As side effects:

• Fotowall has a website!
• I’ve got my shiny domain www.enricoros.com
• My blog is now on ./blog
• I’ve learned PHP5 so I can love Qt4 even more!

See you there ♦

To celebrate this result, here is a short teaser for the upcoming version (that will be out in one week):

A big thank you to the team: Arnaud Tanguy, Alessandro Portale, Andreas Brech, Georges Dubus, and translators Martin Zimmermann and Marcio Moraes.

Once upon a time, a little application called quax was landed on kde-apps. As imagination is better served by a picture, here is what quax looks like:

Quax after being ported to Qt 4.5.

It’s nothing but a simple and easy to use screen magnifier. You can move it around and place it near the detail to magnify. Plus it lets you pick the color of a pixel. It’s the perfect tool when you’re tuning the style, pixels and gradients of your beloved app It had only one drawback: it was build with Qt 3.

Porting Quax to Qt4

It was the first time I had to port an application from Qt3 to Qt4. Here are some stats:

• The porting took 3 hours (you can see it from the git history, dinner included )
• I did the following:
  • create the project files
  • use qt3to4 with that files
  • make it compile disabling some code chunks
  • re-enable the disabled code in a qt4 way
  • disabled qt3 support from the .pro file and fixed all until it compiled
  • update the packaging informations, readme’s
  • published Quax 4 in kde-apps
• The source code dropped from 670 to 547 lines of code. So Qt4 saved 14% of code compared to Qt3 for this project.
• The cleanness went drastically up, the API of Qt4 is a big improvement over that of Qt3.
• I can’t talk about speed or graphics improvements.. but if you look at the old screenshots you’ll see that the font rendering was really improved in Qt4

Conclusion

Finally I have quax back, and you can have it too! The source code is on gitorious, and here is the kde-apps page. A final thank word to Claudiu Costin (the original author) and Vlad Oancea (the skin designer). I won’t maintain quax4 and I don’t plan to add any features, I’m just glad I can keep my box Qt3 free while designing pixel-perfect graphics thanks to this cute little utility

New features in Fotowall 0.7. In the top left area you can see the web image selector.

With this release Fotowall integrates with the deep web. You can now search the network with the embedded flickr search pane and add the images to the canvas. This opens up an infinite can of content, all available for your pleasure .

Other relevant features include:

• support for Qt 4.6 (brings animations and transparencies)
• usability fixes (rubberband selection, deletion, etc..)
• cleaner design (new icons, loadable frames, better property dialogs)
• italian, french and german translation
• check for updates and web links

Don’t hesitate to ask for missing features and always remember that you can rotate pictures by dragging corners with the right mouse button

Enjoy your Fotowall!!

The official homepage remains on my gitorious web-automation project. All the other old clones, like the ones on github are outdated and will (hopefully) disappear in time, as the development continues on gitorious.

Have fun

For those who don’t know me, I’m a long time KDE user (since 2.2), I worked on several KDE projects like okular, kdm, amarok, konqueror, as well as releasing other software like Fotowall and other stuff on various repositories.

This planet is the collector of many of my favorite blogs and it’s an honor for me to be here.

See you soon (with some developer tool hacks…stay tuned) !

Se avete un blog, già conoscete la soddisfazione che si prova analizzando le statistiche.. Si può determinare la risposta all’impulso, vedere da dove viene e dove va la gente, capire quali chiavi di ricerca siano state usate per giungere al tuo blog.

L’algoritmo usato da google non è noto ma la perversione di certe persone è ben evidente, infatti è arrivato su questo blog ANCHE chi ha googlato:

1. pellestrina gay – (miracoli di google…)
2. fattispazio – (levati da li!)
3. produttori di croissant – (cerca “pasticcerie”)
4. design soggiorno – (io so disegnare solo caprette e conigli, vuoi?)
5. meccanismi a catena – (cerca “Wanna Marchi”)
6. appartamento da figata – (sono vietati dalla legge Merlin..)
7. conservanti alimentari per dolci – (conservanti “alimentari” ??)
8. migrazione farfalle 2009 conegliano – (speriamo non ci devastino il raccolto!)
9. la gioia di tutti – (vedi il 10.)
10. luana la zoccola – ()

Il protagonista, del quale non si seppe più nulla (tranne qualche speculazione) è divenuto famoso con il nome di “The Unknown Rebel” (da noi tradotto ne “Il Rivoltoso Sconosciuto“) oppure “Tank Man”.

