With all of my problems and grousing about the iPhone 2.0 upgrade, MobileMe, etc., I do have something positive to say. One 2.0 application I find surprisingly good is the New York Times viewer. Not only is it a clean, simple program (which is mandatory for a small mobile screen), but it turns out to be a great way to read the newspaper. Who knew? Good timing, too, since I just canceled my home-delivery subscription. It appears to be the full stories and there’s just enough of the photos and graphics. None of the other apps I’ve tried are as interesting, and quite a few of them crash rather easily (Movies). One of them even rebooted my iPod Touch (Where). BofA is another handy vendor-specific app. Twitterific also looks pretty good. I may start using that as y primary client. I sort of like having Twitter in a handheld device even while working on a desktop or laptop machine. Something about the immediacy and separateness from my mainstream work.

PayPal is a valuable service, but there are parts of it that just suck. Here at The Conversations Network, we use PayPal (a) to receive membership dues and donations, and (b) to distribute payments to TeamITC. It’s been very difficult to balance the account because of the way PayPal holds onto funds (to their advantage) and only deposits the difference between receipts and disbursements back to our BofA account. So a week ago I decided to open a second PayPal account, tied to a new BofA account. It was easy enough to setup, link the BofA account and make a round of payments to our team.

A few days later, we got an email message that they needed more documentation: copy of a voided check, IRS non-profit determination letter, etc. No problem. We uploaded it to them the same day.

But yesterday I discovered we’ve fallen into a Catch 22. First, Paypal canceled all the payments we made to TeamITC earlier this month. (They’d never gone through, it turns out.) Second, they’ve essentially frozen the account. Third, the funds that should have gone to our team are stuck in the PayPal account. They refuse to return it to the BofA account. So we can’t make payments and we can’t get back a few thousand dollars that PayPal so happily withdrew from our bank. The reason? They say they can’t accept a “starter” check as verification of the account. (Silly, given that they don’t seem to have any trouble withdrawing funds from that account.) Since we don’t plan to use this account for anything other than PayPal, there’s no reason to print checks. As an alternative, they’ll accept a bank statement, but we won’t get our first one for more than two weeks. PayPal refuses to accept anything else such as a letter from BofA.

I’ve spent about 1.5 hours on the phone with PayPal people, none of whom had the authority to solve our problem. The last person said I’d have to email the Compliance department. We’ll see how that works and how long it takes to hear from them.

What we need is a reasonable competitor to PayPal. Neither Amazon nor Google are there yet. Amazon won’t allow us to pay people outside of the U.S. and Google doesn’t (AFAIK) support outbound payments at all.

Update: With a little help from the PayPal executive offices — it helps to have influential readers of your blog — we seemed to have resolved this quickly. It appears we fell under the compliance obligations imposed by the (U.S.) Patriot Act and financial regulators for KYC (“Know Your Customer”) and AML (“Anti-Money Laundering”). It looks as though all non-profits are now, by default, considered terrorist organizations until proven otherwise, and opening a bank account without printing checks is a red-flag warning of intentions to launder money. It’s going to take a while to sort out which payments went through and which didn’t. Some are clearly marked as “Canceled” while others are simply “Uncleared,” but we should be able to get payments to our international team of “terrorist” writers and audio editors in the next few days. (I can only imagine what a field day the NSA’s text scanners will have with this paragraph.) Not that I wouldn’t still like to see some good honest competition for the PayPal near-monopoly.

Guess I should have read the fine print. I downloaded a movie rental from iTunes. 30 days to watch it — okay. My wife and I started watching it yesterday, but she fell asleep. We went to watch the rest of it tonight — afte rthe house cooled off a bit from all the heat here — but no luck. iTunes only allows you 24 hours from the time you start watching the movie, then it just disappears. Seems like a ridiculous restriction to me. I think I’ll just stick with Netflix for now.

Update: I complained to Apple via a web form, and to their credit they gave me another download/viewing without question. And they replied very quickly, particularly considering how busy they must be today (iPhone 2.0). I still don’t like the policy, but at least Apple’s iTunes support works.

