Native App Superiority

I was recently reading an article on the Tao of Mac that said, among other things:

"I’ve long preferred to use iPhone Twitter and Facebook clients over desktop ones..."

It's true, there are a handful of iPhone applications that are actually better than the original apps they replace — Facebook, for sure, and my recent fave, the Zipcar app among them. Of course the original ones — the "Desktop ones" referred to in the article — are actually web applications.

The fact is that, despite the push towards — and what many believe is the inevitability of — the web as the primary source of applications, native apps are still vastly superior in almost every instance. This is why Apple had to finally give developers —  and consumers, of course — the App Store. Web apps just weren't cutting it on the iPhone. And while they do somewhat better inside a full-sized browser on a full-powered computer, I still think web apps have a long way to go — a very long way — before they'll ever rival the experience of native apps. Platforms like the iPhone and continued interest in things like Site Specific Browsers offer very convincing evidence that native apps will continue to thrive for a long time to come. To be honest, I have my doubts that web apps will ever completely replace native ones.

I should also point out the probably obvious fact that there are certain apps that will always be best on a mobile platform because they just happen to be particularly well-suited to mobility. The Zipcar iPhone app is a perfect example. It's an app for finding, renting and controlling cars, for Chrissake. Where better to have an app about travel than on a mobile device? In fact, having the Zipcar app on an iPhone gives it certain powers the web app will likely never be able to match. But I'd venture to say that there are very few, if any, application experiences that are better in a web browser than they would be in a dedicated, native app.

No doubt about it, web apps are supremely useful, especially for certain tasks. But that's like saying the web is useful. The ironic fact that many of us prefer surfing Facebook on our iPhones to using the original web app version on a full-sized computer should give you an idea of just how hard it will be and how long it will take to supplant native apps with web apps.

Snow Leopard Scanner Application

It's been widely reported that the Image Capture app in Snow Leopard can now see and scan from many common scanners. This is a huge boon to those of us who are sick and tired of crappy scanner drivers and software. Image Capture is quite a capable scanner app, and fairly Mac-like.

But there's another, more direct way to access your scanner without opening Image Capture.

Just like with printers, adding a scanner (either via Image Capture or directly in your Print & Fax preference pane) will create a scanner application in ~/Library/Printers. This can be dragged directly to your Dock for quick, easy scanner access.

Or, if you open the scanner directly from the Print & Fax prefs, it will appear in the Dock where you can simply right-click it and choose to "Keep in Dock" from the options.

All-in-all native scanning is an extremely handy feature and seems to work well in my tests. Keeping my scanner in my Dock just makes it that much easier to use.

Snow Leopard Server-Related Changes

That title should give you a hint just how much my responsibilities have changed since I took my new job. Yes, I still run a Mac OS X Server, but I no longer get bi-yearly hardware updates. So my server is running a PPC, as is my workstation. So no Snow Leopard Server for me, at least not for a while.

I have noticed (as have many others) a few changes to how Snow Leopard handles certain server-related tasks, and I thought I'd just jot them down for the record — mine as much as yours.

Directory Utility

The first, and possibly weirdest, change is that Directory Utility is no longer a readily available application. It now lives in the very unintuitive /System/Library/Core Services, which tells me that Apple would rather us not use it unless absolutely necessary, which, generally speaking, it should not be, at least not for binding to Open Directory servers. Much of its functionality has moved to other applications and parts of the OS.

OD Server Binding

Curiously, OD binding now happens in the Login Options section of the Accounts preference pane. Even more curiously, you can open the Directory Utility from here as well:

NFS Mounts

Directory Utility used to have a pane for configuring NFS automounts. That pane has been moved to the arguably more logical Disk Utility application, where you access it under File->NFS Mounts, but it looks pretty much the same as it did before:

Root User

Since 10.5 the root user has also been activated via Directory Utility. I haven't found a new way to do this. It looks like if that's your bag you'll need to either find a way to open Directory Utility, or use the command-line. 'Course, if you know what root is, you shouldn't find either of these things terribly difficult. Especially since I just told you two ways to do the first thing.

