Don's space

The most robust support for collaborative work comes from software development. For years programmers working together on large projects have needed to synchronize their work. Out of this need UNIX tools were development to manage source code repositories. These tools evolved from RCS to CVS and now to the latest SVN (Subversion). These all do a wonderful job of managing text files as source code is usually saved as plain ASCII text. A number of fluffy tools have added GUI features on top of the basic functionalities but they all focus on versioning plain text. One of the most important features for very large projects has been support for multiple programmers working on the same file at the same time. This model is an improvement over the older check-in, check-out process that used locking to ensure only one person could make changes at a time.

LaTeX is a fantastic system for writing natural language documents. It has the advantage of storing the input as plain ASCII text and keeping this input separate from the beautifully rendered document layouts it can generate. Since LaTeX uses plain text it is well supported by the source code version management systems. A great process for collaborative document writing can be built on top of the programmer's toolset by using SVN, LaTeX and a text editor such as Emacs, Vi, Netbeans or Eclipse.

While I am a big fan of LaTeX and the UNIX programmer's toolset I need to enlist the cooperation of colleagues who are engrained in the Microsoft Word and Office toolset. In addition I have to admit that Word has some enjoyable strength over LaTeX and Emacs. I really love the Calibri font and how nicely it and other Type I/TrueType/OpenType (non-bitmap) fonts are rendered in Vista on an LCD display. The live spelling and grammar check is also a comforting feature when used with a sufficiently fast CPU. UNIX diff is a handy tool, particularly when combined with grep; however Word's Track Changes report is very well done when it works. Bibliography management is a must have for any writer of non-fiction and BibTeX does a great job. Microsoft Office lacks support for this but EndNote is popular for filling the gap.

The LaTeX-based electronic writing process comes out on top for consistency with the programming toolset and support for collaboration through version control. The Word based electronic writing process has a few niceties for writing the text and it has the big advantage of the larger user base. This is important if I want to collaborate with others who are not comfortable in the UNIX programming toolset. The problem to solve then is how to setup a strong collaborative process based on Word.

Microsoft offers a few solutions. Microsoft Office Groove appears to be an interesting little product at first glance. It advertises that it is designed for exactly this purpose. It is also part of the Office suite so it should not be too difficult to enlist participation from collaborators who are already using Office. There are a few obvious weaknesses however. First there is not a Mac version of Groove as of yet. Word on Mac is popular among the academic community I want to work with so this could be a big limiter for getting everyone on the same process. Second Groove wants to run whenever my client machine is running. Obviously this will create a performance hit in terms of memory. It also seems likely to be very chatty on the network and running yet another network service creates security issues. On the positive side it advertises support for disconnected use. I consider sandboxing a minimum requirement rather than a bonus feature. It also has a presence system built in. It is unclear if this will be an annoying redundancy to instant messaging or if it will be a useful feature. One predictable advantage might be to encourage collaborators to made presence visible more often as it will be a first-class feature rather than depending on an external messaging program to be started manually.

A second Microsoft solution is Office Live. This is essentially a web site with specific support for sharing Word documents. It has versioning and as a web site does not lock out the Mac users. In going the web site way it is unclear if Office Live has any significant advantage over other basic file sharing services.

It is also worth considering if a Wiki/Instant Messaging solution could be used with some very simple agreements among the collaborators. MediaWiki and Pidgin are popular free open-source systems that provide all of the necessary functionality. Being open-ended however they might be underutilized or cause organization issues among a team. Everyone would have to agree to consistently run the Instant Messaging client manually when working. Everyone would also have to agree to keep the Wiki up-to-date manually. The vision would be to manage the Word document as a versioned attachment to a Wiki page. The Wiki attachment approach would require the check-in/check-out model rather than the multiple editors on the same document at the same time model. This would be quite restrictive, particularly on large documents around deadline times. The Office Live solution seems to suffer from this same problem.

Google Docs is another potential candidate for collaborative document writing. Even with Word's bloat however it provides a much richer experience for the writer than the bare-bones formatting and rendering provided by Google Docs. It is unclear how well Google Docs supports disconnected sandboxes for laptop work. It is also unclear if bibliography management is supported by Google Docs. Without bibliography management it would not be a feasible alternative.

