f13o

PHP5 Unicode support

2010-02-04T09:41:00.000+01:00

Short one on how to setup PHP multibyte support correctly if You need complete (although not 100%) support for UTF-8 in PHP5 applications. Still, some functions (link here!) don't want to know anything about multibyte characters. Hopefully this is going to change once we enter the fantasy world of PHP6 ;)

INSTALLATION
If you are compiling php from source (usually on Linux) you will need to pass following options:

--enable-mbstring
--enable-mbregex

On Windows you will need to enable mbstring support by editing php.ini file. Make sure the following entry is uncommented (remove ";" in begging of the line):

enable=php_mbstring.dll

Also check that the appropriate directory containing the mbstring extension DLL (php_mbstring.dll) is in the path where PHP searches for esxtensions by settings the extension_dir configuration options in the same php.in file:

extension_dir="c:\php\ext"

CONFIGURATION
In php.ini the following options needs to be set:

mbstring.language Neutral
mbstring.internal_encoding UTF-8
mbstring.encoding_translation On
mbstring.http_input UTF-8
mbstring.http_output UTF-8
mbstring.substitute_character ?
mbstring.function_overload 7

While reading Spectrum, November Issue

2009-11-04T21:29:00.000+01:00

What is EGL?

2009-09-09T08:28:00.000+02:00

EGL from IBM, new product, installed as Eclipse plug-in (although they recommend to download id as a bundle), is not all new stuff. There were all ready various solution, to replace the nitty gritty from the developers.
My personal opinion, is that those tools are usually good for one kind of problems, or two. But, when it comes to customizing the solution, this is where the tools usually start smelling.

They say: "The goal is to shield developers from complexities typically associated with Web 2.0 application development". We'll see how will the manage to do it...

Anyway, EGL is a language, new language, they say it simple, and compiles to Java, JavaScript, COBOL.
The language itself hides many complexities from developer. For example, to access Google Map API, just do: map GoogleMap { width = "500px", height = "300px" };

Anyway, I haven't installed it yet, but will, and when I do try it I will demo an application and hopefully show something here.

Reference: http://www-949.ibm.com/software/rational/cafe/community/egl

nWire experience (in PHP)

2009-09-02T16:15:00.002+02:00

Looking to the PHP codebase you inherited is not the very best task a developer can imagine. All those dynamics can make one looking for variables or objects without knowing where it all came from or what is the path included.

Having myself such an experience, I was looking for a tool to help me connect all the dots in newly experienced code base I had been given. I needed a 'thing' to go through all the code in current project, search and relate the classes and files across the project scope.

One again I turned to “PHP via Java” search. Lot of stuff that Java has, already is or can be implemented for PHP. Lot of tools for PHP (such as PHPDoc, Phing...) are counterparts of well known tools for Java (JavaDoc, Ant...). So, I was looking in that direction...

For code analysis and visualization I have been given hint to look at nWire (for PHP, that is) some month or two ago. As I am using Eclipse PDT installing nWire was a breeze and as with most other plug-ins, this was just as simple to start using it.

nWire enabled me (in real-time) to look at variable related context (reading, changing, inheritance, other files..) in a way I couldn't imagine before. This is really nice if you got 100K+ lines of code without knowing much about it, but must to do something with it. In just few hours one can get understanding of the code relation and with help of other tools (xdebugger & kcachegrind) getting to know your code is just more fun than before.

The great thing is that one can see instantly where selected variable is being read or modified. And with just a click you can see the exact line. In dynamic languages such as PHP, this is enormous help, especially on project where the code is not, … you know... pure...

Event better, see you code! Visualizer helps you visualize code relations in context of classes, inheritance, includes/requires with nicely looking graphs.

The down-side of nWire is that is requires some CPU power at the start (to build repository, see their site for details) and sometimes it can be a performance hit, but I am sure they work hard on resolving these issues with future releases. If it bothers you, disable nWire support for a while (after building repository) and later enable it as you move along with your project. I did this while debugging some more complex stuff to save my CPU clocks, but just because I was in a hurry to finish the code that day.

I haven't test nWire for Java (yet) so I can really compare if the tool is more performant for Java. I reckon it could be, as with PHP you must look for a type for yourself, but I don't know really details of implementation (the tool is NOT open source :( ). I know they let the PDT do type resolving, but that is my limit at this moment.

