Alex's Adventures on the Infobahn - google

On the death of Google Reader

2013-03-14T22:17:00+00:00

You can probably tell the sort of on-line company I keep from the deluge of noise on the social networks regarding Google's decision to shut down Reader. However we shouldn't be that surprised. In fact some companies that source content from Reader have anticipated the need to collect content themselves.

I of course will have to make a decision at some point. However I'll not do it today like a lot of Reader users have. The rush to try out alternatives has over-whelmed some open source based projects who were quietly growing organically. I don't envy those that have to suddenly gear up their back-end systems because an Internet behemoth gave us 2593 hours notice to sort out a replacement.

I'm mulling over the difference between self-hosting and having someone else do it. I'm not overly worried about going for convenience if I know I can get my data back if I need to. In fact the knowledge that you can theoretically self-host might be enough. To be fair to Google their Data Liberation team made exporting all my Reader data easy.

Before I make a choice I need to decide what my priorities are. Currently I subscribe to 250+ RSS feeds. Obviously I don't read every single post but I make extensive use of tags to quickly process through stuff I do need to see when I need to see it. Aside from news, blog posts, funny cat pictures I also subscribe to other data feeds like bug trackers, code repositories, and other data sources. I of course want access to all of this data at any point on one of a number of devices. This makes a web hosted solution pretty much a must. There is no point having the data on my desktop when I'm somewhere else. From my point of view I want it to be open source compatible because if the company hosting now decides it no longer wants to I'll only have to move the data and not break my work-flow.

It would also be very useful if it had a public API so others can interact with the data. I don't need the solution to be all provided by one company. It's perfectly fine to have multiple 3rd parties sorting out the Android integration. I might even look to doing something to integrate it with my favourite editor (the name of which even my non-geek readers probably know by now). So far my experiment with moving all of IRC and IM into Emacs seems to be working well and should be a subject of another post.

Are you a Reader user? What are your criteria for it's eventual replacement? Is RSS just a dying protocol or is the need to aggregate and sift through data becoming more important?

There may well be a much better way of solving this problem around the corner. I certainly am open to persuasion. But don't take away my current preferred solution until I'm convinced I'm ready to switch ;-)

Switching buffers and Google+

2012-01-24T08:36:00+00:00

When you do so much of your work in one text editor the efficiency of switching between buffers becomes more important. For a long time I've had two bindings "C-x b" and "C-x C-b" which in days of yore I had bound to bs-show and a hacked up list-buffers that opened another window. These are broadly the "quick switch between working buffers" and "show me all the buffers".

For some time I've relegated bs-show to the longer binding and now use Stephen Bach's excellent Lusty Explorer which works really well when you know the name of the buffer and it's fairly unique. However when you've been going a while it can get un-manageable with a large number of open buffers, especially if you've opened second copy of a file from another source tree. This is what I would use the old classic bs-show for.

A few days ago I discovered ibuffer-mode with it's ibuffer-bs-show buffer navigator. Looking back through the Planet Emacsen history I can see it has been mentioned before and given it's been in Emacs since version 22 I'm surprised I hadn't cottoned on to it earlier.

One thing that might have put me off is the initial buffer list can be quite sparse. By default you only see buffers with files associated which misses out IRC, Edit with Emacs and *scratch* buffers. However hit "h" and you'll see there are a plethora of quick keys for chaning the view. A quick "//" and all filters are removed and you can quickly filter by different criteria. To get the most out of the mode you'll probably want to set up some custom filters ("/r<completing filter name">) to make quickly switching to groups of buffers easy. I currently have "work", "remote", "irc" and "logs" as filters. You can filter by name as well as major-mode.

ibuffer-mode does have the concept of Filter Groups although I'm not sure what they add on top of having normal filters which as far as I can tell can be arbitrarily complex. It also has some quite handy sorting and selection modes e.g. "sv" - sort by last viewing time. Given the amount of space the wiki devotes to the topic I wish I'd re-examined my buffer switching habits sooner. The change is already paying dividends for my productivity.

