs

There are bad developers, there are bad DBAs.

but it really all is about the data.

applications come, applications go. Your list is a perfect example of that....

Lotus 123 - was king, now Excel is. Tomorrow - maybe not. and without the numbers, excel would be a picture of a grid.

Gmail might be your preference today - but email is a very old thing, gmail is the new kid on the block and some would say is fading . And without the content that is email - gmail would be an empty page.

Music predates the introduction of the iPod by - well - at least a couple of weeks I think (I distinctly recall 8-tracks, cassettes, LPs, 45s - I might be mistaken...). Without music, the ipod would be - well, a brick.

Thu Oct 04, 06:47:00 AM EDT  

Blogger Joel Garry said....

... but the data is getting better and better all of the time...

Not what I'm seeing. I remember when every input needed to be validated. I remember when integrity was new, so things had to be validated on entry.

Now I see people validating integrity - or not - with applications, free form fields, all sorts of garbage coming from Excel or wherever unscrubbed. Doubly offensive because the db capabilities have improved. The ubiquitousness of databases has meant lots of people creating apps without proper controls for missing data or accountability for data definition mismatches - this affects people's lives.

$14,000 ipod doc.

Thu Oct 04, 01:22:00 PM EDT  

Blogger Brian Ballsun-Stanton said....

I'm a philosopher of IT and I really like the kernel of thought you've expressed in this entry. Is there any way you could expand it to a proper essay? (I'd love to distribute it to my beginning database classes.)

Fri Oct 05, 01:10:00 PM EDT  

Anonymous Kevin Brady said....

For 29 years, at 6 different companies, I've watched data quality receive inadequate attention and respect. Never time to fix it, never time to prevent garbage in, never the ability to measure ROI for data cleansing efforts, always opportunities for me (and others of a like mind) to be data heroes when it finally "had to be done" because of an audit, or a regulator, or a deep pockets customer. It's been fun and rewarding...and a little sad...

Fri Oct 05, 03:20:00 PM EDT  

Blogger Nagaraj Nadig said....

I do agree with Tom,
what i would like to state here is
whether we are developers or end user, as the data grows it's both who have to understand the data better. it will not matter many times what application but, what and how of data.
Always data matters... it is Data, Essence of life

Mon Oct 08, 05:57:00 PM EDT  

Anonymous Anonymous said....


Nagaraj Nadig said....
it is Data, Essence of life


ummm, data is important, yes, but
NO, not "Essence of life"

Tue Oct 09, 11:09:00 AM EDT  

Blogger Nagaraj Nadig said....

it's Essence throught the life time. (though most of us do not realize it!)

Tue Oct 09, 06:57:00 PM EDT  

Anonymous Anonymous said....

Brooks, "The Mythical Man-Month":

"Show me your flowcharts and conceal your data and I shall continue to be mystified; show me your data and I won't need your flowcharts; they'll be obvious."

This was 1961. Do we just not learn anything in this business? Every one of the problems he describes are still around in one form or another.

Wed Oct 10, 05:44:00 PM EDT  

Anonymous MatteoSp said....

I'm developer: I could think "It's all about algorithms". But I would be wrong, as you are, with your "It's all about data".

That's nothing you can do with data, if you have no application to view/manipulate it. If you loose Word, you still have your documents, but you need another app to use them.

And that's almost nothing an application can do without data.

So (even if I intimately think there's some app that doesn't need data):

"It's all about data and algorithms" Data is not useful without processing capabilities, and apps have nothing to do without data.

Wed Oct 17, 04:25:00 AM EDT  

Blogger Thomas Kyte said....

That's nothing you can do with data, if you have no application to view/manipulate it. If you loose Word, you still have your documents, but you need another app to use them.

but you just made my point, again.

I can reinvent word, in fact - word exists over and over and over again in many forms (there are many "word" like programs)

But, erase my document and I'm hosed. Take my 1,000 page book and remove it - ouch.

Take my 1,000 page book and lock it up in a structure that is so obscure and hard to manage that word is hard pressed to edit it - let alone anything else - and ouch, I'm hurt big time.

I'll say it again, erase word.exe, but don't touch *.doc. *.doc is the data and *.doc will be used and reused and reused over and over again.

Word - comes and goes, it is not the end game, it is something we use to manipulate the most important stuff - the data.