From this preliminary assessment then Groove seems to have the most potential in terms of functionality and if it support multiple editors working on the same document at the same time it would have a big advantage over the simpler Wiki/Pidgin toolset. I'll give Groove a try and see how it works out. At the very least it should reveal some features that might be added to the Wiki/Pidgin toolset for a more open approach later on. Please post a comment if you have any opinions on these tools or suggestions for alternative tools to support collaborative document writing.

I recently purchased a Garmin GPSMAP 60CSx GPS unit. As personal GPS units go this is at least a middle of the road unit. I purchased it with some high expectations. Unfortunately I hit some obvious problems with the most trivial functionality; waypoint management. The customer support communications related to waypoint management are posted below. In future posts I hope to cover the issues related to creating custom maps from USGS, National Park Service and various state GIS datasets with the ultimate goal of developing a custom GIS toolkit for common use cases.

I sent the following message to Garmin support:

I have found the following problems when using Garmin MapSource 6.13.7 to create Waypoints and send them to my device:

1. When I assign a symbol to the waypoint in MapSource it appears as expected. When I transfer the waypoint to the device a blue flag symbol is shown instead of the one I selected. This happens for some waypoints but not all, depending on the symbol assigned. It seems that the symbol sets are misaligned as the GPSMAP 60CSx has symbols that are not in MapSource and vice versa.
2. The comments I assign to Waypoints in MapSource are truncated when shown in the GPSMAP 60CSx. My expectation is that MapSource behave identically to the unit, if the contents are too long for the unit I should be notified in MapSource. Under the current conditions I have to double check everything for fear that the Waypoints I create in MapSource will be incomplete when I need the data in the field with only the GPSMAP 60CSx. 3. In general the process of sending to and from the unit is problematic. It would be much more useful to be able to synchronize with the unit so that changes from the field are brought in to MapSource and changes from MapSource are sent to the unit.

I received the following response:

Thank you for contacting Garmin International.

1. You are correct, not all symbols are available in both programs. When using a symbol on the PC that is not on the unit, it will transfer over as a Blue Flag. At this time, there is not an update for the unit to load the newer symbols.
2. That is correct; you can store larger comments in MapSource than you can on the handheld. You are limited to 29 characters on the handheld.
3. I will pass this along to see if changes can be made.

While not the answer I was hoping for the support response was complete and indicated that my concerns would be "passed along" which is comforting. It seemed that MapSource would be the obvious place to start as I work on constructing my personal GIS / GPS toolset. MapSource is provided by the manufacturer of the unit and one might expect that it would be the most compatible. As it fails in even this simple regard there is a gap in the toolset for waypoint management. Almost every GIS tool that interfaces with a GPS will handle waypoint upload and download. I have yet to find one that fully synchronizes accounting for the character limits and symbol indexes on the unit itself. If you have ideas for GPS waypoint management please post a comment.

Desired Functional Requirements for GPS Waypoint Management:

• Synchronize Waypoints: If the GPS unit holds a modification date of some sort it would be ideal to have the PC and the unit synchronizes changes bi-directionally. If the GPS unit does not have a modification date for local changes and conscious push or pull decision may be required of the user.
• Account for Character Limitations: Waypoints include both a name and a comment. This limitation should be accounted for on the PC so that I user does not experience unexpected data loss.
• Account for Symbols: Waypoints include a symbol. The symbols should be identical between the PC index and the unit.

The following software programs have waypoint management features but none of them meet these requirements:

• Garmin MapSource 6.13.7
• Google Earth Commercial Version
• NASA World Wind with the GPS add-on.

The Singapore Grand Prix will be the "first-ever night race in F1 history and also the first street race in Asia (http://www.f1singapore.com/news/Singapore_promise_to_stage_a_breathtaking_Formula_1_race.shtml#news)." This could be the most beautiful racing event ever; perhaps the killer-app for HD-TV and surround sound. It is scheduled for Sunday September 28, 2008 at 8:00 PM local time (http://www.formula1.com/races/in_detail/singapore_801/circuit_diagram.html). That is Sunday at 8:00 AM in Philadelphia, PA. For other local times see http://www.timeanddate.com/worldclock/fixedtime.html?month=9&day=28&year=2008&hour=20&min=0&sec=0&p1=236. It seems that SPEED TV has exclusive television rights for Formula 1 in the United States so I assume that they will cover the event. A video of lap in 3D made with Google Earth is online at http://www.f1fanatic.co.uk/2008/02/27/video-3d-lap-of-singapore-grand-prix-track/ as is the ".kmz" file. A seat on the Ferris wheel would be amazing.

