Digital Comic Museum
General Category => Comic Related Discussion => Topic started by: John C on March 29, 2015, 11:01:00 AM
-
Hey, All--
In the spirit of DCM's fifth anniversary, I thought it might be a good idea to finally get moving on one of my "one day, it would be nice..." projects. This particular project will hopefully solve two recurring problems that seem to frustrate people.
- Scanners get frustrated when they release a book with minor defects, like duplicate pages, especially when the people pointing them out are...somewhat less than gracious.
- Readers get frustrated with the parade of re-released scans, not knowing if what they downloaded is the latest-and-greatest version, even though that's definitely a great problem to have.
My diabolical plot is to release a program that, on one hand, will check over comic archive files for typical problems, and on the other hand, will archive that information (and allow users to annotate it) for other users.
If you're familiar with the old CDDB (https://en.wikipedia.org/wiki/CDDB) or the more modern and less proprietary freeDB (http://www.freedb.org/) for music CDs, that's sort of what I'm hoping for. (If it works out, I'm hoping to archive the information somewhere permanent, so that people can use it even if the site itself vanishes.)
Interested?
I haven't started on the infrastructure, yet, but if anyone wants to get a sneak peak at the "validation" side, feel free to take a look at the current version on my own website (http://colagioia.net/scan/ComicScanner.html).
You'll need Java installed (https://www.java.com/en/). That's not the same as Javascript, and needs a browser plug-in on (almost?) all web browsers. Chrome and Chromium won't run them at all, due to some pseudo-political issue.
Since it's code on the web that wants to look at files on your computer, you'll need to answer a bunch of security-type questions to make sure I'm not scamming you.
Just like general security advice, if you don't trust the programmer (me, in this case) with access to your computer, don't let the program run. I know the program is safe, but you don't. I'm not making this point as a joke, but rather to make sure that I don't say anything here that leads to a bad habit with someone else's software. The "applet" will have access to your computer, so be careful.
(Side note: If you decide you don't want to trust me (sob!) and you're familiar with programming and compilers, it's pretty easy to put together. See below for where you can find the source code.)
The questions, when I run the program (Firefox, Ubuntu Linux, the IcedTea-Web plug-in), I get the following questions:
- Activate Iced-Tea-Web.
- Allow colagioia.net to run "IcedTea-Web"? [Allow Now]/[Allow and Remember] (Personally, I never let it remember my choices, even for my own code.)
- The application's digital signature cannot be verified. Do you want to run the application? It will be granted unrestricted access to your computer. [Run]/[Sandbox]/[Cancel] (If you're technically saavy enough to set up a sandbox with the files you want to test, you're probably not reading this, so "Run" is your best bet.)
You can select a comic file (only *.CBZ files at this time), run the analysis, and print a report on issues it might have found. The sorts of things it knows to look for, currently, are:
- Duplicate pages
- Mac-style archives, with those .DS_STORE or __MACOSX folders (untested)
- Non-image files, not that they're bad, but maybe worth a heads-up, since some reader applications choke on them
Help!
If anybody wants to lend a hand to get this running sooner, I'd certainly appreciate it. Things like...
- I have a few other problems I want to check for, but if anybody (particularly our scanners) have things they hate wasting time re-checking and re-uploading, or things your favorite comic reader constantly screws up, I'd love to hear about it.
- If you want to test the thing as it gets closer to useful (*.CBR support, for example), drop me an e-mail and I'll tell you when it's ready to go.
- If you're a Java programmer and want something to fiddle with, the code is on GitHub (https://github.com/jcolag/ComicScanner). It's pretty hack-y code, since I've mostly just been prototyping, which could be taken as proof that other people working on the code probably wouldn't be the worst thing in the world. (The service will probably be Ruby on Rails, but I haven't started writing that, yet.)
- If you maintain a comic reader program, somewhere, I'd like the soon-to-be-archived information be useful. I'd love to see notices like "this comic has been updated by Scanner-XYZ" when I start reading a comic, for example, so if this is appealing on your end, contact me to give me a sense of what you'd need to make that happen.
I think that's everything...
-
Thanks John.
I've got it running here now (after installing Java and the hoops they require).
Didn't know about the cbz only rule until after reading the above. I sent you a couple suggestions in a PM that I'll repeat here -
-When selecting a book - have it show somewhere so we know the selection 'took'.
-When a selection is waiting maybe darken or change the colour of the Analize button so we know it should be hit next.
-for your results is it possible to auto indicate pages with the exact same file size? Thinking those are likely dupes.
When I edit I might have a dupe page, but they were created separately so their file size is not the same. The only way to catch it is the visually go through each page.
In such a case your app here wouldn't catch it.
Please let us know as you change things here and want help with testing.
-Yoc
-
Yes, the interface is definitely "just enough so I know it's working," and not the final plan, by any means...
Especially when it starts getting to identifying what comic it is, a lot of what's on the screen will need to get thrown out, anyway.