• (cs)

    The OP is totally right...I gave up on XML long ago. Its much easier to use standard formats for specific datas, and custom formats for everything else.

    Although, for some silly reason I don't mind XHTML too much, except when I run into the well known nobodys-browser-works-right bug.

  • Spike (unregistered) in reply to steve
    Steve:
    s|k:
    "it's so easy that companies feel like they can get away with hiring developers instead of engineers"

    This site gets more full of itself everyday. Seriously, can we drop the generalizations about classes of people from these 'witty' articles?

    Thanks.

    Too bad it's true all too often. Where do you think a lot of the content for WTF comes from?

    Most of the time from engineers I'm afraid. That's the kind of people who learn a lot of different things (chemics, fysics, electronics, ..., informatics) but only a select base. And who also learn that they know best ... always I have a masters degree in informatics only, that makes me a developer... I learn(ed) a lot after (and in) working hours and I check every technology myself before shooting it down. (I can't test everything, there are only 24 hours in a day, so I have to select what could be interesting) That makes I have a much larger base in IT... But these 'engineers' I worked with are so good in informatics that they write above WTF's theirself and tell you the direction to go. No matter how good your arguments are because they are engineers and know...NOT They can't even keep up with the basics (and your arguments)...and so I don't work there anymore

    Summary:

    1. A technology isn't bad because someone uses it badly. Or the same could be said about SQL for just giving another example of a good technology that's been used badly on a regular base.

    2. Judge people on what they do, not on what they say or on their grades.

  • (cs) in reply to Spike

    It would be very interesting to see how opinions about XML correlates with the used development platform. I use .NET, VB6 and vbScript but I have never used Java. In my humble opinion XML is the easiest and fastest way to store and transform structured data. Due to existing, ready solutions such as xsd, dom, xpath and xslt I am able to process data without wasting my time on developing my own text parser. IMO xml is very readable, unless you use goofy abbreviated element names. However, who cares if the file is not human readable, it is the parser's job to read files, huh?

    Is it true that DOM and XPath implementations are just so crappy in Java that they are unusable? Or do I just have bad sense of humour and sarcasm?

  • (cs) in reply to sirhegel
    sirhegel:
    Is it true that DOM and XPath implementations are just so crappy in Java that they are unusable? Or do I just have bad sense of humour and sarcasm?

    Sort of. The whole point of the WTF was that the programmer used SAX (which is rarely the right choice of XML parser, but popular because it was available first) to manually implement the functionality of an XML binding framework (of which there are several perfectly usable ones available).

    It's really just a typical and particularly bad case of reinventing the wheel because someone would rather spend an absurd amount of time using the wrong tool for the job than learn a new, slightly more complex tool.

  • blah (unregistered) in reply to jho

    Thats right. I'm even proud of you; youre showing us correct and proper usage of apostrophe's.

    Oops, forgot to use orange text...

  • blah (unregistered) in reply to jho
    jho:
    DrkMatter:
    Someone forgot their sarcasm detector at home today.

    Is that meant to be English? You American's sure do twist our language some times.

    And I should have hit quote instead of reply. Oops.

  • Freddy (unregistered) in reply to CaptainObvious
    CaptainObvious:
    s|k:
    "it's so easy that companies feel like they can get away with hiring developers instead of engineers"

    This site gets more full of itself everyday. Seriously, can we drop the generalizations about classes of people from these 'witty' articles?

    Thanks.

    [...] Software Engineer, Developer, Programmer, Systems Analyst, Technician; depending on who you ask any or all of these mean the same thing. The dictionary disagrees, however.

    In the case of Software Engineer vs. Developer, an engineer is responsible for design, not implementation. There exists an implication that an engineer is capable of developing a prototype or possibly final version, but no such implication exists for a developer; a developer makes manifest a design, (s)he does not create that design.

    [...]

    Very Funny. I think many a real Engineer (TM) would disagree with you regarding their job as primarily prototyping/designing and sometimes having the know how to make something.

