Thoughts on Google Chrome OS

I’ve resisted writing anything on the recently-announced Google Chrome OS, for a number of reasons… the most significant one being that so far, we don’t know a huge amount about it. This hasn’t stopped reams of opinion being written or spoken about it anywhere else though, so a week on from the announcement, I thought I’d lay down a few opinions of my own.

First of all, I always felt that the Chrome browser itself kind of pointed towards an operating system, since the engineers were clearly thinking in terms that you’d usually associate with an OS – ideas like the threading model immediately made me think of the domain of the operating system. The “Google OS” has been one of those rumours that has consistently failed to die.

So what do we know? We know that Google Chrome OS will be based on Linux and will be largely open source, and that the initial target constituency will be the netbook market but that it has ambitions to the desktop. We know that it will have a new windowing system (bye bye X). We hear that Google has been courting various netbook manufacturers, and we know that it should be out sometime in 2010.

On the threat to the desktop

Scott Bourne was saying on MacBreak Weekly this week that he felt this meant it would be no real threat to Mac OS X, and I guess the ensuing discussion really sparked the majority of this post. Scott based his assertion on a straw poll of people who he’d asked “OS X or Chrome?” (a: OS X) for laptops, and “Chrome or Windows Mobile?” (a: Chrome) for PDAs. I just think that’s an impossible discussion right now. So far it’s vapourware. We don’t know what Google Chrome OS will look like and we don’t know what features it will have. It’s pointless to try to compare it to existing operating systems at this stage.

The other reason the MacBreak Weekly crew decided that Chrome OS wouldn’t be a threat was that it would initially be limited to netbooks but “will it actually be able to make the step up to the desktop?”. This is an interesting discussion, as it assumes that the granddaddy / holy grail of machines is the desktop computer. But… wait a minute. Haven’t Apple spent the past two years convincing the whole mobile market that they have to have fully-capable computing platforms on their handheld devices? Isn’t the netbook market exploding? Aren’t laptops outselling desktops? Aren’t computers and televisions converging via set-top boxes and streaming media? People want computing power and access wherever they are, in a form factor that fits. I think the suggestion that the desktop is still “where its at” is deeply, deeply flawed – the desktop is dying, and has been for a long time. The desktop is a place where people occasionally anchor themselves, but the rest of the time they are moving around and taking their platform with them.

So is it a threat? On netbooks, in my opinion, yes – probably to both Linux and Windows. As for Apple, they aren’t going to be keen to let anyone run another OS on their hardware, and they’re not currently in the netbook market, so it probably is not a big deal right now. Windows XP still seems to be an OS of choice on many netbooks, but Microsoft probably will finally kill that with Windows 7. There are too many Linux distributions around (Linpus, Moblin, Ubuntu) vying for a slice of that market. The Google brand, combined with an experience that is genuinely pleasant, could take chunks out of both sides of that market. From there, who knows… but if it is Linux-based, we already know that that scales from big to small machines, so there are plenty of places it could go. I personally think it will be interesting to see where they choose to go with the user interface, and this is the area that will determine the future.

On the cloud

I think the idea that Chrome OS could be a lightweight system with the majority of content hosted in the cloud is particularly interesting. How secure are we all feeling this week about the safety of the cloud? In the light of the Twitter hack, I imagine that people are rethinking the security of the public cloud, anyway. Paired with a secure behind-the-firewall private cloud, a lightweight OS like this makes a lot of sense. That’s not to say I don’t like the flexibility that cloud/web-based services offer me – I use many of them, including things like Google Docs. Of course, you need a network connection for it to be effective, and it’s true that bandwidth is becoming pervasive, although if you take a look at 3G coverage for various UK mobile operators (hint, look at the PDF and the maps *per operator*, not the overall coverage map), you might want to rethink your dependence on that, too.

On the image

Will Google be able to get away with labelling Chrome OS “beta” to start off with? I think that in order to get people to use it, beyond the Google brand name, it will have to be really good, polished, and / or flashy and convincing enough as an initial experience for people to base their computing lives on it. I think it will probably have to have had more testing than many of the existing Google cloud web products.

Final thought

Right now, all we can really say is: “well, this is interesting”. We can speculate, but frankly, I don’t think we know enough to say anything more. That’s all I’m sayin’.