Before I go I thought I'd put in a quick mention of Google+. There is growing community of fellow Emacs users starting to post on it. One thing that attracts me to Google+ over Facebook (too data-miney) and Twitter (too short) is the concept of "Circles". It makes sharing geeky Emacs posts with people that might actually care easy while sparing them the flood of baby pictures I share with friends and family. If you'd like to follow me or my alter-ego please do mention Emacs in your profile or in a message so I can assign you to the correct circles.

Nexus of possibilities

2012-01-06T22:47:00+00:00

Looking back the last few months I notice every post has been about family and parenthood. Time to redress the balance a little and add some geek friendly contents.

After Christmas I indulged myself with a new phone. When I got the HTC Hero I'd deliberately skipped the first generation of Android phones to give a chance for the hardware to mature. I also made the rash promise it would be my last phone for some time. As it happens I think 2.5 years is a fairly good innings for a piece of technology where the innovation cycle is measured in months. I still have the Hero but now it's not my main phone I can afford to be a bit more experimental with the software I put on it.

There were a couple of changes to my approach to choosing a new phone this cycle. The first was I brought it outright off-contract. Although I suspect I could get it slightly cheaper overall through a phone contract there are some non-financial downsides to the contract approach. The first is the operator takes the view that the phone belongs to them so they can install (and prevent you removing) what they like. The recent Carrier IQ controversy should certainly act as a salient reminder that the mobile phone companies do not have your interests closest to their hearts. The second is lack of flexibility. Although thanks to Ofcom we have number portability most people are only able to take advantage of it at contract break points. The operators know this and off course do their very best to keep you tied in to their deals. Now I'm off contract I'm currently paying £10/month for unlimited data and all the calls/texts I use. They have to work doubly hard to keep me as a customer as I can jump ship at any point. That flexibility is more than worth the up-front cost of paying for the phone in full.

As far as choice there was only really one in the frame. Much as I like Android and it's open-source nature it does suffer from a problem due to that openness. Although it's heartening to see manufacturers are starting to relent and cease the practice of locking boot-loaders to prevent 3rd party firmware there are still problems in their support. They tend to stop updating the firmware shortly after they stop manufacturing the phone. Some manufacturers also have a hard time meeting their GPL obligations which makes open source support for the hardware a lot harder. As a result I decided to wait for the next in the Google Nexus series which ships with the latest release of Android, Ice Cream Sandwich.

Initial impressions of the new phone are good. The phone is a little bigger than the Hero but pretty much 100% screen with more than a few hints of Star Trek data pad ;-). The Super AMOLED display looks very clear and crisp with both photos and video. The phone is very nippy thanks to the accelerated 3D hardware and duel core processor. The camera is certainly an improvement on the Hero and with the LED flash is close to achieving the quality I was used to on my old Sony K750. The A-GPS also acquires location a lot faster than the Hero. In fact a lot of things I thought were app problems have cleared up when presented with faster hardware. All in all I'm very happy with the Galaxy Nexus and think it easily stacks up against any iPhone you might want to compare it with.

Social Wars

2011-09-23T12:48:00+01:00

Well it seems the next major technology war has kicked off on the web. Google has opened up its beta social service to all and sundry. At the same time Facebook have had another major face lift and announced their intentions to become a content hub for the Internet.

For me Facebook's latest move is eerily familiar of the early days of the web when everyone was vying to be the default web portal in the browser. It didn't work back then and companies like Yahoo have never really recovered from failing in the land grab. Back then everyone underestimated the demand for search that just worked and when Google offered something that broadly gave you relevant results the curated model of the web died.

However Facebook have a massive advantage now having captured 750m users who visit their site on a regular basis. They didn't come to Facebook to find cool places on the internet but to interact with their friends. Facebook didn't invent social networking but they certainly demonstrated that executed properly it was something that lots of people wanted. So far people have been very relaxed about Facebook's access to all that valuable social information and its use in advertising. Frankly any company that launches any sort of consumer product these days has to have some sort of strategy about how they will interact with Facebook.