    Why being an Engineer is important, when you're doing things like building bridges, is that real Engineers can sign off approval and they become liable for that thing not falling over and killing people. They are intended to oversee construction to make sure the low-end builders do the job correctly. Hey, everyone likes to sue, and you gotta know who to blame.

    In the UK anyone can call themselves an engineer, thus the Engineer accreditation obtained by professional organisations in North America are redundant.

    Back to Software Industry - we have the same problem that anyone can call themselves anything.

    It's really time to lose the corollary with the building / manufacturing / fabrication industry. Sure there are similarities, but they're just similarities. The problem is because all of these make real world, physical things, and we monkeys have been making these kinds of things for a very, very long time.

    So many people try to replicate the process of manufacturing / fabrication when software is in reality a different beast altogether. Don't do it, and don't use MS project because it warps the way you see software. The waterfall is evil. Trying to pigeonhole people with titles and enforce hierarchies is also just a waste of time IMNSHO.

    Architects/Software Engineers/whatever who don't code are just wankers, and if I had a nickel for every self proclaimed 'architect' 'systems architect' 'application architect' 'software engineer' whatever who was full of crap I'd at least have a buck fifty.

    So lose the titles you self important losers and start being serious about being the best you can be rather than worrying what somebody else thinks about you. There will be better people, and you will better other people. Hopefully you can learn from the former and mentor the latter.

    Anything to improve the abysmal '5 in 6' utter failure rate for software projects.

  • Asd (unregistered)

    The problem with XML in Java is there are too many options. You could uses Sax, Stax, DOM, JDOM, XOM, Castor, XStream, JAXB etc. Plus Xerces and Xalan (the 'standard' parser and XSLT transformer) are horribly written and buggy, do things like write on stderr, and the developers won't fix obvious bugs. I personally like XOM, but it depends on a parser and transformer, and bugs in those can cause a lot of pain.

    That said, if you use a decent library, XML is very easy to use. But there are always idiots who will do the parsing themselves and make a mess.

  • EER (unregistered)

    In fact, XPath IS way too complicated for anyone who doesn't want to bother looking into it (virtually everyone) so I comletely understand why HashMaps are so much better :)

  •  ? (unregistered) in reply to coz

    All that, in fact, is exactly what a chef does. The Chef, as well as organising the rest of the team, designs the menus, and keeps all of his kit clean. Which is why he's the chef de cuisine, not just a 'cook'.

    While cleaning the rest of the kitchen is, of course, delegated, everyone cleans his own equipment.

  • googoo (unregistered)

    Using SAX and related implementations in an enterprise environment is a pure WTF by itself.

  • Leif Arne Storset (unregistered)

    Software Engineer? Developer? Designer?

    I will have none of it. I shall be known as a technomage.

  • Leif Arne Storset (unregistered)

    Software Engineer? Developer? Designer?

    I will have none of it. I shall be known as a technomage.

  • martin (unregistered) in reply to Robin The Obvious
    Robin The Obvious:
    It'd be really nice if someone came up with a transformation language for XML

    Erm.... ever heard of XSLT??? Your post betrays the average XML shills' major downfall.... criticizing something they haven't bothered to fully comprehend. If I hear another person on the web suggesting that XML is 'bloat' and that we should all use JSON/YAML/Insert-your-favorite-half-arsed-serialization-syntax-here I'll go crazy! RTFM, or is that something that only "developers" get to do?

    they were afraid of xpath ... so they can't really make use of xslt now can they ? :p

  • (cs) in reply to Susanna
    Susanna:
    Misuse of XML really gets my goat. I once worked on a project for a large international bank where the XML the contractor's code spit out began with an element named <ROOT>.

