Is XML the Answer?
-
We use XML all over the place not because we need it, nor because it's better than other ways of retrieving data but because 3 years ago it was hot and our developers wanted it on their CVs.
Just a cynical afterthought.
-
YEA! OK, maybe it's a little bit overboard, but YEA!
-
Nassik, I think that you are dead wrong. I know several people that work at MS and other places and I also know Chris Date and Fabian Pascal and all of the best data architects vehemently oppose the use of XML in just about all forms. The decisions that most companies made to support XML were mostly motivated by politics, having to provide what the users wanted, in spite of what was best for them.
I agree with dcpetesron and second his challenge to you to give a real life example of XML being the "only way to resolve a problem". To give this example you will have to be explicit in the details. Of course, as you said, you "wont
of course" give a real life example, probably because there are none where XML is the only solution, or even the best solution. -
Darwin Information Typing Architecture.
It uses XML for content mark-up. The organisation of the content is disciplined and the system is designed to merge topics of information together into a document.
It isn't data storage as a DBA would know it, or transmission but it is what XML was designed to do. Separate content from presentation.
The idea is that it is a cross platform and open standard (I use the term loosely) to allow a disperse group of content authors contribute to a document. Document in this case being something that makes BOL look like a footnote.
-
Of course the publishing world is where XML got its start. As you said David, it is a method to separate format from content.
I won't claim to be an expert in that area, but I would be very surprised if there weren't mechanisms already in place to do that, but they were probably very proprietary to each system (this separation of content from format is nothing new...)
I would also be very surprised if many, if not most, of those other methods are not more efficient than XML at doing the job. Reason would indicate that they would have to be.
I’m certainly not ready to make a lunch out of my laptop yet
/*****************
If most people are not willing to see the difficulty, this is mainly because, consciously or unconsciously, they assume that it will be they who will settle these questions for the others, and because they are convinced of their own capacity to do this. -Friedrich August von Hayek
*****************/
-
May be a small aperetif?
If you look at a postscript file you can almost tell what is going to print. Within certain limits you can say the same about Adobe Acrobat documents.
When it comes to storing XML, HTML, SGML etc in a database I would only do it if
a) I need sourcesafe type functionality
b) The eventual rendering is explicity for the web i.e it is to power a web content management system.
-
Programmers/Developers who do not about Fixed length text file or that a 20-40megabit hard drive was big or FANS of XML.
XML is a buzz word and it is programmers and developers of the new generation are using it. The new programmers
-
Hmm. I look forward to seeing Peterson's new webpage design technology, perhaps entitled HTMRelDB, maybe equiped with eXtensibleStyleStoredProcedures - sure to be light and quick loading on a web page near you.
True, the idea of XML replacing a Relational Database is a bit whacked, and storing XML docs in tables as a part of relational data, would have E.F.Codd rolling in his grave, but that isn't (shouldn't be) what XML is for. It is for packaging data for use by webpages, for communicating between applications - online and off, and for bringing data to and from the database interface. Anyone who's had to parse and develop with his example: Red,L,Hawaiian,Y (which, by the way is missing what type of product it is - typical) between applications or to a webpage, is going to be thankful for (I fixed his XML):
<Product type="Shirt">
<Color>Red</Color>
<Size>L</Size>
<Style>Hawaiian</Style>
<InStock>Y</InStock>
</Product>
The question isn't which is more efficient for a machine to read or generate, it's which is more efficient for Developers to develop with.
Granted, it's always a good idea to stir things up and think thrice about your assumptions, but I think this falls under the "Don't use Hammers for sawing wood" type of admonition. Then again, after having just gone through building a new house, that's probably a good lesson to have spread and repeated. Often.
Van
-
Poor dumb b***ds
- Re. data transport, dcpeterson says "I do acknowledge that XML can be of some use here, however the same thing can be done using ANY agreed on data file format. Unfortunately for XML almost any other physical file format will be more efficient..."
Can you please tell me the name of the data file format, other than XML, that is universally agreed upon and supported across virtually all operating systems and programming languages ?
I would love to use this to solve my enterprise wide interoperability issues. Unfortunately, no-one can tell me what this efficient, universal data format is called.
-
csv?
regards,
Mark Baekdal
+44 (0)208 241 1762
Build, Comparison and Synchronization from Source Control = Database change management for SQL Server
-
...as if XML were a universal data format, PLEASE!!! Get real here. Why does a data file format need to be universally agreed upon (as if XML were)? First off let me nuke your assumption that XML is a universal format.
The format of an XML file is not determined by any standard at all. The format (the tags and hierarchical structure) is unique to each system. Any system that wants to use that XML file must be programmed to look for those certain tags, and in most cases must know what the hierarchical layout of the file will be. If you want a "universal data file format" try EDI on for size...
Now that I have (for the umpteenth time) demonstrated that XML is in no way, shape, or form, a universal data format, it should be pretty self evident--to any rational mind--that all you need is a file whose format is agreed to in advance, by the sender and reciever to exchange data.
As for XML being supported across "all operating systems and programming languages" this is yet another straw man constructed by XML proponents so that they have an easy target to knock down. However, when one bothers to look past they hype it is easy to see that there in NOTHING about XML that magically imparts this benefit. XML is only supported on systems on which you have installed a parser. There have been parsers for almost every other file format imaginable, long before XML was spawned, many of the freely avalible, and almost all of them much more efficient than XML.
So pick a format, any format, install a parser and use it. Whether it is EBCDIC, or ASCII, or some other binary format, it doesn't matter as long as both systems can read it and know what it means (the data in it has been defined such that the recieving system knows where each data element can be found).
Now, there is some benefit to not having to install a dozen parsers on every machine (worst case scenario). So, to the extent that XML has become the standard data exchange mechanisim it has had that much good come of it. However, of all the data exchange file types that could have been chosen, XML is one of the worst because it is so bloated with tags that are neither necessary, nor sufficient to truly facilitate data interchange.
If XML is a standard then there is some small benefit that is accrued by virtue of it being a standard, but it is, and will always be a stupid choice for a standard, given the other options.
/*****************
If most people are not willing to see the difficulty, this is mainly because, consciously or unconsciously, they assume that it will be they who will settle these questions for the others, and because they are convinced of their own capacity to do this. -Friedrich August von Hayek
*****************/
-
csv = comma seperated values. Normally it denotes an ASCII file, but not always...
/*****************
If most people are not willing to see the difficulty, this is mainly because, consciously or unconsciously, they assume that it will be they who will settle these questions for the others, and because they are convinced of their own capacity to do this. -Friedrich August von Hayek
*****************/
-
Thanks for correcting the XML, but since I don't claim to be an expert in XML (as I avoid it like the plague) I pulled that example right off of XML.org. Although I don't remember the exact article and can't seem to find it anymore...
I did make one small change though, originally the Style was "Polo" and I changed it to "Hawaiian" mostly as an inside joke for those with whom I work.
/*****************
If most people are not willing to see the difficulty, this is mainly because, consciously or unconsciously, they assume that it will be they who will settle these questions for the others, and because they are convinced of their own capacity to do this. -Friedrich August von Hayek
*****************/
-
The Cache database has always been hierarchical in nature until early 1990-s, and have always been the fastest and very reliable. Nowadays it is called Post-Relational. You can read more about it at http://www.intersystems.com
Viewing 15 posts - 61 through 75 (of 144 total)
You must be logged in to reply to this topic. Login to reply