.Mac's Missed Opportunities

Since comments have reached a degree of uncivilized behavior that I will not tolerate, I refer people to both my Disclaimer and my Comment Policy. If you're not up to being constructive, I suggest you vent at your own blog (preferably sticking to the point). This is not your playground, and insulting me won't make my views go away.

Like a good deal of people, I've been wondering whether to renew my .Mac subscription for next year. I held on until 10.4.7 to see if iDisk syncing improved, but I've come to the conclusion that .Mac's problems regarding IMAP access and iDisk aren't just due to Mac OS X bugs - at this point, I squarely blame Apple's servers.

You see, I have two different ISP connections at home - Netcabo, which goes through the usual Marconi link, and, of course, Vodafone's own, which currently bounces off Open Transit. Both are 4Mbps or above at my end (sorry, can't disclose the correct figures, there's just too much curiosity from my employers' competition right now).

Both go straight to Level 3 (to which Apple has a direct connection at Sacramento, if the naming is to be believed), and even if I were to allow for a slow last-hundred-miles on my end, there's absolutely no network-related excuse for .Mac to be so pedestrian.

For added brownie points, traceroute idisk.mac.com and other Apple-related site names and watch as the traffic courses through Level 3's network.

iDisk - Less Bang For The Buck

I rely on my iDisk to keep stuff I'm working on consistent across my Macs - I store my main VoodooPad file on it, copies of drafts, snippets of code I haven't yet committed to Subversion, etc., so it's the one .Mac feature (aside from IMAP mail, of course) that I use every day.

And it's dog slow (not to mention expensive as hell considering that I can buy a hundred times the storage space on a hosting server for the same cost). Om Malik quoted me on this, and will readily confirm that it's slow as molasses even in the US.

Now, I know a thing or two about WebDAV. Even if it were encrypted, it's not supposed to be this slow - even with the added overhead of HTTP, there's just no excuse.

Oh, you noticed I mentioned encryption? Yes, that's true - your iDisk traffic isn't secure. Apple doesn't use SSL for it, they only encrypt your password. Some people are too relaxed about this, so I'll highlight it -

Everything you store on your iDisk is sent (or retrieved) in cleartext over the Internet. Storing sensitive data on an iDisk is asking for trouble.

So you're getting the most expensive network storage deal in the market, and letting your data loose on the Internet, to boot. I've long switched off personal data Backups to it.

Why do I use it, then? Because it's convenient. But even that is fading away, considering that I am unable to shut down my iMac cleanly now and then because MirrorAgent gets stuck (the Finder becomes unresponsive, and after a while I just press the power button - even in 10.4.7).

Spam? What's That?

I don't get much Spam these days, largely because I pipe most of my accounts through a centralized SpamAssassin/procmail setup that filters Spam and files things away neatly in folders using server-side rules.

All my ISP accounts go through it, and Gmail goes through it too, even considering it's probably got the best built-in Spam filters (I need the server-side filters to separate mailing-lists by folders), but I kept my .Mac account out of the loop.

Over the past two years, it has become the account I get most Spam in - and it isn't even the one I use for most mailing-lists.

Other folk have complained more loudly, and, quite frankly, it's getting pretty damn annoying - I have a free Gmail account with four times the storage space (I split my 1GB evenly between IMAP and iDisk, so I only have 512MB for mail) that seems to be able to stop every new form of Spam after a few hours, and am paying US$100 for no protection?

For that, I'd rather stick to free accounts and use anti-Spam service from people who know what they're doing.

Full disclosure: I know these people. They're good, and the only reason I'm not using their service right now - despite their offering me a freebie - is that I've been wanting to shut down one of my domains and keep postponing it.

But there's more to .Mac than iDisk and mail. Or is there?

I started out meaning to use .Mac to publish my photos (and some of them are still there, somewhere), but iPhoto's web albums are simply hopeless - they are sure to be great for a lot of people, but I needed something that did not fail on upload every other time, so I rolled my own.