    I evaluated reporting engines for my company a few years back. One requirement was that the engine should be able to emit reports in XML format for consumption by external systems. At least two of the leading reporting engines out there would take a printed report like this:

    Phone Book Report                     March 14, 2007
    Printed by: Rob Freundlich
    
    Name        Address      Phone Number
    Rob         1 Main St    555-1212
    Fred        3 First St   555-9876
    

    and produce XML similar to this:

    <report>
      <header>
        <headerLine line="1">
          <field align=left>Phone Book report</field>
          <field align="right">March 14, 2007</field>
        </headerLine>
        <headerLine line="2">
          <field align=left>Printed By: Rob Freundlich</field>
        </headerLine>
      </header>
      <section>
        <columnHeaders>
          <column index="1">Name</column>
          <column index="2">Address</column>
          <column index="3">Phone Numbr</column>
        </columnHeaders>
        <line index="1">
          <column index="1">Rob</column>
          <column index="2">1 Main St</column>
          <column index="3">555-1212</column>
        </line>
        <line index="2">
          <column index="1">Fred</column>
          <column index="2">3 First St</column>
          <column index="3">555-9876</column>
        </line>
      </section>
    </report>
    

    USELESS!!!!!! How about something like this, instead:

    <report>
      
        <name>Phone Book Report</name>
        <date>2007-03-14</date>
        <printedBy>Rob Freundlich</printedBy>
      
      <data>
        <record>
          <name>Rob</name>
          
    1 Main St
    <phoneNumber>555-1212</phoneNumber> </record> <record> <name>Fred</name>
    3 First St
    <phoneNumber>555-9876</phoneNumber> </record> </data> </report>

    and perhaps even a nice embedded or sidecar schema?

  • Tragomaskhalos (unregistered)

    Looks like a job for ... Commons Digester. This is one of those great WTFs when you try and imagine what they were thinking of - "hmmm, XML is ubiquitous these days, so I guess everyone is just writing gobs of hideous Java like this to parse it all, eh?"

  • thogard (unregistered) in reply to CaptainObvious

    You know that many of the b-29s built in Wichita during ww2 were built by former short-order cooks and one of Boeings leading engineers was in fact a historian.

  • thogard (unregistered)

    Is there a single coder who understands at least most of "The Art of Computer Programming" who takes XML seriously? I've always figured XML is for the guys who can't understand real parsing and failed out of CS 201. When Knuth can prove (using that mathematical stuff that gets ignored by the amateurs in CS) why xml type parsing is pure evil, why is it still considered?

  • Ryan H (unregistered)

    Whew.....

    As a WTF submitter of lunacy I had experienced at a previous company, I was starting to get worried that I had found myself sitting in the middle of another WTF-y company after reading this story. However, judging from the differences in domain vocabulary, I can almost be certain that this isn't MY Fortune 500 railroad. It must be one of the other 3.

  • Valentin (unregistered) in reply to EmVee
    EmVee:
    The field of ICT development is packed with people who want to make their job sound that much more interesting. I'm not a java programmer, I'm an Architect! I don't write documents, I'm a functional requirements analyst, etc...

    Apart from cooking, a cook will often clean his own equipment, write the menu's and go out shopping for ingredients. In IT, this would require 4 roles for one reason or another :(

    If you develop software, you are a software developer. Period.

    The same way that if you cook food, you are a chef. Oops they are some chef, baker, head of pastry, saucier (responsible of the sauce), ... By the way, a head chef very rarely writes the menu (its is printed) and absolutely never clean his own equipment (that's the job of the junior employee).

    Let's take another analogy. If you work in the building trade, you are a break layer.

    What you are an architect not a brick layer. What do you mean that carpenter, plumber, electrician, roofer are not brick layer. Back to the drawing board.

    The real WTF is the comment.

  • Duhveloper Employer (unregistered) in reply to EmVee

    Amen! Titles are only useful for people who swap business > cards... You know the kind...

    Do you mean the kind who sign your paycheque?

  • (cs) in reply to PSWorx
    PSWorx:
    xtremezone:
    I'm definitely no XML expert. I took a 4 month course on it last semester and after the first week I realized I should stop paying attention for my own good.

    I prefer the standard syntax that has been made so popular in Unix-like systems' configuration files:

    #Comment name=value until newline

    I'm way against XML documents. Explain to me what you can do with an XML document that you can't do with the traditional configuration syntax..

