Am I Supposed To Be Excited About Streaming….Today?

It’s all about streaming these days. Netflix, Amazon Cloud Player, Spotify, Pandora. Everyone wants you to stream and, I agree, that the model makes a lot of sense. The problem is that other parts of the infrastructure have not currently caught up to the end point technologies.

I’ve been a long time proponent of subscription type services like Netflix and Zune that give you access to large libraries and the ability to stream them instead of storing them locally. Recently, I have been bugged to the extreme with buffering and drop outs during streaming. The problem is simple. Depending on your connectivity, you may not have the available bandwidth to support the stream. That’s it.

The reason that I’ve been so frustrated about it recently is that we (collective we) overcame this issue over the last decade by innovating chipset speeds up to the intended usages like full screen, full motion video playback. It works. I never have to endure a hiccup of video stream or dropout of audio during playback save for the occasional defective disc. With streaming, I can rarely make it through a Netflix movie or Amazon Cloud Player song without a buffer or drop.

I am fully aware that my results are dependent on my connection speed, but if it’s happening to me in multiple environments, I have to assume that it’s happening to other as well. I asked my early 20’s daughter about her Netflix watching experiences online from school. She said, “yes, I get drops and buffers.” I’m not sure why that’s OK.

I am super excited about the future of streaming, but I just think that today’s hype outweighs the realities of the experience.

Here’s another example. I use a Windows Phone with with a Zune Pass music subscription service. Zune/Zune Pass allows for both local music and streamed music. Microsoft, by it’s actions in the marketplace, is obviously nudging users to a streaming only model. Xbox 360 integrates Zune/Zune Pass, but does not really consume local music (e.g.: your PC based music collection). So, even though I had a bunch of music and podcasts local on my phone, I decided to live ‘streaming only’ for a week or so. This meant in my car driving, walking around town, at the gym, etc. I was amazed at how many times I entered dead zones where my music/podcast just die and tried to buffer. Buffer…buffer….buffer. Sometimes it grabbed the stream and sometimes it didn’t. I ask again, “am I supposed to be OK with this?”

Now, in my opinion, the solution for this situation at least as far as my music goes is to marry the local music with the streamed music. Which, BTW, is exactly what Zune/Zune Pass does, but this is not a commercial for that service so I’ll let that one go for now.

I look forward to the future and I’ll embrace the evolution of today, but get real folks, it’s not great….it’s just acceptable right now.

Infrastructure note – I understand that streaming is dependent on bandwidth. I neither think that I have exceptionally great nor exceptionally poor bandwidth so I feel that I’m a reasonable test case for this rant. I have a home cable connection of ~20Mbps down and ~5Mbps up. OK, but not great. I have a work connection of 3Mbps synchronous. I would like it to be more, but the economic realities of business data pricing and office location specifics won’t allow it. I own a 3G Samsung Focus. Typical 3G speed.

Calendar Print Assistant–Object Reference Error

Here is a weird edge case related to a null reference error thrown by the Outlook Calendar Print Assistant. If you stumbled upon this page there is a very good chance that you do not use Outlook, let alone the very application specific Calendar Print Assistant free tool from Microsoft. On the other hand, if you arrived here via a search engine then there is a very good chance that you do use both of the previously mentioned applications and are receiving an error message.

First, make sure that you are using the most current application for CPA. It’s not something that gets rev’d all-to-often, but there was an important update in mid-2010 to add support for 32-bit Outlook 2010. That information is readily available on the web from Microsoft. The reason that I am writing this post is that I eventually was able to track down the simple fix information, but I didn’t find it on Microsoft or Microsoft Answers. It came from those shady looking, ad driven support sites. I hesitate going to most of them because 1) information is hard to separate from ads and 2) I never feel like support is the primary purpose of the site. I figure that I will republish the fix on my site to get it out there and away from the junk support sites.

After installing Calendar Print Assistant and running it, I received an error that Outlook could not connect to the mailbox. The more specific error was a standard, “Object reference not set to an instance of an object.” Not exactly a lot to go on.