Somehow, I never took to Flickr. It smacked too much of Social Networking for my taste (and I had just left Deviant Art). But if you don't want to roll your own, you'd have to have a pretty good reason to use .Mac instead of Flickr for posting photos - I surely can't think of one.

iWeb has recently come along, and I was... underwhelmed. The HTML it generates, although not as bad as FrontPage's, is certainly on the far side of "crufty", and even though I get the point that it is aimed at making web page creation easy for everyone, it is still hideously slow at updating your web site. Even putting up your own (static) HTML is a hassle.

I know this because only last month I was called to help a friend of mine who simply couldn't upload the simplest site layout to his .Mac account. I managed to duplicate his problems - flawlessly, in every regard, including slow as molasses WebDAV traffic.

I wonder if that's really a coincidence.

I use .Mac syncing now and then, but gave up on most of it after a short and tempestuous honeymoon with the .Mac preference pane.

To this day, even after the 10.4.7 update, it hangs on every action - what's the deal with something that needs to talk to a server when you use a scroll bar? It can't just be poor design - it has to be some form of idiocy.

My first attempt at syncing my Mail.app accounts across my three machines turned out to be disastrous - .Mac syncing destroyed the cached copies of my non-.Mac IMAP accounts, and I had to reconfigure everything on my iBook and mini, which included waiting an entire afternoon for Mail.app to re-cache my mammoth mailboxes on each machine.

I have been working up the courage to try to sync mail rules, signatures and search folders, but don't want to risk my finely honed Mail Act-On setup - so I guess I will never be using that, either.

I don't use iCal, so syncing that is a bit pointless for me. Nevertheless, attempting to sync a mostly blank calendar took 15 minutes or so.

Syncing Safari bookmarks was the only thing that sort of worked, although I lost my neatly arranged toolbar - then again, I only have ten bookmarks or so in my browser, since del.icio.us and this site hold the rest.

So all that's left is Address Book, which I re-sync every month or so, religiously performing a backup beforehand. So far, I've had to recover from backups three times in 10.4.6 alone. I am (understandably) reticent to try syncing contacts in 10.4.7, but will eventually get around to doing it.

iChat

I take a very dim view of iChat's features. I use Adium as my main IM client, and never use my .Mac account on it.

What really irks me in iChat (and since it isn't really a .Mac core service, I won't go into painstaking detail about the rest of the things I don't like about it) is the gall of Apple to make it completely incompatible with existing H.323 video conferencing gear, some of which has been able to do H.324 video for a hear and a half now.

Working at an office where Tandberg equipment is regularly used for meetings and knowing I can join a video conference with my 3G phone, iChat's disdain for standards is almost insulting.

As with most Apple products, the things that .Mac doesn't do are almost as annoying (when you need them) than the things it does poorly. But what about the things it could do?

There have been suggestions for .Mac enhancements from day one, and I'm not going to rehash every one of them. Instead, I'll stick to two that have some bearing on what I know best - mobile services.

Ever since the iPhone rumors started, I wondered what Apple could bring to the table in terms of mobile services. Even without becoming an MVNO (something I never believed to be worthwhile for them, and that has of late become much tougher in the US), Apple could deliver some pretty neat services on "Mobile .Mac".

Being somewhat skeptical where it regards content (maybe due to overexposure to the mobile portals, with their ringtones and downloadable games sub-markets) I never thought they would go for content. Downloading stuff from iTunes over a mobile network would be utter madness at the time, and is probably still only viable here in Europe (in case you're not aware of that, UMTS and HSDPA service is a usable reality here, and at accessible rates).

But services, that would be an entirely different thing. And what are the services you could extend to a mobile phone, straight from .Mac?

IMAP Mail

Nowadays, just about every Symbian or SonyEricsson-based mobile phone out there has IMAP mail support. Better yet, there are a lot of phones with IMAP IDLE support, which means Apple would need zero investment to offer a push-e-mail-like service (excluding carrier charges, obviously).

