I'm going to run into toolserver some simple python + djvuLivre routines to test the possibility to obtain a "wikicaptcha", built to be useful for wikisource activity.
I'm far from sufficiently skilled to write all the project, in particular the final user interface; but I'm not far to implement something like a "voluntary wikicaptcha", t.i. to select controversial OCR interpretation of words of a djvu file, and to present their image, extracted from image layer, into a html form, so that a willing user could upload their "human interpretation" and fix djvu text layer.
While asking if any of you is interested about, I wonder there would be any wrong, or hurting toolserver politics, or raising safety issues, in publishing the python code of such layman tries into toolserver wiki into a subpage of my account, or otherwhere, so that any toolserver user could take a look if curious or simply could take inspiration to develop the idea as the idea IMHO deserves.
Alex brollo
Hi,
I would be interested in helping / contributing to this project. Although I understand (read) basic python, I'm most likely not able to help out there. Instead I'd like to help out with the user frontend (Graphical user interface and building this in HTML, CSS, and some nice JavaScript enhancements etc.)
Let me know how/whatwhere when you're up to the point that you know what functions the backend will provide and I'll get started at that point on the interface.
Looking forward to this project.
-- Krinkle
2011/2/9 Alex Brollo alex.brollo@gmail.com
I'm going to run into toolserver some simple python + djvuLivre routines to test the possibility to obtain a "wikicaptcha", built to be useful for wikisource activity.
I'm far from sufficiently skilled to write all the project, in particular the final user interface; but I'm not far to implement something like a "voluntary wikicaptcha", t.i. to select controversial OCR interpretation of words of a djvu file, and to present their image, extracted from image layer, into a html form, so that a willing user could upload their "human interpretation" and fix djvu text layer.
While asking if any of you is interested about, I wonder there would be any wrong, or hurting toolserver politics, or raising safety issues, in publishing the python code of such layman tries into toolserver wiki into a subpage of my account, or otherwhere, so that any toolserver user could take a look if curious or simply could take inspiration to develop the idea as the idea IMHO deserves.
Alex brollo
Toolserver-l mailing list (Toolserver-l@lists.wikimedia.org) https://lists.wikimedia.org/mailman/listinfo/toolserver-l Posting guidelines for this list: https://wiki.toolserver.org/view/Mailing_list_etiquette
Alex Brollo wrote:
I'm going to run into toolserver some simple python + djvuLivre routines to test the possibility to obtain a "wikicaptcha", built to be useful for wikisource activity.
I'm far from sufficiently skilled to write all the project, in particular the final user interface; but I'm not far to implement something like a "voluntary wikicaptcha", t.i. to select controversial OCR interpretation of words of a djvu file, and to present their image, extracted from image layer, into a html form, so that a willing user could upload their "human interpretation" and fix djvu text layer.
While asking if any of you is interested about, I wonder there would be any wrong, or hurting toolserver politics, or raising safety issues, in publishing the python code of such layman tries into toolserver wiki into a subpage of my account, or otherwhere, so that any toolserver user could take a look if curious or simply could take inspiration to develop the idea as the idea IMHO deserves.
Alex brollo
Looks interesting. Does the language matter for you? Because if python does not offer an advantage over php, I would recommend the later so that it'd be easier to merge into wikisource.
Looks interesting. Does the language matter for you? Because if python does not offer an advantage over php, I would recommend the later so that it'd be easier to merge into wikisource.
Yes, it matter; I hardly can write something in python, I have only very vague notions of php and javascript, and I just posted my cgi "Hello world". But I encourage any of you to catch my scripts and to feel free to develop them as you like; I'm fully satisfied and bold from preliminary results, I know that I can't develop the whole project, so please don't expect running results by me.
As I told into WMI, the Italian chapter, it's interesting that I found source, some years ago, following the stumps of reCAPTCHA, then going into Internet Archive and finally into it.source. A good example of a circular path....
Alex brollo
Alex Brollo wrote:
Looks interesting. Does the language matter for you? Because if python does not offer an advantage over php, I would recommend the later so that it'd be easier to merge into wikisource.Yes, it matter; I hardly can write something in python, I have only very vague notions of php and javascript, and I just posted my cgi "Hello world". But I encourage any of you to catch my scripts and to feel free to develop them as you like; I'm fully satisfied and bold from preliminary results, I know that I can't develop the whole project, so please don't expect running results by me.
As I told into WMI, the Italian chapter, it's interesting that I found source, some years ago, following the stumps of reCAPTCHA, then going into Internet Archive and finally into it.source. A good example of a circular path....
Alex brollo
I considered that as an option. I code better in python is a valid answer. :)
toolserver-l@lists.wikimedia.org