The Great Address Book Cleanup

This was fairly tedious, so I thought I’d share it with you, my loyal blog readers

Over time, my address book has become unmanageable. Until today, what I had managed to achieve was a curious fusion of useful family postal address data, mostly from the pre-social network, pre-email, pre-mobile days of 10-15 years ago when I had loaded it all into Evolution on Linux; several personal and work mobile phone SIMs worth of numbers in various formats, covering probably a 10 year period judging from the names and numbers in there; and lots and lots of email addresses, covering a similar period, but recently bloated by the habits of Gmail and Apple Mail of storing some data automagically.

In short: a total mess. I had multiple entries for several of my close friends and family; original home addresses of university friends I have not been in contact with for 12 years (!); and outdated information for many other folks.

The iPhone was really the final catalyst for sorting all of this out. More than once recently, I’ve tried to find a number for a friend, and had several entries for that person, several of which were out of date. I’ve called a good friend’s brother on a phone he gave him more than seven years ago, twice in the past 6 months. That’s how useless I’ve become at managing this stuff.

Since the iPhone was core to the issue, and since it syncs with the Address Book, I decided to use that as my core tool for fixing the problem.

Here’s more-or-less what I ended up doing:

• Cleaned up multiple entries in Address Book, merging into a single entry for any one contact, including work and personal data, and adding company names where possible.
• Deleted redundant entries such as the university addresses which I knew were going to be wrong.
• Read through the various hints on this page for information on how I could get contact data from LinkedIn and Facebook and elsewhere…
• Exported my LinkedIn Contacts as a .vcf file for import into Address Book… imported and checked the merge information.
• Grabbed my Facebook Contacts phone details using a Greasemonkey script and a tool called AddressBooker (see the above linked hints page for details). At this point I had an issue. Downloading from AddressBooker as .vcf caused all the names to be exported as just FirstName fields, so when I tried to merge into Address Book everyone came up as a “new” contact. A better solution turned out to be to import directly from AddressBooker into Google Mail Contacts first; do any data merging and cleanup there (GMail Contacts has a nice “merge contacts” capability if you have multiple entries for one person); and then export as .vcf, and import into Address Book, checking the merge information as before with LinkedIn. Sounds long-winded, but it would have taken longer to manage manually in Address Book.
• At this point, Apple Address Book contained my “authoritative” contact list. To get it back into GMail, I scrubbed my whole GMail contacts list (slightly scary, but necessary), and then played around with the Automator action described in the Maciverse tips page, trying to use Numbers instead of Excel. Ultimately that was a disaster as it just couldn’t encode the tab-separated to CSV files correctly. The final answer turned out to be a brilliant little (free / donationware) utility called A to G. Get it. Use it. Donate.
• Result! And next time, I’ll try not to let this get so out of hand…

I see another wrinkle in this whole space. Although tools like Google Voice tend towards centralising all your contact details behind a single number you control (great idea – dear Google, when can we get it in Europe? hint hint!), the proliferation of social networks means that we all have multiple profiles and IDs. In particular, I find Skype and Twitter addresses hugely useful. Apple’s Address Book, of course, doesn’t provide “official” spots for those… there’s no entry for Skype in the IM drop-down list, and no space for arbitrary URLs, you have to edit and maintain a list of them. Something more extensible and easy to keep in sync would be very useful.

Update: the broken part
Google Talk appears now to have completely lost my contact list. Gah. So if you were connected to me on Gtalk then you may need to reconnect, or you may get a new connection request from me. Sorry about that…

On the Chrome bandwagon

Well Twitter and the interwebs sparked up with discussion of Google’s new browser, Chrome, late last night (UK time) and I thought I’d add some ill-formed thoughts of my own. Don’t expect reasoned, complete analysis at this point: the morning coffee is sitting beside me unsipped…

First thing to say is that I enjoyed the web comic that leaked out a day before the formal announcement, although I did wonder at some of the logic at the time.

