GPF News Archive

Hey, there, folks. I hope you're enjoying Scylla and Charybdis so far. It's going a lot slower that I would have liked, but that's to be expected when I'm only updating three times per week. There's a good chance that "Year Fifteen" might end up taking longer than an actual year, but I hinted about that fact in my previous News article. I'll try to keep you posted on this front as the story progresses.

That, however, is not the intended focus of this post.

I just wanted to drop a News item in your lap to let you know of a minor change to Archive. For a while now, we've been using Oh No Robot as both a transcription and search engine for the comics in the Archive. This fabulous service (which was gifted to the webcomics community—for free—years ago by Ryan North) allows volunteers to transcribe comic images (which are notoriously not searchable) into simple text (which is notoriously easy to search). As such, webcomic readers could contribute to the comics that they love by volunteering a little bit of time here and there to transcribing a few comics, then reap the benefits of using the search engine to find that one comic where that thing happened so they could share it with all their friends. We've been very happy with Oh No Robot for years now, and we genuinely believe that it's been invaluable to us and to many, many other webcomics since its introduction in 2006.

That said, there have always been a few things that have bothered me about Oh No Robot. My concerns about it stem mostly from its implementation rather than its core functionality. I've never been a big fan of JavaScript, and I'm also reluctant to install third-party scripts on my site for both security and performance reasons. While we've never had a problem with Oh No Robot in the seven or so years we've used it, I have noticed a time or two where loading their scripts has slowed down the GPF site. I have enough frustrations dealing with bloated advertising scripts as it is (which aren't a problem for our ad-free Premium subscribers, by the way), so eliminating as many third-party scripts as I can would make me one happy cartoonist.

So with a mixture of reluctance and enthusiasm, I set out to create our own in-house transcription system, identical in spirit and similar in approach to Oh No Robot's system. The advantages to the new system are numerous, but the primary incentives are (a) less reliance on JavaScript, (b) less reliance on an outside service that could go away at any time, and (c) a quicker turnaround between when transcriptions are added by volunteers and when they become useful to search engines and other downstream consumers. From a functional standpoint, it should work in an identical fashion as the Oh No Robot system, only it should work for those of you who insist on turning JavaScript off. (There's a little bit of JavaScript used in our validation steps, but those are also backed up by server-side validation.) The pages will look a little different as you'll remain on the GPF site the entire time, but it will still follow the same fundamental process. Just look for the "Transcribe this comic!" button under certain strips in the archive, click on it, and type in your transcription. That's it. Once submitted, it goes into my queue to approve and you can go right back to the place in the Archive where you left of. If you happen to catch an "Improve Transcription" button, you have an opportunity to improve on a transcription someone else previously submitted but which we feel needs a little bit of work. The process is similar.

I can't fully express how important the transcriptions many of you have already provided over the years have been. These transcriptions have a myriad of uses, from providing raw data for own internal search engine, to helping external search systems like Google and Bing index our site, to even helping the visually impaired enjoy the comic via screen reading technology. As of this writing, approximately 87% of the GPF Archive has a transcription of some sort, meaning that the core of our site's content is already available for searching. However, of that 87%, many of those transcriptions could use a bit of tweaking, and the remaining 13% will only continue to grow as we add new comics on a regular basis. It's an ongoing, never ending process, and it's a lot bigger than just me and a handful of our most dedicated readers.

To find out more information about how the new transcription system works, please see the new transcription hub page. If you'd like to help out but aren't in a good place to do a comprehensive archive dive, we've also added a handy list of comics needing transcriptions and transcriptions needing improvement. Detailed instructions on how to actually perform a transcription are available on the transcription page itself, once you've selected a comic to transcribe.

I wish to express a huge thank you to Ryan for creating and hosting Oh No Robot for all these years, and to all of you who have fed your transcriptions into that system since then. We wouldn't have that 87% base to start from if it weren't for you. And if I can find a way to feed the data we collect locally back to Oh No Robot so they can keep their database updated, I'll gladly implement it. I'm pretty sure they've fed a fair number of new readers our way over the years and I'd love to return the favor, even if we're no longer officially using their service.