Currently @  ::  Compendium  ::  Business blog  ::  Personal blog  ::  Gallery  ::  by Svein-Magnus Sørensen

Computing Archives

All entires posted in the Computing category.

Tags: conference, defrag, innovation, social, technology November 28, 2009 Experiencing Defrag 2009

Defrag conference 2009 logoThe past year I've been speaking at several both large and small conferences in Norway, but this month I also went abroad to speak at the Defrag social technology conference in Denver, Colorado. This conference is one of the most interesting I have attended, so to share my experience I've written this piece about the experiences and insights that I got out of Defrag.

Now if you'd like to start off by getting an impression of what went on during the conference before I get into my analysis, then go have a look at the Defrag 2009 liveblog that Graeme Thickins did throughout the event. Another good starting point is to look at the twitter-talk that took place with the #defrag and #defragcon hash-tags, which is all documented at Defrag's EventVue page. Finally there is a guerilla video stream covering most of the conference that were being created and put online by ReussDesign. My talk on open data was also filmed by Reuss and can be found about 12 minutes into the recording titled "Defrag Conference Clip 4".

Pre-conference dinner
The conference experience began with the pre-conference dinner on the night before the official opening. Unlike most conferences this dinner wasn't just a meet&greet, but also a conference session in itself where all attendees were given a choice of groups to join for dinner, each group being sponsored by a different company with a set discusson topic for the night. I had selected a topic titled The Inbox as Filter hosted by Gist, a web-service that aims to improve email through improving your connections and relationships with the people of your inbox. Of course we didn't stay on topic during the entire meal, but combined with the presentations by Gist it was a great way of to start off conversations among a bunch of like minded guys that touched onto a range of interesting topics throughout the italian tapas dinner we were served. Absolutely a great way to kick start a conference!

Defrag Conference Day 1
The next morning the conference itself began with a short opening remark by organizer Eric Norlin before moving on to the first keynote by Andy Kessler, accompanied by digital gadgets in every hand of the audience. The Defrag crowd would easily out-tech any of the developer-conferences I usually attend, well helped out by the gracious availability of both WiFi and power-strips everywhere at the conference. I'd estimate that nearly half of the audience sported laptops during the talks, with most of the remainder was keeping a keen eye on their smart-phones. This was especially interesting during Andy's talk as it caused quite an uproar among the audience, but all of it happened in the silence of the digital back channels so Andy himself might have been entirely unaware of it until leaving the stage.
One of the effects that this had on the conference might have been somewhat unsettling for some speakers, as only a fraction of the audience were paying attention to the stage at any time. The rest were mostly busy enhancing their conference experience by following Twitter and the other back channels, the chatter of which in many cases turned out to be as interesting as the speakers themselves. For myself I absolutely felt that this secondary stream of information was a good addition to the conference experience, and while it took some of your focus away from the talks it also provided fresh perspectives while the topics were still relevant in your mind.

Defrag%20stage%20IMG_6076.JPGNext up on the agenda were presentations of four current problems, as well as the afternoon topical explorations, neither of which I found very interesting as many were nothing but poorly disguised marketing pitches for a company or product. Due to this I found the largest value of this conference to be the small-group interactions and networking with all the amazing attendants, something that came especially well to light in the very high quality open space discussion session held just before lunch. This was unfortunately the only such session in the program, and something the conference could only gain from adding more of next year!

Crowding back into the main hall after the topical explorations we were greeted by a series of very interesting 10min "fragments", of which the presentation on Atlassians use of 20% time were especially inspiring. Then to sum everything up at the end there was a lively panel discussion with Chris Sacca and Chris Shipley, the two 'douchebags' :-) doing a great job of setting the mood and giving everyone a bunch of laughs before the evening reception with the sponsors. Unlike some other conferences I've been to there was no organized dinner in the evening of day one, so during the reception people clumped together in groups more or less at random to go out for dinner separately at various restaurants around central Denver. Afterwards those most eager to network randomly reconvened back at the hotel bars to squeeze the final bits of action out of the day, which let me have some great chats with amongst others Robert Scoble and Eric Knipp.

Defrag Conference Day 2
The second day of Defrag proceeded in much the same way as the first, with a few talks being very interesting while many were a bit light on content. Especially interesting were the fragment on the Synaptic web by Khris Loux, as well as the collaborative keynote on Discovery vs. Search where Robert Scoble among other things showed one of his FriendFeed-streams on-screen, a literal torrent of information flowing past as you can see for yourself beginning at 5:25 in this clip from the panel. Other than this the chats with all the great people attending were still the highlight of the day, and at the stands in the lobby I was introduced to a bunch of upcoming technologies including both brand new ones like Lijit and promising ventures that have been around for a while such as Xobni. Amongst these I got a great demo from Atlassian on their next version of Confluence, a tool I'm working with a lot these days, and I also got turned onto as a serious Sharepoint-competitor in the hosted space.

Defrag%20speaker%20IMG_6077.JPGDay two was also when I did my own talk in the topical exploration session called 'Leveraging the Open Web'. In brief the session started out by an introduction to the industrialization of content creation Peter Sweeney, after which I covered the basics of Open Data from where I handed over to Paul Miller to talk about Linked Data. After the talks the session was then rounded off with a short debate facilitated by Ben Kepes. I must say that I rather liked the format of the session that began with four lightening talks around a central topic leading into a catalysed group discussion where both the speakers and the audience got engaged, however I believe that it could have been made even better by setting aside more time for both the speakers and the facilitated discussion, as the constraints forced the exploration of the topics to be cut a bit short. For a more detailed summary of the discussons in this session see the coverage by CMSWire, where you can also find writeups of the other topical explorations at Defrag.

