But fun thing, i’ve accidently noticed,  that different people .. like different patterns.And most of us probably have the most favourite one. Its fun to notice that in the same situation different people tend to reuse his favourite in the first place When some start with the facade, the other try to make the realy strong builder and the third tries to suggest to start with the singletone.

But what if it doesn’t feet (after all, the patterns mentioned above have different purposes)? well, we can implement something strange and probably ugly, but reminds us the pattern we love

PS I’ve tried to find psychological aspect in that preferences. This is crap, for sure. But the guy, i know, who likes builder is a kind of workoholic. And the one that likes facade.. plays the most important role in the project, and it is his concern to communicate with the managers of the project and then to give tasks to the coders in the team:)

Well, the exam was quite easy. Training was much harder and required some effort.

The fun thing is training materials (i mean the book for preparation and mock exams) look VERY scary.  After first attempt to pass mock exams i feel myself being totaly stupid. The exam is so f.. simplier!

In fact this is the reason why some people think SCJP (and probably other sun exams) useless. If you have 1+ year experience of java development, there is no way for you to fail it. And if you just try to pass it, you’ll be thinking “what’s a crap. that was to easy”.. But if were preparing and trainig before it.. it’s for sure that you will find a lot of white spots in your knowledge..  (after all, now i don’t find java being so elegant like  i used to )

Bottom line is 65% passing score is too low. Think It should be 80%.. yeah.. if i’ve reached 90%, i’ve been saying now that passing score should be 90%  :)

Probably my google is not so good, but I could find adequate explanation through all that buzzz around properties in java

This year wasn’t very nice. Though, I have something to be proud of, I can’t really say, that I did the best of my self..

The year started with a funny trip to Siberia. Wow, that was fun and stupid! Being curious about how people live and work there, I’ve leaved a job in a large development team with about 150 coders. Well, I’ve seen Siberia It was imperssive after all. Its really cold there For those who thinks they are real travellers couse they’ve seen New York, London, Paris, Rome etc, etc, etc.. you suck!

This year I’ve learned to drive cars! Without fear, being relax and having fun. That’s not very hard. If you have smb to teach you. Well, I had not. I did it myself. And I’m proud of it. The sad truth is that I don’t have any car… and can’t even think about when I could buy it. But nevertheless, cars are great symbols of freedom. Cars are symbols of the man, after all. or the woman… if it’s daewoo matiz ))

I digg a lot in coding, development. I’ve reached understunding of such cool things like unit-testing and aop. I’ve completed not very light project on j2ee. I’m kind of grown up in all-that coding stuff.. No more lame spaghetti code, that just pretends to be object-orientired. The sad truth is that I’m 22 now.. and perhaps I should have done all this just about 2-3 years earlier. In the time when I was concerned just about “money”… if one can call so a “40-50 $ per month” a salary.

Well, I’ve completed my unversity education. That was lame and waste of time. Though my certificate of degree doesn’t has any B or C marks…That wasn’t hard. I’m a geek. Studing is my main skill The sad truth is that sometimes i think that it’s the only skill, i have.

I’ve get acquanted with lots of people.. Smug people, grown ups, disappointed people, successfull people, lame people, geeks… I’ve learned a lot about humans The sad truth is that I’ve lost almost all my connections with old ones.

Next year i have to change my life cruelly. I have to change almost everything. I’m too stucked in the place where I’m now. Somedays before I’ve tried to count my life-cost.. If I would continue to live like I do it now, It wont succeeds 150 000 $ per life. Not very impressing.. not at all. That doesn’t worth living

I give a lot of promises to myself. My goals are cheeky and my objectives are hard and large.

Well, I’ll see how it will be at the next new year’s eve

The very significant thing that there wasn’t much buzz about delphi for php. Though, buzz in my country is rather strong. Couse delphi traditionally was the most popular development platform.

I DON’T LIKE BORLAND! Though, i was using Delphi and C++ Builder for a long time. I maked up 2 rather strong projects with them. But I DON’T LIKE them. Couse they don’t encourage me to grow, to find more elegant solutions.

They keep silence about refactoring, unit testing, profiling, about design patterns(MVC first of all), version controll, team work and all that buzz-stuff. They have their own complete framework VCL and design editor. That’s all. And they think that is enough for most of the projects. And perhaps they may be right. That’s almost enough for most of individual light business projects. Though I remeber the time of my trying to make up new component combining TDBGreed and TComboBox. That was awfull and disgusting. Because of limitations of VCL. The main aim of the delphi was to replace visual basic. And they get a great success, in my country at least.

Delphi for PHP is totally the same. It’s designed for so-called Rapid Application Development. It’s fun and it’s exciting when you watch demos (helo-world and database-access). But when you’ve installed it, the very first thought is “wait a minute, am I in 2000 again? What’s a hell?? Where is all that cool stuff that all IDE have nowadays?”

Where is:

1) version controll integration

2) php doc

3) unit testing

4) refactoring patterns (through, zend studio neon is rather poor in refactoring too, but it has 3 patterns)

5) class templates

etc

Where? I can’t find it.