    Web pages for starters?

    Not knowing a technology - no problem Stating a technology is bullshit despite having no clue of it - questionable Stating a technology is bullshit and being proud of having no clue - please leave.

    XML cannot 'do' Web pages. XML defines data. That's it. Anything other than that is done with parsers of XML - which could be done too with configuration files if you parsed them and rendered them in a Web page.

    Actually, now that I think of it this is sort of done with PHP's phpinfo() function, isn't it? If not directly, than indirectly.

    Conveniently, data can be defined and brought into Web pages using databases, which gives you control of data that tends to be lacking in an ASCII text file.

    You've heard of databases, right? The right way to use data?

  • (cs) in reply to thogard
    thogard:
    Is there a single coder who understands at least most of "The Art of Computer Programming" who takes XML seriously? I've always figured XML is for the guys who can't understand real parsing and failed out of CS 201. When Knuth can prove (using that mathematical stuff that gets ignored by the amateurs in CS) why xml type parsing is pure evil, why is it still considered?
    Maybe because people who understand real-world IT and aren't so enamoured of their own eliteness that they judge tools by how "hardcore" they are have figured that it is extremely useful?
  • Adrian (unregistered) in reply to xtremezone
    xtremezone:
    PSWorx:
    xtremezone:
    I'm definitely no XML expert. I took a 4 month course on it last semester and after the first week I realized I should stop paying attention for my own good.

    I prefer the standard syntax that has been made so popular in Unix-like systems' configuration files:

    #Comment name=value until newline

    I'm way against XML documents. Explain to me what you can do with an XML document that you can't do with the traditional configuration syntax..

    Web pages for starters?

    Not knowing a technology - no problem Stating a technology is bullshit despite having no clue of it - questionable Stating a technology is bullshit and being proud of having no clue - please leave.

    XML cannot 'do' Web pages. XML defines data. That's it. Anything other than that is done with parsers of XML - which could be done too with configuration files if you parsed them and rendered them in a Web page.

    Actually, now that I think of it this is sort of done with PHP's phpinfo() function, isn't it? If not directly, than indirectly.

    Conveniently, data can be defined and brought into Web pages using databases, which gives you control of data that tends to be lacking in an ASCII text file.

    You've heard of databases, right? The right way to use data?

    XHTML is how a web page is defined. XHTML, if you didn't realise, is XML.

    Write me even a simple web page using a configuration file, then come back and post it.

  • (cs) in reply to steve
    steve:
    s|k:
    "it's so easy that companies feel like they can get away with hiring developers instead of engineers"

    This site gets more full of itself everyday. Seriously, can we drop the generalizations about classes of people from these 'witty' articles?

    Thanks.

    Too bad it's true all too often. Where do you think a lot of the content for WTF comes from?

    Oooh! A trick question.

  • (cs) in reply to sinistral
    sinistral:
    ... Texas has passed this law. Only people who pass the PE (Professional Engineer) exam [and early on, some folks who were grandfathered in] can call themselves Software Engineers. Works just like Civil Engineer, Mechanical Engineer, etc. Other states don't have this requirement.
    What about Sanitation Engineers or any of the other inflated job titles?
  • (cs) in reply to thogard
    thogard:
    You know that many of the b-29s built in Wichita during ww2 were built by former short-order cooks and one of Boeings leading engineers was in fact a historian.
    And what's wrong with historians, exactly?
  • (cs) in reply to Adrian
    Adrian:
    xtremezone:
    PSWorx:
    xtremezone:
    I'm definitely no XML expert. I took a 4 month course on it last semester and after the first week I realized I should stop paying attention for my own good.

    I prefer the standard syntax that has been made so popular in Unix-like systems' configuration files:

    #Comment name=value until newline

    I'm way against XML documents. Explain to me what you can do with an XML document that you can't do with the traditional configuration syntax..

    Web pages for starters?