I can (and do) view the doc on my pda, I convert them to html for the web, I use them in various word processing tools.

The applications are secondary, and may the BEST APPLICATION (at the time) win.

But remember, that applications are transient, they come, they go. They are popular one day and not the next.

the data lives on.... forEVER.

and if you mess up the data, you have quite simply - messed up, for everyone.

mess up a new user interface to the data and you have just wasted your time - no big deal.

Wed Oct 17, 06:29:00 AM EDT  

Blogger The Underboss said....

OK.

All of you application fanatics above, tell me if this looks familiar: v1.0, v2.0, v10.9, Pro, Lite, 2000, 2003, 2007, XP.

I bet none of you will think of the Information/Data that is accessed via that vast array of "applications".

So lets not confuse "an application" with a "the application".

Of course application is as important as data for some people, but the point here is, data is irreplaceable where as applications are often available on a dollar sale once the new version comes out. Hell you sometimes don't even need a new version if you just change the underlying application, your OS.

Get the point?

There is no substitute for uniqueness.

PS: I am too a developer AND a DBA.

Thu Oct 25, 12:04:00 PM EDT  

Blogger The Underboss said....

This comment has been removed by the author.

Thu Oct 25, 12:23:00 PM EDT  

Blogger Saager Mhatre said....

@Tom Kyte
I guess I'm a little late with this reply, but here goes.

My point was that, to an end user, it only really matters that their data (tax return, inventory records, manuscript, music, photos, etc) are stored in SOME format so as to be able to be retrieved as and when they need to. For instance, you don't care wheter Yahoo stores your mail as text, binary, Base64 encoded, Hex or Kingon(!) as long as they can send it back to you formatted as an IMAP response whenever you request for it. One could very well argue that since users change several application over time to manipulate the same data, that's where the innovation and USP is. But that would be one-sided too.

That a lot of people in the software community (developers and DBA's alike) actually store data in absolutely horrendous formats is something I wont deny. My comments to Chets post hint at some of my personal experiences with poor physical models.

It's just that I don't think we can counter the "it is all about the application and persistence and objects and classes and user interface" argument with "It's all about the data...". Maybe we should all go read some Eric Evans

Thu Nov 01, 07:52:00 AM EDT  

Blogger Thomas Kyte said....

@Saager

... . For instance, you don't care whether Yahoo stores your mail as text, binary, Base64 encoded, Hex or Kingon(!) as long as they can send it back to you formatted as an IMAP response whenever you request for it. ...

you make my point so well. You are right that I don't care how they store the data - as long as they can respond to me and give it back to me.

The application here - is the email client.

That email client might be a web interface. A handheld device. A thick client on a pc. A (new thing) on (whatever comes next).

You see, it is not about the application at all - it is all about the data and what you just wrote drives that point home.

Now, if yahoo stored it in klingon, that would make it horribly hard to return it to us - and hence, they should read those books (so as to avoid the complexity they imposed on themselves) for the future.

Thu Nov 01, 08:16:00 AM EDT  

Blogger Saager Mhatre said....

@Tom Kyte

Ahh... but that is not the point! I make a point of larger scope than yours. My point is that, IMHO, although it is not ALL about the data, it is ALSO not ALL about the application. I believe @matteosp already made that point above.

It is the combination that has to make the most sense. Without Yahoo (or whatever email client you use) presenting it to you and allowing you to manipulate the information, the data is worthless.

I admit the data is invaluable, but if you remove the capacity to view/manipulate it in a meaningful way, all that value is applied a factor of ZERO! You can store it on paper or parchment, but if you cant find the papers or can't read off them all the value is lost.

I'm asking you to look at the golden mean as opposed to taking an expreme view. I don't get it; I like you, but sometimes your views come out horrible one-sided.

Fri Nov 02, 12:40:00 AM EDT  

Blogger Thomas Kyte said....

... but sometimes your views come out horrible one-sided. ...

they always appear that way when you don't agree with someone. I see you through the same glasses basically.

We can generate an application in a short time. They come, they go. Of course you need them.

However, they are not the end game. They are not the top of the stack. They are not what should be the sole focus.

But, almost every development team I've seen presumes they are the pinnacle, end game - the biggest thing ever

And they make a complete mess of the data - ruining everything for everyone.