Cut to the chase. In addition to the executable standalone application there is also an Outlook dll Add-in installed to manage information transfer between Outlook and the Calendar Print Assistant application. In my situation, this dll had disabled itself due to the error condition or preceding it. I’m not sure which.

From within Outlook,

• Outlook 2007 – Tools | Trust Center | Add-ins
• Outlook 2010 – File | Options | Add-ins

Look for a Calendar Print Assistant Add-in. It may be in the Disabled Add-ins section. Enable it. If it’s not there you can add it by clicking add and navigating to

• c:\program files\microsoft office\office12\cpaoaddin.dll

Once it’s enabled, restart Outlook and then try Calendar Print Assistant.

Hopefully it works now.

Stamps.com Printing Offset Fix

I run across this situation semi-frequently and, because enough time lapses between events, forget the location of the solution nearly every time.

Every so often Stamps.com offsets too far to the left when printing envelopes. The fix is not contained in the print configuration wizard. The setting is controlled via a hidden dialog from within either the Print Postage/Print Sample or the Print Configuration Wizard.

The dialog can be accessed by holding down both the Ctrl + Shift keys while clicking on the Configure button. From the Print Configuration screen it can also be accessed by holding down both the Ctrl + Shift key and clicking on the image above the words “Short Edge.” My particular left offset error is always solved by negating the L90 checkbox. By negating I mean turn it on if it’s off and turn it off if it’s on.

There are other options on this hidden screen so obviously other issue abound and this dialog is intended as a tech support only screen. Personally I think that as a matter of UI design the whole issue of a hidden support only screen is sort of amateur and harkens back to a less mature day of software development. Alternately, the support options should just be built into the appropriate menu items or within a support screen sort of like Sys Info is for Windows apps in the About screens.

I spent 30 minutes on the phone (25 of them on hold) just to have the tech walk me through a procedure that, in theory, I already knew. It doesn’t seem like a great use of time. I searched the Stamps.com online knowledge base prior to calling and didn’t find any reference to the issue or the hidden fix.

2011 Year End Recap

Merry Christmas to our dear family and friends!

2011 has been a quiet year of seemingly little change for our crew, but like the small bud on the branch when reviewed in time lapse, there is remarkable evidence of blossoming and growth all around us!

Peg is in her third year as Principal of Kinoshita Elementary School in San Juan Capistrano and is hitting her stride. Andy is in his ninth year as VP of Information Technology for Highridge Costa Investors. Don’t let the name fool you… it changes with the seasons, but remains the same company and has managed to stabilize well throughout the tough economy.

Andy and Peg enjoyed the first half of the year as empty nesters and took frequent short trips up the coast and through the wine country. Of course, the annual trip to Maui was a “must” and a trip to Jamaica is just months away.

However, the most significant change for Andy and Peg is that we have made Saddleback Rancho Capistrano our church home. Together, we have been welcomed into a weekly small group and are increasing our participation in church and community events. We feel blessed to be able to use our gifts to help others and it has, in turn, blessed our relationship.

Now for the good stuff… We couldn’t be more impressed with the way in which our girls have become responsible young adults!

Kyra, age 22, graduated in May from SDSU with a double major in Public Relations and Political Science. Immediately after graduation, she flew to Ramstein Air Force base in Germany where she worked for 2 1/2 months as a camp counselor with the US Air Force children. The cool part is that she traveled throughout Europe every weekend and visited Berlin, Munich, Trier, Heidelberg, Paris, Salzburg, Interlaken, Brussels, Amsterdam, and Luxembourg. She moved home to Dana Point in August and landed a full-time position as Marketing and Sales Assistant at EMS Safety Services in San Clemente. Not a bad year for our Kyra!

Kacey, age 19, is in her second year at CSUF. She moved from the dorms to an apartment near campus with two roommates. She recently changed her major to Psychology (just like her mama!) with a minor in Art (NOT like her mama). She joined Zeta Tau Alpha sorority and has quite a busy social calendar. Her parents are so impressed with her maturity and independence. Our baby’s all grown up!

