In the software world, there are two primary systems for organizing information: folders and tags (also called labels).  The difference is most noticeable as people transition from Outlook to Gmail, but I want to talk about the more general uses of tags and folders.  Specifically, I’m going to go over when tags aren’t the right way to organize things.

First, let’s make sure everyone understands the difference between the two.  While there are a lot of superficial differences, the thing that really separates the two is that a file can only exist in one folder, but a file can be associated with multiple tags.  For example, let’s say you just bought an airplane ticket for your mom and you want to file away your confirmation email.  You often classify your emails as “family” or “travel”.  In this case, the email applies to both categories, but with folders you have to pick one.  With tags, you can assign the email to both categories.

People often think there are other differences, such as a hierarchical structure (i.e. subfolders), but both folders and tags can support that (it’s just less common with tags).

In the example I gave above (organizing emails) the tag system is obviously much better than the folder system because it doesn’t make you choose between the two different categories. The problem is that people seem to think that tags are always better, but that’s just not the case.  They are generally better, but sometimes organization requires mutual exclusivity between different categories.

I’ll use the Less Annoying Software CRM as an example of when tags just aren’t as useful.  When you’re dealing with sales leads, you generally want each lead to belong to exactly one status.  If someone fills out a contact form on your website, you would categorize them as a “lead”.  If you call them up and decide that they aren’t ready to buy, you would categorize them as “inactive”.  “Inactive” and “Lead” are mutually exclusive.  It’s one or the other, never both.  For this reason, we use a folder system (we call them “statuses”, but they’re basically the same thing as folders).

If you’re a hardcore tag supporter, you might wonder why it matters that the two statuses can’t both exist at the same time.  The best reason I can come up with is because it messes up reporting.   If I want to see a report of the current status (or folder) of all my leads, I can’t have overlap between inactive and active leads.

This is obviously a somewhat obscure example, but there are all kinds of different scenarios where folders actually do make sense.   When I finish writing this blog post in Google Docs, I’m going to want to move it from the “draft” folder to the “published” folder.  This means adding it to one and then removing it from the other, because Google uses tags for just about everything (they call them folders in Google Docs, but they’re really tags).

So enough rambling.  What’s the point?  The point is that in many cases folders are outdated and ineffective, but don’t disregard them entirely.  They still serve a purpose, and if you find yourself designing software, I suggest you put some thought in before going with tags just because that’s what’s cool right now.

Over the past few years in graduate school, I've somehow ended up managing the computing resources for our lab. Prior to joining the lab, I didn't have any real system administration experience, so while I've learned a lot over these years, I'm still learning plenty whenever something breaks or we decided to upgrade our resources. As you might expect, there's a huge amount of complexity in managing a linux cluster, and one often has to just trust various suggestions at face value without fully understanding the underlying logic. Regardless, I always try to make a point of having at least some idea of why various best practices have been established. The following is a story about my not doing that, and the minor annoyance that it's causing.

As you may have heard, Apple is expected to announce a new tablet tomorrow (January 27th).  All the information so far has been based on rumors, but everyone seems pretty sure that the tablet (or whatever they end up calling it) will basically look like a larger ipod touch and it will work like a netbook.  Several people have asked me why it's getting so much hype, so I thought I'd post my thoughts here.

I spend too much time talking about bad software.  It's easy to point out the wrong way to do things, but that's not very useful unless I also give examples of the right ways.  With that in mind, I'm going to try to periodically post about software that I really like.  These posts will be a blend between concepts (what makes good software) and practice (you should use this specific software for this reason).  Today, I'm talking about my favorite image editor.

As promised, I'm back with more analogies to chemistry and physics. If you enjoyed yesterday's post, you're in for a treat. If not, this post is completely different.

Technology is cooperative
It's no big revelation that change is difficult but important. We've talked about it before, and we've talked about how trying to keep up with the bleeding technology front at all times is probably not worth it. But there's another aspect of the upgrade path that is easier to overlook if you're not careful. To set the analogic* stage, I briefly want to talk about the concept of cooperativity. Basically, cooperativity is the idea that an initial event can lower the activation barrier for following events (see yesterday's post if that doesn't make any sense). The classic example of this in biology is oxygen binding to hemoglobin. Hemoglobin is the main oxygen transporter in your blood carrying oxygen from your lungs to the rest of your body. Each hemoglobin molecule can bind up to 4 oxygens, and there is some energy barrier to bind each of them. The cool thing is that once the first oxygen binds, it changes the shape of hemoglobin such that it's easier to bind the second, which makes it easier to bind the third, which makes it easier to bind the fourth. The result is that hemoglobin almost always has either zero or four oxygen molecules bound. This is an important property for hemoglobin because it lets it pick up oxygen easily in the lungs, and then dump it all in the tissues of the body. That's cooperativity.

Technology has a tendency to be cooperative in similar ways, both by it's nature and by design. Apple originally released the iPod to try and get people to buy their computers; the iPod experience was much better on a mac, encouraging users of one to get one (many people would either have both or neither). Now this plan didn't work out exactly like they planned (plenty of windows users with iPods, obviously), but that was the original idea. Microsoft has done similar things with Office and Windows. Until recently, Outlook didn't even exist on the Mac, and Office in general was slow and close to useless. As such, if you wanted to use Outlook or the rest of Office, it was much easier to use Windows. To finish the software giant triumvirate, anyone who has ever used Google's services knows they exhibit extreme cooperativity. Gmail, Google Calendar, Google Docs, and Android are all pretty great services on their own, but using all of them together is much more useful than any individual product by itself.

So cooperativity is everywhere in technology, and there's nothing necessarily wrong with that. As I've said, the Google suite of software is much more useful all together than it possibly could be in separate chunks. The extreme success of Office (which I still think is a good product, despite what Tyler would likely say) is largely due to similar properties. The problem with cooperativity, however, is that it can make upgrading even harder. If all of your software is codependent, you basically have to simultaneously overcome the activation barriers for all of them at once. I think that a large amount of the backlash against Vista had to do with the fact that most people who upgraded also got a new version of Office and weren't able to change so many things at once.

As for how to combat the problems of cooperativity in technology, I think open (adhered to) standards is the answer, and are worth sticking to whenever possible. The increased trending of web software is a huge plus. For most online software, any browser provides a valid platform for the software. If you decide to switch to a mac, or upgrade to windows 7, or just use your smart phone full time, you can still get basically the same gmail (yahoo mail, hotmail, etc) experience. That's possible because HTML is a well-defined standard (even if Microsoft doesn't care). Similarly, the fact that I can export my gmail contacts as a csv, or my calendar events as ical, or my docs as rtf, means I can decouple the services if I want to. If a better document editor comes along, it's not hard to switch just that (without upending everything else I use through Google).

Cooperativity is great when used right, but don't let it be used to lock you in to a specific combination of software without a clear upgrade path.

*not a real word