I’ve been using different forms of computer “chat” for over ten years now, starting with operator-to-operator communications over a 9600baud satcom circuit in the Navy. Over time, I’ve become used to using certain forms of “emoticons” to convey subtle nuances in the conversation that are unnecessary in face to face communications. I even have friends with whom I communicate with entirely over chat.

Over the summer my cousin appeared on chat, and I tried to have a conversation with her. She was not familiar with the conversational tone and rhythm of chat, which made the interface difficult and frustrating, to the point where we both simply decided to go back to email.

Apple understands the human element of their devices possibly more than any other company in their field. Their advertising plays to your emotions, and their products are designed to elicit an emotional response; an appreciation for their beauty. Technology like Skype and Apple’s FaceTime video chat removes one layer of abstraction between you and the person you are trying to communicate with, and allows the emotional facial queues that are so important in communication to come through.

If my cousin and I were chatting face to face rather than over the keyboard, I imagine our stunted conversation would have lasted a bit longer than it did.

Another point I’d like to make about emotion and computers is that even if you do see your computer as simply a machine, a tool to accomplish a task, it is difficult to use a tool for any serious length of time, with a serious financial investment without an emotional connection to the tool. A carpenter is likely to have his favorite hammer, a mechanic his favorite ratchet, and anyone who uses a computer to create something will have their favorite brand, and strong reasons for choosing that brand.

Last night I did my civic duty by casting my vote at the local community center. I walked down since it was not far from my house, and enjoyed the crisp night air. Once I arrived at the community center I noticed that the voting process was being run by a group of elderly women, two of whom had Lenovo laptops, which were curiously tied together by an ethernet cable. Each of the laptops had a label printer attached to it via USB, with the other USB port occupied by a mouse. As I approached one lady noticed me and asked me to fill out a form, which I did, and then asked if I had voted there before, which I had not. That turned out to be a bit of a problem, one that was easily resolved, and one that was caused entirely by the laptops.

The laptops were labeled “Primary” and “Secondary”, and each had stickers on it showing which ports on the side to attach the mouse and sticker printer to. They were each running some kind of database software that had all of our names and registration status. When I was asked if I had voted here before, I said that I had not voted in that town, but I had voted early at the county seat during the presidential elections of ‘08. They wanted to make sure I was in the database, but since they were continually having problems with the computers it took some time.

While I was waiting, a man next to me needed to be registered, so one lady asked another, who was apparently in charge, to come and help put him in the database. I overheard the two of them ask questions like “I’m not sure what the difference is between ‘accept’ and ‘apply’”, and “Ok, I don’t know what to do here, where do I go next?” I couldn’t help but wonder who had designed this system, knowing that its intended users were going to be elderly women who had little to no computer experience. One of the ladies rebooted her computer twice before she was able to get it to work again.

I leaned over to take a look at the screen, and confirmed what I had previously thought. It looked like an application left over from the Windows 3.1 days, multiple screens, buttons everywhere, seemingly random labels. How much simpler and easier could the entire night have gone if they would have given that application to a UX designer first, before sending it out to be field tested.

People have become used to computers behaving this way. They are incomprehensible, confusing machines that if you look at them wrong they break. I wanted to tell the ladies that the problems they were having with the computers were not their fault, but the fault of the people who designed the computer, the operating system, the application, and the process that must be followed to glue them all together. I wanted to tell them that it doesn’t have to be this way, and that computers are meant to make things easier, not harder; simpler, not more complicated. If they don’t, then why do we continue to use them? It would have been easier to stamp everyone with a rubber stamp last night than deal with those machines.

I wanted to tell them a lot of things, but there were people behind me in line, and it had been a long day already. So, I smiled, said thank you, and cast my vote. Then, I walked home and enjoyed the starry night.

I like to keep both my desk and my computer desktop clean and clutter free. I’ve found that when there is less visual noise, I’m able to better concentrate and focus. In the article “The Proximity Compatibility Principle: Its Psychological Foundation and Relevance to Display Design”, Wickens and Carswell outline scientific principle’s that back up my personal preference.