    Not knowing a technology - no problem Stating a technology is bullshit despite having no clue of it - questionable Stating a technology is bullshit and being proud of having no clue - please leave.

    XML cannot 'do' Web pages. XML defines data. That's it. Anything other than that is done with parsers of XML - which could be done too with configuration files if you parsed them and rendered them in a Web page.

    Actually, now that I think of it this is sort of done with PHP's phpinfo() function, isn't it? If not directly, than indirectly.

    Conveniently, data can be defined and brought into Web pages using databases, which gives you control of data that tends to be lacking in an ASCII text file.

    You've heard of databases, right? The right way to use data?

    XHTML is how a web page is defined. XHTML, if you didn't realise, is XML.

    Write me even a simple web page using a configuration file, then come back and post it.

    Isn't this a little late? I mean, Tim Berners-Lee fucked up the entire Internet fifteen years ago (inadvertently) by inventing the WWW.

    The fact that, fifteen years later, you're better off using shitty little tools like SAX, DOM, XSLT or indeed anything developed by the Apache crowd doesn't mean that you have a good reason to criticise somebody who bemoans the lack of a (simple) alternative.

  • Stilgar (unregistered)

    and of course catch(Exception e){} is always good

  • Brett McNally (unregistered)

    XML cannot be suspicious, it can be something that you are suspicious of.

  • (cs)
    Adrian (unregistered):
    XHTML is how a web page is defined. XHTML, if you didn't realise, is XML.
    Just like XML, XHTML is merely defining data (hense, it is XML). It's data happens to represent Web elements. It needs to be parsed by a Web browser's XHTML parser to be rendered as a Web page.

    XHTML is XML, but XML is not XHTML; so to say that you're creating Web pages with XML is like saying you're creating Web pages with SGML.

    It's true, to a point, but SGML is more broad than Web pages. If you had an SGML parser/renderer and used it to parse an HTML document I'm guessing you wouldn't see your pretty little Web page, unless it supported HTML. At the same time, a purely XML parser would not render an XHTML document as a Web page.

    For the same reason, XML documents are usually rendered as raw data in a Web browser, or sometimes are formatted to be collapsable, etc.

    XML was designed to define data and nothing else. As a result, it does not mark-up Web pages. Some derived langauges, such as XHTML, do (although again it's just meaningless data without a parser/renderer).

    At the same time, HTML is used to mark-up Web pages, which is not a derivative of XML at all. Therefore, you do not need XML to render Web pages and I still fail to see it's usefulness.

    A database stores data, which can be pulled into a Web page. So arguing that it is used to define data, which it is, is again a failed argument in my mind. It does so in a poor way.

    I don't even like SGML syntax, so every language based on it is stupid. HTML has it's use (mostly for evil), but I still dislike it (as many developers do).

    Adrian (unregistered):
    Write me even a simple web page using a configuration file, then come back and post it.
    Even if I wrote a parsing engine to parse and render a configuration file I wouldn't be writing a Web page using a configuration file, regardless of how you see it. I would be writing a configuration file, or at least using the same syntax to define data, and my parsing engine would be generating a "Web page" with it.

    To humour you, here you have a simple Web page:

    #This is a Web page defined in a configuration file syntax:
    heading1=A Web page defined in a configuration file syntax
    font=Comic Sans MS,Arial,Impact
    size=10
    weight=bold
    text=\n\nI am a text string drawn on a "Web page"
    I'm not suggesting we do that. I'm suggesting we stick with using databases to store and fetch data to embed in applications of any type, regardless if it is a Web application or a Desktop application.

    Also, XML is not a Web language, rather is used to define data. As a result, many other applications use it to define data, such as their configuration data. Something I believe to be unnecessary since configuration file syntax does the same thing with fewer characters and is just as, or more, user-friendly.

  • hardly (unregistered) in reply to CaptainObvious

    Wow, quoting the dictionary! but hardly useful here. These terms/titles were hijacked from the construction industry. to make it even more confusing, at my company - architects are engineers, engineers are really developers and engineers have their own lackys called web technologists

  • Anonymous (unregistered) in reply to anon

    All your sarcasm detectors are belong to us!