In my research planning our new site for finding and sharing recordings of spoken-word events, I came across Soundflavor, a site doing something similar for music. I’m still learning my way around Soundflavor, but it seems to me that it has nearly all the features we want in a spoken-word site, and it seems to be very well implemented. Looking at Soundflavor, what would you do differently for a spoken-word finding/sharing service?

Now that we’ve got OpenID running as a login/registration option on The Conversations Network, I’m concerned about a particular weakness. Maybe the Identity Gurus can help me out.

I’ve taken the advice of others and allowed registered members to attach multiple OpenIDs to their CN logins. It’s very convenient. For example, I can login with http://dkaye.myopenid.com, http://rds.com (a delegated OpenID) or a variety of others. But this isn’t solving an important problem that I think it should. What happens when I, as an OpenID owner, change my email address? I’d like to just change it in one place (my OpenID provider’s site) and have that change automatically propagate to the sites where I use my OpenID the next time I log into them (if not before). The service providers allow me to change my email address and that address is transmitted to the sites when I use my OpenID.

The problem is that because we receive those email addresses from potentially multiple providers, they can be different. And when we receive an email address as part of an OpenID authentication transaction, we have no idea whether we’re supposed to change our database to reflect that new email address or not. Bottom line: We have no choice but to ignore the email address we receive except the very first time when we can use it as the default for a registration-form field.

I thought OpenId credentials were like the old wallet concept, but how is a web site supposed to deal with an individual who supplies multiple wallets? Am I missing something here?

I can’t tell you why, but Tony Hirst has use the clip/excerpt feature from The Conversations Network as part of a pad-triggered JavaScript audio demo. All sorts of things are possible.

I’ve spent the last three days trying to implement OpenID for The Conversations Network’s web sites. I’m familiar with the concepts and protocols, so I figured it wouldn’t take much effort. I guess that’s what I get for being so self-confident. I still don’t have a solution to the current snag. Maybe someone else knows a workaround or at least some search engine will find this so others can avoid the traps I’ve fallen into.

All our sites are built on PHP5. I develop on OS X, check my code into Subversion, then checkout onto our live servers running RHEL. I chose the OpenID Enabled library. It’s really the only choice as far as I know.

I don’t know whether it’s a bug, poor documentation, or just our peculiar architecture, but there was an incompatibility between OpenID Enabled and my code that took about 12 hours to find. (If the URL of your response receiver includes a query string, you’ll have some extra work to do.) I finally got past that, coded new pages for login, registration and managing multiple OpenIDs for a user, and checked it into svn. I installed the library on the RHEL server and updated the code there. Didn’t work.

It turns out that the OpenID Enabled library needs an XML parser, either domxml (which is no longer available) or the newer/better DOM, which comes with the PHP 5 core code. But we use yum, and the yummy version of PHP5 was built with the –disable-dom option. So no XML parser and hence no OpenID. Aargh!

Sysadmin Tim doesn’t want to mess with another PHP5 build, and after years of working with him — and sometimes ignoring his advice — I’ve learned he’s always right. Surely, replacing our PHP5 (or even rebuilding it) would likely break something else. So I’m stuck until I can either find another OpenID library that doesn’t require the DOM XML parser, or some replacement for that parser. Too bad. OpenID is sort of cool.

Update: Problem solved. Just ran “%yum install php-xml” and restarted Apache. We now have OpenId across all The Conversations Network’s web sites.

Windows and OS X users can now download a new version (1.4.0) of The Levelator. A few small fixes and minor new features. If you’ve been stuck on a bug, it may be fixed in this version.

Podcasting pioneer and friend, Dave Slusher, is reviving his Reality Break radio series as a podcast. Dave honed his interviewing skills well before the advent of podcasting, and I’ve always appreciated his research and preparedness, which help make his programs a cut above the rest. For a while, Dave hosted the Voices in Your Head series on IT Conversations. The majority of Dave’s interviews are with authors, most notably those of science fiction.