Unless I’m actively working on a project that requires papers, my desk has nothing on it except my notebook, my computer, and a pen and pencil holder. Likewise, my desktop on my Mac is normally free from files, icons, and distracting wallpaper. If there is a file on my desktop, or if the trash needs to be emptied, I find that my attention is drawn to those cues, and I wind up dealing with them right away.

I always thought it was just because I was picky, but in the article it says

“It is clear that the negative influences of confusion and clutter will be enhanced to the extent that the contributing elements are both salient (bright, distinctive) and cannot be easily discriminated from the relevant ones. (In the visual search literature, this is known as target-distractor similarity.)” (Wickens, and Carswell 473-494)

To me, this is saying that when there are distracting visual elements, like a bright and colorful wallpaper, it takes additional mental effort to concentrate on the task at hand. Furthermore, the article also states

“Added clutter is known to disrupt visual scanning, whether this scanning is carried out by movement of the eyeball or by movement of an internal ‘attention pointer’ (Thorndyke, 1980). It is evident that the costs of scanning over (or filtering out) this clutter will be greater when there are added burdens of integration.” (Wickens, and Carswell 473-494)

When writing, or entering commands into a terminal, I find it much easier to concentrate on the task at hand when there is a single window open on my display, and the background is a neutral grey, and all non-essential visual elements are removed or hidden. I was very glad to read about Information Access Cost, it is good to know that there is actually hard science backing up my personal preferences.

So, last night, when I should have been sleeping, I had an idea. What if, instead of going to a bank to get a loan, you could ask a few of your Internet friends for a loan instead? Say you want $1000 for a new iMac, you go to some imaginary website and tell it how much you want. This website does a quick credit check to get your credit score, and then determines your interest rate based on that score. You agree to the rate, and your request is posted anonymously to the site.

Next, people with money to invest come to the site and see your request. Let’s say your agreed upon interest rate was ten percent. One person might see your request and agree to loan you $100. This person would then make nine dollars off of your loan when you pay it back. On a $1000 loan, at ten percent interest, the web site would take one percent for itself, and then split the remaining nine percent evenly between individual loaners based on how much they loan. If one person loaned the entire $1000, he would make a profit of ninety dollars off of that loan when you paid it back. If ten people each loaned $100, each of the ten would make nine dollars apiece. If one person loaned $900, and another $100, the person who loaned $900 would make $81, and the person who loaned $100 would make nine dollars.

I’m sure this has already been done somewhere, but it was a new idea to me so I thought I’d make a note of it. I’m not sure how this compares to micro-loans for businesses in third-world countries, or how it compares to Kickstarter, but I thought it was interesting. I imagine there would need to be government regulation, and some kind of federal insurance as well, for cases when someone defaulted on a loan. I’m also sure there would have to be some kind of protection against fraud, but again, this is just a fresh idea, it’s not well thought out at this point at all.

Who knows… maybe there’s something there.

A long, long time ago, what seems like a different life now, I was a Sailor. Towards the end of my teenage years, I came to a point where I knew I had to do something with my life, and at the time, that something was not college. My adoptive father was in the Navy, so I decided to follow in his footsteps and joined the Navy myself in October of 1995. From June of 1996 to July of 1999 I was assigned to the USS Platte, an oiler. During this time I made the best friends of my life, met my wife, and travelled across Europe and even into the Middle East.

It was a different world back then, back before 9/11. It was a brief time of peace, a period of national calm that came after the cold war was over, and before the war on terror began. I went on two six-month deployments to the Mediterranean, Med cruises we called them. We would travel from port to port, spending anywhere from a couple of days to three weeks in port, followed by a week or two underway.

In port, my friends and I would make a point of going out and seeing the sights during the day, before taking in the local beverages at night. I loved the architecture, I loved the age of some of the buildings and castles that we found. Back here in the States, if a home gets to be one hundred years old, its an amazing thing, but over there structures built by man could last for hundreds and hundreds of years.