DuPont has some very enlightened policies on ergonomics and safety. Some of the most useful are their recommendations on traveling with a laptop. I remember the following two guidelines:

1. Use a bag that distributes weight evenly, such as a backpack.
2. Use a bag that does not advertise the fact that you are carrying a laptop.

As my new commute includes a one mile walk from my house to the train station followed by some moderate city walking upon arrival, and a return trip each day, I have come to appreciate the first guideline. When I purchased a new laptop Office Depot threw in a courier style bag for $0.90. The Case Logic Canvas Messenger Bag (http://www.caselogic.com/canvas_messenger_bag/product_detail/index.cfm?modelid=96836) has a nice feel to it and it holds a laptop just fine. Unfortunately even after significant effort I could not find a way to walk with it comfortably. The weight was not even so my shoulder or back always hurt by the end of my walk and it either bounced on my back or rubbed against my side when walking. It held everything I needed fairly well and it had good protection from the rain, but it just isn't useful if I can't walk with it comfortably.

Since I am a student again it is appropriate that I return to the staple of academic life, the backpack. Although almost every university student seems to have a laptop these days the bookstore did not carry any backpacks that purposely accommodated one. After some shopping around I found that backpacks seem to fall into three categories: no accommodation for a laptop, laptop sleeve in the main compartment, and a dedicated laptop compartment. My old backpack is in the first category. Over the years debris collected in the main compartment such as dirt and sand. I feared that my USB ports and fan outtake would quickly fill with such debris if I used the classic backpack. The sleeve design does not seem to offer much improvement as the equipment is still left to float around with everything else that has collected in there. The dedicated compartment is clearly the way to go. The best of these use a top loading zipper as well which is great for traveling on the train. I can lift the laptop out and slide it in all without needing to get the bag up off the floor at my feet. I settled on the North Face Surge Computer Backpack (http://www.ems.com/catalog/product_detail_square.jsp?PRODUCT%3C%3Eprd_id=845524442592283&FOLDER%3C%3Efolder_id=2534374302876305&bmUID=1199823681908) and I am very impressed with it so far. The front pocket with the pen holders and such is a bit of a hassle to access but other than that it is perfect. The black color is also nice enough that it doesn't look loudly out of place in business casual dress. The bag is comfortable and durable. The only thing wrong with it is the high price but I have to accept that you get what you pay for with such products. The Case Log Canvas Messenger Bag was $0.90 wasted while the North Face Surge Computer Backpack was money well spent.

It is curious that the standard bag sold with laptops tends to be a branded; single should strap style which violates both of the two guidelines for travel. With laptops as standard equipment these days, hopefully the backpack will reinvent itself as appropriate business attire so that business travelers don't end up lopsided in their old age.

Drexel University offers a 10% discount on the purchase of SEPTA passes. With a Zone 3 TrailPass costing $142.50 this seems like a nice savings of $14.25 and possibly worth the hassle of having to pre-order the pass online and pick it up at the student center. Unfortunately the passes must be ordered two months in advance. Anyone with access to a savings account should quickly recognize the opportunity cost of such an early prepayment cost. Today Citibank is currently savings accounts with an interest rate of 3.75%. Over the two month period, with interest paid monthly, the money lost to paying for the pass also loses $4.81 interest in the first month and $4.99 interest in the second month.

Therefore the actual savings from the program is $4.45 if the money used for the purchase could have been sitting in a Citibank savings account rather than prepaid to SEPTA. Since the solver add-in is missing from my copy of Excel I played with the interest rate manually and found that a rate of 5.41% would be the break even rate for comparing a savings account to the SEPTA prepayment.

TrailPasses for the cheaper zones would yield even less savings than the $4.45 shown above, making the hassle of ordering the pass online and then picking it up at the student center even less worthwhile. Such savings would be lost very quickly if the student uses a credit card to make the purchase and does not pay the credit card in full each month. Any interest paid on the prepayment cost would quickly exceed the meager savings.

In conclusion, a Zone 3 TrailPass ordered two months in advance with a 10% discount on the prepayment price is actually a 3.12% overall discount versus buying the pass immediately and having saved the money in a Citibank saving account over the same two month period. If a credit card is used to make the purchase and interest is incurred then the scenario would result in a net loss versus not using the program at all.