After finishing season one of Leverage on Netflix, we’ve started watching season two on TNT’s website. Netflix’s streaming video has been great, and TNT’s has been decent enough aside from dropping out of full-screen for commercials…until yesterday.

Last night, while watching “The Order 23 Job” on our MacBook, we got to the final commercial break — and TNT popped up an error saying that the content required Windows to play. The episode played fine. Previous commercials played fine. But this one? The DRM wasn’t compatible with the player on the Mac.

Yeah. The DRM for the commercial wasn’t compatible.

It wouldn’t have been so bad if TNT approached it the way Hulu does when a commercial fails to play, which is to blank the screen for the duration of the ad (typically 30 seconds) and admonish you for not watching the commercials. Unfortunately, the episode didn’t pick up again.

As near as I can tell, the player was set up to continue the episode when the ad finished, and didn’t account for the possibility that the ad might not play. To make matters worse, the scene selection thumbnails don’t work right in Safari, so we couldn’t jump straight to the final act.

Because neither of us wanted to spend a lot of time troubleshooting, we just went into another room and brought up the Windows box to finish the episode. I suspect the scene selection would have worked in Firefox on the Mac, but haven’t tested it yet. I did go back later to see where I could report the problem to TNT, but the wording in their FAQ suggests to me that they’ll just ignore any reports of Mac problems.

I don’t mind watching reasonable ads to get a free service, but if the ad breaks, it shouldn’t take the actual service down with it. You don’t kick people out of a movie theater because the previews didn’t play, and you don’t send them home part way through an event because one of the sponsors’ banners fell down.

Vienna Teng: Inland Territory.

  • Hard disks should not sound like buzz saws.
  • Slashdot article “FOSS Sexism Claims Met With Ire & Denial”…gets met with ire & denial. *headdesk*
  • Listening to lightsaber sounds from across the office. I think my coworker w/ the new Android phone found an app for that.
  • Vertical Horizon’s Burning the Days is growing on me, but I think Vienna Teng‘s Inland Territory is my favorite new album this year

Spammers have been using misspellings, synonyms and malapropisms for years now. Lately I’ve been seeing a lot of Viagra/Cialis/etc. spam using the word “pilule” instead of “pill.” At first they’d just find misspellings for the drug name, but I guess some filters are blocking or scoring on “pill,” so they’ve substituted words for that…including the hilariously ironic “soft” as an abbreviation for “soft tabs.” (Comments on this post are going to give Akismet a workout, aren’t they?)

Anyway, I found it odd that so many different spams would use the same obfuscation, particularly since it looked like it was just adding letters. So I looked it up.

It turns out that pilule is a real word. According to Merriam-Webster, it entered the English language from French around 1543. Sadly, it doesn’t refer to a cute magical creature, but to a small pill — which means that (wonder of wonders) the spammers are actually using it correctly!

One question remained: was it simply an obscure word, or an archaic one? I did a search on Google Books and came up with mostly medical texts dating from the 19th century. Just about every match in the first 15 pages was either:

  • An English-language medical text published between 1830 and 1930.
  • French.

The few cases where I thought I’d found a more recent reference turned out to be reprints of older material.

So it looks like the word died out (in English, anyway) during the 20th century until spammers exhumed its corpse and pressed it into service.

Side Note: Twitterspam

On Friday, I posted the discovery to Twitter on @lol_spam, then retweeted it on KelsonV. Within 15 minutes, lol_spam picked up 45 new followers and KelsonV picked up 40. They were all obviously bots:

  • From the time that the second post was made, each of them followed both accounts, making it obvious they were automatically following based on a keyword search.
  • They all used the same scheme for the user name (first name + first 2 or 3 letters of last name + short number).
  • Many of them shared name components, as if a random generator were taking a list of first names and a list of last names and mixing them together.
  • None of them had posted a single tweet. I suspect that if I’d been foolish enough to follow any of them back, they would have started spamming me with links via direct message. (I caught a subtle one last week: someone had posted a series of inane tweets for the first couple of weeks, then switched to all tooth-whitening links.)
  • Several profile photos appeared on more than one account.
  • Many of them were following upwards of 1,000 users. (After the first few, I stopped looking at the numbers.)
  • All of them claimed to be women. (A majority? That I could believe. But every single one of them?)

I will give them credit for using ordinary-looking snapshots of women with a wide variety of appearances, rather than going for the lingerie, downblouse, outright nude (the spam filters are going to be busy, aren’t they?) and other sexy (or “sexy”) poses that usually show up on these. They actually looked like photos real people might use on their profiles.

Nice try, spambots.

»All pages site-wide with this tag