• Chrome installed very smoothly on Vista. For a beta, it’s remarkably stable and well-featured. It has been no real secret that Google have been at least tinkering with browser technology for a while, but they’ve made a very credible entry to the space by coming along with something so (relatively) complete and solid as a first release.
• I didn’t quite get the reasoning behind the inversion of current UI paradigms with the tabs along the top of the window. I’ll be interested to see how this looks across the major OS platforms. Phil mentions that it’s probably easier when each tab represents a separate process, and Phani points out that it lets them associate the address bar and tools with a tab rather than the window. I guess I’m more used to a platform-native look-and-feel now.
• Some of the touches in the UI are quite nice – sites with insecure SSL certificates appear with a crossed-out line through the https, and in common with Firefox 3 it initially warns with a big red page when the site is not trusted.
• It seems fast, but I typically browse with a lot of tabs open in Firefox so I could just be “used” to a slower browsing experience brought on by bad habits.
• On that note, the concept that a “bad tab” won’t kill my browser really, really appeals to me. The design of this feature leaves me wondering whether we are going in another technology loop (processes good, threads bad) but I’m willing to be convinced.
• I like the idea of the start page with thumbnails of favourite pages. I’ll be the other browsers adopt that soon.
• I like the idea of “application shortcuts” which can go straight onto the desktop. It makes sense. Lots of applications are webapps now.
• Flash “just works”. Java does not.
• On the subject of plugins… I’ve come to depend on a lot of the plugins that I can get for Firefox. Given the heritage of the Chrome development team, and the model that the comic describes for isolating plugins / scoping them to a tab, I assume that it will be possible to extend Chrome… but the user-visible extension points are currently limited (no menu bar, no status bar) so it will be interesting to see how they approach this and what the take-up is likely to be.
• The developer tools that are built-in are very impressive… nice source viewer, element inspector, and the “task manager” with “stats for nerds” gives an interesting glimpse into the way the browser and task isolation is working.
• RSS feeds don’t appear to display. I half-expected an RSS link to whisk me off to Google Reader – I have no doubt that all my base WILL belong to Google here. Maybe that’s a “todo” feature.
• Another technical thought – are images in Chrome rendered with colour profiles? I assume so since it uses WebKit, but I’m not sure.

I’ve been saying to people for a while now that Apple has been worth watching… using WebKit as the basis of Safari, Dashboard widgets for OS X, and the iPhone. With this (re)use of the same techology by Google, I’m intrigued to see where all of this will take the browser. With Google’s brand recognition and reach, this has the potential to be a very disruptive move.

Update: WOAH, major licensing issues. And Mark Cathcart brings some interesting perspectives, too. For now, I’m removing Chrome.

Update: as Mark and Justin note below, the EULA has been fixed. Chrome gets a reprieve and can come back to my system, for experimentation purposes at least.

Rationalising email: Gmail, IMAP and Mail.app

The current setup

Over the years, I’ve accumulated a lot of email addresses… and a proportionately large amount of spam to go with them.

For the past five or so years, my mail processing system has looked like this:

1. Linux server running fetchmail which fetches from a total of… um… 8 (!)POP3 accounts.
2. All mail run through SpamAssassin which catches probably 90% of the junk, and scripts run every month that clear down and learn from the spam.
3. SquirrelMail on the same server to provide web access.
4. Dovecot IMAP on the same server to let me manage the aggregated mail from Thunderbird on various laptops at home.

This has worked well, but it has also meant that I’ve had to maintain a Linux server at home, and I’ve not opened up IMAP access to it over the Internet. So, with Google’s announcement of IMAP support in Gmail, I thought I’d give it another go.

How am I using Gmail now?

It took about a week for Gmail IMAP to appear on my account, as those who followed my increasingly-frustrated Twitterings will confirm.

It’s a progressive process, but I’ve decided to try to use Gmail’s ability to suck mail from my other accounts. The problem is that I have 7 of them (the eighth is Gmail itself), and Gmail will only let me pull down mail from 5. That actually turns out to be OK, since a couple of them were essentially unused or spam-only accounts, so I’m cutting down on those too.

Using Gmail as the front-end to all of my mail is good for a couple of reasons, and bad for another:

Good - I will eventually be able to decommission the Linux server.

Good – Gmail has good spam filtering, labels and all that good stuff around search, and is mostly accessible.

Bad – it isn’t accessible from everywhere, and my last client actually blocked access to Gmail explicitly, whilst I could still get to my home server very easily. I think this is likely to be the greatest annoyance.

I was pleasantly surprised with how easy it was to configure Gmail to pull my other POP3 accounts… generally I only had to name the provider and my account details, not enter all of the server information manually. Good stuff.