Directory

There used to be an app called Directory in the Utilities folder, but it too is gone. I'm assuming some of its functionality has been added to Address Book, which now has its very own Accounts preference pane:

And I've read that some of its functionality has been moved to the iCal Server Utility app now included with the 10.6 Server Admin Tools:

I've also read that there is some functionality that is completely gone now.

MCX Cache

A fellow SysAdmin has posted his own groovy list of Snow Leopard changes as well. My favorite:

"New command, mcxrefresh, used for refreshing managed preferences on clients"

Hallelujah! I've bitched frequently about Mac OS X Server's overly aggressive cache. Having a way to clear it makes all the difference.

Conclusion

So we have a bit of a shuffling around here, but overall it looks to me like Apple is trying to keep simplifying the OD binding and setup process in Snow Leopard, as they have done with each iteration of Mac OS X. The most obvious features are in obvious places, whereas the more obscure features have been moved to more obscure locations. Most of these changes make sense, too, though dedicated apps for OD setup make sense on some level too. Must everything be another preference pane? In any case, it's just good to know that all the same stuff is there, it's just been moved around a bit.

On a personal note, it's a bit of a bummer to not get to play with Snow Leopard Server. I may never get the chance, actually. It could be long gone by the time we get new hardware, and we just don't rely on Mac OS X Server like we did at my old job. Ah well life goes on.

If anyone has any Snow Leopard Server stories to share, I'd love to hear them in the comments. As far as reportage goes, though, I'm gonna have to sit this one out.

The Google AdSense Experience

Well, as promised, I performed my little experiment and have come to the conclusion that, for this site at least, Google's AdSense is an abysmal failure. Here's how it all went down.

Getting an AdSense account was pretty easy — just sign up and wait for approval. And approval was quick as I'd submitted my site to Google's crawlers in the past. I was approved the next day.

Adding the ads was pretty easy too. The first thing you tell the AdSense webapp is the sort of ad you want to create. There are a few options, those of interest to me being AdSense for content, feeds and search. Once you tell the webapp what sort of ad you want to make — I started with a content ad, an ad that appears on the main site and in posts when viewed from a browser — it will lead you through a series of customization steps that determine the look of the ads — colors, size, etc. — then generate a bit of Javascript for you to insert on your site anywhere you want these ads to appear. I put my code in one of Worpress's sidebar text widgets, which work very nicely for this purpose. Before I knew it I was seeing ads on my site.

A few days later I was checking my AdSense account and noticing that I wasn't getting nearly as many impressions (page views, in AdSense parlance) as I expected. This, I realized, was because a good amount of my traffic comes from feed readers. So I decided to add another ad for feeds. Doing so is largely like creating ads for content. But getting them to show up in my feeds, easily and reliably at least, required the use of a Wordpress plugin. This was also relatively easy to install and set up, and once I had done so, I had ads in my feed.

A few more days pass with disappointing results. I'm getting a decent amount of impressions but the money just isn't flowing, so I make one last ditch effort in the form of a custom Google Search bar. What this search bar does is allow users to search Google right from my site. Any AdSense links they follow from the results will, in theory, generate revenue.

It's now been a full 10 days since I began my little test. I have a total of 2,483 page impressions according to my Google AdSense report and I'm averaging about 220 per day. (For some reason these numbers differ wildly from the number of page views reported by my Urchin stats, but whatever.) After receiving well over 2000 page impressions I've made a grand total of — wait for it — five cents. And I'm pretty sure that at least one of those cents was my girlfriend poking around on the site.

Ten days, 2400 impressions and only five pennies to show for it. Clearly this has not been a success. So what went wrong?