That is all - it is all about the data - give me good data and I'll give you an application to present it, manipulate it shortly.

Further, if you secure the data close to the data (eg: in the database, not playing the fake game of "security is application logic" which is utter bullocks) I'll deliver the application faster, cheaper.

Further, if you put data integrity where it belongs (see rant about application logic and the misplacement of it above), I'll be even faster.

To the point where applications become a dime a dozen, disposable.

It truly is all about the data, you have good data.

Ok, analogy (dangerous ground, analogies are)

Data is the foundation and superstructure of your house.

Applications are the elevation, the exterior look and feel. In my neighborhood - there are three models of homes only. However, it looks like dozens because the elevations are so different looking. Yet, when you get into the house - they all look pretty much the same (except for the furniture - you know, the customization of the 'application').

So, without good data (foundation and superstructure), all of your pretty elevations and customizations are useless.

Give me a good foundation and superstructure and I'll give you as good a look and feel as you are willing to pay.

Undersize the foundation (don't think about scale and such) and you'll have a small house forever (and would likely try to cookie-cut many small houses and say "we scale at the application layer!!!"). Goes with developers that say "don't do it in the database, they don't scale" - because they always build tiny foundations.

Actually, I think I like this analogy - it seems to work.

Fri Nov 02, 06:56:00 AM EDT  

Blogger The Underboss said....

@Saager,

"That a lot of people in the software community (developers and DBA's alike) actually store data in absolutely horrendous formats is something I wont deny. My comments to Chets post hint at some of my personal experiences with poor physical models......."

That depends on individual capability of that person designing the data. Not arguing here about the skill-set but value of data.

Ahh... but that is not the point! I make a point of larger scope than yours. My point is that, IMHO, although it is not ALL about the data, it is ALSO not ALL about the application.

Ahh, but not true.

Application = means of transport = 100s of options available = transports "data".

DATA = Reason to develop applications = can define many applications in many different ways.

Without data, your application doesn't exist = useless. But without a particular application, such as yahoo, outlook, thunderbird, etc., you can still pick another one which has a reason to exist because of your and my data.

I guess what Tom is trying to say, and i agree, that "your"|"my" data can not be substituted but how you look at it , can be. And how you look at it can be designed even better if the data is organized well.

So what is of essence here? Something that just exists or something that gives others a reason to exist in many different ways, at many different costs and transporting in many different formats.

Fri Nov 02, 07:15:00 PM EDT  

Blogger Saager Mhatre said....

@Tom Kyte & @The Underboss

Apparently we keep making each others point.

TK- "... leave me my data, take my applications but please just leave me my data, I'll be able to find *something* that can deal with it..."
(*something* = another application)

TK- "... when a better IMAP client comes around, I'll [...] use it..."
(better IMAP client = another application)

TK- "I can reinvent word, in fact - word exists over and over and over again in many forms."
(applications...)

TK- "I convert them to html for the web"
(using an 'application', no doubt)

TU- "you can still pick another one"
(another application)

The fact that they can be replaced doesn't disount their value. As @The Underboss said:

... lets not confuse "an application" with a "the application".

(I'm focussing on "an application")

One of my biggest rants about this post was about being typecast as anti-data/anti-database. But you already agreed with me and dispelled that.

But, I still stand by what I (and anonymous and @MatteoSp) said earlier. I particularly liked the analogy in the anonymous post. If it's all about the data, we could do just as well with paper-based, index-card filed systems.

And BTW, I think your analogy has too many holes. The most important one would be that-
applications are used to work (retrieve/manipulate) data; houses don't provide mechanisms to alter foundations, the foundation is final.

Come to think of it, after that analogy I wonder if by 'data' you mean 'data structure'? That's where the whole discussion could start to make sense.

Mon Nov 05, 01:22:00 AM EST  

Blogger Thomas Kyte said....

@Saager

I don't see how I was making your point. I keep saying 'applications come, they go, they come again'

You help by listing them out and showing them coming and going.

I short, not being the relevant component. In short showing them to be the transient piece of the pie. In short making the point (especially with the filing cabinet bit - that user interface was *better than good enough* for a really long time - and when the card catalog - the physical one - disappeared, what remained? Ahh, the data, the data remained - but the application, irrelevant. It was replaced.

You cannot replace the data.
We do replace the application over and over and over again.