Then again, Mail.app still doesn't support IMAP IDLE after all of these years, so the notion probably never hit them.

Imagine reading your e-mail on the go on a KG800. Tempting, huh?

Calendar and Contact Syncing

Another thing that would be a nearly-zero-deployment feature would be over-the-air iSync. You see, like IMAP, there are many phones out there with SyncML support, and that support isn't limited to Bluetooth or USB connectivity - most of them can sync remotely with a SyncML server over HTTP or HTTPS.

And Apple has shown they know (even if somewhat belatedly) support syncing a Mac directly with several models of mobile phones - doing so over HTTP/HTTPS would not be the same, but that codebase and know-how could be re-used almost entirely.

Moblogging

Those were the kick-ass, no-frills features that (to me) show how Apple could have made .Mac go mobile with a minimum of fuss. Now for fancier, more media-friendly features that could have tied in nicely with their approach to podcasting...

It never ceases to amaze me how people like the iPod's single-purposeness. I hate to think I'm indirectly giving credence to the many rumors regarding new iPods, but what if the iPod was bi-directional? i.e., if you could record your own video media on it? What if you could post it on the web directly as well?

Well, I don't have to use an iPod. Any recent mobile phone worth having is able to record 3GPP video and AMR audio, and act as a recording-studio-in-your-pocket (my K610i even has little applets to cut, paste and re-mix audio and video). And guess what, the Mac is the only platform that supports 3GPP video without any add-on software.

Can you spell "Duh"?

For instance, let's take a flight of fancy - Apple could have added basic YouTube-like features to .Mac, and they could have done so by supporting direct mobile phone uploads (or, if your video is larger than the typical 300KB MMS/WAP limit, they could have made it easier to do through iWeb).

As Carlos Andrade has recently demonstrated, all it takes is a few likes of code and a Flash player.

Well, quite honestly, I don't know. I do know that I am not just whining, which is the way the Zealots on the Mac "fanboy" support base will surely try to dismiss this - These are real problems that I (and others) experience, and just because they don't happen to a (lucky) few that doesn't mean they don't exist.

Also, I can analyze the issues involved - this is not just about comparison shopping or attempting to "roll your own" .Mac-like shim. To make this absolutely clear, I find hackish efforts to duplicate .Mac rather pointless, since the effort involved makes it a zero-sum - or negative - endeavor. Sure, you can set up your own WebDAV server with little effort (and SSL-wrap it within 15 minutes) but what's the real use of that?

Not that I won't roll my own WebDAV server if I leave .Mac, or if I start needing to store sensitive stuff on a machine I trust without risking sending my data in the clear. Again, the peanut gallery should take note that that is not the point. If you can get past your own bias, that is.

My main point is that .Mac isn't the benefit it's touted to be, and that I can come up with a number of enhancements and added features that would make it worthwhile for me to keep my .Mac subscription and push Apple's online services ahead of the pack where it regards mobility.

But other people will surely have other views (and must-have enhancements that would, in their view, motivate them to keep their subscription). Me, I will happily grant that my suggestions are clearly not aligned with whatever it is Apple apparently wants to do with .Mac. Or even my requirements (since I mostly want WebDAV that works, and less server-side cruft to get in the way of hosting static files).

On the other hand, if Apple thinks their target market for .Mac is the I-don't-want-to-understand-computers folk, they may be falling quite short of the mark - because it doesn't work consistently right even for those people (if you've never had issues with .Mac, you're very lucky).

Maybe they just don't want to do anything much with it. Or maybe it will take Leopard to, say, have an SSL-enabled iDisk (which is just their style, even considering Tiger has all the technical bits in place).

I do know that I'll be re-visiting this in a few months' time, and that they'd better have something to show for the cash I'll be tossing them - otherwise, it's just a "Mac tax" for those of us who like to have the full Mac OS X experience.