Marsden is Here

If you hadn't heard already, LJ is giving the Scrapbook a long needed makeover. (I don't think it's changed significantly in the 10ish years I've been using it). Unfortunately in the process they are doing away with (reverse order of importance to me)
- the ability to have an image in multiple galleries
- I'm not clear on sub-galleries but the screenshots aren't promising
- TAGS, which are not being preserved in any form

In my case this last means hours of manually entered naming information simply evaporating. (The template I was using already seems to have broken somewhat - although that could have happened ages ago. I don't go in there very much myself any more.)

So what can you do to save your information in some form, no matter how obtuse?

a) Go to the Scrapbook export page which will give you an xml file containing your gallery structure and thus what images are in each gallery including tag galleries*. It also gives you caption and description data for each image and the current url for each image - which will be redirected to the new location of the image after the change.
* LJ creates a gallery for each tag and then associates images to that gallery rather than associating tags to images. Back-asswards but I can sort of see the logic so maybe it's just half-pie.

Do this now. Once you have that information it may later be possible to parse the xml and download the images to your computer while associating the tags with the images in a human-readable way. Which is to say I'm about a third of the way through a script to do exactly that.

Update: since the changeover you can no longer get the XML file, but if you have it I finally finished the script.

b) you can use the Semagic client for LJ to sync your Scrapbook galleries with a directory on your computer. I did this yesterday. Semagic creates a copy of the file in every gallery it is associated with - so my 2Gb of files online grew to 3.75 GB locally (this took some time to download). Do this by going

Journal (menu) > Pictures > Open LiveJournal Scrapbook , _Export_ (link)

Semagic does not appear to retain caption/description information nor is there any way of figuring out which set an image was in originally except by searching for all instances of the filename in the archive. Problem - this is the original filename as uploaded - which doesn't have to be unique. I don't know what semagic does if two files with the same filename end up in a tag gallery.

- I know I have multiple files with the same filename from times when my old camera reset itself or clocked. Whether any are uploaded / same-tagged I have no idea.

My own script is probably also going to download an instance of the file into every gallery it is associated with just in case, and will handle identical filenames somehow.

You can also use Semagic to backup your LJ entries locally - as an alternative to LJArchive. The advantage over LJ Archive is that it will also make local copies of images linked from your entries, which prevents them being lost into the ocean of dead links.


c) Chebe has some instructions for using wget to get your scrapbook images but I couldn't make them work - I would have thought just mirroring the galleries with wget would have given you a locally viewable copy with all the tags etc as they are displayed in Scrapbook galleries. That's what I'm aiming to do with my script.

Update: a new version of Chebe's instructions can be found here. I haven't tested them myself.
http://andlauer.de/biologie/apps/livejournal.html

Unfortunately writing the captions and tags into the file metadata with a script is still beyond my abilities :(

So I have a webform which posts to a processing script. Intemittently and apparently based in part on whether diagnostic comments in the script are commented out or not (based on F5/Yes resubmit data after uploading the edit), but becoming more frequent the POST data seems to get lost as far as the PHP is concerned (empty print_r of the array, yadda yadda) yet the mysql which uses the data is updating fine with the correct info. - even though the query is being written by the php and outputting the query text shows empty fields.

The sql working shows up another problem - even though the error condition is being triggered the only thing an exit; does is stop the output to the screen - the rest of the script is merrily running through - sql statements, everything except exits and breaks! (The switch statement works until it reaches the appropriate condition - and then the default runs as well! )

Replicated IE/Firefox.

In desperation, after a couple of hours of this I tried the code (I was only making some minor tweaks) on the 'production' (client's) server and lo, it works fine. Which leads me to conclude that my webhost is doing something bass-ackwards with their php config.

When I have a little calm back I'll set up a copy of the DB and try it on my local WAMP - Right now I'm just pissed off.