Sometimes, when we were out to sea, the water would be so calm it looked like glass. Other times the waves splashed over the weather decks and would threaten to wash an unwary sailor overboard. During those times I’d have to take medicine from the ship’s doctor to try to ward off the seasickness that would invariably come. I never got over it, in three years I never got my “sea legs” like most of the guys did. But mostly it wasn’t a problem, most of the time we avoided the rough weather and stayed in the calmer seas to do our refueling of other ships. Most times the waves were small enough that I could stand on the edge of the ship and watch flying fish dart between the crests of the waves. And sometimes, I would close my eyes and smell the salt in the air.

We’ve been having a months long discussion at work around which Linux OS to use. It’s all come to a head recently, and it looks like the winner is going to be Red Hat. The decision leaves a slightly sour taste in my mouth, but over the course of the past year I’ve gotten used to having it around. While trying to understand why I’ve got such a dislike for this particular flavor of Linux, I thought it might help to take another look at OpenBSD.

OpenBSD and I go back a long ways. It was the first Unix operating system that I really got to know well. I had worked with HP-UX a little, and Solaris a little more, and a few flavors of Linux, mostly Mandrake and SuSE, but OpenBSD was different. OpenBSD doesn’t give you any room to not be an expert. You need to know what it is doing, and why it is doing it. The demanding nature of OpenBSD begins with the installation, which is actually a shell script with a few prompts thrown in. Back then, you couldn’t get the OS in an iso download like Linux, you either needed to dd an image to a floppy disk and install off the network, or create your own iso and boot from that. Once you booted the install disk, you used fdisk to partition your hard drive. Next, the script would install it’s boot image, and after formatting the partitions, simply unzip the OS into the root partition.

The entire process of installing OpenBSD takes about six minutes. When finished, the server is ready to be a web server or a firewall, but not much else. Everything else is build from the ports tree.

If you accidentally muck up some file that OpenBSD needs, you could just unzip the tarball that holds that file from the root partition, and you’d be back where you started again. I’ve seen OpenBSD servers run for years without a problem. They are the kind of server you setup once, and then forget about. OpenBSD is simple and powerful, and simply feels like quality engineering.

Looking at OpenBSD I discovered the root of my annoyance with Red Hat. Its not just that Red Hat has continued to add more and more junk to their operating system for no need. Its not just that they have gone out of their way to charge their customers more for a more confusing setup. Its that they feel just like every other modern product on the market today. Full of features, and lacking in quality.

Its not just a problem with geeky computer operating systems. Its not just a problem with computers. Its a problem with everything that is on the market today. Air conditioners, refrigerators, hardwood floors, cars, coffee makers, the list goes on and on. At some point in our recent past, consumerism took a turn for the worse. Companies decided that they could make more money if they made a worse product, and we went along for the ride. So instead of a washing machine lasting for fifty years, it might go for six before some part in it goes out, and fixing that part is more expensive than buying an entirely new machine.

In the 40’s and 50’s, American companies made some great products. I went to pick up a mixing bowl for my brother in law the other day that he got off of Craigslist. It was a classic, built in ‘48 or so. The seller wanted to show me that it still worked fine, so she plugged it in and gave me a demonstration. Sure enough, it fired up and worked perfectly. That little device is at least sixty years old. What are the odds of any of our small appliances or electronic devices lasting even half that long today?

My wife’s grandmother has a Maytag washer that she’s had since she can’t remember when. It has a hand crank on the top of it, and she still uses it to wash clothes today.

There is simply not enough great stuff in the world anymore. Companies produce crap that’s just good enough to get by, offer it up as being worth it because they’ll warranty it for three years. Then it breaks on the fourth year and you buy another one.

I don’t have room in my life for crap.

Things are only expensive if they don’t last. Quality lasts. A few companies get it. Apple gets it. They don’t build crap. They could start pushing out $300 netbooks and flood the market with substandard junk that’s slow and throwaway breakable, but they don’t. They build quality machines, and improve on them year over year. They build quality software, and improve on it slowly over time. Saddleback Leather is another company that seems to get it. The specialize in quality leather goods that are built to last. Not the “quality leather” that you can get from a TV infomercial, but real tough leather that only starts getting broken in after the first few years of daily use. They have the only warranty I’m interested in. 100 years.