  • Adrian (unregistered) in reply to xtremezone
    xtremezone:
    XML was designed to define data and nothing else.
    You keeping saying this. XSLT, XQuery, Ant scripts, MathML, ...
    As a result, it does not mark-up Web pages. Some derived langauges, such as XHTML, do
    So XML is used to mark-up Web pages - just a certain type of XML, right?
    I'm not suggesting we do that. I'm suggesting we stick with using databases
    (And not XML databases I bet). Ok, so how are you going to do your simple web page with just a database, and no XML or configuration file?
    I don't even like SGML syntax, so every language based on it is stupid.
    ...I'm not even going to bother.
  • Adrian (unregistered) in reply to real_aardvark
    real_aardvark:
    Isn't this a little late? I mean, Tim Berners-Lee fucked up the entire Internet fifteen years ago (inadvertently) by inventing the WWW.

    The fact that, fifteen years later, you're better off using shitty little tools like SAX, DOM, XSLT or indeed anything developed by the Apache crowd doesn't mean that you have a good reason to criticise somebody who bemoans the lack of a (simple) alternative.

    Nope it doesn't, but crap like 'I stopped listening for my own good, 'configuration files can do it all' does.

    Anyway I don't know why you'd bemoan alternatives: JSON, YAML are both popular. But in the end they're really not that different (http://www.megginson.com/blogs/quoderat/2007/01/03/all-markup-ends-up-looking-like-xml/), and not as well supported.

  • (cs) in reply to xtremezone
    xtremezone:
    Adrian (unregistered):
    XHTML is how a web page is defined. XHTML, if you didn't realise, is XML.
    Just like XML, XHTML is merely defining data (hense, it is XML). It's data happens to represent Web elements. It needs to be parsed by a Web browser's XHTML parser to be rendered as a Web page.

    XHTML is XML, but XML is not XHTML; so to say that you're creating Web pages with XML is like saying you're creating Web pages with SGML.

    It's true, to a point, but SGML is more broad than Web pages. If you had an SGML parser/renderer and used it to parse an HTML document I'm guessing you wouldn't see your pretty little Web page, unless it supported HTML. At the same time, a purely XML parser would not render an XHTML document as a Web page.

    For the same reason, XML documents are usually rendered as raw data in a Web browser, or sometimes are formatted to be collapsable, etc.

    XML was designed to define data and nothing else. As a result, it does not mark-up Web pages. Some derived langauges, such as XHTML, do (although again it's just meaningless data without a parser/renderer).

    At the same time, HTML is used to mark-up Web pages, which is not a derivative of XML at all. Therefore, you do not need XML to render Web pages and I still fail to see it's usefulness.

    A database stores data, which can be pulled into a Web page. So arguing that it is used to define data, which it is, is again a failed argument in my mind. It does so in a poor way.

    I don't even like SGML syntax, so every language based on it is stupid. HTML has it's use (mostly for evil), but I still dislike it (as many developers do).

    Adrian (unregistered):
    Write me even a simple web page using a configuration file, then come back and post it.
    Even if I wrote a parsing engine to parse and render a configuration file I wouldn't be writing a Web page using a configuration file, regardless of how you see it. I would be writing a configuration file, or at least using the same syntax to define data, and my parsing engine would be generating a "Web page" with it.

    To humour you, here you have a simple Web page:

    #This is a Web page defined in a configuration file syntax:
    heading1=A Web page defined in a configuration file syntax
    font=Comic Sans MS,Arial,Impact
    size=10
    weight=bold
    text=\n\nI am a text string drawn on a "Web page"
    I'm not suggesting we do that. I'm suggesting we stick with using databases to store and fetch data to embed in applications of any type, regardless if it is a Web application or a Desktop application.

    Also, XML is not a Web language, rather is used to define data. As a result, many other applications use it to define data, such as their configuration data. Something I believe to be unnecessary since configuration file syntax does the same thing with fewer characters and is just as, or more, user-friendly.