Google have know that "social" is going to be a big thing for some time now. They have had several attempts to capture a segment of the market and Google+ is their most serious contender to the space so far. Their reasons for running the service are much like Facebook's although their principle advantage is not having to worry about how to monetise the service yet. Google's history has very much been developing ideas and seeing if they become popular before working out how to monetise it. Meanwhile it's a dangerous time for Facebook as they move towards an IPO where their long-term investors will be looking to cash-out on their investment. If Facebook start hemorrhaging users to alternatives its market value will literally disappear before the investors eyes. In the web-based world where a competitor is literally a click away the last thing they want to do is repeat the experience of MySpace. This would be especially ironic as Facebook was one of the major reasons MySpace went the way it did.

Who will ultimately "win" is an exercise in futurology that is a futile as predicting anything related to the fast changing world of the internet. I personally can't wait to move away from Facebook and to a service that gives me more control over my data. For now that is Google+ but it's by no means certain that's where it will stay. However one thing I'm sure of is the process of taking my data from Google and moving somewhere else will be a lot easier than the exercise has been from Facebook. Coincidentally it's probably the main reason why I'll be staying with Google for the time-being.

What do you think of these latest moves? Do you care who "wins" or what the services turn into? Or is the age of social networking the latest in a long line of internet fads that has already peaked?

Ponies and other requests

2011-07-21T13:30:00+01:00

So Google Plus is not even in beta, it's still a limited trail but they are soliciting feedback. Following on from Rich's suggestions here are some things I would like to see Google+ do.

Allow nesting of Circles

Circles are great. They are a pretty intuitive way of arranging your friends into groups. It's certainly way easier than Facebook's rather clunky friend lists interface. However I have a lot of people in multiple circles but it's a little inconsistent and it's hard to check people are in the right place. To illustrate here is a subset of circles I've defined:

* Fellow Geeks
* Cambridge Geeks
* Googlers
* Transitive Alumni
* #blue

Really the last four circles are all sub-sets of fellow geeks but sometimes it makes sense only to post to a sub-set. For example the Cambridge Geeks are those that are more likely to make a Geek Beer shout out. The Transitive crew are a special set of geeks that would appreciate some really obscure posting that might be irrelevant to the wider Geek community.

What I really want to do is declare Fellow Geeks as a super-set of a the other circles. It's not going to be a one to many relationship either. Cambridge Geeks also should belong to the Local People group because I don't spend all my time socialising with people just like me ;-)

Support Huddle in the browser

I like the concept of Huddle which is a group based messaging system in the mobile version of Google+. However currently it seems the only place to access is on a mobile device and that is currently Android only. To make it useful you should be able to access the Huddle from the browser as well.

Allow rolling up of comments

This is especially a problem with some of the higher visibility users that I follow. They can often have hundreds of comments attached to a post and there doesn't seem to be any way of hiding the comments. All you can do it "Mute" the whole post. Sometime G+ does hide "older comments" but the only thing you can do with them is expand them.

So like a good trail user I'm also raising these issues via the feedback option. I'm still enjoying the Google Plus experience and once I can post from Gwibber I'll be a lot happier. So what niggles have you found so far? Have you reported them?

Plus One

2011-07-17T10:20:00+01:00

Well I've been playing with Google+ for a week and a bit now so I thought I would add some punditry to the swirling ether of the 'net. For a product that hasn't even earned the famed Google Beta tag it's already been rolled out to around 10 million people. A good percentage of my social graph have managed to find their way onto the network. Invites now seem to be permanently open so do let me know if you want one.

So the first thing that struck me about G+ was the circle implementation. It's not surprising that it's such a fundamental part of the experience. I recall seeing a presentation by a Googler some time ago where they discussed the types of relationships people have, the sort of things people share and with whom. While Facebook does allow you to set up friends lists and restrict posts to certain groups the G+ setup is a lot more intuitive and built into every aspect of the service. For example I'm fairly careful about posting my location on-line. Even though I use services like Latitude most people are restricted to City level information unless it's likely to be useful for them to know where I am (for example on a stag weekend). While not many people use Latitude being able to "check-in" to a location that may be relevant to a subset of people is quite a nice ability.