Quality. Volkswagen didn’t change the basic design of the bug for over two decades. They made small improvements over time. OpenBSD looks almost the same today as it did when I first fired it up seven years ago. Its not that they haven’t been busy building awesome software, it’s that the initial product is timeless in its simplicity.

Simplicity is hard, admittedly. But what if there were more companies that took Apple’s approach to building consumer products? How about a washing machine with one button on it that says “wash”. Who needs a digital display and sixty-four buttons on a washing machine? Looking around my kitchen right now I see I have four clocks. One on the stove, one on the microwave, one on the radio, and one on the coffee maker. Why? How about a coffee maker with one button that I press after I’ve put in the coffee and water? Does everything need to be digital?

When I look at our server infrastructure I think that it needs to be simpler. Complexity is the enemy of security, and uptime. Too many companies have taken the easy way out by not considering how to simplify their products. Linux, all flavors of Linux, are most guilty of this. The more complex a product is, the more parts there are that could fail, or at the very least, not work as well together as they should.

Which brings me back to OpenBSD. I’ve loved this system for years because of its simplicity. Which happens to be the same reason I love Macs, and the same reason I love things made out of leather and oak and metal and glass. Simplicity and quality.

I don’t have room in my life for anything else.

I’ve been writing online since around 2000, starting with a geocities site. After that, I had a 50megs.com site that offered gasp 50 Megabytes of online storage! (Just noticed, they still do!) After a couple of years, I noticed a company named 1and1 that was offering three years of free hosting and the domain of your choice. That was a deal I couldn’t pass up, and I grabbed the domain name sourceport.org. I used the domain, but didn’t get a whole lot accomplished with it. I used it for a while to sell PCs as firewalls with OpenBSD, but that didn’t end well, so then I started a Wordpress blog with that domain name. Unfortunately, the person who owns the sourceport dot com address had a beef with me, so I moved the blog over to a new domain: jonstechblog.com. While I was using this blog, I moved away from wordpress and started using a Mac application named RapidWeaver. I switched to RapidWeaver because it made the task of posting videos of Linux distros that I was testing out easy. I was testing the distros anyway, so I figured, why not share it with the world.

One day, I got an email from the guys at OSDir, asking about a “link exchange”. It turns out he thought I was on to something, and that he was toying with the same idea. He donated all of his videos to me, we exchanged links, and osvids.com was born. OSVids was a big hit, and got very popular very fast, starting with an interview I did with Tina Gasperson from NewsForge.

It turns out that the site grew too popular too fast. When I first started uploading the videos, I was encoding them with quicktime, which turned out to be a very bad idea from a usability standpoint for my target audience. The Linux crowd demanded an open format, so I offered downloads in Ogg-Vorbis format in addition to switching to the more widely deployed flash. Starting out with Quicktime had another very bad side effect: bandwidth. The quicktime files were huge, and I went from a negligible bandwidth use to 1.5 Terabytes in one month. This took a toll on my wallet in a big way. I noticed an advertisement for a hosting company that claimed “Unlimited Bandwidth!”, so I signed up and moved OSVids over to their servers. Another bad idea. I learned quickly that there is no such thing as “unlimited bandwidth”, as OSVids quickly brought there servers to a screeching halt. I complained, they tried to respond, and eventually I moved the site back over to 1and1. Even on shared hosting, 1and1’s servers never skipped a beat, and I never had any downtime with them. But, like I said, I paid for it.

I kept up the site for a few months steadily adding videos of new distros as they were released. Then, several things happened in my personal and professional life that caused OSVids to begin a slow but steady decline. The most major item was the loss of hardware. Through some unfortunate events, I lost the nice IBM laptop that I was using to create the videos. I tried using my Mac, but the resolution was too low on the iBook, and even after upgrading to a MacBook, the resolution was far too low. In addition, we sold our house, moved to a new state, I had to find a new job, and we had some semi-serious in-law issues. The biggest news was finding out that we were expecting. So, I tried several different approaches to make OSVids work, but in the end I had to shut the site down. Not because it didn’t get any traffic, it did, but… it was an open loop that needed to be closed, and I couldn’t find anyone who wanted to take the time to do all the video encoding and manual coding of the web site that made OSVids run. I lost most of the data from the site, but what I found I uploaded to YouTube.

