Writting an entry is easy, it basically follows the rule of what you enter is what you get(WYEIWYG). If all your contents are plain texts without any style, then just enter them like you are using a text editor.

To customize your displays like font color, background or insert images, sounds(wma, mp3) or movies(wmv), you need to use some self-defined TAGs with my weblog, check the guidence for details.

Below are some features with my webblog:

1) multiple ways to Readmore

In the main page, all blog entries are trimmed to contain only a given number of words(roughly), if the content of the entry is longer than this size, a '...More' link will be used to cut the content. There are four methods to read the full information:
    * Click the subject bar
    * Click the arrow at the right-top cornor of an entry box
    * Click the ...More link.
    * Click the Read Article link at the right-bottom side.
If the node contains comments, you must used the last method to open the entry and read the comments, check the figure below.

2) Preview before submission

For users using a lot of customized TAGs, it's always recommended to preview before submitting. My weblog provide a full--featured preview system that allows you to check the result on the preview-window and refresh and unload it later on.

3) Search engine

Currently, my search engine supports searching content, subject, author fields of the blog entries. By default, the search is conducted on the blog contents with case-seneitive keyword. Press here to refine your search standards.

The searching results are highlighted and some more information is shown on the top of the resulting window, see the following sample:

4) Login system

to be added

5) User Interfaces

to be added

You must register to write your own entry node. You can write comments anyway without registration, but only registered users can modify or delete an entry(of their own for sure:-)).

...More

Perl 5.10 new features

Read from a Perlmonks thread and some suggested links. my first impressions about the coming Perl 5.10:

1) Some new features:
  * new keyword say, an addon to print
  * new operator //, fixed definedness problem with || operator
  * Stackable File test, -f -w -z $file with one shot
  * new State varable with state keyword, works like static in C, previously implemented with closure and is not necessary now.
  * new lexically-scoped feature %H
  * Smart matching with ~~ (looks cool)
  * given + when.. finally, Perl has its own switch...case
  * Inside-out Object (no experience)
  * field hash: use reference as hash key and more..

2) Regex:
  * named captures (?<name>pat) (?'name'pat) and \k<name> \k'name'
  * new hash %+ (nice)
  * possessive quantifiers /(?>x+)/ to /x++/ ..
  * new back-reference syntax \1 => \g{1} backreference \g{-1}
  * dynamic pattern to recursive pattern (?1), (?R), (?0) <-- PHP feature
  * (?&name), (?(DEFINE)...) (nice)
  * relative referencing (?-1) (?+1) (nice)
  * backtracking control verbs (nice)
  * /p modifier with using $`, $&, $'
  * internal:
      * recursive engine to iterative engine
      * trie optimisation on regex alternations
      * [..] treated as literals
      * and more....

For Regex engine, besides adding some features that already existed in the progeammin world(i.e. named captures, recursive pattern), more fine controls like backtracking control verbs are quite attractive..

A short-phrase: no drama, more convenience or tricks for the programmers.

References:
[1] http://www.slideshare.net/rjbs/perl-510-for-people-who-arent-totally-insane
[2] http://www.slideshare.net/acme/whats-new-in-perl-510
[3] http://www.regex-engineer.org/slides/img0.html ...More

Some Progresses

From the middle of September till now (Fall 2007), I've written 195 various files(currently in use) which contain totally 12K lines of code. The above figure shows the lines of code corresponding to different types of the files and illustrate how different languages/tools contribute to my current work.

The following table and figure (the number of files vs the type of files) show more information about the overall code distributions.


From the above statistics, about 80% of the contents are HTML::Mason (file ending with .html, .mas, .pm, handler), 20% are DHTML contains CSS, JavaScript, and AJAX code.

Among 195 files, 184 files (94%)are under the control of HTML::Mason engine. note: three special files are under Mason control, they are:
    1) style.css: contain dynamic styles, i.e. prefered screen width, color scheme and any state information the users configured. (11037 vs 6537 bytes) [code size vs. real transfered size through HTTP]
    2) mctrl.js: functions and variables containing dynamic controlled styles. (3127 vs 2681 bytes)
    3) query.mtxt: send query to display different manuals, can be switched to html anyway. ...More

this blog is kind of like my working diary which means I wrote down stuff right in my mind at a specific time. Good or bad, it records a process of the growth.

For me, it's not bad to learn from my previous mistakes as long as I think it over and do better next time.:-))

--XC@Cornwell