As of January 5th, I have written 269,558 words in this journal. This equates to 40 straight hours of typing at 110WPM, or 696 pages in Microsoft Word (12-point font, single spaced, Times New Roman). I feel horrible for all those wasted electrons as they are used to support this drivel.

I spent most of today hammering out the backup feature. The problem with the previous version's was that it didn't factor in the longevity of accounts (perhaps I subconsciously thought that Tabulas wouldn't last, who knows). So with accounts like mine (900+ entries), creating one huge backup file was a very big mistake.

The new version partitions your entries into groups of 50 entries each; it creates a separate file for every 50 entries. It then zips up the files and stores the zip file on your personal data storage so you can download it.

Gone are the RSS formats for backups; there is no real set standard for archiving large amounts of posts for transition between journaling services (partially because true support for total exporting is severely lacking on all sites). I have used a simple HTML file and will probably add support for some sort of XML format in the future, but any developer should be able to parse the entry HTML file without a big problem for the time being.

Some things I'd like to do in the future for this feature:

  • Better templated support for each entry; each entry is placed on a page haphazardly and there is no moving between. Perhaps I can add in a HTML file that serves as an index file; this can be a frame or something that links up to all the backup pages.
  • Automated backups. Paid users should be able to set a threshold of time where backups are created and e-mailed to an e-mail account automagically.
  • And as I mentioned before, some sort of XML format (maybe RSS *is* best here) or something. Maybe it'd be more preferable just to use a random XML that allows for storage of all the metadata (book of the moment, etc. This data is actually currently not stored in the raw HTML versions ... yet).
  • I'd really like to do gallery backups as well.

Forr those of you on the new Tabulas CP, you can try out the 'backup' feature to see it at work from the control panel.

. . .

Oh yeah, to the person who suggested rollovers for the CP tabs ... thanks. I finally got around to doing that and they are TOTALLY 100% BETTER!

. . .

You also notice how the add entry page automatically selects the title screen? Coolness. Because of that javascript, I couldn't implement the Javascript Standards-Compliant Image Rollover script ... there was some issue where the onLoad attribute in the body was overriding the script. Boo. I really liked that script, too.

Posted by roy on January 5, 2005 at 01:01 AM in Tabulas | 12 Comments

Related Entries

Want to comment with Tabulas?. Please login.

Comment posted on January 6th, 2005 at 12:53 PM
But if people keep replying to a comment instead of posting a new one, the text is going to be all squeezed; at some point it's going to look like one long column of single words, no ? There's a handy PHP function for word count also.
Comment posted on January 6th, 2005 at 01:36 PM
Yes. But what can you do? It's not something I can do much about (unless I want to keep track of threads).

The MS one is better... I don't really want to use a public server to wordcount my journal ;)

middle_aged (guest)

Comment posted on January 6th, 2005 at 08:58 AM
hey roy, quick question about archiving/backing up the journals. are we able to take the archived entries and re-import them into tabulas? say, another tabulas account. ie. like moving entries between accounts? also, is there a way to import entries into tabulas in general? like say from a RDF file?
Comment posted on January 6th, 2005 at 12:52 PM
As of right now, the answer is no to both questions. I used to export Tabulas accounts into a standardized format, but it just didn't work out too well because I don't think anybody used it.

As for importing, again this answer is no (for now). The LiveJournal import works, but their export format is just random XML . There hasn't been a clear standard established for exporting journal entries, which makes supporting this feature very time consuming; I would have to (in essence), find all the different export formats and support each one.

However, with that said, I do have it in mind to support both import and export in *some* XML format in the near future. Right now though I'm more worried about getting 2.0 out into the wild before worrying about the smaller details :D

Daniel (guest)

Comment posted on January 5th, 2005 at 10:02 AM
hey super roy! happy new year! hey, do you happen to know why bloglines can't get the feed to your journal? it's displaying a "feed error". just wondering if it has to do with all that new fangled coding you're doing to the site. :P laterz.
Comment posted on January 5th, 2005 at 02:24 PM
Hey Daniel,

Yes, it's the new work I'm doing for the site which is killing the RSS feeds (actually more specifically it had to do with the server hackery; I just didn't have the energy to set-up the RSS feeds again from the old CP for two weeks...).

The new feed will be located at <a href=""></a>
which is active right now for me because I'm using the new control panel.

(Yes, this is the *last* time I'll ever change locations of RSS feeds!)
Comment posted on January 5th, 2005 at 01:15 AM
how did you calculate how many words you typed?
Comment posted on January 5th, 2005 at 01:30 AM
originally when i was working on the feature, i had it generate one huge file instead of fragmented files. i just copied that into MS word to get the statistics above.
Comment posted on January 5th, 2005 at 01:27 AM
he counted those suckers, one by one
Comment posted on January 5th, 2005 at 01:30 AM
or i did that.
Comment posted on January 5th, 2005 at 10:02 AM
I knew it, you're no programming genius, you're just another ordinary word counter like me
Comment posted on January 6th, 2005 at 03:05 AM
you are not an ordinary word counter, sir. you are the word counter MASTER. just like you are the google master.