I wrote once before about the existing WebOS systems. I planned to follow that post up, but time did not allow. Now I will follow it up twofold.

Using
I have come to use WebOS in a broader sense than I once did, which seems to be the ‘catching’ way to use it these days. That is, defining WebOS to include things such as Netvibes and BoxtheWeb, Omnidrive and Box.net, Zoho and Google Docs. I have been asked by some how I find WebOS useful at all, or if it is just an interesting experiment. I will attempt here to answer that.

Online storage is amazing. Work at home, save online, work at school, save online, etc. It is far more convenient than carrying flash media everywhere I go. I have used YouOS, Google Docs, and Gmail for this, but the result is basically the same.

Online office is similarly useful. I can work on and access the same document across computers at my house, or on campus, without any real hassle. Open, edit, save. Collaboration features just make it that much more fun! SVN for documents

AJD (ala BoxtheWeb or Netvibes) is something I really love. I use BoxtheWeb, being the project originator, but there are many out there. I have many feeds, but I still like to be able to glance at them all at once. See what my contacts are reading these days, access my del.icio.us, and search, all from one page. There’s something to be said for that convenience.

Last but not least (and I’ve likely forgotten others) : YubNub. I couldn’t live without it. The amount of time saved being able to type ‘g singpolyma’, ‘tet microformats’, ‘hwdial singpolyma.net’ is amazing. And the development stuff is fun.

As for full-scale webtop integrations. I haven’t been using the full-fledged features yet. There’s something there though… just not ready yet (or maybe I’m too geeky to see it past the GUI ).

Standards
I mentioned in my last post the need for standards. If I could run Netvibes widgets on my BoxtheWeb page while integrating my YouOS storage, that would be very most awesome. The companies themselves seem to be organising, but it’s private and they’re not taking input. So, as always, the community needs to get their foot in the door before the industry runs on its course and we have many too many products and it takes years to create standards.

Drawing inspiration from the Microformats process, research into existing practices/standards should be present before suggesting something new. To aid the community in organising such research and development, I have created a wiki and a discussion group.

Perhaps not all of us are geeky enough to do the research and formal stuff for the wiki, but anyone who uses any of the products listed above, listed on the wiki, or related products, as well as those interested in Simile and related projects can contribute to the discussion on the discussion group. Anything from observations about how things work, suggestions as to how things should/could work, or even pointing out projects that may have been missed is welcome. Feedback from users as well as geeks is necessary to make this project work.

So if you read any blog besides mine (and surely you do) you’ve by now head of Yahoo’s Pipes application. Mashups without programming, and a team that’s promising more and better things to come.

One of the immediate uses to the Blogger community occurred to virtually all hackers at once. Sorting the feeds. This has never been a problem for me (I screen-scrape my feed via hAtom), but for others the fact that Blogger feeds sort by when they were updated is annoying.

Aditya suggested creating individual pipes, but I wrote a sorting pipe, as did Ramani (who beat me to blogging about it and has a nice how-to written). Ramani discovered an issue that causes this solution to be a bit buggy just yet. It has to do with ATOM being stupid and RSS 2.0 being cool (yes, that’s a partisan statement and not entirely true ). Basically the publishing date is not being copied from the ATOM format to the RSS format correctly. Vote on Ramani’s suggestion to get this fixed. I also discovered a less critical issue with the UI that may confuse some less geeky users. Please vote on my suggestion to get that fixed.

I also wrote a pipe for mixing together Google Calendars (for those of us who track events from more than one) into a nice, sorted feed of upcoming events. The email alerts system provided by Y! is dumb though, at least for this application. I want the next 5 events emailed to me every day… likely gonna have to write my own emailer for that…

Tired of working with pages that do not have Microformats support yet? The Microformats Proxy is the tool for the job. Individual or globbed URLs can be set up via XPath to extract data and format it as microformats. In this first version only hCards are supported.

All existing hCards and XOXO data are also maintained.

JScripts, my Ning app for storing JavaScript includes, has got some upgrades! At Johan‘s prodding I have added the following:

• Include multiple JavaScripts with just one request to save time! It even has a form!
• Tack ?minify on to a script to have JScripts compress it before transfer (more time saving!)
• Tack ?callback=whoever on to have whoever() called after the script has loaded

OpenID is meant as a distributed single signon protocol. Unlike corporate-only systems (ie, Google Accounts) however, you have to sign on at every single site. When I sign in to Gmail I am automatically signed in to Blogger, etc.

While the inability to store ‘cookies’ or similar across domains makes this not 100% possible, I have a suggestion that will make it easier.

Make every page on an OpenID-enabled site accept the openid_url GET parameter. If it is there, authenticate the user using that OpenID and then redirect back to the page the GET string was passed to – minus the openid_url parameter. Thus if I click a link from one OpenID-enabled site (where I am signed in) to another, I can be signed in automatically.

A browser plugin (ie, Firefox extension, or built into the browser as may come in Firefox 3) could be used to store the last-entered value into a form with text input named openid_url (dependent on user settings of course). This value would then be added to the GET request on any URL (or pertinent URL, depending) the browser goes to. OpenID-less sites won’t care, OpenID-enabled sites will automatically sign you in.