Christian Weiske [2006-12-09 09:01 UTC] - The top level comments should not be indented
- Use protected - private is not necessary most times
- You don't handle errors when using "@file()" in apiRequest
- Make your debug information either XHTML compatible (<br/>), or better use newlines so that it can be used on console, too.
Apart from this issues, I question the usefulness of the whole thing.
Walter Hop [2006-12-09 15:54 UTC] Thanks for the comments Christian.
First, I would argue that the class is useful. I agree it is not algorithmically complex, but Nabaztag forums and blogs are littered with bad and buggy code snippets, so I see a need for a simple class. The HTTP API is simple to you and me, but perhaps not for the Nabaztag's target group. CPAN and Rubyforge have Nabaztag modules as well.
A new version of the Nabaztag rabbit is imminent which has more features. It is unclear how Nabaztag.com is going to expose these features in their API but I plan to keep it up-to-date.
I am in agreement with the style comments and will address these.
Christian Weiske [2006-12-10 12:03 UTC] Walter, I don't question the usefulness of your package but the one of the rabbit.
Christian Weiske [2006-12-10 12:03 UTC] But that's not the question here
Stephan Schmidt [2006-12-26 00:08 UTC] I need this in PEAR. My wife bought me a Nabaztag/tag for X-Mas and I'm desperate to build new services for it.
BTW: Code looks good.
Walter Hop [2006-12-26 00:19 UTC] Hi Stephan, thanks for the positive comment.
Unfortunately Nabaztag have changed their API when they migrated to their new platform in december. I haven't been able to connect for a week, but yesterday my rabbit came back to life.
If you pull the class from my SVN, you should have basic functionality again. Api.nabaztag.com (the new url) uses a XML format for return messages which I'd like to parse properly instead of just finding a text, so it might be a few days until I've got everything worked out.
Arnaud Limbourg [2007-12-31 07:35 UTC] I know people who would use this !
|