There are a couple of gripes about the circle implementation. The first is duplicates. I have a number of people duplicated because they were originally dragged from my Gmail contacts but have then subsequently joined. It would be nice to have an easy way to group them together and have Google Do The Right Thing (tm).

The second is the default circle screen needs a little work for small screen setups like my netbook. Having created quite a number of circles the screen real estate gets quite tight when trying to add more people to them.

Other nice features include the group video chat known as "Hangouts". It works out of the box on Linux although it does require a proprietary plugin to work. I haven't had more than a cursory play with it so far but it certainly seems slick enough from what I've seen. It's also nice to have an alternative to Skype who's Linux support could disappear any day now Microsoft have taken it over.

The mobile app offers the "Huddle" which is a simple group chat facility. I can see this probably being of most use when catching up with a group of people on some sort of shared activity like a pub crawl.

The photo facility is fairly slick although I don't think it's quite got the edge on Facebook's auto-tagging feature. It remains to be seen how else they will pimp it up. Certainly it's nice to be able to upload everything from Shotwell as it basically re-uses the Picassa infrastructure. It sure beats fighting with the basic upload facility Facebook offers.

The one major omission and probably biggest reason I use Facebook is for event management. For all the problems associated with inviting people only in the walled garden Facebook's event creation/invite system is pretty slick. I suspect Google are planning some sort of event/calender integration. I hope when they do they make it easy to add emails so people don't have to be in G+ to receive invites.

So is it time to close down my Facebook account? Well for the time being probably not. Facebook does have the incumbents advantage given almost everyone who wants a social networking account will have one. However I'll certainly be using G+ in preference for sharing photos and private status updates. Google's Data Liberation Front give me more confidence that my data won't be kept away from me. It will also be interesting to see what APIs Google release for the platform. Perhaps the biggest problem with Facebook apps is the privacy implications of sharing your data with a third party. It will be interesting seeing how they balance the privacy-centric nature of the G+ ethos with the ability to do cool stuff in the social sphere.

Baseline Codecs for web video

2011-01-17T13:15:00+00:00

A lot of 'net comment has been generated in the last few days following Google's announcement that they will be dropping support for H.264 in future versions of their Chromium browser. They expound on their decision here.

In making this move they join the ranks of Mozilla in supporting Open Video in web-pages. The trouble with H.264 is that while it is a documented standard it's not a free one. To implement a H.264 codec you have to use patented techniques that are controlled by the MPEG-LA. Despite claims that licenses will be given away freely for decoding non-commercial video it doesn't meet the requirements of Free Software which needs to be able to distribute implementations freely and without additional restrictions. It's a problem the Open Source world has been talking about for years and it's good that Google are bringing these concerns to a wider audience.

For all practical purposes the change will have little effect for my day to day browsing. I can already view pretty much any video format on my Linux Desktop and if Chromium has to pass decoding to a plugin instead of doing it within the browser I probably won't even notice. This will have more of an effect on Apple's iOS devices (iPhones and iPads) when places like YouTube switch to using either WebM or fall back to a Flash based player. Even today you can't browse a large amount of media on Wikipedia which favours free software compatible audio and video formats. That is because Apple refuse to add support for these alternative codecs to their mobile OS. Not surprisingly they also have a vested interest in H.264 becoming the de facto format for video on the web being one of the many licensees represented by the MPEG-LA.

It's noticeable that the majority of the comments on Google's blog involve people throwing their Chrome shaped toys out of the pram and returning to Safari. Windows users will I assume have the same ability as me to handle formats via plugins instead of native browser support. It does lead me to think that their anger is somewhat misdirected when aimed at Google.

A number of people have pointed out there is a degree of hypocrisy in Google's stated support for open web standards while their browser (in non-free software form) still supports Flash and MP3s which have similar issues. I'm willing to concede this isn't ideal from a software freedom perspective. However I'm happy to accept their pragmatic point about this being a statement on support for open codecs for the emerging <video> tag without wanting to cripple the browser for the amount existing content out there. Now is very much the time to make stand over video lest we repeat the mistakes of MP3's and GIF's in the rush of companies to stake out their own little revenue generating area of what should be the free and open standards that underpin the web.

