S M F G · A Counter-Clockwise Trip Around The World

About this blog

Jan 4th 2010
No Comments

This page is about the work-flow and the technology involved in maintaining this web site up and running. I hope there will be some discussion about better solutions or tips-and-tricks, but in the worst case it will just be my on-line version of the scrap book I keep for this purpose.

If you are not into computers, please stop reading now :-)

Blogging software

When I first started to think about writing a blog about my trip, I already had a Google account, so the natural thing seemed to go Blogger. I was writing on blogger, linking the photo galleries from Picasa and the travel map from Google Maps. But I had a couple of issues with this approach: the scarce, if any, customisability of the software, the need of updating using three different interfaces and, more important, the fact that my texts, photos and info were completely out of my control.

As some of you may remember, in October I switched to WordPress as blogging software. It is free software, and it has an embedded photo gallery.

I still have the problem of updating the map on Google, but this is of marginal importance to me at the moment. By the way, I am planning to get rid of the Google map and post a clickable map of the world — probably split in continents — that will link the countries to the photos and articles on the site. Unfortunately, this requires a lot of work on Gimp and a decent resolution world map — as a side note I’d love to use SVG, but I am not aware of the extent of SVG graphics support in decent web browsers.

This web site is mostly based on two connected pieces of software: WordPress with the Unstandard theme, and the NextGEN Gallery plugin for an easy managing of picture galleries. I have slightly modified both, especially the CSS, but also other bits — and changing how Nextgen handles the singlepic photos (i.e. the pictures in the posts as opposed to the galleries) required me to learn a bit of PHP, which is good. If you are interested in those files please drop me a line :-)

The WP-Highslide plugin takes care of the display of the photos, and Hana FLV Player of the videos.

Web Hosting

As part of my move to a free blogging, I bought a domain, web space and hosting through DreamHost. This way I have — almost — full control on the software and the material. This also means that if something goes wrong, I can still re-publish my work using some other provider or even at home — although this will be of little use. And yes, I have daily backups of the database emailed to me :-)

For the hosting I am paying around seven dollars per month, and I have unlimited space and bandwidth, so I think this is a decent deal. The server is a Linux box running Debian, and I have full shell access. This means that I can modify every little piece of the software involved, as if it was running from my home computer — this is not completely true, because I can’t install system-wide software, for instance changing the underlying database server, but you get the gist.

The Work-flow

This is the most interesting piece.


I saved a random article on my little baby and erased the content of the post. Then I made it point to the local version of the CSS file. When I want to write a new post, I open this file and I put the new text in the empty space in the middle, so I can preview it with a browser without the need of being connected to the Internet. For writing HTML/CSS I use vi(m) or Bluefish.


Every week or so, I download the photos from my camera to my little baby and I rename all the photos following the Exiflow schema. This is something I really have to do, because I am collecting pictures from people I meet during my travels, and very often I end up having same-name photos, typically coming from same-brand cameras. The problems arise when you pick photos from different folders and you put them together in a gallery. With the Exiflow schema the file name is composed by the date, a camera identifier followed by the original file number, then the author and the revision number, making it very hard to have same-name files and making it easy to sort and archive the files without having to open them.

After the renaming I erase the unwanted pictures — roughly the 50% of the total, but I plan to reach a 75% of deleted photos — and I start tagging the location, subject and keywords. I usually do the deleting/tagging in Geeqie due to the great speed of this software, and to the fact that it lets me do the tagging with «marks» in the photos ((I can associate for instance the mark «1» with the tag «Food», «2» with «Landscape», «3» with «India» and «4» with «Travellers», then I go through the photos pressing the corresponding key and having the tag being added automagically.)).

As a photo manager I use F-spot, that is very slow and has some incredibly annoying bug ((It kind of mangles the Exif tag of the time the photo was taken, depending on which time zone you are when you download it. Check here. A note on the note: as of 2010-05-14 it looks like the bug as been fixed, but I still have the old version. Let’s see what happens …)) but it has two great features: support for photo versioning and for the Exiflow work-flow ((I don’t like mono applications, so I am still considering the switch to Digikam, although the massive amount of kde libraries just scares me. If they add versioning support — and the Hidden tag feature I will probably do.)).

Every time I have a stable Internet connection, I launch the synchronisation between the photo archive in my little baby and the one in my web space, using rsync ((In case you are wondering the options I am using, here you have the full command: rsync --archive --human-readable --progress --partial --partial-dir=partial --copy-unsafe-links --copy-dirlinks --keep-dirlinks --delete-after --exclude=tn/ --exclude=.html /home/pt/foto/???? piergi@traversin.org:foto/.))

To be continued


To be continued

Leave a Reply