Federico Poloni's profile - activity

Federico, looks like you'll need to modify skin template askbot/skins/default/templates/questions.html

<h2><a title="{{question.summary}}" 
    href="{% url question id=question.id%}{{question.title|slugify}}">
    {{question.title}}</a></h2>

to something like:

<h2>
    <a href="{% url question id=question.id %}">
        Problem {{question.id}}: {{question.title}}
    </a>
</h2>
<div class="question-body">{{question.html|safe}}</div>

The main thing is to display .html field of the question object and pass it through safe filter.

note: if you do not yet use git - it's a very good idea to start since it helps a great deal in maintaining forked code in sync with the updates of the parent project. Use

git clone http://github.com/ASKBOT/askbot-devel.git math_problems

Then stick to using the math_problems (or however you call it) repository. If you move files around, remember to use git mv instead of plain unix mv. This mistake cost me a lot of wasted time before.

In addition

Possibly you might want to modify search to restrict searching in question body only. That lives in the function askbot.models.question.QuestionManager.run_advanced_search - that would be a two-liner change.

Other than that - you'll need to change locale messages. For example - edit msgstr strings in file /askbot/locale/en/LC_MESSAGES/django.po. It is much better than extensively changing messages in the template files directly, because that way chances are you won't have to even touch many templates.

You can also create your own special locale:

cp -r askbot/locale/en askbot/locale/math

and in settings.py use

LANGUAGE_CODE='math'

Then take a look at the Django localization manual.

One last bit of localization (with javascript string translations) lies in the file:

askbot/skins/default/media/js/com.cnprog.i18n.js

Ping is a very cheap operation - it just tells that your site has changed so the robot can re-crawl it.

Looks like your site will be internal - I'll add a "enable/disable" google ping option into the settings for the cases like yours.

Thanks a lot. I'll let you know how everything goes --- though probably it's going to take some time, since we are moving to a new server right now. If we do adopt Askbot, I will surely try and contribute an Italian i18n back to the project. :)

The documentation for Askbot says:

Google will be pinged each time question, answer or comment is saved or a question deleted and the site verification key is required for this to work.

Why does this happen? What information is sent to google in the process?

I see in http://qa.nmrwiki.org/ the beginning of the text of each question is reported in the default view. This looks already very similar to what I'd like to have. How is it achieved starting from a default install? Can one configure the number of characters displayed?

"Competitions" are just sets of 20-50 questions that are selected and approved. I didn't plan to manage them through Askbot: once the problems have been selected, we compose them by hand in a single (LaTeX) document and print them. A tag could say "used-in-competition-x", for bookkeeping.

"answers" could come from the collaborators trying to answer the problem (to make sure it's well-posed). I don't need a call for proposal for every competition: proposals stay in the system and can be used for later ones.

Sure. I'll describe you my use case first. I am part of the organizing committee for the Italian mathematical Olympiad --- hey guys, please go on reading and don't be put off by the m-word. :)

Each few months, we assemble a text for a competition, composed of 20 to 50 different problems. Each problem is defined by a short text (~two to ten lines) --- the solutions are not published in this phase.

Our collaborators submit proposals, tagging them to denote the problem field (algebra, geometry, et cetera), its difficulty, and the competition it's being proposed for. Other collaborators may comment on the proposal and up- or down-vote it.

For their convenience, they need to be able to print the problems, so that they can try to solve them without being in front of a computer. Each proposal may take time to analyze --- especially since they come without solution. :)

After this discussion, the responsible for the final text selects the questions to put in the competition, according to the votes and comments received, but also autonomously based on his/her judgment, and produces the "official" text of the competition (~2-4 pages of text). This last step doesn't need to be automated.

What we would need in the perfect application for our job is:

• a standard login system for our collaborators
• ability to submit proposals (short text snippets), and to later edit them for errors
• ability to enter mathematical notation <--- not a big issue, we can add it ourselves using jsMath.
• ability to comment, up- and down-vote proposals
• ability to search easily among the archived proposals, using also tags (e.g. tagged [unused] and [geometry], containing the text "Pythagoras")
• ability to print down the texts of a given subset of problems (e.g. all [unused] and [geometry] problems)

Currently AskBot satisfies all this bullets but the last one --- we'd just need to replace the word "question" with "proposal". Moreover, it has other interesting features like the ability to vote for comments, and a reputation system to let us know who are the most active collaborators.

So I asked if there is any possibility that the last bullet gets implemented.

I hope this was clear, feel free to ask for more details.

And thanks again for trying to help me and for reading all of this post. :)

The new "search results with their texts expanded" page could also be a simple static page (i.e. without any javascript at all). I'll reply to the other thread too, thanks for your consideration!

No, actually the best thing for me would be if the search stayed the same, and there were a small link somewhere saying like "show all the texts of the matching questions". So one can have the normal search behavior, plus the possibility to show their texts as well on request.

No, actually the best thing for me would be if the search stayed the same, and there were a small link somewhere saying like "show all the texts of the matching questions". So one can have the normal search behavior, plus the possibility to show their texts as well on request.

The new "search results with their texts expanded" page could also be a simple static page (i.e. without any javascript at all).

I'll reply to the other thread too, thanks for your consideration!

I am looking for a Q&A system to use in our organization to manage a list of proposal. While they are not technically "questions", the voting, comment and tagging system is great and fits exactly our needs.

However, there's one more feature that we would need and that all Q&A systems seem to lack: the possibility to obtain a page containing the text (not the title, but the full text without comments) of the items matching a search query/tags. This would be great especially for skimming through the list of questions and seeing if something has been asked already. It would be great for printing and offline viewing, too.

Does Askbot have anything similar? Is there any hope that this could be implemented in the near future?

I was considering using Askbot as a means to store proposals for exercises. Ideally, there are several proposals, and the most voted get into a problem sheet.

For this, it would be useful to have a sort of a "print view", that would display the texts of the questions (not only their title) matching a given tag/search all in the same page (without comments and answers).

I understand it is not its primary purpose, so I don't expect this to be high on the TODO list, but if it is eventually implemented it would be a really neat management system for my needs. Thanks!