I started this blog a few months after loosing OSVids. I needed to completely reboot my online presence. I found that what I really needed was more time offline, to really think about why I wanted to write, and what I wanted to write about. After some thought, I’ve decided on a simple blog hosted here on github.

I’ve been inspired by Daring Fireball, and Dive into Mark, and a host of other great sites. I hope that over the years as my writing voice develops I’ll be able to look back on what I’ve written here and see my work progress. I do not write this blog for it to be popular, I write it because I love to write.

The loss of OSVids was unfortunate, but really inevitable. Even without my life going nuts at the time, the rise of YouTube, Google Video, and all the other video sharing sites quickly made my hand-coding and inability to accept submissions from fans obsolete. Perhaps there is still room for OSVids. I still own the osvids.net domain, and I’ve been contemplating bringing it back. For now, I’ve really come full circle, but as a much more experienced writer. In the end, its just me and my blog, but if you’d care to join me, I’d love to have you.

As a carpenter has his tools, so do I, as a “knowledge worker” have my computer. I spend many hours a day with my Mac. I have my workflow honed and finely tuned, and I know when something is wrong with my computer, when there is more friction than there needs to be. I am a Systems Administrator, so my knowledge work is to ensure that other people can get their work done. My job is to keep the servers, services, and systems I support up and running 24x7. The tools I use to get this job done mean a lot to me, and over the years I’ve tried many of them with varying levels of success. I know exactly what my ideal setup is, and I’m working towards filling out my toolbox with the very best as I strive to bring my craft to the next level of mastery.

At home, my computer is college, entertainment, finance, photos, blogging, and fun. Mostly fun. There have been times when I’ve walked away from even owning a computer at home, seeing it as a distraction more than anything, but I always come back to wanting one around, if for nothing else than as an outlet for creativity.

In the years that I’ve been using computers, I’ve found that I desire simplicity more than configurability. Favoring fewer options over more. The machine I use needs to be beautiful to look at, because I spend a lot of time looking at it. It needs to be simple to use, because I have work to do, photos to edit, words to write, and I don’t want to have to mess with anti-virus updates or X windows crashing because of some beta driver bug that made its way into the mainstream release. I just want my computer to let me do what needs to be done.

Beautifully designed and crafted, simple to use, powerful… my computer needs to be a Mac. No one else on the market can release a computer that matches a Mac. I’m not sure why, it’s like they don’t know how. They try, but they fail.

Open source operating systems like Ubuntu are not as good because there are far too many cooks in the kitchen. Ubuntu is not an operating system like OS X is an operating system. It’s the Linux kernel, the ext4 filesystem, the Xwindow system, the Gnome desktop, the Firefox browser, and thousands of other open source packages and applications that work loosely together, and are developed by different teams. OS X is developed by Apple.

Windows XP is a suitable operating system to work with at the office, but I am far more productive on a Mac. With tools like Yojimbo, Spotlight, and Quicklook, Macs are far better suited for information management. I hear Windows 7 is nice. My wife has it on her PC, and so far, it is still just a PC.

As much as I love Macs for their design and ease of use, I also see the faults of some business decisions Apple has made in the past few years. The App Store is either a resounding success or a horrible failure, depending on who you talk to. In sheer volume, 200,000 apps is a lot of applications, but like Windows was last decade, most of them are crap. Apple’s decision to approve each app in the store is admirable to a degree, but they are not executing well at all. Some people are philosophically opposed to the app store, saying that the iPod/IPad/iPhone ecosystems should be open for any application to run on them, as is the case on the Mac. I do not care about this aspect, but I do wish that Apple would fix their approval process to make the system much more transparent. There should be clear cut guidelines on what is acceptable and what is not, and those guidelines should be applied across the board. Random app store rejections are the running gag of the current implementation. Its wrong, and it needs to be fixed.

Is the Apple today the same as the Apple so many fell in love with in the ’80s and ’90s? The scrappy underdog that just won’t die? No, and I couldn’t care less. I find it interesting that the era that some romanticize is actually one of the worst in the history of the company. Back when Apple was allowing clones and releasing crap with the Apple logo on it. Good riddance to bad rubbish. OS 9 was not interesting to me. OS X is.

