So yes, here they are. Google has all the search data--the web cached, your searches stored, and now it's added your recommended search results. Sounds a bit like Jawaya, except we don't have nearly the data they do. Or the market cap :)
But all is not lost. As we get closer to launch, I'm getting more confident that what we're doing is deeper and more valuable to people. I'm not sure we can stay ahead of the goog, but I do know we can serve people very well with what we're doing.
I don't know how far they can go; when you start sharing people's search results, you better be very good about privacy, and Google has its issues there.
So far, this is just a toe in the water. But you can bet they've got team of people working to go deeper into making search and sharing a fundamental part of their model--for those who want it.
And we, the scrappy, unfunded startup, we will keep pushing that stone up the hill. We're a tenacious lot and we'll continue …
Last year I worked feverishly on a product to help me and you focus more and cut out the things that distract us. I called it Focus. I never released it because I ran into a few tech glitches in C++ and decided to take a break. That gave me time to work on Buckets, through which I discovered the joy of Jawaya. Buckets will be back--it's a fluid way of organizing your thoughts quickly, saving them to buckets, and sharing the buckets.
Seth Godin blogged about distractions today. He basically says that if you're reading an article or playing a game, you're getting satisfaction but you're not working. You're not making. You need to make. Be a maker.
It's not a new thought, but it's an important one to remember.
Close the browser. If you need it for work, only use it for work. The internet never sleeps. It's always there, and will be the same big black-hole time-sink when you get back. And you will, of course, come back.
Results are polluted with SEO-targeted sites, yes. Content farms create mediocre content that gets significant play on Google, yes. Google has been creating "personalized" versions of results that end up sending us to Page 2 more than ever before, yes.
But it's not broken. It does what it intends to do.
Some things it does very well:
allows searching over multiple types of media. News is different from Aunt Mae's website, blogs are different from video, etc, and the results in each are reasonably good. Google Maps as directory. I use Google Maps as a way to get basic contact info for businesses all the time. It's my number 1 local app, and yes I have foursquare, which isn't as much about finding basic place info. Autocomplete. Lawsuits aside, Google's autocomplete is really amazing. I wonder if autocomplete is effected by personalization? Commerce. If I want to buy something, I search for it on Google, which is the entryway to a ton of online and offline stores…
I was able to speed up the load time of the Jawaya plugin by a whopping gagillion seconds--something dramatic. I was able to achieve this stunning bit of engineering mastery by simply changing a value in the manifest so it would load before the page load. Clever.
When I realized later that there was something awfully wrong, I had forgotten what I had done. The bug wasn't obvious at first--everything appeared to be working, but later I found the data was simply not getting stored because the extension was losing its authentication, or simply not getting it in the first place. I hadn't written that code, so I didn't know how to track it down immediately.
Frustrated, I took the evening off.
And this morning, I started by tracking it down, and realized a few things, including that if you want to grab a value from a page, you damn well better try to do it after the page loads, not before. I look forward to bringing on smarter people than I.
I like Wordpress and will continue using it for some blogs.
But I like Disqus more, and it only works with self-hosted Wordpress, which I'm not willing to do because of the maintenance and spam management.
Which means it's become an important communications tool. I've had 5 people tell me they'd post here on the blog if they could simply use Disqus to do it. Over the top? Well, no, people are busy, and the Disqus comment system is very good and getting better.
Twitter's decision to change the goal posts for its APIs is, in effect, stealing. After building--to its own benefit--an ecosystem that helped to spread its use, make it relevant, make it ubiquitous, and make it money, Twitter has decided to pull an Apple and screw part of its developer community.
Twitter's apparent arrogance is unfounded; it doesn't produce the best client for its customers, even on its own site. But that's not even the issue--it's that they've represented one thing for years, and are now screwing developers who depended on them, trusting them to not be Apple.
Reparations are in order. Any developer/company that has invested in building software that is now "banned" by Twitter should be repaid, with interest, for the time, effort, and innovation they brought to the Twitter ecosystem.
Now we just need a US District Court to aid in the effort, and some plaintiffs. I won't be among them; we haven't developed that dependency though …
That said, I'm looking forward to integrating it with Jawaya :)
For as long as I can remember, I've had trouble sleeping on Sunday nights.
I'll guess that my habitual procrastination left me with a pile of homework due on Monday, still untouched until after dinner on Sunday, and of course there was always something else more interesting. I hated homework.
I typically finish the week with more on my plate than intended, and usually work through the weekend. Starting a company can be overwhelming, especially when you have limited resources and tough choices among 10 top priority items. Try prioritizing 10 #1 items that are interdependent--never fun. When you have a full-time team (I don't yet) you can spread the pain. When it's just you and part-time help, well, you slog through, and your weekends are eaten up quickly.
