Today was a good day for import•io – we won the judge’s vote at the Startup Showcase at Strata, Silicon Valley edition. I thought I’d take the time to explain what we mean when we say import•io is in Developer Preview.
Spoiler: it basically means that the import•io builder is not 100% successful yet. But, by using it you are helping us make it better, especially if it fails, and we will always help get failed connectors working.
You know what a beta is, right? When software isn’t necessarily “production ready”, but is ready to get out there for feedback and for early adopters to start playing with. As Wikipedia says:
“Software in the beta phase will generally have many more bugs in it than completed software, as well as speed/performance issues. The focus of beta testing is reducing impacts to users, often incorporating usability testing. The process of delivering a beta version to the users is called beta release and this is typically the first time that the software is available outside of the organization that developed it.”
Well, we chose the title “developer preview” rather than beta, but because we’re a platform, we wanted to let developers know that we may release new versions of APIs as we get feedback, and that we are looking for devs to try out and feedback on the APIs, client libraries and integration.
I guess it is also about expectation setting. Before I go any further let me recap our terminology, quoting from the builder splash page:
The Builder is designed to extract and connect to all data on the web. At the moment it works on most of the sites that it is used on. The Builder learns and as it is used on more sites we are able to make it better
Extractors structure web pages into tables of data that can be copied into a spreadsheet.
Extractors can be used on the import•io website and over the import•io API but you need to know the URL of the web pages that you want to extract data from.
Connectors automate web browsing and fetch web pages for Extractors to work on.
Connectors allow you to use any web search-box on the import•io website and over the import•io API.
The import•io technology for creating connectors and extractors using our GUI is reasonably mature, but we are on a curve of improvement. Currently the extraction works on around 90%of sites, and building full connectors works on around 60% of sites. Those figures will continue to increase over time, and it’s important to us to collect cases where it specifically doesn’t work so that we can improve the training set to make it work on those sites, and other sites. Getting this data is important to us to make our service better and better. And this should be more transparent to users.
So, one of the things that we’ll work on in the coming weeks is our messaging to users when a connector doesn’t publish. It s important for users to know that we are going to look at all of those that don’t publish, try to fix their connectors, and let them know when they are published and working.
So thanks for all your support, and I look forward to import•io going from strength to strength.
P.S. Here’s Andrew Fogg, my co-founder and Chief Data Officer of import•io accepting our accolade.
by Matthew Painter, CTO
This blog post was originally posted on 26th Feb 2013.