A Week with Google Wave Testing GitHub as Blogging Platform
I've had my Google wave account for several weeks now, but I've only used it regularly for the first week.
The google provided front end is a pretty ambitious web application that merges instant messaging, e-mail, and collaborative documentation (aka wiki) and discussion. Extensions allow for all sorts of other forms of collaboration - video conferencing, voting/polls, turn-based games, etc. Wave uses standard protocols on the backend (extensions to XMPP/Jabber) that will allow "federation" or organization-hosted wave servers that interconnect with others - similar to how organizations host e-mail and web servers today. To date, my experience is with the Google provided front-end and here are my impressions:
Real Time Editing Can Be Confusing
Imagine five... or thirty of your colleagues all sitting at the same desk attempting to author the same document at the same time. No matter how talented the individual, the model eventaully breaks down and everyone starts stepping on each other's toes. The realtime document editing concept needs to be managed better in part by software and in part by proper netiquette. The software may be able to facilitate the netiquette using a moderator system found in other sytems. This feature should be optionally enabled or disabled on a per-wave basis.
Revisions Tracking (aka Playback) Needs Some Work
Playback/revision tracking needs some work both in its implementation and in its design. Maybe its the developer in me, but a simple revision log with side-by-side diff is more obvious than finding the edits while "playing back". Playback is the only mode that caused the wave ui to "crash" requiring a reload. And was a times slow to respond.
A New Dimension to Instant Messaging
Wave adds new dimensions to instant messaging. No longer do you have to cut and paste part of what someone said to make it clear that you are replying to that bit. You can click on an IM line and reply to that bit - kind of like a threaded discussion forum. The only danger is that the recipient might miss the new message if it is off their screen. Wave UI designers might want to consider a way to keep the conversation linear and still allow replying to past IMs - or perhaps some sort of visual cue to indicate new messages above/below the current visible portion of the screen.
Got RAM?
The web front end is memory intensive taking anywhere from 100M on initial load to 500M on viewing a large wave.
All of these measurements are taken from my Mac Pro running Snow Leopard. Pictured above: "Wave" a Fluid SSB, Waveboard a webkit-based applications, and Firefox with initial weighty overhead of 86M was the least resource intensive. Safari, not shown, shows a similar foot print to the other webkit-based apps. While RAM is cheap, I do enough RAM hungry work in a day to NOT want this bloated Wave app open all the time.Helloooo... Is This Thing On
Perhaps everyone feels the same about the RAM requirements because only once or twice did I ever catch someone on at the same time. Further, because wave accounts are scarce, so is the collaboration.
Google, Build Some Bridges
Google wave is NOT ubiquitous like e-mail and IM and its current UI is RAM heavy. No one is going to be willing to keep th
e RAM eating wave client open all the time like they do for e-mail, the web, or IM. Google needs to be able to push notifications of wave activity to users via e-mail and IM. Likewise, wave should support sending plain-old-e-mail and IMs from the wave client. With these hooks in place and as the UI gets better, people will spend more an more time in the wave UI. In short, people need a migration path that is less disruptive than switching to a whole new tool and protocol.
Internets, Federate
"A Federation is multiple computing and/or network providers agreeing upon standards of operation in a collective fashion. ..." -- wikipedia
Organizations are likely the first to find real value in what wave offers - more so than the end "mom and pop" users who chat with their grandma via IM or shares photos via flickr. Organizations and businesses, are in constant collaboration and as business becomes more and more global more of this collaboration is happening electronically. For the same reasons business host their own e-mail servers, privacy, identity, etc. they will want to host their own wave servers. If the wave client is still in preview, then the federation software is pre-preview. As of this writing, reference/prototype modules exist for the OpenFire XMPP server software. This is all free and open source. A quick glance at the installation instructions and it looks pretty straightforward to install. Unfortunately, there is no wave graphical wave client that will work with it. (Someone correct me if I'm wrong on this point.) All that exists is a simple, limited, command line client. For more see: http://www.waveprotocol.org/Hey, Its a Preview!
Some people love Google to a fault. Some people hate Google because some people love Google to a fault. I love technology. Email, instant messaging, wiki's and other forms of online collaboration and discussion are a part of my daily life. None of them are perfect and some of them are getting long in the tooth, or at the very least being abused. I work in an environment where e-mail and IM is the medium by which almost all collaboration is carried. I welcome a change from the overload of static discussion threads, MS Word attachments etc.
While I welcome a re-imagining of e-mail, IM, and wiki's for the 21'st century it isn't clear that Google wave is it. Based on the preview, It has a ways to go but lets not declare it an overhyped flop before it even makes it to beta. I think it shows some promise simply for its use of existing and extended open protocols. Further, if Google can't get the UI right, I'm sure someone else can. You could be looking at the next big thing since the browser, but a lot of maturing and adoption needs to happen between now and then.Want More?
blog comments powered by Disqus