Well, there are two kinds of ad models employed by Google AdSense, colloquially referred to as CPC and CPM. CPC, or Cost Per Click, is pretty much what you'd expect: someone clicks a link and you get money. The other form, CPM, or Cost Per 1000 Impressions, is a type of ad that pays you per 1000 impressions. So if 1000 people view your page, you get money.

Google AdSense is very easy, but it's also almost completely opaque. You have minimal control over which advertisers are allowed to use your site; you have no control over how much ads cost, nor does Google tell you what your rate is; you have no control over what ads appear; and most importantly, you do not control what type of ad — CPC or CPM — gets posted. All of these things are decided for you by Google and their advertisers.

In my case, CPC ads are all but useless. My site is read mainly by very tech-savvy users who tend not to click on Google ads ever, even if they're not already using an Ad Blocker of some sort in their browser. And if they want to do a Google search, they either use the browser's search bar, or they just go to Google. I have reached a certain milestone in terms of traffic, however. According to my webstats I get around 1000 page views per day, give or take. Now, if Google's AdSense engine was in agreement with this, and I featured ads that use the CPM model, I might be able to make some money (though I have no idea how much). But unfortunately, the CPC model is the default, and CPM ads only show up at the request of advertisers (contrary to how it may seem in the literature). That, coupled with the odd difference in page view statistics means I will not be able to make any money with Google AdSense. And it would seem to me that the only way to effectively do so is to litter my site with ads and to get gobs and gobs of general, consumer-level traffic, two things that are really quite unlikely to happen here at TASB.

I do believe it's possible, in theory, to make money off this site, with targeted, prepaid or per-impression ads of the sort offered by The Deck and Fusion Ads, but those services are invite-only and seem to cater to a particular clientele, and I don't really see them inviting me into their networks anytime soon. In my initial post about this, a commenter mentioned a company called Federated Media Publishing. Although I've only perused their site, Federated seems to offer a service similar to The Deck and Fusion, but their invite system seems a bit more open. If I read it right, you can apply to them for an invitation. See if there's anything about your site they might be interested in.

Of all the approaches I've seen, this targeted, site-specific one is the most to my liking. It's very non-intrusive for the viewer and yet more accurately targeted than Google's algorithms could ever hope to be (I frequently had ads for HVAC and landscaping appear on my site). Down the line I may make another attempt at advertising here using a system such as this. But for now I'll be removing Google AdSense. It does not work for me, and I don't think it ever will.

802.11n

Recently a friend was commenting on how much he enjoyed his 802.11n wireless speeds, and with the arrival of my new MacBook Pro I realized I was finally in a position to take advantage of those speeds.

I resolved to buy an 802.11n Airport Base Station.

However, upon further examination of my existing Airport network, it turned out that two of my Airport Expresses were already 802.11n-capable. But they were the Remotes in the network, and the Main — the pipe to the Internet and the controller of the whole network — was an old 802.11g-type unit. Not 802.11n capable!

Setting up an 802.11n network, however, was merely a matter of rearranging the Airport Expresses such that an 802.11n unit was the Main. Doing so was a bit of a pain, as usual. I even tried some of the wizards folks had mentioned with little luck. In the end I used my trusty instructions for wireless network setup and basically started from scratch, resetting all the units. I was saved a bit of time using this option in the Airport Utility wizard:

This allowed me to at least get the new Main set up and running with the old Main's network settings, though I still had to set the replacement unit as the Main on a WDS. Although not all settings transferred, this was still a bit quicker and more convenient than if I'd started over.

I have to say, the speed boost is really nice. When it comes to download speeds, at least, I'd say they're comparable to what I see on my wired system. Very fast! I'm curious to see what file transfers are like on the LAN, but that's for another day. Suffice to say, I'm very pleased with the upgrade. And even happier it was free.

One extra bit of good news: The 802.11n range is better than my old unit, so it looks like I'll be able to take the older Airport out of the mix.

Our household is now fully 802.11n. And loving it!