An Index of Frequently Asked Questions

Is there a way of mining the forums and figuring out an index of the
top 10 or 20 frequently asked technical questions on ROR ?

Top 2:

1. Why is scaffold broken aka "undefined method scaffold"
2. Why can't I load sqlite3 aka Why is it trying to load sqlite3 aka
"no such file to load -- sqlite3"

Ryan Bigg wrote:

Top 2:

1. Why is scaffold broken aka "undefined method scaffold"
2. Why can't I load sqlite3 aka Why is it trying to load sqlite3 aka
"no such file to load -- sqlite3"

hahaha :slight_smile:

Patience my friend ... we shall tackle this all :slight_smile:

but seriously we should compile a list of frequently asked questions and
take this as a hint of what to tackle next. If they are having problems
we could give them REALLY REALLY descriptive error messages with links
to pages that help them tackle such errors. Some of these errors I am
sure could be tackled by some rewrites and making the framework more
"nooby proof".

Thank god I didnt have such questions when I was starting out but you
have to also see that people who want to give ROR a bad name will
attempt to raise the barrier to entry (which we must lower) so more may
take advantage, contribute, improve the framework...

That's a job for the wiki, when it's eventually cleaned up :slight_smile:

Ryan Bigg wrote:

That's a job for the wiki, when it's eventually cleaned up :slight_smile:

As for keeping count of the indexes and top searches and top questions,
could someone managing this site have an index like this one for
example:

http://isohunt.com/stats.php?mode=zg

Thanks!

Ryan,

How could you ever forget the absolute No 1???

    Why is Rails better than PHP?

Cheers, Sazima

Sazima wrote:

Ryan,

How could you ever forget the absolute No 1???

    Why is Rails better than PHP?

Cheers, Sazima

Seriously, I know loads of people would want to harrass the community of
developers for their own reasons. We should however be able to combat
that by positive steps and solutions.

I just cleared my response when it got to about 7 paragraphs of ideas
- that no doubt have occurred to everyone else. But in summary.

1. I think mining the data in the forum is a good idea if there is a
way of generating some sort of report that would be a start. When I
have a random question, I nearly always do a forum search first
(google search in the group is great, i wonder if they have any
analysis tools)

2. I was pleasantly surprised when I looked at the wiki the other day
for some help on testing. It is much cleaner than it was and I found
the help I needed. (be encouraged if you have anything that could
usefully be added there)

3. To me, it seems that Rails documentation is pretty spread about
and so what we need is not necessarily something new, in a different
place, but something adding to a known place. So my vote goes for
putting a faq on the wiki, unless there is something clever that can
be done within the google forum area itself.

4. It is a shame that the rails manual is not being updated. Has it
been supeceded by one of the new documentation ventures or is there
scope for editing/updating it? It has quite a nice format.

Well it nearly became seven paras again!!

Tonypm

All great ideas

tonypm wrote:

1. I think mining the data in the forum is a good idea if there is a
way of generating some sort of report that would be a start. When I
have a random question, I nearly always do a forum search first
(google search in the group is great, i wonder if they have any
analysis tools)

I was asking for an index of frequent searches and index of frequent
questions because we should target the most COMMON questions first if we
want to help the other developers. If we target the most common
questions and put up responses , help , bug fixes, tutorials on those
questions we will be doing what the rest of the industry does.

The rest of the industry follows this rule : target the most common
questions , problems, bugs etc. and the rest will follow (I forgot the
name of the law but there is a law in EE/CS that asks you to target
these most common flaws first).

Perhaps the authors of http://www.ruby-forum.com could have something
like that?