Google Gate-gate

2010-10-25T14:13:00+01:00

Google was on the news this morning for the latest piece of fall-out from their WiFi survey.

For the uninitiated for some time the Google Street View cars have been scraping WiFi data as they roam the streets. The intention has been to build a database of SSIDs as an alternative way of locating position than using GPS or Cell Tower information (which I assume is out of Google's control). It's a testament to Google's resources that these sort of mass data collection exercises are seen as a simple exercise. I do have to wonder what a geo-location database of thousands of "linksys" WiFi SSIDs will achieve though*.

However it seems the configuration of the information gathering was a little over zealous. Instead of just recording WiFi information they also recorded live network data, i.e. what was being transmitted when the Street View car went past peoples houses. This has rightly gotten Google into hot water with the Information Commissioner for the sort of mass surveillance that GCHQ just wish they could afford. A lot of noise has been made about the potential capture of private emails and passwords. There is more than a hint of hysteria building up with MPs getting in on the act and rent-a-quote heads turning up on TV. However as per-usual there is a distinct lack of perspective.

For one thing the window of data is very small. Unless you had a street view car parked outside you house the amount of potential privacy busting information have on you personally will be quite low. It also ignore the fact that when sending data across the Internet you don't actually need to be outside someones house. The Internet is made up of a collection of networks run by a disparate group of corporations and individuals. None of these organisations have the security of your data at the top of their list of concerns and any one of them could actively be snooping on every packet of information you send. In fact as far as networks go it's about as untrustworthy as you get.

As far as passwords are concerned if your not sending them over an encrypted channel then anyone can see what they are. At the very least you should ensure all web-sites your log-in to over a SSL link (usually signified by a padlock or similar icon in your address bar). Nowadays the padlock may turn another colour or show a cross if there is some other problem with the connection that may indicate your connection is secure but maybe not to who you thought it was too. None of these precautions addresses the issue of if you should trust the remote end not to tell anyone else your user name and password. This is why you should have different passwords for each service you use and not share credentials between sites (i.e. only give your Google logon details to Google).

When it comes to email not many people seem to be aware that it's less secure than sending a postcard. Everything in the email is readable by any system between you and the person your sending it to. Even worse any system could tamper with the message and make alterations the message and neither you or the recipient would know it. If you really want to send email that can only be read by you and your recipients you really should use encryption**.

So while I applaud the ICO for bringing Google to book for not thinking through what they were doing please realise it's not the new Big Brother. Big Brother has been quite happy reading everybody's data before the first Street View car took to the roads and ultimately you are the only person that can actually do anything to stop him.

* - slightly facetious, if they were recording MAC addresses they would be guaranteed unique.
** - my key can be found here

A few paragraphs

2010-07-26T10:33:00+01:00

There are a number of potential topics for blog posts swirling in my brain which I'm not going to have time to write. I shall therefor attempt to address each on in paragraph form (Twitter style is a little too brief for me).

We were at Maelstrom this weekend. We splashed out for a van so we could take our proper IC tent to the event which we also slept in. Although our normal sleeping tent is quite spacious the IC tent is a lot more so and benefits from not broiling you in your sleeping bag when the sun rises.

I tried Google's Navigation mode on my software pimped up Hero. I can see why the shares in TomTom and Garmin shares fell so sharply on it's announcement. I particularly liked the street view picture that's presented at the end of the journey for extra confirmation. I predict the era of the dedicated GPS road navigation unit is coming to an end - to be replaced by multi-functional smart phone type data pads.

I awoke this morning to the latest Wikileaks data dump as reported by the Guardian. I can understand why the armed forces are upset about this but it's really an OpSec fail on their part that someone can extract that much data from their systems. Going after Wikileaks will ultimately be a futile exercise. Also baring a live twitter feed of current manoeuvres I'm not sure there should be an expectation of secrecy for historical military data. Transparency is coming to government which I think is a good thing, why not the military? At the same time we need to take some steps to educate the public on the limits of data. If people consider an entry in a database as equating to actual truth on the ground they are going to find themselves disappointed.

