I haven’t commented very much, if at all, on most of the new Google projects, which include Chrome, Chromium and Chrome OS; today since I’m currently waiting on a few long-running tasks to complete, I’d like to spend my two eurocents on it.
You can already guess from the title of this post that I’m really sceptical about Google entering the operating system marked; the reason for that is that I haven’t really seen anything in Google strategy that would leave us expecting a very good product from them in this area. While Google is certainly good in providing search services, and GMail is also my email provider of choice, there are quite a few shortcomings that I see in their software and that does not make me count on Chrome OS being any more good that Windows XP is.
First, let’s just say that Google Chrome is not the first software that Google released for the desktop; there has been quite a few other projects before, like for instance Google Talk. Since I have a personal beef with this, I’d like to go on a bit about it. When Google launched their own Instant Message service for the masses, through GMail and a desktop, called Google Talk and base on the XMPP protocol, there has been quite some talk around because, while using the same protocol we know as Jabber, it didn’t connect to the Server-to-Server Jabber network that allows for multiple Jabber servers’ users to communicate; with time this S2S support was added and now a GTalk user can talk with any Jabber user, so as a service, it’s really not bad at all, and you can use any Jabber client to connect to GTalk.
The Windows client, though, seems to be pretty much abandoned, I haven’t seen updates in a while (although I might not have noticed in the past month or two), it lacks quite a few features like merging of multiple usernames in a single contact and stuff like that. Now, at the same time as releasing the Windows client, or about the same time, Google released specifics for their extensions that allow audio (and video?) chat over XMPP-negotiated connection, and a library (libjingle) for other clients to implement this protocol.
The library, unfortunately, ended up having lots of shortcomings, and most projects decided to import and modify it, then it was forked, at least once but I think even twice, cut down and up and so much mangled that it doesn’t probably look anywhere like the original one from Google. And yet, the number of clients that do support GTalk audio/video extension is… I have no idea, Empathy does support it if I recall correctly, but last time I tried, it didn’t really work that well. As far as I know, libpurple, that is used by both Pidgin and Adium, and which would cover clients for all the major operating systems (free or not) does not seem to support them.
Now, why do I consider GTalk a mediocre software does not limit itself to the software that Google provides, it’s a matter of how they played their cards. It seems to me that instead of trying to push themselves as the service provider, they wanted to push themselves as a software provider as well, and the result is that beside Empathy (which is far from an usable client in my opinion), there is no software that seems to be implementing their service properly. They could have implemented, or paid to implement or something like that, their extensions in libpurple and that would have given them an edge; they could have worked with Apple (considering they are working with them closely already) so that iChat could work with GTalk’s audio and video extensions (instead iChat AV from Leopard uses a different protocol that only works between Macs), and so on.
What about Google Chrome? Well when it was announced and released I was blocked in hospital so I lost most of the hype done in the first days; when I finally went to test it, almost a month later, I was surprised at how pointless it seemed to me. Why? Because for what I can see it does not render text as good as Firefox or Safari on Windows, it’s probably faster than them, but then again most people don’t care (at least in Italy, Internet connections are so slow you don’t notice), and there is one important problem: the Google bias of the browser.
I think lots of people criticised the way Microsoft originally treated Internet Explorer and their Internet services before. to the point that now Microsoft allows you to set Google as provider for search in the default install. Well, I don’t see Chrome as anything much different: it’s a browser that is tailored to suit Google’s services, and of course the development of it will suit that too. Will it ever get an advertising block feature, like is available for Firefox, Konqueror and Safari? Probably not because Google takes a good share of revenue out of Internet-based advertising. Will it ever get a delicious extension? Probably not because that’s a Yahoo! service nowadays, and Google has its own alternative.
Now, I don’t want to downplay the important technical innovation of Google chrome, even when they are very basic like the idea of splitting the tabs by process; and indeed I think I have read that Mozilla is now working on implementing a similar feature on the next Firefox major change; this is what we actually get out of the project, not Chromium itself.
Then there is Android; I don’t think I can really comment on this, but at least for what I can see, there is not really much going on with Android: nobody asked me yet if I develop for Android, while I got a few requests for Symbian and iPhone development in the past year or so. Android phones does not seem to shine with the non-technical people, and the technical people at least in Italy are unlikely to pay the price you got to pay to get the Android-based HTC phones with Vodafone and TIM.
By contrasting with Nokia, Google fragmented the software area even more. While Google already provided mobile-optimised services on the web, and some Java-based software to access their services with J2ME-compatible phones, they also started providing applications for Nokia’s Symbian-based phones. Unfortunately this software does not shine, with the exception of Google Maps, which works pretty well and integrates itself with Nokia pretty decently; in particular the “main” Google application for Nokia, crashed twice my E75!, I ended up removing it and living without it (the YouTube application sort of works, the GMail application also “sort of” works, but with the new IMAP client is really pointless to me). So we have mediocre software from Google for Nokia phone, and probably no good reason for Google to improve on it.
But there are also things that haven’t been implemented by Google at all, for instance there is no GTalk client for Nokia phones, or a web-based version for mobile phones, which would have been a killer feature! Instead Nokia implemented its own Nokia Chat, which now became Contacts for Ovi, which also uses XMPP, which also has S2S, but which does not allow you to use GTalk accounts requiring you to have two different users: one for computers and one for the mobile phone. And similarly, with just partially-working Google Sync for Nokia phones, in particular with no support for syncing with the Google Calendar, and with a tremendous loss of detail when syncing contacts, Google loses to Nokia’s Ovi sync support as well.
Now, I’m not a market analyst and I really like to stay away from marketing, but I really don’t see Google as a major player for Software development, I’d really have preferred they started improving the integration of their services with Free Software like Evolution (whose Google Calendar integration sucks way too much, and whose IMAP usage of GMail causes two copies of each sent message to be stored on the server, as well as creating a number of folders/labels that shouldn’t be there at all!), rather than having a new “operating system”.
There are more details I’m sceptic about, like hardware support (of which I’ll leave Mathew Garrett to explain since he knows the matter better) and software support, but for those I’ll wait to see when they actually deliver something.