Changing mail clients

I was listening to the MacFormat weekly podcast the other day and discovered that the new version of Apple’s Mail.app has some very interesting features. Amongst them are some very cool data scraping capabilities (called Data Detectors) that allow todos, addresses and iCal entries to be intelligently created from analysis of the message body. Here’s an example, featuring Roo’s IET lecture next week:-)

My default mail client up until now on the MacBook Pro has been Thunderbird, but that has been largely a matter of familiarity… I decided that it was time to give the Apple alternative a try.

So far, it has been an intriguing experience. I can’t say I’ve found Mail to be the most intuitive application. For a start, configuration for Gmail IMAP was not very easy (here is some useful additional information that wasn’t on the Gmail FAQ). Not only that, but in Thunderbird and Gmail, I’m used to hitting a key for the next unread email, but Mail inexplicably doesn’t allow this. Two solutions:

1. An Applescript that causes Mail.app to jump to the next unread. I used Quicksilver to bind this to Option-` and it now pops up Mail and switches to the relevant Space as well as moving to the right message.
2. Probably an even simpler option, that Andrew Webb suggested via Twitter: a Smart Folder which only shows the Unread messages.

Now that I’m getting used to it, I’m quite liking Mail… particularly the ability to jump straight into Quick Look to view images and documents.

Trials and tribulations will be reported as the experiment continues.

Technorati tags: Gmail, IMAP, Google, Apple Mail, Mail.app, Thunderbird, SpamAssassin, SquirrelMail, email

Google gobbling

Google’s acquisition of Jaiku came as a bit of a surprise to me.

Robert Scoble says it shouldn’t be a surprise to anyone, and that combined with Google’s other properties, this will all translate into some super Facebook-killing announcement next month.

The thing is, I’ve been a bit amazed by some of Google’s acquisitions. I can’t say that I see all of them as totally successful. Blogger in particular hasn’t seen much love, apart from a spruce-up earlier in the year when they finally started to catch up to WordPress (and WP has continued to jump forward since then). Picasa? Sorry, but IMO Flickr still wipes the floor with that… Yahoo! has been pretty clever about not pasting logos all over the site and has quietly and neatly integrated Flickr, Upcoming, del.icio.us and other properties. Continuing the Google theme: Jotspot, YouTube… not sure they are examples of sites or tools that have been well-integrated into the Google “family”.

As Mark Cathcart pointed out, this doesn’t mean that all of Google’s acquisitions have been unsuccessful. For instance, Google Earth (Keyhole) is superb, and has great integration with Maps, Panoramio, SketchUp etc.. Google Docs is great too, although I don’t personally use it.

So will the mighty new Googaiku kill Twitter? Personally, unless they do something startling as Scoble suggests, I don’t see why it should. My network is all on Twitter, and despite trying both Jaiku and Pownce, I’ve not been tempted away despite various the outages and wobbles Twitter has gone through. It is simple, multipurpose, and remarkably useful… I don’t know what the business model is, but I know that I like the tool.

The social software space remains extremely interesting. Times change. I wonder what’s next?

Technorati tags: Google, Jaiku, Twitter, social software, Robert Scoble, Facebook

Not feeling the Google love

For some reason, it seems like my blog has largely disappeared from Google. I was previously getting significant daily traffic from searches on all kinds of topics (Message Broker, TomTom, and car tax being the favourites). As of yesterday, that fell to a trickle. I took a quick look this morning, and it is nearly impossible to get my site to come up on a search.

I know that the Google Dance happened recently, and it looks like I’m still listed on their datacenters – but for some reason my posts are no longer hits for specific search terms.

Here’s the irony – yesterday the blog appeared in a presentation at Lotusphere based on the stats growth, and also broke the 40k ranking mark on Technorati for the first time.

Back down to earth with a bump – from ~1000 hits/day to just over 300 yesterday. After all the excitement, it looks like I’m in for a lean period

Technorati tags: Google, Google Dance, search, traffic, blog stats, statistics

Plazes meets Google Earth

Over at the Plazes blog, there’s a cool post entitled Plazes + Google Earth

My favourite location-awareness application (aside from Sametime 7.5) now enables you to use Google Earth to browse the location data. So, for instance, if you were to look at it right now – this second - you’d see my head hovering over a hotel somewhere north of Regents Park. Very cool.