We watched the latest Grand Moff/Gatiss collaboration on TV last night. Although the final third had us shouting at the TV for the slowness of the protagonists to put two and two together overall we liked it. The text overlays were fun and the writing has done a good job of bringing the characters forward into the 21st century. We shall be watching the rest of the series.

Adding Google Juice to mutt

2010-03-17T07:35:00+00:00

As I've been mailing out invites I discovered a minor problem with my data.

My main email client is the fantastically functional mutt. It's terminal based but incredibly flexible. When it comes to mass sorting/searching your email it leaves GUI based clients standing. However now I'm a roving around with a Google Phone the majority of my contact data is in the cloud. While I have a small address file used by mutt it only has a few oft-mailed addresses in it.

Luckily thanks to Google's data APIs your information is only a few RESTful requests away. The goobook program provides a handy mutt compatible address book interface to this cloud data.

There is one wrinkle however. The configuration of the script involves putting some rather valuable login details in a plain text file on your home partition. While I like to think my machines are pretty secure and maintained you can always do more. Good security is defence in depth. A quick patch later and I can store those details in an GPG encrypted file that can be decrypted on the fly when required.

The final piece of the puzzle is creating these encrypted config files in the first place. Although you can do this by hand from the command line I find the best method is using EasyPG (now part of Emacs 23). This will automatically cause any files with a .gpg extension to be encrypted. You can control the Emacs mode selection and default encryption key to use by using file variables in the header comments of the file.

It's not all perfect though, when enabling EasyPG I had to do the following:

(if (maybe-load-library "epa-file")
    (progn
      (setenv "GPG_AGENT_INFO" nil) ; gpg-agent confuses epa when getting passphrase
      (epa-file-enable)))

The problem seems to be that when GPG agent runs in terminal mode it confuses Emacs/EasyPG. By suppressing the GPG_AGENT_INFO environment variable EasyPG will fall back to requesting your passphrase in the mode line. While it takes care to flush the value as soon as possible it does open a small window of attack if an attacker can cause emacs to crash and dump core.

Google's change of heart

2010-01-13T08:34:00+00:00

So Google are finally having second thoughts about their relationship with China. We may see then end of google.cn as a search engine that filters search results that might embarrass the authorities.

Google got a lot of stick at the time when they first created their Chinese presence. In reality they were doing what every other western Internet company had to do to get access to the massive growing market that is China. I suspect the reason they got more stick than most is because it was one of the early tests of their (in)famous "Don't be Evil" motto, a statement of purity that is never going to be easy to live up to while operating a profitable business.

Now that Google seem to be resigned to not being able to operate inside Chinese law I wonder if they would be willing to go the next step. There may be many arguments for censorship but secret censorship is very hard to defend. If you don't know what you aren't allowed to see how can you be sure of the motives of the powers that be that create that list. What I would love Google to do is share that list of censored results so we know exactly what it is the Chinese authorities don't like. For bonus points Google could rank the list in order of search popularity so we know what is the most requested censored pages are.

I suspect this won't happen as Google, despite their new approach, will want to maintain as cordial relationship with the government of the largest growing market in the world. While google.com does periodically drop off the radar in China every now and again (as it does not filter results) the temptation must surely be to ensure it has at least some visibility rather than none at all.

EDIT TO ADD: More information about hacks used and speculation about why. I'm just glad I don't use Adobe Reader...

Going Mobile

2008-01-08T14:33:00+00:00

I spent bits of yesterday getting final meter readings in and cancelling my Sky account. The rest of it was taken up with a fair amount of cursing at Call of Duty 4 as I struggled my way through act 3. Having the rest of the household point out the danger of hand grenades every time I bit the dust again didn't help that much. I'm really quite enjoying the game but it does get quite hard in places. I'm tempted to say it's one of the best FPS games I've played and I'll stick my neck out and say the experience is probably better than playing on the PC. After all not many PC's have the wide-screen display my PS3 does. Although a lot of PC gamers pour scorn on the controller vs the classic mouse and keyboard I think it's actually a better control method. The two sticks give quite fine control when you are maneuverings and of course all the buttons for grenades and weapon switched are in ergonomically helpful places.