Both girls have very polite and kind boyfriends named Michael, which makes things easy for us to remember.

We want to thank you for being a source of support and friendship to our family throughout the years. As such, we ask for your prayers for Peg’s dad, Fred, as he struggles with esophageal cancer.

We wish you an exceptional Christmas season and a blessed 2012!

Love, Andy, Peg, Kyra, and Kacey Baerst

Complete set of 2011 Photos

How to Merge in Traffic

I was unaware that this needed to be clarified, but based upon repeated experience I think that it’s worth my effort.

Let’s say that two lanes of traffic are merging down to one due to a lane ending or maybe some roadwork. Here’s the process.

One car from one lane and then one car from other lane. That’s it.

Here are some unacceptable, but very common mistakes.

• Come hell or high water I am not going to let you merge. I am going to ride the car’s bumper in front of me so closely that you couldn’t slip a piece of paper between the two vehicles.
• Repeatedly flip off the correctly merging vehicle and mouth unspeakable words at them through your raised pickup truck’s windshield.
• As the merger, pull the I’m going to buck the merging protocol and ride the cone or shoulder lane for about six more car lengths until I butt my vehicle’s nose in and then refuse to look at the person who I’ve just cut off.
• Throw your hands up in disgust at the valid merger to show how this person has now deeply offended you by moving into a position one car length ahead of you even though you are both now sitting stopped in traffic. 

Scale Computing Drive Failures

In March 2011 my company purchased a 12-disk, 6 TB disk array from Scale Computing from PCMall (not my usual vendor). This is known as the S2 product line. To date we have experienced 5 individual drive failures. Each SATA drive is 1TB for a total raw capacity of 12TB. The 6TB is the usable storage after striping/data protection.

We had one drive fail in each month of June, September, October, and November. Each time a drive fails the available storage reduces temporarily by .5TB until a replacement drive arrives via ground shipping. Ground shipping usually takes approx 5 days based on shipping location to my location near Los Angeles. That’s a long time to be without available storage.

A drive is supposed to automatically remove itself from the pool when it fails, but the most recent drive failure hung during the removal process and took down all of the CIFS (NetBIOS) shares until the Scale Computing support staff manually intervened. The OS is a Linux implementation.

I find the Scale Computing support and engineering staff to be very knowledgeable and helpful, and I like the array when it is functioning properly, but I feel that the drive failure rate is unacceptable and leads me to not have confidence in the array as a primary storage solution.

My implementation is a combination of CIFS and iSCSI. It is meant to handle daily file sharing (CIFS) and virtual machine hosting (iSCSI). I do not believe that my implementation is anything unreasonable or out of the ordinary.

I’m not sure where I will end up with this unit, but as of now with so many affordable storage solutions available I cannot recommend this product with good conscience. 

Update 12/5/11 – I received a call from Doug Hooton (VP of Operations) at Scale Computing to talk about my drive failure issues. Doug seemed up front and acknowledged that the my failure rate seems excessive. We discussed this specific issue and my overall opinion of the Scale Computing array. He offered to ship me a ready spare to mitigate any down time issues on the chance that I have any further failures. Their engineers are going to perform additional forensics into the array’s log files.

The offer of the ready spare was appreciated and accepted. I had looked into a spare drive upon earlier failures, but didn’t pursue it for two reasons. 1) The drives individually are not inexpensive and 2) I already pay a maintenance fee for the array and feel that replacement drive should be provided in a timely manner. As mentioned above, ground shipping can be slow and I was planning on addressing that issue in due time, but I didn’t really have a good case for expedited shipping given that my maintenance agreement already spelled out ground shipping. I think that that is an incorrect choice on Scale Computing’s part and that they should, by default, overnight replacement drives immediately upon confirmation of a drive’s failure. Doing so lets the customer know that array stability is a top priority to Scale. The issue of additional shipping cost should be a moot point because, if what Doug Hooton told me is correct, failure rates are not as high as I’ve been experiencing. If that information is not correct and overnight shipping costs start eating into Scale Computing’s profits then they have a bigger problem on their hands related to flawed equipment.

