Hello everyone! As some of you know, I've developed a photo manager application for Ubuntu phones, called Imaginario (which in interlingua means something like "collection of images").

Imaginario on the Ubuntu phone

Last February Imaginario has reached feature completion, meaning that I'm quite satisfied with the features it offers (but if you find it lacking, then by all means please let me know by filing a bug report). I recently bought an Ubuntu tablet, the BQ M10, and verified that Imaginario works fine in the tablet form factor, even though it could be improved to use the adaptive page layout widget in a few places.

Anyway, the relative quietness the project had reached was starting to scare me, therefore since March I've been working on a desktop version of Imaginario, which shares the same code base but renders its interface via the QtQuick Controls module, which provides native-looking widgets on Linux, OS X and Windows. The goal is indeed to create a cross-platform photo manager, heavily inspired by the fantastic F-Spot which has been hibernating for the last 6 years but which I'm still happily using.

Imaginario is under development for your desktop!

I'm in no way done. I've got the application in a stage where I can start playing with it and some basic functionality is working, but I'm not yet ready to entrust it with my photo collection ­— though indeed, so far I haven't experienced any data loss or data damage.
That's why I think this is a good time to share the news, in the hope that I could find someone else (besides me) who could test the software and report bugs.

What should work:

• Import your photo database from F-Spot
• Import your photo database from Shotwell
• Import photos from a folder
• Edit photo tags
• Search by tag (with OR/AND operators exactly like F-Spot does)
• Search by geolocation

What does not (yet) work

How to install and test it

sudo add-apt-repository ppa:mardy/imaginario
sudo apt-get install imaginario

If you are interested in building it for other OSes, you can find the source code here, but please don't distribute any binary packages yet, as the software is all but ready to be used in the wild.
As for the other point, it's likely that split views, treeviews and menus are coming to the Ubuntu UI toolkit too, but of course I'm developing Imaginario now, and I have to use what is available at this very moment.
All in all, none of the above is a strong reason to chose one toolkit over the other; and it's actually not unlikely at all that in the future I'll decide to go for the Ubuntu UI toolkit on all OSes and all form factors; I've just made a contingent choice, which suits me fine for now but which I'm open to revisit anytime.

If anything, this double face of Imaginario shows the strength of Qt and QML: the same core C++ classes get exposed as two totally different user interfaces, and the speed of prototyping in QML makes this effort very cheap — I'm not afraid of literally throwing away one user interface or starting a new one from scratch.

Solo un breve aggiornamento: siccome qualche tempo fa ho scritto sul blog che non avrei rinnovato il mio abbonamento al Fatto Quotidiano, mi sembra giusto scrivere anche che, dopo aver notato i recenti miglioramenti al sito, ho deciso di supportarli ancora, acquistando un abbonamento come sostenitore.
Adesso il sito è molto più leggero e leggibile, la pubblicità nel sito è quasi assente (benché la situazione non sia ancora perfetta perché, essendo un utente sostenitore, di pubblicità non dovrei vederne affatto), ed ho quindi voluto supportare quello che ritengo essere uno dei migliori siti di informazione sulla politica italiana.

Update: I was informed by some reader that there exist a 1080p version of the Zuhres video, so I repeated the analysis below to that version of the video. Here's the result: https://www.youtube.com/watch?v=dr9cfzo-pBE
Indeed, the artifacts are much smaller in this version. However, they are still there, and again they apply only to the cabin of the truck, and not to the rest of the convoy.

Last night I happened to stumble again on the video of the BUK being towed by a track in Zuhres, and I noticed some video artifacts as the truck moves behind the building: on the white wall, you can see parts of the truck. I loaded the video file in Blender, and watched the magnified video in slow motion, and indeed those artifacts were there. So, I've prepared a demonstrative video which shows the effect more in detail, and uploaded it to Youtube:

I'm not an expert in digital video analysis, but I don't recall having seen any similar artifacts in other videos. You can see some artifacts on the image right before the truck disappears behind the building (at 00:06-00:07 of my video), and these might be the result of the video compression algorithm; however, at 00:09-00:10 you can actually see parts of the truck cabin being drawn on the building's wall. In my experience, digital video compression sometimes causes residual parts of a previous video frame to be shown over the current frame, when the movie is heavily compressed. However, in this case what we see is a completely new image: none of the previous video frames show the truck's cabin in that position. Also, the video is moderately compressed and of pretty good quality overall; and last, given that the building's wall is uniformly grey and square-shaped, it would seem weird that the compressor has decided to put some image elements in it, and made the image more complex: usually, compression algorithms simplify the video frames.
But, as I said, I'm not an expert on this matters. If you are, or know someone who is, please ask him to have a look at this video, and share his opinion, along with his CV, blog or linked-in profile (just to be sure that we are actually talking with some authoritative expert, and not just another wannabe expert like me :-) ). Also, examples of similar artifacts in other raw videos are very welcome.

Le referendum resultava in un clar victoria del partito del "NO" (circa 61% contra le 39% del "SI"), e le percentual del votantes esseva 32,2%, superior al 30% qui esseva requirite pro conferer validitate al voto.