I switched my mobile tariff a few days ago to one that included a decent amount of data. I subsidised this by dropping the number of free minutes as I wasn't quite using all 900 every month. As a result I've been playing with the mobile web.

First up is Opera Mini. It's a little Java app that runs on your phone and runs as a clipping browser. This means it takes normal non-phone optimised web-pages and attempts to render them in a vaguely sensible form. Generally it works very well although I personally preferred the older versions rendering of the BBC new site. The new version added a panning display instead of forcing the text to flow down one column which was a little simpler to read. The only website that has so far failed is the ever present Facebook and that could just be down to finger trouble.

I've also played with some of the Google services. Of all the services Google Calender and Google Maps (separate application) are the most mature. The calender functionality is basic but enough to check your appointments while on the go. Adding a simple events with a natural language (e.g. "Meeting at 1400 tomorrow") works well enough. Given the growing ubiquity of mobile data services I can see my days as a loyal Palm user coming to an end. I've been using the calender a lot over the last year and especially like the ability to have different ones with different collaboration profiles. It's extremely handy for Fliss and I to have one shared calender for all out activities (even though I suspect I rely more on electronic memory augmentation than Fliss does). If the integration with Evolution pans out it will pretty much solve all my calender requirements.

The Google Maps application warns you it is data-intensive but on modern 3G networks it still runs pretty quickly. They have taken care with the design of the UI, especially in the zoom control. You can very quickly get a map of the right scale for navigation in the limited space of a phone display. The only disappointment is the non-GPS "My Location" service isn't supported on my phone. The list of supported devices is a little vague.

There is not much to say about the mobile version of Google Reader except that it works reasonably well. However I suspect the reading of large numbers of RSS feeds is going to hard to optimise for a mobile phone sized screen.

Google Apps is also available in a limited mobile flavour. You can't edit documents but you can certainly view documents and spreadsheets (if you select desktop mode, the default spreadsheet view is one column at a time which isn't all that useful) with Opera Mini. I don't expect to be using it much but it certainly could be useful if you just want to check the last time you called a certain utility company while in the pub.

The last app I played with was MidpSSH which is a GPL'ed Java ssh client implementation. I doubt you will be wanting to do much with the hassle of text-typing to the command line but it may prove useful for an emergency server reboot or service restart. Having said that I haven't managed to log onto one of my boxes yet (I haven't tried very hard yet).

Web 2.0 apps that are worth it?

2006-10-18T14:29:00+01:00

Web 2.0 is a lovely little marketing buzzword that is bandied about a lot. It refers to the concept of applications running on your web browser. It used to be that people thought Java apps would be what everyone would want but it turns out that AJAX can provide you with quite a powerful application experience.

The one I've been playing with quite a bit over the last few days is Google's RSS Reader. RSS readers are an exercise in saving the time poor from too many clicks. I used to use Liferea and then when that stalled I started using the Fixefox extension Sage. However I've finally come across a web application that is truly comfortable to use.

It seems a natural fit of course. RSS is a product of the web - why would you want to leave your web-browser to read stuff? Much as I find the Live Bookmark capability of Firefox useful it doesn't give the status view of what is new. However Google Reader allows you to set up aggregate feeds really easily which are an ideal compliment of Firefox's live feed view. The natural integration of feeds into the web environment is one of the key selling points of the reader for me.

Another thing which is very well done is the keyboard shortcuts. I've not seen them used so widely in a web application before. I'm definitely a keyboard person. Compared to navigating with a mouse it's far more efficient to use a couple of well placed keys. On Reader they have made the process of browsing through articles pretty painless. Switching between feed "tags" is a simple 2+n key sequence which auto-completes to the best fit tag within 2 or 3 keys. "Staring" and sharing items are similarly easy operations.

The one big bonus of all of this of course is that it's independent of which machine I'm on. I can pick up from were I left of if I'm on the laptop, desktop or work machine. For this reason Google's Reader picks up the first Bennee award for a truly use-able web application.