Past EED rants


Live leaderboard

Poker leaderboard

Voice of EED

Wednesday 22 July 2009

Google Wave

Back in May, Google announced that they were rethinking the whole dated concept of communications that we hold so dear. With Google Wave they claimed to have gone back to the start and invented electronic messaging from the ground up, with none of the baggage of SMTP or the various IM network protocols.

Indeed, their keynote presentation at I/O 2009 was almost universally praised. Wave looked to be just as revolutionary as they'd claimed - and, more enticingly, almost finished. It was genuinely exciting to see this working, and so polished and the concept seemed to just work as well. Meanwhile, for the developer, the prospect of a rich API and bot specification was almost too good to be true. The possibilities were endless - both for communication, for collaboration and for fun.

So, I put my name down for the developer programme, indicating that I'd happily put up with bleeding edge APIs and shakey service. And, finally, this week I got my access.

I've not yet tinkered with the bot APIs yet, I'll maybe tackle that in a different blog, but I've had a couple of days of intermittent usage. So what's it like? Is it as revolutionary as we hoped? Yes, kind of, maybe...

First off, it's worth mentioning that Wave is very early. It is slow, half of the features don't work, and half of the working ones don't work properly. It is memory hungry and you really don't want to leave a tab open with Wave running in it - not unless you like 500Mb Firefox processes that is. It pretty much kills my EEE901 stone dead, but runs okay on my desktop. This is very much a developer preview of a product that is probably six months from a proper launch, so being rough around the edges is to be expected.

How is it to use?

In this developer programme, every participant is subscribed to the Wave-Discuss list - like an old-school mailing list, but with Waves - so the first thing you notice about Wave is that it seems alive. Your inbox writhes like an organic thing as active Waves jostle to the top constantly. Individual Waves ripple with activity in the right hand pane as the much-talked-about realtime typing going on around you. The consequence of all this activity is that it can initially seem overwhelming - Waves slip out of your control if you neglect them and you never seem to be on top of your inbox.

I suppose this is an artefact of thinking like an emailer. We feel the need to absorb every bit of information, and information should persist for longer. Wave is different, it is sort of like a cross between email and twitter - conversations yes, but conversations that are time dependent. If you try to reconstruct the flow of information in such a thing, you'll soon get lost - even with the replay feature. Waves are both persistent and disposable - a strange contradiction.

It's this aspect that I'm not sure I like about Wave, and one of the reasons why I'm not as convinced as I was that it will supplant email.

More positively, this model works very well for collaboration. There's a number of Waves on the developer list that contain FAQ information, and are free for everyone to edit. These end up like a cross between a Wikipedia article and a forum thread. This kind of thing is one of Wave's strengths and it should find a corporate home if nothing else.

Wave is also home to a number of bot entities. Bots are currently hosted on Google App Engine and can be added to any Wave simply by adding the bot as a participant. From there, they can manipulate existing "blips" (snippets in the Wave), add new ones and pretty much do anything that a user can do. As such, they're open to massive amounts of abuse - especially since anyone on the Wave can add a participant.

It's quite weird to see Waves that have been abandoned because a runaway Elizabot has spammed it into uselessness. It's even weirder to see a counter-measure bot deployed to take care of unwanted rogue bots. What on earth will happen when the Chinese start coding for it??

And this is what I think is a major problem for Google - security. Both in terms of phishing/spam and informational integrity. At the moment, there is nothing to stop anyone from doing anything. If I start a Wave, then you can edit any part of it to your heart's content. I'm not sure I'm comfortable with the idea of someone changing one of my posts from "I dislike Nazis" to "I love Nazis". Okay, so there is replay feature, but it is virtually useless for long running Waves, with no timeline to skip between periods. Wave desperately needs a "padlock" icon to lock blips.

My final gripe is with organisation of Waves. Google got Gmail's organisation dead right. But Wave, which arguably needs it far more, has it dead wrong. At least at this early stage. Waves currently land in your inbox. You can then send them to a folder (currently not working) or define a search, which you can save and which can also be applied as a filter. The search thing is horrible to be honest, it is arse backwards. Come on Google, let's have labels like Gmail please.

I sound quite negative don't I? I'm maybe being a little harsh on a product in the very early stages of development. All of the things I've mentioned are issues that can, and probably will be sorted.

In fact, using Wave is very cool, and very addictive. Actually being engaged in a conversation within a Wave is kind of like IRC, but with persistence and threading. And gadgets. And like IRC, you can find that the topic rambles on till it no more resembles the original point of the Wave than any conversation resembles its first sentence. For this alone, Wave is still worth watching. We just have to hope that Google attend to the really quite obvious gotchas before they release.

1 comment:

  1. Re editing others, that was something that I thought watching the I/O presentation. It's all very well doing review of history but it just won't work like that in practice. You must need the ability to post a wave contribution as a locked piece versus an open one.

    This is just a microcosm of what I'm pretty convinced is the major battle ground of the interweb in the years to come - revisionism of history. Already wikipedia is falling apart due to factionalism, the comments pages of even the best broadsheets are nuked by extreme points of views rather than rationale commentary and yet all of the above are used increasingly as 'definitive sources'. Technology is like any other medium but the speed and proliferation of bias and deliberate misinformation the internet is facilitating is seriously worrying.