About Piperka

FAQ located here.

What is Piperka?

It works as a plain old web comic list and there's a top list built from user preferences too. But those things are not the site's main focus.

It also helps you keep track of comics. It tells you what comics have updated since your last visit and works as a bookmark for you. Here the help page or a demonstration of the interface might be illuminating.

What kinds of comics does Piperka keep track of?

It works with most comics that are arranged to have a single URL per comic page and have a link on that page pointing to the next page. Anything that uses flash, javascript or redirects for navigation is unlikely to find itself in here. Also, Piperka wants to have a consistent archive. Comics that hide old archives won't be listed in here either.

Frames are a bit problematic too.

If you want to make my job really easy, as an author, use a separate tag like <link rel="next" href="..."> or do something like <a rel="next" href="..."> in the usual tag pointing to the next page. Please don't make "Coming Soon" pages, either, or if you do, don't link to them from the regular archive pages.

Does Piperka do hotlinking?

No.

All links are to the comics' own archive pages, through a redirect that updates the user's data along the way, when appropriate. All images are from Piperka itself.

Is there anything I could do to help Piperka?

If you have anything to advertise, you can use Project Wonderful to place a bid on Piperka. That's a referrer link.

Sending me an email about crawler issues is always welcome. Knowing where to look is often the hardest part of maintaining the database.

If you really want to get involved and have the skills for it, then you can download the source code and see if you could do anything with it.

I have some banners available too.

Banner size 468x60

Banner size 200x40

Banner size 88x31

Links to Piperka are welcome, though I won't claim that getting your readers to register as users and increasing the ranking at Piperka's top list would get you dozens of new readers.

I don't want to ask for direct donations.

I don't want to be listed on Piperka!

Are you sure? Piperka doesn't copy any of the webcomics' contents (save for the URLs of their archives). I would expect that having your comic listed on Piperka would only bring you more readers.

That said, just send me a message if you don't want to be listed in here and I'll remove your comic. We don't keep anyone here against their will.

What do you do with users' email addresses?

They're there mostly for sending a new password if you lost your old one. I might also send you email if you submit a comic and I can't include it for some reason. I won't give them to anyone else.

Other than that, some aggregate statistics are generated from users' subscriptions.

What's the deal with public profiles?

You can have a public user profile page (like this) that lists your subscribed comics and whatever you'd like to say about yourself, life or anything.

This feature is optional and disabled by default.

How does the related comic algorithm work?

There's a list of comics related to a comic on its info page. These statistics are gathered from users' subscriptions.

It uses a convolutional algorithm. Basically that means, if user 1 reads comics A and B, raise the score for B on A's results. If user 1 reads A and C and user 2 reads C and B, slightly lower the score for B on A's results. That second stage is in there to weigh down popular comics. Plus some random multipliers and functions thrown in to tune the results. It's all terribly ad hoc.

If you're interested in these kinds of things, check out Jaccard index too.

How often does the site update?

The update job is ran hourly, though only a portion of the comics are checked on each run. It tries to act intelligently and be efficient yet timely. With varying success.

The update job may be thwarted by badly timed net outages and the comic sites may change their archive structure. The scripts are not infallible. Normally an update to a comic should show within a day at Piperka, but if they seem to be absent for longer than that let us know and we'll investigate. I've written a blog post that covers more detail about what the crawler does.

What about RSS?

Piperka does not offer an RSS feed. If you want that you may find other sites that offer feeds for those comics that do not have them themselves. But there are problems with using RSS for this kind of content.

RSS is more suited for news, where there's little interest to read the whole history of the feed or every item in detail. Comics are more often about telling a story instead of delivering news. When you start reading a new comic you start from the first page instead of just starting to watch for updates. How should a feed handle that situation? Offering a feed with potentially thousands of items on it gets very cumbersome.

Also, it's unclear what the feed content itself should be. Some comics incorporate text to their pages or have multiple images per page. Getting the right combination of IMG tags and such to reproduce the content for the feed may be tricky. Not to mention that most comics rely on ad revenue and frown upon showing the content outside their own pages. They may block downloading the images if the referrer field isn't correct, too.

The RSS feed may have a link to the archive page instead. This approach has problems too. If you have five updates since the last time you read the comic and follow the link in the oldest RSS feed item and read the five pages, using the "next" links on the archive itself, you'll still be left with four RSS items pointing to content that you've already seen. Alternatively you could ignore the next links and just use the links in RSS feed instead, but that can be cumbersome too.

What does Piperka run on?

It uses Debian GNU/Linux on a Hetzner EQ 4 dedicated server. The pages are served by Apache, using mod_perl and Mason. The database is PostgreSQL. Don't worry if none of this said anything to you.

There's a statistics page with some random statistics if you're curious.

If you're even more curious, just grab the source from the repository.

Piperka? Paprika? Pepper?

When I started working on this site, I just called it "subscribe". I had a quick look at the respective domain name, subscribe.com and saw that there was no such server yet. Later on I had a closer look and found out that the name was actually reserved, by yet another domain hoarder, just not one that set up a DOMAINFORSALEOMGLOLZ web page for it. Hence I was sent to look for names elsewhere. And I looked. And I looked.

Finally, I was looking for various terms around paprikas and finally found one of them free. "Piperka" is paprika in Bulgarian.

I had a friend take an image of those paprikas long before I thought of a name for the site. The yellow one was cut into slices and eaten with bread and cheese, the rest went into a pizza.

Piperka.net copyright Kari Pahula <kaol@piperka.net> 2005-2013. Descriptions are user submitted and Piperka claims no copyright over them. Banners copyright their respective authors. Fnord.