Technology - 6. page

Find out at the Electronic Product Environmental Assessment Tool.

Along the lines of Squeet, which I mentioned previously, it looks like Craig2Mail does a good job of converting RSS feeds to emails.

I found this paper (PDF), while a bit old (from 2002), to be a useful analysis of on the fly compression (a la mod_gzip or mod_deflate).

Updated 1/7/2008: Note that Squeet appears to be non functional at this time.  Craig2Mail.com still appears to work.  Please let me know (via the comments) if there are other sites that provide this valuable service.

craigslist is an online classified ad service, with everything from personals to real estate to bartering offered online. I’ve bought a table from Denver’s craigslist and I know a number of folks who have found roommates via craigslist.

If you have a need that isn’t available right now, you can subscribe to a search of section of craigslist. Suppose you’re looking for a used cruiser bike in Denver, you can search for cruisers and check out the current selection. If you don’t like what you see but don’t want to keep coming back, you can use the RSS feed link for the search, which is at the lower right corner. Put this link into your favorite RSS reader (this is a simple application that manages RSS feeds, which are essentially lists of links. I’d recommend Bloglines but there are many others out there) and you can be automatically apprised of any new cruisers which are posted.

(You find tons of stuff via RSS–stock quotes, job listings, paparazzi photos… The list is endless.)

If you don’t want to deal with yet another application, or you’re not always in your RSS reader (like me), you can set up an RSS to email gateway. That way, if your cruiser bike search is so urgent you don’t want to let a good deal get away, you receive notification of a new posting relatively quickly. If you want, you can even email it to your mobile phone.

The basic steps:

1. Go to the Squeet signup page. Sign up for a free account. Don’t forget to verify it–they’ll send an email to the address you give them.
2. Open up a new browser window and go to craigslist, choose the city/section you are interested in, and do a search. The example up above was ‘cruiser’, in the bike section of the Denver CL.
3. Scroll down to the bottom of the search results and right click on the RSS link. Choose either ‘Copy Shortcut’ or ‘Copy Link Location’, depending on your browser.
4. Switch back to the Squeet window, and click in the ‘FEED URL’ box. Paste in the link you just copied. Choose your notification time period–I’d recommend a frequency of ‘live’, since cruiser bikes in the Denver area tend to move pretty quick. Then click the subscribe button.

That’s it. Just wait for the emails to roll in and soon enough you’ll find the cruiser bike of your dreams. Just be aware that it’s not real time–I’ve seen lags from 30 minutes to 2 hours from post to email. Still, it’s a lot easier that clicking ‘Refresh’ on your browser all day and night.

Google now offers geocoding services. Up to 50,000 addresses a day. I built a geocoding service from the Tiger/Line database in the past. Comparing its results with the Google geocoding results, and Google appears to be a bit better. I’ve been looking around the Google Maps API discussion group and the Google Maps API Blog and haven’t found any information on the data sources that the geocoding service uses or the various levels of precision available.

A friend has been working on a startup which looks like it’s been focusing on identity management on the internet. That startup, Naymz, launched today. I just joined–check out my page. It’ll be interesting to see if this site fills a need.

Check out spreadsheets.google.com. Limited time look at what javascript can do for a spreadsheet. I took a quick look and it seems to fit large chunks of what I use Excel or calc, the OpenOffice spreadsheet program, for. Just a quick tour of what I such spreadsheet programs for, and what Google spreadsheet supports:

• cut and paste, of text and formulas
• control arrow movement and selection
• formatting of cells
• merging of cells and alignment of text in cells
• undo/redo that goes at least 20 deep
• sum/count
• can freeze rows
• share and save the spreadsheet
• export to csv and xls

On the other hand, no:

• dragging of cells to increment them (first cell is 45, next is 46, 47…).
• using the arrows to select what goes into a formula–you can type in the range or use the mouse

Pretty decent for a web based application. And it does have one killer feature–updates are immediatly propagated (I have never tried to do this with a modern version of Excel, so don’t know if that’s standard behaviour). Snappy enough to use, at least on my relatively modern computer. I looked at the js source and it’s 55k of crazy javascript (Update, 6/9: This link is broken.). Wowsa.

I’ve never used wikicalc but it looks more full featured that Google spreadsheets. On the other hand, Google spreadsheets has a working beta version…

This and the acquisition of writely make me wonder if some folks are correct when they doubt that Google will release a software productivity suite. (More here.) Other interesting comments from Paul Kedrosky.

I know more than one person that absolutely depend on gmail for business functionality, which spooks me. And in some ways, I agree with Paul, it appears that Google “…takes a nuclear winter approach wherein it ruins markets by freezing them and then cutting revenues to zero.”

Personally, if I don’t pay for something, I’m always leery of it being taken away. Of course, if I pay, the service can also go away, but at least I have some more leverage with the company–after all, if they take the service away, they lose money.

I moved from my own personal RSS reader (coded in perl by yours truly) to Bloglines about a year ago. The main reason is that Bloglines did everything my homegrown reader did and was free (in $ and in time to maintain it).

But with over 1 billion articles served as of Jan 2006, I always wondered why Bloglines didn’t do more collaborative filtering. They do have a ‘related feeds’ tab, but it doesn’t seem all that smart (though it does seem to get somewhat better as you have more subscribers). I guess there are a number of possible reasons:

• It’s easier to find feeds that look like they’d be worth reading (I have 180 feeds that I attempt to keep track of)
• blogrolls provide much of this kind of filtering at the user level
• privacy concerns?
• No demand from users

But this article, one of a series about data management in well known web applications, gives another possible answer: the infrastructure isn’t set up for easy querying. Sayeth Mark Fletcher of bloglines:

As evidenced by our design, traditional database systems were not appropriate (or at least the best fit) for large parts of our system. There’s no trace of SQL anywhere (by definition we never do an ad hoc query, so why take the performance hit of a SQL front-end?), we resort to using external (to the databases at least) caches, and a majority of our data is stored in flat files.

Incidentally, all of the articles in the ‘Database War Stories’ series are worth reading.

Tim Bray gives a great write up of Grid Infrastructure projects. But he still doesn’t answer Stephen’s question: what is it good for?

I think the question is especially relevant for on demand ‘batch grids’, to use Tim’s terms. A ‘service grid’ has uses that jump to mind immediately; scaling web serving content is one of them. But on demand batch grids (I built an extremely primitive one in college) are good for complicated processes that take a long time. I don’t see a lot of that in my current work–but I’m sure my physics professor would be happy to partake.

Owen Taylor, who I pleaded (in person and in this blog) to start blogging, has done so. This is Owen Taylor’s Weblog. If you’re interested in jini, javaspaces or random technological musings, give it a look. Welcome, Owen!