Just go back to your original post, you made this point hugely clear. "Why prefer spreadsheet X over Y?" Why application application A over B?

Because applications are a choice, sugar on top of the fruit.

The only non-choice is the data, the think that provides the true sustenance


Houses do not alter foundations - this is true (and so holds the analogy). Houses are redecorted, houses are rearranged, houses have walls being moved around, houses are painted. Houses are applications making somewhat comestic changes to the appearance of the foundation.

And hence, you are limited in the application by what your foundation can do, the application can only do so much with a poor foundation.

Redecorating, pretty cheap and we do it all of the time.

Reorg'ing your foundation - pretty huge, lots of engineering, and expensive.

Hence you are stuck with your foundation for a long long time - get it right.

It is all about the foundation. We can fix the elevation.

Mon Nov 05, 09:43:00 AM EST  

Blogger Thomas Kyte said....

Let me wrap up here...

My point is - the data is crucial, vital, relevant, important. It is the very foundation. Without it, you can build whatever you want - application wise - and you'll have nothing. Nothing but pretty screens.

But the focus in business IT development? The data? Nah, bit buckets those databases are, it is all about object-ty stuff, persistence of "objects", SOA is sooo cool, XML is perfection. The focus is 99.99% on the application - and 0.01% on the data.

The ratio needs be reversed.

Applications are replaceable, applications ARE replaced - constantly, as new things come along.

But the data, the data lives forever.

Mon Nov 05, 10:45:00 AM EST  

Anonymous Kevin said....

Tom - it seems that if I replace your term 'data', in the previous statements, with 'data architecture', or 'data design', it seems to fit the scenario 100x better than the generalized term 'data'.

I think some people are confusing the matter, by assuming by the term 'data', you mean the information itself. When in effect, your arguments are with regard to the 'structure' of the information.

The two can be really broken down into three elements:

Raw Information
Organization of Information
Manipulation of Information

No amount of Organization will make bad or incomplete information good. However, good, solid information can *always* be equitably improved with a good organizational design. This is due to the facet that information is only useful when it can be easily accessed, manipulated, and modified.

Applications generally comprise the outer-most tier... Manipulation of Information, but can often times hide poor organization.

What's critical however, is in reality, the interaction of the three elements.

Good, solid, useful information, coupled with a reinforcing organization, allowing for the easy and painless manipulation of that information into a cohesive whole, implementing work.

Databases, help a system organize and facilitate the manipulation of information. A good database design, can significantly improve the accessibility and extraction of knowledge and usefulness from the information.

So, in essence, you're both right. What you both differ on however, is a matter of 'valuation'.

A well designed application can mask a poor data design.

A well designed data design can be destroyed by a poor application.

It takes both to work effectively, however, a good data design can assist/ease the creation of a good application.

It just makes it 10x easier, to build a good application utilizing a good data design... than it does making a good application off of a bad data design.

Hence where I think Tom's argument comes in.

Good data design is a 'facilitator' of a good application.

Mon Nov 05, 05:01:00 PM EST  

Blogger The Underboss said....

This comment has been removed by the author.

Mon Nov 05, 09:43:00 PM EST  

Blogger The Underboss said....

Kevin,

Data design/architecture may fit well but the topic is targeted at just plain "DATA" such as: Your assignment, board report, financial figures, etc.,

No amount of Organization will make bad or incomplete information good.

Here you make the point again. It's all about the information. ergo Data.

Of course it is vital for both to work smoothly but here one (application) doesn't exist at all without the other (data).

Stop of you're going to say, "if not one than the second one can do the job". Because you'll be making the point again, applications are disposable, good or bad.

Data on the other hand, good or bad, is most valuable weather it's viewed by word, excel, Apex, XML or Bat-Data-reader in Batmobile.

I reckon this thread needs to be closed.

It'll take ages for everyone to be educated. I thought that's what the universities are for.

Cheers,

Mon Nov 05, 09:44:00 PM EST  

Anonymous Kevin said....

Underboss:

You're missing the point of my statements entirely. I simply restated both Tom's and "the Application developer's" tenets in a single form.

Neither one is incorrect. It's not just about the data, since if the data is in some form that is unrecognizable (say, in French if you only speak/read English) then it is useless.

