New layout under testing…

I was mentioning this a few days ago: the actual layout definitively need to be updated or changed.

Let’s change, but let’s test it first.

As this area is a workshop, it’s hardly ever finished. My goal today is to prepare something I can tweak and enhance later on. That is, this is mainly a wire‑frame, even though I’d love to keep it “like this” somehow.

I played a bit with HTML5 and since I’m kind of going crazy with liquid layouts these days, it should be readable on multiple resolutions. You should then be able to zoom in or out from whatever screen resolution and still have no scrollbars. You can try to use it on a 200 pixels wide device, but you might be willing to switch off positioning — though it might work.

Forms are definitively part of the game since all the header is a big ol’ form. Good thing with having a navigation based on forms is that you finally can navigate the menus with Opera without having to use crazy shortcuts: Tab is enough, and I love this one.

So for example, if you want to reach articles about Paris‑Web, you can either click on the months and/or days when those events happened, or just click on “Paris‑Web” tag.

I need to make the colors or behaviors reflect the additive or substractive behaviors of the menu for sure. Any tip on this is welcomed. Other problem with these rules is that you might end on a dead end, which is not so nice.

Planned are nicer URL — that won’t be difficult to make it at least better, ahem —, better graphics, and possibly a resurrection of the myah the day canvas officially starts to support 3D via WebGL.

And there’s a bit more… but there’s things that should never be explained.

Any comments are welcomed and encouraged, of course! But most of all, play around and enjoy it, please!

restore lost icon after installing Thunderbird 3.1 from Ubuntuzilla package

I installed Thunderbird 3.1 from Ubuntuzilla packages a few minutes ago and lost the icon in Docky. I wasn’t so happy with that, so I checked around to find a solution.

The reason appears to be that the system is looking for /usr/share/applications/thunderbird.desktop whereas Ubuntuzilla creates a file called /usr/share/applications/Mozilla Build of Thunderbird.desktop; this is not right.

Just rename the culprit file to thunderbird.desktop and you’ll be a happy bunny as me.

That might not be the best solution ever, but it seems to work pretty well so far.

Just a tip…

If you find yourself in a dead end with some code you don’t seem to be able to solve: get off your screen, sit on your toilets, and think about it with no help whatsoever (paper, pen…).

You might even end traveling in time.

Source of the logo is http://bttf.wikia.com/wiki/File:Flux-Schematic.jpg and what I’m doing with it might harm kittens…

:first-child pseudo‑class and the <caption> element — :first-child et <caption> sont dans un bateau

case solved, thanks for the answers!

So… I was answering a rant from one of our readers, and I came across something I can’t find any answer to, but “WTF?” (Excuse my French).

When using the following codes, the <caption> element looks like being ignored by browsers as the first child of <table>; I made a test case in the codeplay section:

CSS

table tr:first-child {
  color: #f00;
}

HTML

<table>
  <caption lang="it">è un'ingiustizia però</caption>
  <tr>
    <th>code</th>
    <th>comment</th>
  </tr>
</table>

Since I can’t find any reason why this would be the case, I’ll ask the Web: Dear lazy Web, do you have any answer to this question? Not “WTF?”, but “Why is <caption> ignored as the first child of <table>?

The DOM in Firefox says it’s children[0] of <table>, which is fine with me! But why would the rendering behave an other way?

Happy birthday!

This website is now nine years old today. Well well well…

Still have around 1400 articles to put back online, once they’re cleaned up from the previous home made CMS tag’s and typos. Oh, and something new to come…

Dear blog, I hope you’ll make it to your 10th!

The photo comes from Flickr Commons, with unknown license (so far).