Where is smart code analyzer/intellisense? Why I see only dumb, clumsy peace of smth?When i’ve created a new so-called Unit (а plain file with only “<?php ?>”-content) it doesn’t even recognized my class members after all, and couldn’t even complete “public function __constr” line.

It’s the same dumb delphi.

I agree that php needs RAD tools. I agree that php needs frameworks like VCL. They may be very convinient sometimes. Php needs WYSIWYG editor, like jsf and asp.net have.
But in the times when zend struggle to make up php more professional-like(i would rather say more java-like), borland may cause a lot of harm to young php society(i mean society where there are a lot of not very experienced students).

In my country there are still a lot of so-called php-coders, who are stuck in “php 4″.

When people migrated from borland pascal (using it without any knowledge about OOP) to delphi that wasn’t very cool. Couse most of them started to use delphi in some strange ugly half-procedural half-object orientired manner. Borland Delphi doesn’t encourage coders to be more professional, to grow up. It encourage them just to find complete solutions(new vcl-components) and use them. For those who just yesterday struggle to write “helo-world” program, delphi brought sweet feelings “look, mom, i’m making real Apps!”(i felt totally the same, i was 14) When i see some projects of that time(even made up by teams).. wow. they are ugly.

And now if people will start migrating from plain “php 4″ to “delphi for php”. that will be the same thing. I can quite imagine that soon will start “look, mom, i’m making sites!” disaster. and that will be “return of the era of darkness” but in web

Googling aroung didn’t show up any serious projects base on the platform. Some examples look like a bad joke. But strengthen the “déjà vu”-feelings.

PS here is a good review of VCL for PHP framework drawbacks

unit testing KICKS ASS, guys. it really does

For me, as junior developer, it was very hard to imagine how to test the code, i used to write.. i spent a lot of time thinking how to do that. the answer was so simple, that i can’t get it the answer was “you should write it in another way”. the pragmatic principles about weak coupling and strong cohesion are the keys to the world when everything become simple and clear.

To be able to test a code, one should write rather small classes with sometimes tiny methods. if a method has a deep path of nested conditions, you can probably die trying to test it.

And I had a wrong criteria, which i’ve learned at university desk, of how long a method should be. I thought the main aim is to put all lines of it in one screen, with no scrolling. But it’s a very-very old criteria, coders used to consider perhaps in dos times, when one screen was about 25 lines long. Though, i think that a method with 25 lines of code isn’t tiny at all. Well, “do one thing at a time and do it right” is much more progressive.

Unit testing brings a sweet feeling of satisfaction by your work. Cause you can see result of you work very early. even before you complete a single method of the very first class. That’s very sweet indeed.

Well, i’m feeling to much delight in this tool now i need a time to calm down and to write some examples of how fun unit testing can be

But as for me i think that the best revisions of core java and j2ee principles could be found in “Java/J2EE Job Interview Companion” by K.Arulkumaran. It’s well orginized and covers large amount of core questions. Just go and look through table of contents. There is about 400 questions.

Well, it’s revision! That’s why it’s real short, but that’s great, i think. 237 pages can be looked through in one night just before your interview

By the way, recalling my try to get .NET developer position.. most question that i didn’t answer are covered in this book

Unfortunately, the book isn’t cheap, but it’s worth its cost.

And by the way, blogg I mentioned above may be useful too. But it’s more specific

Yep, there is still no namespaces in php. And if you open source of Zend Framework, you can see that they suggest to combine you namespace path in class name.. this leads to names like Zend_Gdata_Calendar_Extension_AccessLevel

Well thank you, zend-guys.. but i don’t want to spent most part of my life typing class names..

static method and static fields.

Yeah, there is static keyword.. and yes it have some functionality. But i suggest that in one of two cases you’ll face some little bit strange behavior. Especially if you want to use overriding through inheritance.

Well, the most enthusiastic discussion you can find here.

Cann’t remeber any case of using static keyword whithout pain..

No refactoring tools

What’s a hell? I need to rename stuff! I need to extract methods! I need to remove safely! Why nor Zend Studio, nor Eclipse For PHP don’t have refactoring? Can’t understand. Do they think that php is superior language and the development doesn’t require refactoring?

Well, there are more tiny tings that make me sick. But these hurt me almost every time when i have to code in php.

Well, talking about views. Of course when you have you database in normal form you have no chance not to use views. but i have strange habits for using them for:

1) DRY principle when i have to use the same WHEN-expression more than to times i’d make a view for that stuff. lacking perfomance, i know.. but it’s so convinient

2) usage with ORM . sometimes i need to produce some ManBearPigs (espessialy when working when statistic stuff) and to communicate with it. most ORM tools doesn’t support stored procedures and functions (neither do ejb 3.0). so creating view is the only solution for that case. i can be totally wrong though

This book is rather old it was published in september of 2005. And A LOT OF DISCUSSION can be found with google. it’s very polemic. well, Bruce showed himself as a master of dramatic statements Sometime you feel that it’s Bruces very own tragedy that java’s leadership may end someday.

I liked that book, especially for the point that i keep my eye on opposition from .NET, but didn’t realize that a real threat probably will come from the very different side..

Didn’t make all my own conclusions yet, but certanly will.