Even with a little performance issues, from now on, I am sure nothing could replace it in my growing arsenal of SD tools.

See more details at: nWire for PHP (http://www.nwiresoftware.com/)

lebac

2009-07-10T01:08:00.000+02:00

Surreal sky

2009-05-22T23:28:00.002+02:00

Surreal sky with rain drop on lense.

Catching thunder strikes

2009-05-22T23:25:00.002+02:00

Catching thunder strikes over Belgrade... #1

Chrome and delicious

2009-02-06T20:33:00.009+01:00

If you are like me, want to use this speedy Chrome more often, but are to tightly tied to Firefox for its more-than-usefull add-ons , there is one little hack to at least help you in daily surfing.

I don't want to save local bookmarks, since I discovered delicious . So, in FF and in IE there is delicious bookmark plugin, but for Chrome there is not (yet).

But here I've found interesting trick how to help bookmarking your favorite URLs (just the address inside address-bar).

Sweet Linux

2008-12-22T13:07:00.002+01:00

Nice new video for today...

Happy Esperanto Day!

2008-12-15T09:13:00.002+01:00

You may not know it, but today is a day of celebration for up to 15 million people around the world, who speak Esperanto. To them, we say: “Saluton. Hodiau estas Zamenhof tago!” (Hello. Today is Zamenhof Day!).

Go and learn more: lernu.net

Enterprise PHP

2008-12-11T15:35:00.001+01:00

Enterprise PHP (php|works 2008)

View SlideShare presentation or Upload your own. (tags: php phpworks)

ACT NOW!

2008-11-30T18:24:00.001+01:00

ZendCon 2008 slides in one big download

2008-09-30T13:53:00.007+02:00

As You are "lucky" as I am You probobly haven`t been on ZendConference this year. As for the rest of us, they posted slides on slideshare, and that`s all very nice. There are some really interesting ones and I am eager to view them all (well, not 100% all ;-P ).

There is a little catch, tough. My personal preference is to have them offline, so I can read in bus or nature, where wifi is not available.So, I have downloaded them all and place them in one .zip file, so every one else can just download them in one click. Hopefully I am not violeting any copyright laws, as I never touched any contents.

Enjoy! Here is the .zip (hopefully)

Google videos on Android-Powered Phones

2008-09-26T10:14:00.008+02:00

I would like to share with You some YouTube videos about new Google's mobile platform - Android. There is a full web experience and instant synchronization with Your Google account and all well known Google apps: Maps, Contacts, Calendar...

... and onther one...

World Day Against Software Patents

2008-09-24T14:47:00.003+02:00

Five years ago, on 24 September 2003, the European Parliament adopted amendments to limit the scope of patent law and thereby protect small software companies from the harmful effects of broad and trivial software patents. A global petition asking to effectively stop software patents worldwide will be launched on 24 September 2008, together with specific additional requests for certain regions such as Europe, the United States or India.

What to do?

Please sign or comment on our Draft Petition
Write to your Patent Office, Senators and Deputies
Spread the word in your blogs, mailing-lists, news and social sites
Join our Facebook group and invite your friends to join the cause
Put a banner on your website and help making new ones
Buy a T-shirt
US: Ask your presidential candidates McCain and Obama about software patents
Observe a minute of silence for the death of software innovation on the 24 Sep at 12h00
Donate money (coming soon)

Reference: http://stopsoftwarepatents.org/

Google Chrome - comic book

2008-09-02T13:12:00.001+02:00

I just spotted this interesting comic book drawn by no less than Scott McCloud, creator of the classic Understanding Comics.
The book points to www.google.com/chrome, but I can’t see anything live there yet. In a nut-shell they picked-up from all around (InternetExplorer anonymous browsing, Opera SpeedDial...). Revolutionary concept they are announcing is multithreaded bowser. Every tab is separate process with separate memory space and management. It should allow browser to crash more gently - just one crach per tab :)

Go, scan the commic book for yourself: http://www.google.com/googlebooks/chrome/

Next Generation Mozilla JavaScript Engine

2008-08-31T20:53:00.010+02:00

The Mozilla Foundation has developed TraceMonkey a trace-based JIT compiler that pushes the envelope on JavaScript performance. With plans to be incorporated it in the 3.1 release of Firefox, it delivers near C performance and promises to ‘leap frog’ RIA development to a new level.