    Indeed.

    Bravo, several times over.

  • (cs) in reply to Adrian
    Adrian:
    real_aardvark:
    Isn't this a little late? I mean, Tim Berners-Lee fucked up the entire Internet fifteen years ago (inadvertently) by inventing the WWW.

    The fact that, fifteen years later, you're better off using shitty little tools like SAX, DOM, XSLT or indeed anything developed by the Apache crowd doesn't mean that you have a good reason to criticise somebody who bemoans the lack of a (simple) alternative.

    Nope it doesn't, but crap like 'I stopped listening for my own good, 'configuration files can do it all' does.

    Anyway I don't know why you'd bemoan alternatives: JSON, YAML are both popular. But in the end they're really not that different (http://www.megginson.com/blogs/quoderat/2007/01/03/all-markup-ends-up-looking-like-xml/), and not as well supported.

    Well, I'm not bemoaning alternatives, am I? I'm objecting to people who do so without cause or thought.

    Thanks for the "crap" comment. I agree. Unix configuration files are crap, and always have been. Period. (For further reference, feel free to check out the various Python attempts to build a generic "Unix configuration file" parser. Cute, but doomed. Made even worse by trying to incorporate MS ini files ... Python is great, but it can only do so much.)

    Personally, I like Stratus VOS configuration files. They all adhere to the same easily parsed microlanguage (I know it's easily parsed, because I've built a parser myself. And I'm "crap" at this sort of thing). They all sit in the same directory -- none of this init.d rubbish. Generally speaking, they don't need tweaking between VOS 6 (1987 or so) and VOS 13 (1999 or so). They're human-readable, they're computer-readable, and they're consistent.

    I have no idea what JSON might be, and I'm not going to look it up. I've been forced to work with XML, and I hate it. YAML appears to be a pile of shite.

    AFAICS, the only good reason to use XML is as an interoperability tool. It's thoroughly useless on its own, and (like many badly-designed tools) attracts a whole bunch of enthusiastic incompetents who will defend its inadequacies to their dying breath. And it just goes on, and on, and on ... like the Energizer Bunny. Makefiles are horrible, but after twenty years, I've got used to them. Ant is just obscene. Are you really suggesting that it's easy to fix a problem that sits three files down an include chain, and is dependent on every single thing above that chain?

    XML is a necesary evil. Mostly it's necessary for marketing purposes. Let's not kid ourselves that it has anything to do with good practice in software.

    Thankyou and good night,

  • Gabriel (unregistered) in reply to W
    Can you say: JAX-B?

    jaxb has pretty much made xpath superfluous in the company I work for.

    Too bad most people who work with xml have never heard of schemas, and most of those who don't are of the kind that think a doubling in bandwidth or parsing speed is always a disaster.

    I think part of the issue also is that at this point, those of us with little or no exposure to XML don't have a clear idea of where to start. Knowing about the existence of JAX-B, or XPath, etc, and (more importantly) what the differences are takes more than a little research.

    When there are a bunch of people out there saying "Use My technology-flavor-of-the-month!", it's hard to sort the cruft from those that have a clue if you don't have any domain knowledge. It'd be great to have a "Technology X is better to use than Technology Y for small problems, whereas Z is better for complex problems" from someone who can give a good explanation of them. Kindof like the sort of things Martin Fowler gives us in his Patterns of EAA book.

  • Adrian (unregistered) in reply to real_aardvark
    real_aardvark:
    I have no idea what JSON might be, and I'm not going to look it up. I've been forced to work with XML, and I hate it. YAML appears to be a pile of shite.
    Ok, so you hate everything. A lot of people will at least point to specific faults in XML, and say YAML is better because of XYZ, or whatever. But you seem opposed to any sort of structured document? What do you propose using for exchanging information in a language independent way?
    AFAICS, the only good reason to use XML is as an interoperability tool.
    And interoperability is pretty damn important & pervasive.
  • Adrian (unregistered) in reply to Gabriel
    Gabriel:
    I think part of the issue also is that at this point, those of us with little or no exposure to XML don't have a clear idea of where to start. Knowing about the existence of JAX-B, or XPath, etc, and (more importantly) what the differences are takes more than a little research.