Una delle immagini più forti del nostro tempo.

Su Google immagini sono presenti molte foto, anche in alta risoluzione (nel caso vogliate anche voi stamparne una formato poster, ed appenderla al muro).

Il Foscolo sostiene che il ricordo dei grandi, ispira grandi imprese. E voi cosa vedete in questa immagine?

Please enjoy this HQ stream of deceptive video pixels while I explain this fine new piece of trickery

As you can see, this other frustration-driven program lets you climb the ladder of a typical facebook game while drinking coffee or feeding your pet!

So, before starting you need:

1. a dictionary file for your language of choice (1 word per line)
2. a dictionary file with the names of your facebook friends (for the ‘bonus’ levels)
3. a compiler, because I’m not making binary packages this time
4. the source code of AutoGram, published here and available as a tar.gz here.

Build it with your favourite compiler (tested on Linux with gcc and on Windows with msvc) or your favourite IDE (you can download QtCreator if you don’t know what an IDE is and click on the autogram.pro project file). Then Run it and embarass your friends

Here is how the program works:

1. grabs the screen image and cuts off the 6 images representing the letters.  [ScreenCapture.h]
2. recognizes each character by comparing the image to the alphabet and picks the match that has less differential energy. (is like xor-ing 2 pictures and computing the energy of that)  [Ocr.h]
3. performs all the permutations of the characters, to find out all the words that can be made out of the given alphabet. [Scrambler.h]
4. validates with a dictionary to filter out unknown words. the dictionary is the “main language dictionary” if the screen is yellow-ish (indicating normal game) or the “bonus game dictionary” if the screen is red-ish (indicating a bonus game, where all the words are picked from the names of your friends).  [Scrambler.h]
5. injects the words (as keystrokes + return) to the browser window.  [InputUtils.h]
6. there is a timer-based loop that performs the operations 1-5 while updating the GUI.  [WCGame.h contains the logic, AppWidget.h handles the GUI]

If you run AutoGram on a virtual machine, please remember to switch it off after a while! I take no responsability of any score overflow

Again here is the source code, stay tuned for more web automation articles, and enjoy

update: XP/Vista download here

You are happily wasting time on facebook (that site with all the ignore buttons, with noisy stuff from people who call themselves ‘friends’ of you.. you know..) when you unfortunately get hooked up with Biotronic. This flash game is quite nice, but there is a rank ladder on bottom and you’re last. So you play and play … still last, play even more.. still last.. can you see the frustration?! The layman’s answer is to play (aka waste hours) to get better and climb the ladder..

And then there’s the programmer’s answer: I’ll write a Qt program to get me on top!! (yes, and still waste hours ) — Here is the result: BioCheat

The game is in FLASH, so no HTML or DOM tricks, you have to emulate eyes, brains and hands (only 1). Or in other words you have to see, think and act. Or in other words you have to recognize the image, solve the puzzle and send the clicks back.

Not only I became the first among my friends, I was the world #1 for a month (with 5Million points) Then they wisely reset my score to zero (was too high for a human player) and many people started using and modifying BiCheat!

Here is the simple work flow of the program:

Plus when there is near 60% of recognized blocks, a condition telling that the level is finished, we press the “next” button (a mouse click at an absolute position) to go to the next level. Then you can go grooming your cat while BioCheat plays for you.

In the end it’s a machine vs machine fight, not a human-to-machine one

Again here is the source code, stay tuned for more web automation articles, and enjoy