As Andreas Gal says:

Our JIT generates code that is roughly equivalent to the performance of unoptimized C code (gcc -O0). ...
TraceMonkey is a trace-based JIT compiler and it pushes the envelope on JavaScript performance. On average, we speed up Apple’s popular SunSpider benchmarks by factor 4.6 over the last release of Firefox. The overall runtime of SunSpider improved by about 1.83x (parts of SunSpider excercise things like the regular expression engine which is outside of the scope of JIT compilation, hence the lesser overall speedup). For the SunSpider ubench suite, which focuses on core JavaScript language features, we achieve a speedup of 22x. Whichever metric you chose to apply, Firefox now has the fastest JavaScript engine in the world.

How this tracing works? He explains:

Traditional just-in-time compilers (like Sun’s Hotspot VM) are in their design and structure very similar to static compilers (like GCC). They observe which methods get executed frequently, and translate the method into native machine code once a certaint threshold has been reached. While such methods often contain performance-critical parts (such as loops), they often also contain slow paths and non-loopy code, which barely if at all contributes to the runtime of the method. A whole-method compiler, however, has to always analyze and translate the entire method, even if parts of it are not particularly “compilation-worthy”.

Trace-based compilation takes a very different approach. We monitor the interpretation of bytecode instruction by the virtual machine and scan for frequently taken backwards branches, which are an indicator for loops in the underlying program. Once we identify such a loop start point, we follow the interpreter as it executes the program and record the sequence of bytecode instructions that get executed along the way. Since we start at a loop header, the interpreter will eventually return to this entry point once it completed an iteration through the loop. The resulting linear sequence of instructions is what we call a trace.

Traces represent a single iteration through a loop, and can span multiple methods and program modules. If a function is invoked from inside a loop, we follow the function call and inline the instructions executed inside the called method. Function calls themselves are never actually recorded. We merely verify at runtime that the same conditions that caused that function to be activated still hold.

Without details this means:

We have, right now, x86, x86-64, and ARM support in TraceMonkey. This means we are ready for mobile and desktop target platforms out of the box.
As the performance keeps going up, people will write and transport code that was "too slow" to run in the browser as JS. This means the web can accomodate workloads that right now require a proprietary plugin.
As we trace more of the DOM and our other native code, we increase the memory-safe codebase that must be trusted not to have an exploitable bug.
Tracing follows only the hot paths, and builds a trace-tree cache. Cold code never gets traced or JITted, avoiding the memory bloat that whole-method JITs incur. Tracing is mobile-friendly.
JS-driven <canvas> rendering, with toolkits, scene graphs, game logic, etc. all in JS, are one wave of the future that is about to crest.

"So what does this all mean? It means that JavaScript is no longer confined by the previously-challenging resource of processing power. With this improvement it's leap-frogged any sort of traditional and has gone head-to-head with computationally-powerful languages like C." (John Resig on these gains)

If you want to try out TraceMonkey for yourself, you can download the Firefox 3.1 nightly, open about:config, and set the javascript.options.jit.content preference to true. Although it is still work-in-progress and buggy, John Resig claims that the nightly build should be good enough to handle most web sites.

How to merge trunk into branch

2008-08-05T12:00:00.003+02:00

The point is to note the revision numbers range from the tree you want to merge from. The merging happens into the current working copy, so take a note where you are when merging. Probably the most important thing is to write merge info into log message, as Subversion does not track this.
As I am developing on Windows platform and using TortoiseSVN, I have ended up merging in the GUI. TortoiseSVN has a nice feature of interactive merge. When a conflict occurs on a merging file, user is presented with a dialog to decide on next step. User can use a local file, use remote file or directly open&edit conflict or postpone resolving conflicts, i.e. leave the file conflicted. This is a nice feature, as I had a big merge (~2k files) and something around 24 conflicted. As lazy as I am I never would resolved conflict in one row. But, by using "online" resolving I had a time to rest between two conflicted files ;) (And wrote this post, also.)
Happy merging!

References:

PHP 5.3 alpha 1 released

2008-08-02T16:32:00.004+02:00

First PHP 5.3 alpha released yesterday and here are some links with relevant detailed information about this alpha version and 5.3 version in general:

Wiki page(s) about PHP 5.3 upgrading notes
NEWS file CVS
Official documentation has some things documented

The files for download can be found here: http://downloads.php.net/johannes/

At least, do a test run installation to support early bug catching. Next step would be to run your app (not in production) and/or write some test.