Data + Organization + Presentation = Work Done

Raw information in a format which is unorganized can be next to useless. The effort required to do real work from that data set can be well beyond the benefits incurred from the usefulness of the information.

Each is a facilitator of the other. Data itself is useless without some means of producing work from it.

For example: Say I have a database of information regarding the purchases of 4 million people sitting in a system. I can get little to no work done with the raw data itself. It's just too large to manage individually.

In addition, without a good organizational structure to the data set, trying to utilize that information in any form of work effort, would be an effort in frustration and exasperation. (Think of tossing 4 million index cards into a pile with each index card having their purchases recorded on it...)

On top of that, with out any means to adequately manipulate that information (graphing, editing, sorting, etc.) that 4 million purchases in a database becomes so much a lump of bits.

Each is the facilitator of the other.

We use data to do work. So, in reality, it's all about the work done. Data is the storage of information produced from work effort and is manipulated via applications.

We need both. Tom's argument values data, and rightly so. While the application developer's value the application. Each one has valid arguments, but in reality, both are required.

Tom's argument is that while we generally have multiple ways of 'presenting' information, we generally have very few methods of 'storing' it. The organization and 'storage' of that information is what is 'prized'.

That's somewhat of a flawed logic tree, simply because there is a built in weighting of 'information' over 'presentation'.

In reality, the whole purpose of both 'information' and 'presentation' is to do work. so they are both needed. We just tend to have typical methods of organizing information, and myriad of ways of presenting it.

Tue Nov 06, 11:28:00 AM EST  

Blogger Thomas Kyte said....

... Kevin said...

Kevin, I think you hit the proverbial nail on the head for the most part.

I would go further with:

The organization and 'storage' of that information is what is 'prized'.

to say

The organization, securing and protecting (data integrity) of that information is what is 'prized'.


so many times the presentation of some data is prized above all else (oh look, pretty shiny graphs)

so many times the securing of the data is given no thought (until it is compromised and then security is grafted on).

so many times the integrity, quality of the data is given no thought (oh, we do that in the application, yes we know that it sometimes lets bad data in because we cache the world....)

so many times the organization of the data is given no thought - DBA's will make it go faster later, that isn't a development problem...

Wed Nov 07, 08:08:00 AM EST  

Blogger Gary Myers said....

"the data is crucial, vital, relevant, important. It is the very foundation. Without it, you can build whatever you want - application wise - and you'll have nothing. Nothing but pretty screens."

We have applications without data. They are called 'games'. Space Invaders etc. Okay, they have high scores, but that's just fluff.
New 'recreational apps', like Second Life and World of Warcraft, is probably more 60/40 application vs data. Lose the data and some customers/users won't bother reinventing themselves. But more would be lost if the app suddenly became text-only or otherwise ugly.
There's some 'real-time applications' where the historic data is less important. For example, I want to know whether the engine temperature is in safe limits NOW. If I don't know that, having last week's reading in a file somewhere really isn't important.

Thu Nov 08, 11:36:00 PM EST  

Blogger Thomas Kyte said....

space invaders is just a pretty screen, nothing more nothing less.


And I'm pretty sure that if you took someones second life/world of warcraft persona and "accidentally erased it", they would freak out. People spend years getting their online persona's to the 'right level' (so much so that a search for world of warcraft child endangerment on google turns up quite a few hits)

Those applications are all about the data - lose someones "level" in world of warcraft and you'd have someone going loopy on you.

And lose someones second life online persona and all of their second life money and such - and you'd have the same thing.

Fri Nov 09, 07:57:00 AM EST  

Blogger Franco said....

No matter which instrument you use to play J.S.Bach, it's still and and it sounds always as J.S.Bach. .

He has been first getting it ;-)

Thu Mar 27, 09:15:00 AM EDT  

Blogger Gary Myers said....

Just to emphasise, check out the JournalSpace story. Their concept of a database backup was RAID. Disk got wiped (bug / sabotage ?). RAID copy wiped. Business wiped.
Doesn't say what the DB was, but they mention they were running on OSX which gives a pretty short list.

http://lifehacker.com/5122848/hard-lessons-in-the-importance-of-backups-journalspace-wiped-out

http://journalspace.com/this_is_the_way_the_world_ends/not_with_a_bang_but_a_whimper.html

Tue Jan 06, 11:20:00 PM EST  

POST A COMMENT

<< Home