Kirill Grouchnikov's Blog

Developing Java project - a team effort

Comments

• I think the entry's observation was a little one sides. The list is too concerned with horizontally layering the application concerns where you can take frameworks now like JSF (component based frameworks) and manage concerns vertically. It makes development tons easier, and compensates for confusion over AJAX functionality and client-server communication.

  Posted by: jhook on October 14, 2005 at 05:39 PM

• Nice observation Kirill.

  As new developers, and managers, we were all guilty of falling for the 'silver bullet' pitch; hoping to quickly attain sophisticated ability, without experience.

  As we mature, we realise that only through thorough understanding, and practise, with the technical fundamentals; combined with specialisation and teamwork; does any significant accomplishment possible.

  Unfortunately, as you are experiencing, once you have attained this vantage; the 'snake oil' pitches become positively irksome!

  Posted by: cajo on October 15, 2005 at 11:48 AM

• Yes, and every now and again, some jaded developer will make cynical observations, lamenting that his fellow developers are falling for "the next big thing" and are being fooled. Little does he know that he's just become a calcified old codger. When new and useful tools come along he blithely ignores them, and mutters to himself as he yearns for the good old days of punch cards and serial terminals.

  Posted by: gbarton on October 17, 2005 at 08:09 AM

• @cajo - I think that Kirill was aware of the phrase 'silver bullet', as you can see by his play on words with 'golden hammer', ie something supposedly even better than a silver bullet ('the best thing since the subroutine' hype that these frameworks often come with from breathless admirers) - that once you have it you start treating every problem as though it can be solved by your framework du jour (eg: 'to a man with a hammer, every problem looks like a nail'). Possibly with other subtleties thrown in - eg gold being a soft metal would make a particularly poor hammer. (etc)

  @gbarton - more tools is not necessarily better. Just because these frameworks are shiny and new does not automatically promote them to being the best. We should examine them with a critical eye. You shouldn't attack Kirill for being critical. We should be rational about our tools. Indeed, I have yet to see the web framework that measures up to the power, ease of use, and scaleability of 'plain old servlets'.

  @Kirill - I don't think that list is fair. Having to learn Struts should *not* be on the list of things required to learn Ajax. Lets cast a critical eye at that list:

  HTML 4.0
  JavaScript
  XML / AJAX
  DHTML
  HTTP
  JSP/Servlets
  Tag Libraries
  Struts, Tiles, Validation (these have now been combined)
  WebSphere / Tomcat

  Should really be:
  HTML 4.0 / DHTML
  JavaScript
  XML
  AJAX
  ... and the cgi of your choice

  And how hard are Html or Xml really? Javascript can be quite tricky because of versioning/browser problems. You can choose an 'easy out' for the cgi, eg servlets (or 'insert random scripting language' of your choice), without mucking around with the pain and agony of JSPs, tag libraries and struts. If we remove those things the difficulty drops by a factor of 4 at least. I wouldn't want to inflict weblogic on anyone, but tomcat isn't too bad (the one built into netbeans is quite easy to use). Websphere wasn't too bad either (disclaimer: was using ibm toolset and someone else set it up).
  

  Posted by: rickcarson on October 18, 2005 at 04:07 AM

• Rick,
  I am not going to argue about whether the linked list is correct or not. It was just an illustration. But thanks for the insightful comments.
  gbarton,
  You say "new and useful". The only thing new in AJAX is the acronym. Web developers have been using the asynchronous calls to server from web pages since 97. I remember using a hidden IFrame to retrieve contents of a combobox based on the selected item in a previous combobox. There was no blocking, and the contents of the hidden frame were filled with XML. That was in 2001, three years before AJAX term was coined. The only thing missing back then was XmlHttpRequest.
  I remember one post of a delighted AJAX developer who showcased a progress bar that talks to server and updates its contents. Wow, what a great "new" component. I wish Sun adds it as a Swing component in Mustang :)

  Posted by: kirillcool on October 18, 2005 at 08:51 AM

• The thing is that all new technologies are marketed as Revolutions but most of them are purely evolutionary,
  so only if you set aside the marketing you get the real picture.

  As I see it most of the 'new' technologies are just an integration of the old ideas in a more slick and usefull package, and what is what it's all about, making old problems go away more easily.

  Posted by: melhiah on October 19, 2005 at 03:21 AM

• 
  Kirill mentions several software solutions that promise to cut development cost, time and productivity in half (in worst case).

  
  Add development methadologies to that list as well. It was a short time ago everyone hammered the effectiveness & benefits of XP programming over the tradational SDLC. Nowadays its hard to justify elements of XP, such as "never code for future needs, only the needs of now" versus the now heavily used design patterns.

  
  Sure there was good elements of XP, like pair programming. But XP's mantra of "do not program for future requirements" is a direct conflict with many patterns I can think of, i.e. Strategy (why extrapolate behavior if nothing changes in the future), Factory (why isolate instantiation if we have no classes in the future), etc.

  
  But our fads are fun.
  

  Posted by: phlogistic on October 19, 2005 at 03:28 PM

• kirillcool, I was mainly reacting to your dissing of AOP.

  And, yeah, sounds like you were ahead of the curve on AJAX. Great. Now the technique has been formalized and named. So what's with the old man "there's nothing new under the sun" routine? Yeah, the young'uns coo in delight at stuff we old codgers find old hat. Thet's their job. Our job is to steer them in the right direction, but also learn from their example as we forget, in our old age, to be open to the new stuff... Like AOP... :P

  Posted by: gbarton on October 20, 2005 at 07:41 AM

This work is licensed under a Creative Commons License.