I was using Linux and OpenBSD when I first heard of OS X, my first iBook was a revelation. Finally, someone had put a decent GUI on a Unix box. Apple has only gotten better from there.

I know there are a lot of very smart people who disagree with me. Lets let the next twenty years decide who is right.

Along with the app store debacle, there is Apple’s stance on Flash. My personal feeling is that if Flash were a true open standard, if anyone could create Flash applications without relying on Adobe, it’d be a whole different ball game. As it is, Flash is controlled by Adobe in its entirety, and that seems wrong for the Internet. The Web is the great leveling field, a mechanic in Kansas has the same chance of creating an awesome web site as a multi-billion dollar corporation. All the tools to create amazing web sites are free, and the specs for building the sties are readily available. All you need is a computer, Internet access, and a text editor. With Flash, you need some pretty expensive software. Also, having run a video serving site in the past, I can tell you that HTML 5 would have been a Godsend back then. It would have been so much simpler to just drop a .mov or .ogg file enclosed in video tags than the junk code I had to put in.

I’d like to watch Hulu on my iPad. Netflix already rocks on it. Flash is not a necessity.

Finally, there’s Google. I used to love Google, back when it was a search engine. They could have been happy with just being the best search engine in the world, and making billions, hand over fist, but no… they had to go and get greedy. Eric Schmidt sat on the Apple board of directors and saw what Apple was doing, and thought to himself… Google could do that. So, they “stabbed Apple in the back”, and released Android, and then the Nexus One, a direct competitor to Apple’s iPhone. Bad form, old boy, bad form indeed.

Also, I don’t like Google’s business model any more. I used to be fine with it, when they would show ads on the search results. Now though, Google wants to watch everything you do online, and figure out a way to monazite your activity. Your email, calendar, RSS feeds, photographs, friends, chats, videos, music, there’s even Google Health where you can put your medical record in Google. It all goes into the big black box that is Google, to be analyzed for who knows how long. Me, I like to be a little more honesty with my transactions. That’s why I pay for my email service. I give Apple money, they give me an email address, and a few other perks. It’s as simple as it gets.

I think that about does it for the major topics of the day. Of course, in all these things, I might be wrong. However, if I am wrong, and you want to call me out on it, I suggest you do your homework first. I’ve done mine. I have several years of experience, and a finely honed sense of craftsmanship.

I am, after all, strongly opinionated.

I was asked on Twitter the other day why I disliked IBM’s enterprise software. This, in addition to my previous TWS rant, is my answer to that question.

We wanted to do two simple tasks. Add a new user to the system, and have the system send emails automatically when there’s a problem. Seems easy enough, unless you are using the Tivoli Workload Scheduler. Then it’s an entirely different matter.

Add A User

Some new websites like Postulous create a new user for you when you send them an email. Others like Tumblr need only three the username and password to get them setup. To add a user to TWS, you would think that there would be a nice GUI with a menu option that says “Add User”, but none exist. Instead, you have to log into the command line on the server, run the command “dumpsec”, and redirect it’s output to a file. Then, you have to vi that file, and edit the XML to add the username to the correct group. Save that file, and run “makesec filename” to load the new user into the system.

Then, restart the TWS application server. IBM is not sure if this is a required step or not, or a least the help I had on the phone wasn’t sure.

Then, you need to go into the web interface for TWS, and add the user into WebSphere as well.

It’s like the creators of TWS got together for a brainstorming session one day and asked “What is the most difficult, unintuitive way we can add a user to the system?” Congratulations, folks… I think you nailed it.

Send An Email

The heart of TWS is scheduling jobs, and then acting on the results of those jobs. One task we wanted from it was to let us know if it couldn’t run a job by sending us an email. For two years we worked with consultants and IBM to get this to work. We wound up having to get some of the original developers on the phone from Rome, and with their help we finally found the problem.