Yesterday I took a break from coding and just did some light research around revenue models. I have a very clear idea of the model, but it's time to nail some of the basic assumptions down and build the revenue model …
As usual I worked all week including most nights. I spent a lot of time refactoring some code, testing, etc, and a lot of research. At some point you need a break, and last night I had a nice break after serving my poor, sick wife some miso soup while I ate some really great sushi (yes, Lancaster has sushi!).
Saturday. Saturday is my gear day. This is when I break out the Guild instead of the Pensa, clean off the desk, reorganize the mics in the studio (really just my office), look for cool stuff on the web, maybe even order something. Like the iPad 2 :), which, of course, is something I need for work (one of the blessings of being a software developer).
So likely very little Jawaya today (some API tests later), and a lot of geeking out.
I just came across a blog post (won't attribute) that sparked a thought: so much of the web is automated now, yet the relevant web is not. It's produced by people for people.
The means of publishing, distributing, and sharing might be automated, but the generation of quality content is, for the most part, owned and tailored by humans. Created by humans.
I quickly searched (yes, using Google) for "Social Automation", and there are now of course a bunch of tools for automating social interactions. I've known about these tools for a while, I just didn't know it had a name for it. Social Automation. Sorry I asked.
Hootsoute, TweetAdder, TweetDeck, Echofon are all in this category. Generally they do the following for their happy customers:
scheduling updates. So you might write 24 tweets in a single setting, but schedule them an hour apart so you cover the day. I imagine you can scatter more randomly. managing multiple accounts. I can see that, though I only have two t…
Well, we missed a few important tests. So I'm cranking on testing for the Chrome extension, which is not what I would call a test-friendly environment because you effectively have to add alerts around every relevant function. That's an issue, but not the problem.
The problem is trying to make the software do beyond what it's currently designed to do using workarounds, which is never a great thing, but it's a question of timing and intent. Minor bug, major effect, and it will be fixed shortly.
I'm in Lancaster--an hour by train from Philly, 2.5 hrs from NY, and a world away from the Silicon Valley. The online scuffle between Vivek Wadhwa and some Bostonians about whether Boston has a vital scene is a bit boring and ridiculous to me. Why?
There will likely never be another Silicon Valley. 95% of the world's venture capital sits on Sand Hill Rd, and it's not going anywhere, so the world goes to it. And when the world comes knocking, more often than not the VCs almost require a relocation to somewhere within spitting distance.
No, that's not always the case. But it's not unusual at all to require at least a strong high-level presence in the Valley. Why?
It's where the deals get done. If you're not in the conversation there, then the business that flows from those conversations won't include you. That doesn't mean you can't build a competitive business somewhere else, but the likelihood of the good biz dev deals, the good distribution, the g…
Yesterday my new wifi-n router arrived, along with two adapters. Our old Netgear wireless-g kept getting hung up and needed about 3 resets a day.
Worse than that, though, was the speed and throughput. When you're developing web apps, latency during testing equates to unnecessary time invested. If you sit for a second instead of 5/100 of a second, well that adds up to a lot of unproductive time.
But we humans tend not to notice what we don't measure until it's a real problem. We treat symptoms instead of root causes. I'm human, and treated the symptoms by moving the router closer to my studio, daisy chaining other weak routers, etc.
For just over $100, I got two good adapters and a new router, the Linksys e2000. My wife noticed the difference right away, even without the new adapter. Both of us are feeling like we just left dial-up.
I hate spending money on gear anymore. But this was a long-needed change, and I should have been more strategic a long time ago. With Jawaya, …
I typically define Beta as "code complete, internally tested". The Jawaya beta has a number of parts, and most are code complete, internally tested. A few parts are still moving toward beta, but the core functionality is not dependent on that.
Beta is also the ugly duckling period. The user experience is rougher, the user interface isn't as pretty as it will be in a month.
Finally, there are open betas and closed betas; this is a closed beta. We have close to 1500 people signed up, but we're rolling it out slowly and progressively so we get it right for the majority of them. Today's a day for some final tweaks, and we'll start rolling it out to more than friends and family, who have been helpful.
I'd love to hear from someone about great beta management. We're going to try our best to serve people well through the software, support, and communication. Finally, the Jawaya blog will be live later today as well, and I'll move most of the Jawaya posts ov…
I've always bought into the argument that staying stealth is important for a startup with something groundbreaking, and that's the approach I've taken with Jawaya.
But the reality is that very little new in tech is really new; sometimes it's the timing, execution, deals, or delivery that makes the difference between success and not.
Today I decided to talk about it, because it's clear there's competition right around the corner. It's time, even though we're at first alpha.
Jawaya is shared search, and through that, it's a search network. It's a way to get better search results through the discovery of results that others have found from the same search and shared to the world.
The quality of those results has been determined by the person sharing the results; the relevance of that person to your interest and search is key to your trust in the quality of the results. That's a big part of the work we're doing.