Looking at storage projects like Walrus, I increasingly feel that they shouldn't be evaluated with a short-term mindset. Storage infrastructure fundamentally isn't driven by hype; it's about the frequency of being repeatedly called upon. It's okay if there's no discussion today; it might not be reliable even if the next trend arrives—what matters most is whether it can truly be put to use.

So I now have a habit: when evaluating, I consider three time horizons for myself: What will happen in one month? Six months? What will it become after a year?

Let's start with the one-month cycle. During this period, the most realistic development isn't a killer app suddenly appearing, but developers starting to test the waters on a small scale. What does testing look like? It’s not about grand product launches, but about someone beginning to integrate it into existing applications, starting with the smallest features—like migrating data of large objects from makeshift solutions to see if stability, access control, and referencing methods can run smoothly.

I pay more attention to whether the ecosystem starts to see more small actions like "conveniently storing blobs in the storage layer." These accumulated small actions are more indicative than a big news headline.

Moving to the six-month mark. Six months is a watershed; at this point, it reveals who has truly stuck around. Many projects look great early on, but once incentives stop, their true nature is exposed. So, my focus after six months becomes very specific:

First, whether genuine paid usage has appeared. It doesn't need to be large-scale, but there should be ongoing actual consumption. Second, the node operators. Can those running nodes provide stable services? Will they start to drop out as profits decline?

The one-year period is about whether it has become an "invisible standard" within a certain ecosystem. Not necessarily everyone discussing it, but quite the opposite—good infrastructure should be naturally called upon, just like no one specifically discusses TCP/IP. If after a year developers reflexively use it for data storage when building applications, then it has truly become alive.