TWS stores some of it’s settings in a DB2 database. That right there is enough for me to toss the entire application in the trash. In Unix, configuration settings are stored in a plain text file, one file per application if possible. And if that wasn’t bad enough, we found that one of the binaries was modifying the configuration settings in the DB2 database when it was launched, changing the port that a certain daemon was supposed to be listening on. This daemon was responsible for listening for incoming configurations from the main server, including the configuration telling it to send the email. It’s hard for me to express how wrong this is, but I’ll try.

Any daemon should read its configuration from a file under the /etc directory. That’s how it works in Unix, and for the past thirty years its worked out pretty great. No daemon should have access to modify the configuration of any other daemon. Also, if listening on a certain port is central to the communications and functioning of the application, don’t make that configurable, just hard code the daemon to listen on that port. I suppose it would also be acceptable to allow configuration that would override the default, but only if the daemon reads its configuration from a plain text file and only if in the absence of the overriding configuration the default port is chosen to listen on.

Again, $30,000 and the application is held together with duct tape and silly putty.

If the only computers you use are Macs, or if you only use Microsoft Outlook on a PC, you may not have spent a significant amount of time in the web interface to MobileMe. The web interface occupies a curious spot in Apple’s portfolio as one of their few products that tries to do everything, and gets very little of it right. One piece in particular that gets most things wrong is the web interface to MobileMe Mail.

MobileMe Mail is built on SproutCore, a library of Javascript and HTML 5 frameworks intended to build desktop like applications for the web. The Mail web app looks like the desktop Mail application, and has certain features that try to mimic the feel of the desktop application, but falls short in functionality. The Mail web app tries to bring the desktop experience to the web, but in doing so ignores the web paradigm that makes Gmail such a success. I’ve narrowed down the problems with the web interface to five main areas.

Speed. Too often I’ve seen the dreaded “MobileMe Mail is loading” message. That this message even exists tells me that the application is too slow. The message is a web equivalent to an app splash screen; I don’t want to see a splash screen, I want to see my mail. Mail displays another spinning gear and a “Loading…” message when switching between folders, or sometimes just between messages in the same folder, instead of text. These messages do not tell me that Mail is working, they tell me that something is wrong with the applications internal design.

Search. In the top right hand corner of MobileMe Mail is a search field. Typing in the search filed will allow you to search in only the currently selected folder. This is wrong. When Apple introduced spotlight in Tiger, they made a huge deal out of it, and rightly so. Search on the desktop changed the way I used my Mac, and the speed improvements in Leopard made it the default way I find anything. In placing this search field in the web interface, I expect to be able to search not only all of my mail in all of my folders, but all of my contacts, documents, appointments, and media available in MobileMe. Instead, it searches in one folder in one application. Not only that, but it appears to only search basic fields in each message, not through the text of the message itself.

Rules. There are no rules for automatically processing incoming mail. This is also wrong. Rules should sync from the desktop Mail app. For some users who may be using an iPad, and only and iPad, there is no way to filter incoming mail. The lack of rules and filtering severely limits the usability of the web interface, and is another big advantage Gmail has over MobileMe Mail.

Folders. While Gmail has no folders and instead uses labels and search, MobileMe Mail has neither good search, nor good folder management. The only options available for folders is to create, delete, and rename the folders. Since Mail is unable to use search as an organizing feature, folders are the only option.

Notes. It’s almost hard to include notes as a part of MobileMe Mail that does things wrong, when it actually doesn’t do notes at all. iPhone OS 4 is supposed to (finally) bring notes syncing between the desktop Mail client and the iPhone over IMAP, and therefore MobileMe, but I haven’t seen an announcement of the Notes being added to the Mail interface, or as an additional web service.

Apple has made strides to improve their online services since the rocky conversion from DotMac, and I hope that they continue to make the MobileMe web interface better, faster, and more usable. Critics often call MobileMe Apple’s answer to Google’s, Yahoo’s, and Microsoft’s online services, offered at a premium price. However, it is important to keep in mind that MobileMe’s Mail is not a competitor to GMail, Yahoo Mail, or even Hotmail. MobileMe is meant to enhance the experience of owning multiple Apple devices: Macs, iPhones, iPads, and iPods, by providing a synchronization service between them for things that matter. The web interface to MobileMe is an addition to the MobileMe service, not the service itself.