I appreciate the call from Doug and look forward to hearing what the Scale Computing engineers find. I’ll update when I hear something back or in two weeks if I hear nothing.

Update 12/28/2011 – I was contacted last week by Scale Computing tech support. The tech was very helpful and ran a series of health test that showed that my array was configured correctly with ample throughput. Aside from a bunch of common sense suggestions there was nothing new. Which is kind of what I suspected.

So now I am kind of left where I was before. I like the array and it really is working well for my iSCSI connections. The CIFS stuff is just SAMBA and a little weird with its implementation for ACLs if you are used to how it’s done in Windows, but that’s another issue. There have been no drive failures since the one that initiated this blog post. All I can do now is keep a hawkish eye on it.

Mobile Phone Battery Issue

I currently use a Samsung Focus as my primary mobile device. The Samsung Focus runs the Windows Phone operating system. About a month ago there was an OS upgrade from 7 to 7.5 (aka Mango). Around that time I became aware that I was burning through the battery charge much faster than I remembered prior to the Mango upgrade. I researched a little and found other mentions of increased power usage after a Mango upgrade so I noted it and figured that it would be addressed with a future service patch.

I started using the Samsung Focus in Q4 2010 when it was first released. At the time one of the selling points of the phone was the ability to add additional memory in the form of a MicroSD card. In reality, a lot of controversy regarding the addition of more memory ensued because of the lack of guidance and clear specifications about what type (and more specifically what brand) of memory was required. It seemed crazy that there should be any challenges with the ease and abundance of MicroSD cards, but that became the reality and my solution was to just table the issue until others, hopefully the vendors, flushed out the issues and provided clear guidance.

One of the challenges here is assigning responsibility. There are numerous vendors in this distribution channel. There is the Operating System vendor (Microsoft), the handset manufacturer (Samsung), the handset retail vendor and service provider (AT&T), and the manufacturer of the MicroSD memory (various options).

Based on the fact that I am attempting to phase out my Zune HD from daily service and focus exclusively on my mobile, I decided (late November 2011) that I would now add 16GB or 32GB to the Focus to hold an acceptable amount of media content. The default 8GB was woefully inadequate for the amount of podcasts, video, and audio that I want in an always available format. I picked up a 16GB MicroSD card for pocket change after verifying that it was one of the cards that was reported as successful and added it to the phone’s memory pool.

Additional memory on the Samsung Focus behaves a little differently than we are used to with PCs or cameras. With PCs or cameras, the memory is additional memory on top of the devices base memory and can therefore be added and removed at will. The additional memory on the Focus integrates with the manufacturer’s memory and increases the base pool, but can never be removed or changed without a full hard reset back to the “factory new” configuration. This is where I found myself when adding the 16GB MicroSD card. It’s a little traumatic to fully wipe your mobile phone, but if you’ve structured things correctly, it’s not the end of the world.

So here’s the crux of this post. I added the 16GB of RAM, performed the hard reset, and reconfigured my basic services including:

• Live (Hotmail) email account
• Exchange email account
• Facebook app
• Flickr app
• WiFi connectivity

Guess what? My battery life is now outstanding again. Prior to the hard reset I was burning through battery in about 10 hours. That was about 3pm given my schedule. I’d have to charge it in the car on the way home if I had any hope of using it in the evening. After the hard reset I still have about 70% battery life available at the end of the day after about 16 hours.

I was floored when I realized this. I originally ascribed the decrease in battery performance to the Mango upgrade, but the hard reset with the associated app wipe clearly points to an app issue. I now suspect that one of the installed apps was aggressively using power via wireless updates or CPU interaction.

My next step is to add back apps in a controlled manner to identify which one causes a dramatic decrease in battery life. Moral – Don’t assume and don’t blame without the facts to support your assertion. Try a hard reset. It’s not the most pleasant activity, but you might be pleased with the results.

I will update this post when I track down the power consumption culprit.