Check the owner and use the logs

2008-08-01T02:47:00.000+02:00

I was setting some home backup system using tar and rsync.
As it was late I stuck with remote authentication on rsync. I generated my RSA key and updated remote authorized_key file, but something was missing...

Then I realized somebody invented /var/log/ and something called "permission" ;)

So... When setting up something like rsync and you don't want to type a password but want to play it on the safe&secure side - check the perms on ~/.ssh/authorized_keys! This file must be owned by the user and writable/readable only by this user, because ssh checks for this.

And in case this isn't the case - loggile should be written in the log when you attempted to connect (in my case /var/log/secure).

UTF8 in MySQL server

2008-07-31T14:33:00.002+02:00

He who dares - wins.

Edit /etc/mysql/my.cnf
Locate the section that starts [mysqld]

And add these lines under it:
# utf8
init-connect='SET NAMES utf8'
character-set-server=utf8
collation-server=utf8_general_ci

Moving databases from host to host

2008-07-31T14:31:00.003+02:00

I needed this the other day: move some DBs from one host to another.

An easy way to move a number of databases from one host to another is:

$ mysqldump -uSRC_USER -pSRC_PWD -a --databases DATBASE1 DATABASE2 DATABASE3 | ssh TARGET_USER@TARGET_HOST "mysql -
uTARGET_USER -pTARGET_PWD"

100 push-up challenge, week #2

2008-07-09T20:10:00.003+02:00

It's been a while since I posted on this. Being busy ;)

I've done my week #2's day #1 and #2:
week2day1: 12, 12, 9, 7, 10
week2day2: 16, 13, 11, 11, 10 (plan was to do: 16,13,11,11,at least 15)

The day 3rd I have missed due to very tight schedule I had. Wanted to do on day after, but I was helping my girlfriend to move (that counts as work-out!).

For any reason I have decided to go trough week #2 again. This time I hope not to miss it :)

In the mean time I read a little about how to do proper push-up (breathing and stuff). I think my push-up improved...

Anyway, week #2, 2nd take, day #1 results: 12, 12, 9, 7, 15

Hopefully, my push-ups are improving and, what's more important, I am feeling better ;)

phing basics

2008-07-09T15:35:00.004+02:00

Short notes for me...

INSTALL
pear channel-discover pear.phing.info
pear install phing/phing

BUILD FILES
Build files consists of:

TASK - built-in specific action
TYPE - data structure used by task
TARGETS - grouping of tasks (to perform certain actions)
PROJECT - root node of the build file

Properties: variables for build scripts.
Dot-path notation for class names: path.class.name = /path/class/Name.php
<fileset> - ability to match set of files with certain pattern
<mapper> - adds filesystem transformation ability (copy, move...)
<filterchain> - adds data filter/transformation ability
<pdo> - allows to execute SQL statements from the buildfile or external file(s)
<propertyprompt> - basic user input interaction
<input> - more feature rich version of <propertyprompt>: yes/no, multiple choice, input validation
<hpdoc> - wrapper for phpDocumentor application (use values from phing build file)
<tar> and <zip> - ability to bundle packages/folders/files
<pearpkg>, <pearpkg2> - creating PEAR packages

Extensive support for PHPUnit and SimpleTest.
Support for Continous Integration (XINC, CruiseControl).
Embedding PHP using <php> and <adhoc> tasks.
Deplyment Tools: <svn*>, <scp> tasks
Validating: <jslint>, <xmllint>, <phplint> - 'php -l' ;)
Tidy support: <tidy>

hundered push-ups challenge, day #3

2008-06-28T11:15:00.005+02:00

The week #1 is over. There were three days of training, and I haven't done all that was planned, but as I haven't been doing anything to improve my physical condition (Apart from riding a bike for 1h per day, but even that I stopped since I moved, as I haven't bought a new one.) I am satisfied.

Here is little summary of my journey:

Drawing the line - I sucked... but I am not stopping... :)

Day 3 stats:

planned: 15, 13, 10, 10 and min. 15

pause: 120sec

result: 15, 4min, 13, 3min, 10, 2min, 8, 6min, 7+7