[Free-sklyarov-uk] Bad CD database

[Jim Peters]

Peter Clay wrote:
> What's this CD database going to look like? Do we want to make it
> "moderated" (ie we approve anything before it's seen by the world)? Do we
> want to require people to create an account and log in before they submit
> things? (probably not)
> 
> How do we verify the accuracy of the information submitted? I suppose
> recieving a number of independant submissions is a good metric for that ..

Okay, the original plan (as thrashed out on the CD Task Force
forwarding address) was to have it unmoderated, and not to require
accounts.  However, I *would* like to have verified E-mail addresses,
as this give a good safeguard against abuse, and also gives us a way
to contact people.

I have a suggestion for a solution, which is a bit low-tech in places,
but it can be built quickly to start with, and it can expanded as time
becomes available.

This is my plan:

- We have a web page into which people can enter reports.  We'll use
  Yes/No/Don't know selectors for a bunch of standard questions, with
  a comments section below for everything else.  They will also be
  able to get to this page from existing reports, which fills in some
  of the CD details (artist/title/etc) automatically for them.

- An E-mail address is always required on a report (although this is
  never shown on the site), and when the report is submitted, it is
  E-mailed in text form to the person who reported it, for them to
  confirm by replying.  This verifies the E-mail address without
  needing a user database.

- On receiving the confirmed report, the report is automatically added
  to our database.

Initially, the database can be a simple TSV file, and the web display
of these reports can be a simple list of links to static pages
regenerated on demand.  I can develop all of this quite quickly.

As volume increases, and people have time to implement and test a
MySQL database and PHP front end, this can all be transformed into a
fully searchable resource, and made a lot more user-friendly in the
process.

There is an additional advantage - that I'll have thrashed out a lot
of the weird problems and got an idea of the shape of the data before
the full-blown front end is developed.

The submission-confirmation-by-E-mail approach also allows
keyboard-lovers to skip the web interface and make submissions
directly by E-mail.  ;-)

Admittedly, it's not as nice as a proper site with accounts and all
that, but it's a whole lot quicker to implement to start with.
(However, if someone disagrees, please let me know - I'm just judging
this based on my own skill-set).

An overview of the contents of the database, and a sample submission
report are attached below.  (This is all based on previous discussions
within the CDTF so I can't claim credit for much of this at all).

This would all be prettified for the web-submission form, but you can
at least see the kind of information I'm thinking of storing.

What do you think ?

Jim

- 
 Jim Peters                  (_)/=\~/_(_)                        Uazú
                          (_)  /=\  ~/_  (_)
 jim@                  (_)    /=\    ~/_    (_)                  www.
 uazu.net           (_) ____ /=\ ____ ~/_ ____ (_)           uazu.net

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

Here's an outline of one possible design for the `Bad CD' database
------------------------------------------------------------------

Okay, I think it's important that we get people's E-mail addresses.
This gives some credibility to the reports, and we can contact them
later on if necessary.

This usually suggests using accounts on the server, and the E-mail
bounce-back to confirm the E-mail address.

However, I have an idea which is a bit more low-tech, but will get us
the same result.  Whenever someone enters a report on the web-site,
they give us their E-mail address.  We then E-mail them the details
they've entered, with a little header to tell people who've received
it in error to delete it (like list subscriptions).  A "Reply-To"
header field gets all non-automatic replies redirected to our
submission address, at which point it is added to the database.

Effectively we're doing the E-mail bounce-back for every submission,
but at the same time saving with all the hassle of logging in and
keeping user details.

Also, those who prefer typing can fill in their reports and submit
them without using the web interface at all.

We store report records as follows:

- Artist name
- CD title
- Type (single/album)
- Other info to identify the CD (like label/whatever)
- Publically visible name of person reporting (or "Anon")
- Country code, of country in which CD was purchased
- Category Yes/No/Don't-know flags (see below for categories)
- Comments (free-form text)
- E-mail header info for tracking abuse
- Submitted date/time
- E-mail address of person reporting


Here is a sample of the submission file, which is either generated for
the user to confirm by replying, or is filled in by the user directly
(for keyboard lovers):

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

# Put here your name as you want it to appear on the site (your E-mail
# address will never be shown, only this name)
MyName: Jim Peters

# Put here the country-code of the country in which the CD you have
# was bought.  Use normal country codes, as used in internet domain
# names (use US for the USA): 
SourceCountry: UK

# --- First, the details of the CD

# Add here the artist name and CD title
Artist: Heather Nova
Title: South

# Put here the letter 'A' or 'S', for album or single
Type: A

# Put here any other details of the CD, if required (like label, etc)
CDinfo: Zomba / V2 records

# --- Now, the details of how it works or fails to work.  

# For each of these categories we need either a 'Y' for yes it works,
# an 'N' for no it doesn't work, or '-' for don't know, didn't try, or
# cann't tell you.  Leave settings as '-' for all equipment you don't
# have access to or didn't try.  If you want to give us additional
# details, put them in the comments area at the bottom.

# Does it play okay on a normal CD player ?
PCD: Y

# Does it play okay on a Windows PC ?
PWIN: -

# Does it play okay on a Mac ?
PMAC: -

# Does it play okay on a Linux box ?
PLIN: -

# Does it play okay on a DVD player ?
PDVD: Y

# Does it play okay on a PlayStation 1 ?
PPS1: -

# Does it play okay on a PlayStation 2 ?
PPS2: -

# Does it play okay on some other equipment ?  (Tell us what equipment
# you're using in the comments area)
POTH: -

# NOTE: In the following questions I'm using "rip/extract" to indicate
# the process of moving the audio data from the CD onto your computer
# hard drive.  This is sometimes known as `ripping', sometimes as
# `audio extraction'.

# Can you rip/extract the audio successfully using a Windows PC ?
XWIN: -

# Can you rip/extract the audio successfully using a Mac ?
XMAC: -

# Can you rip/extract the audio successfully using a Linux box ?
XLIN: Y

# Can you rip/extract the audio successfully on other equipment ?
# (Tell us what equipment you're using in the comments area)
XOTH: -

# Add here, between Comment-START and Comment-END, any other
# information you have about how the CD works or fails to work, like
# details of your equipment, details of how the sound is distorted,
# and anything else you feel is important.
Comment-START

Doesn't seem to be protected at all.

Comment-END

# Thankyou for your report.
#
# -- The Campaign for Digital Rights <http://uk.eurorights.org/>

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

The processing back end will have to cope with some unintentional
abuse and send back rejection messages with additional error comments
for things that need correcting in the submission.  It would have to
be able to cope with reply-quoting and some wrapping, but hopefully
that shouldn't be too hard.  Initially I can get it to send queries to
me first for checking/fixing the script before complaining to the
user.

Comments ?

Jim