    When there are a bunch of people out there saying "Use My technology-flavor-of-the-month!", it's hard to sort the cruft from those that have a clue if you don't have any domain knowledge.

    That's true.

    At least Java's JAX* API's have improved & grown over time. Although I guess the w3c DOM API will forever suck.

    SQL integration in Java faced the same problems though - a ton of different libraries, and ton of ways to achieve the same thing. Over time some best practices settle out: the same thing should occur with XML, and with the advantage that XML and the related technologies (XPath, XQuery) are much better specified & interoperable than SQL ever was.

  • (cs) in reply to Adrian

    After reading these comments I have concluded, that I should give up useing XML & XSLT for processing data since it is crap. Instead I should begin using UNIX configuration files.

    Could someone (not to mention any names) give me an example of unix configuration file containing structured & nested data, and how that data can be transformed into various other formats? Perhaps by writing an unix configuration file describing the transformation to be done? Is there an unix configuration file which could be used to describe the data structure so I could validate my data so I can find errors easily? Are there any ready-to-use components available so that I don't need to write my own?

  • Tobe (unregistered) in reply to W

    People haven't heard of Schemas? With all the XML-related garbage being spewed out from vendors, w3c and people like yourself, most people I talk to think Schemas are a pre-requisite for XML.

    The people who really understand what XML is about avoid schemas like the plague. Ever heard of EDIFACT and UNSM:s? Well, schemas are making the same mistake again, forcing XML into the same constrictive path.

  • (cs)

    Can we please keep sarcasm to a minimum on the posts. Either it doesn't translate or you suck at writing it. Let's save everyone the trouble of parsing it by dropping it.

  • docca (unregistered) in reply to Jan

    One word: XStream (http://xstream.codehaus.org/)

    Saved my bacon when I had to deal with XML monstrosities in Java.

    Captcha: atari

  • codefish (unregistered) in reply to anon
    anon:
    slashbot:
    jho:
    DrkMatter:
    Someone forgot their sarcasm detector at home today.

    Is that meant to be English? You American's sure do twist our language some times.

    What's the "twisted English" in the original?

    you can "forget something".

    or you can "leave something at home".

    But "forgetting something at home" - that's a linguistic WTF.* At least to English English speakers - perhaps it has evolved as a common/acceptable idiom in American English, I don't know.

    • It does make sense, it just doesn't make the sense the speaker intends ;-) as it tell you where they did the act of forgetting (at home), not where their sarcasm detector is (undeclared).

    Arghhh...

    Clearly the real WTF here is the idea of using the same sarcasm detector at home and work. I find the required settings to be COMPLETELY different.

  • mightymacx (unregistered)

    the fact that the whole thing is also wrapped in an empty try block...amazing!!!

    if any of those threw an exception, then what? hahaha...

  • Patrick (unregistered)

    NYAGH! MY EYES!

    NINJAS! THOUSANDS OF THEM!

  • prabu (unregistered) in reply to EER
    EER:
    In fact, XPath IS way too complicated for anyone who doesn't want to bother looking into it (virtually everyone) so I comletely understand why HashMaps are so much better :)
  • <xsl:jump-off-cliff /> (unregistered) in reply to Unomi
    Unomi:
    Or wat is wrong with XSLT? I'm not into XSLT that much, so if anybody can point me to any drawbacks, please do.
    <xsl:choose> <xsl:when> <xsl:otherwise> <xsl:for-each> to name a few. The syntax was clearly invented by retarded monkeys.
  • Dieter H (unregistered)

    Engineers don't work in IT. No, a software engineer is not an engineer... It's a misnomer. Read Alistair Cockburn on the subject.

Leave a comment on “Shooting XML ”

Log In or post as a guest

Replying to comment #:

« Return to Article