Technorati tags: Google, Google Earth, Plazes

Google integrates its webapps

In case you missed it: Google has brought Writely into the fold by integrating it with Google Spreadsheets (here’s my earlier mini-review of Spreadsheets). The result is Google Docs & Spreadsheets.

NB the old Writely URL www.writely.com now redirects to docs.google.com.

Logging in with your Google account leads to a welcome screen. An interesting aspect of the screen is that in the bar at the top you can just see the links to Google, Gmail, Calendar and Photos (i.e. Picasa), so they are increasingly bringing all of their services together.

I thought I’d try creating a new document. That was trivial. Clicking New Document produced an editor screen. The editor itself is very usable, even down to the “standard” keyboard shortcuts working within the editor within the browser. The File drop-down menu provides options to save locally in a variety of formats, including PDF… I tried this and it offered me a PDF download version of the document, which was exactly what I’d typed on-screen.

Next, I saved the document. I was curious to see how it would cope with importing a Word document, so I uploaded a trivial example. I was never a Writely user, so had no idea of what to expect. It worked fine, although I didn’t test out any of the more advanced features of Word. It looks like GD&S can support a range of document types including OO.o and HTML. It also supports uploading via e-mail, which I didn’t test.

There’s a nice, simple list view from which you can open and edit documents.

One interesting surprise was that Google allows you to publish the documents – either to the web via google.com URL, or to your blog. I tried the latter option. I configured GD&S to know about my blog, and then hit the Post to blog button - results here. I can’t see this displacing my normal blogging client, but it is a neat feature and “just works”.

Finally, I took a look back at Spreadsheets, which I haven’t had cause to use since my original mini-review. Nothing much has changed.

Other features that I didn’t try out were the Collaborate and Discuss tabs on the UI.

Overall, it is a pleasant experience. I think there’s a change of mindset required to move across to working with all my documents online, and there are clearly privacy / confidentiality issues for corporates. However, the interface is simple, consumable, and smooth. I’m impressed – but I’m not sure that it will become a central part of my computing experience.

Technorati tags: Google, Google Documents, Writely, spreadsheets, word, excel, office

Hello World, testing Google Documents

Hello World, this is a test of Google Documents.
You can have different styles and fonts (although only a limited range of fonts).
There are keyboard shortcuts like ctrl-B for Bold.

this is a quotation

Very interesting!
I can save as OO.org, Word, PDF, RTF, HTML.

(NB full post on the way…)

New Google Blog Ping

Reviewing the announcements from the Google Blog last week, I came across a note about their new blog pinging service. This post is to see whether Windows Live Writer will work with it “out-of-the-box”.

Update: I tried configuring the ping URL in Tools -> Preferences to be http://blogsearch.google.com/ping/RPC2 and my URL does indeed appear in the changes feed (for the past 5 minutes), so I assume that worked. Just add that Google URL to the list of pingable endpoints in WLW.

Technorati tags: Google, blogging, ping, Windows Live Writer

Google makes Talk better

Google Talk has just been upgraded to add file transfer, music trends, voicemail and buddy icons – basically it is catching up with a number of the other instant messaging offerings. They also have a “chat theme” feature which allows you to customise how the chat window displays the chat (with or without buddy icon, with or without dividers or chat bubbles, etc.). It is not quite as simple as it was when it was launched, but the interface is still clean and tidy. The upgrade was really neat, it didn’t force me to download a new binary, I just clicked Check for Updates Now on the popup menu, and it upgraded itself silently.

I admit that I haven’t used Talk much, but I’ll try to log on to it more often and give it a try. There’s a Google Talk Blog which you can follow if you are interested – in fact it turns out that there are blogs for pretty much all of the Google offerings..

In other IM news, I’m now running one of the release candidates of IBM Sametime 7.5 on my work laptop, and it is really good Not long to wait before the release, now…

Update: Sametime 7.5 is available early – great news!  More here and here. The full release is next week.

Technorati tags: IBM Sametime Lotus

Trivial Google annoyance

Google have changed their page so that the order of search shortcuts is now Web, Images, Video, News, Maps – and then a “more” item which pops up a box with further search types.

My default behaviour is that if I can’t find the answer to my query on a Web search, I immediately go for Groups. That used to be one extra click, now it is two. It’s trivial, but annoying. I suppose I should write myself a Greasemonkey script to put things back the way I want them… actually somebody has probably already written one… [goes off to check]