After these sessions the conference wrapped up with a keynote discussion that looked back on the 10 years that have passed since the publication of the Cluetrain manifesto. It was a very interesting summary of the thoughts that have shaped the American internet economy, but it didn't feel very relevant to me due to the limited influence this book has had on the IT business in Norway thus far. Hopefully this will change a bit over the next couple of years.
After this final panel the conference was a wrap, and everything died down pretty quickly with most people understandably being tired from two long days at the conference, so we said goodbye to everyone except those of us with later flights out of Denver that ended up going out to dinner together and keeping the conversation alive for a few final hours.

The digital back-channels
As I mentioned the digital back-channels were a very important part of this conference, much more so than at other conferences I've attended. This meant that a lot of people spent much of their time following the continuous stream of tweets and other web-content about the conference that were published during the talks, some instead of following the actual talks themselves, which might have been just as interesting in some cases, as shown with Andy Kessler's talk mentioned above.
According to one summary the Twitter-hashtags clocked in at nearly 5000 messages during the two days of the conference among the 400 participants, meaning an average of 5 tweets per minute if you only consider the actual conference hours! In addition there were quite a few blog posts and articles published along the way, something one could easily keep on top of using the OneRiot real-time social search engine, so obviously the things going on at stage weren't the only thing holding the attention of Defraggers :-)

Not only did the back-channels provide a very interesting set of insights and commentary that you otherwise wouldn't get, but through these channels the conference itself can even be influenced directly by those only following the online streams, as described in this post about twitter and the ricochet effect. Such live coverage as this not only gives the attendants an extra level of insight and connectedness at the conference, but it also provides a lasting reflection of the event through all the tweets, blogs and videos that are published online for posterity. This is especially valuable for those that didn't attend, as the insights gathered can be found in the many blogs that has been written about the conference. For those interested in more on Defrag, these posts provide a good overview:

Last but not least an overview of more Defrag posts can be found at Lou Pagilas Defragging the Defrag Coverage.

Posted by Svein-Magnus Sørensen at 22:20

Bookmark/Share Email this Reddit Digg! Stumble! Technorati Facebook Linkedin Ma.gnolia Google

Tags: browsers, hack, preferences, social media, twitter October 6, 2009 Getting around the real name length limit in Twitter

Twitter logoSince I started using Twitter last year I've been especially annoyed with one thing, and that is the arbitrary length limit on the real name field in settings. The field is limited to 20 characters, but my full name unfortunately is 21 characters including spaces. This means that I've had to either truncate part of my name or remove the spacing between my first and last names, neither of which are good solutions when considering one of the main reason for having the real name field at all, namely search engine findability.
As can be seen from the forums at Get Satisfaction, I'm far from the only one having a problem with this, but despite this there isn't any satisfaction to get as Twitter themselves are dead silent on the issue. Luckily I have now discovered a work-around for this limit, published here for the benefit of all those with a "long" name of more than 20 characters!

I got the idea for this work-around when I was once again checking up on the name-issue in the Get Satisfaction forums to see if there had been any new developments. As usual there were no news from Twitter officials, but I noticed a post from @McoreD who claim to have spoken with the Twitter developer who implemented the change that limited full names to 20 characters. That this limit was a late addition is the key bit of information that put me on track to finding a work-around.

This info was significant for two reasons: First of all it meant that the backend database would likely still be able to support longer names than 20 characters, and second it made it a possibility that all of the validation code wasn't properly updated to match the new limit, which is a common programming error when modifying field limits. As it happened both of these assumptions turned out to be correct with the consequence that it should be possible to input a longer real name by simply removing the field length limitation in the browser, which is something you can easily control yourself. As I had hoped this worked perfectly, and I am now enjoying my 21 character real name on Twitter!

Please note that while the fix appear to work perfectly for me with my single extra letter, it might not work with longer names or this could cause weird problems with Twitter or especially with third party clients or services, so use the method below only at your own risk!

Now onto the details of the workaround, which is actually a bit of a hack. There are many possible ways of unlocking a form-field in your browser, but this is the easy way I did it:

  1. First run Firefox with the WebDeveloper add-on installed.

  2. Log in to Twitter and go to the settings page for your account.

  3. While there select "Forms" and "Remove maximum lengths" from the WebDeveloper menu.

  4. Then type in your long real name in the correct field, and simply click save!
After saving you'll see a validation error about the long name, but despite this the name is still being stored for me, so hopefully you too can now enjoy Twitter with your full real name!

UPDATE: It appears that while the long username-change seem to work for a while just after-being applied, it is reverted back to the old name within a day or so. This has now happened to me twice, so the hack unfortunately isn't very permanent. Sorry :-(

!!! WARNING !!!
While this work-around could probably also be applied to the Username field, I would strongly advise against doing this as it could cause serious problems with API-clients and SMS-services that depend on the Username-field being of limited length. Ultimately messing with this could get your account deleted!

Posted by Svein-Magnus Sørensen at 18:46 | Comments (1)

Bookmark/Share Email this Reddit Digg! Stumble! Technorati Facebook Linkedin Ma.gnolia Google

Tags: avatar, hacking, site, troubleshooting December 2, 2008 Updated Gravatar plugin for Movable Type 3.x

My GravatarWhen launching this blog I started out using Disqus for the comments functionality as it seemed a good choice that would be more interactive than regular comments. However I soon discovered its many downsides, including a lack of search-indexing and the instabilities caused by having the comments section generated in Javascript, which currently is the only option when using their Movable Type plugin. There is a v2.0 plugin for Wordpress that avoids these downsides, but there has not been any word from Disqus on whether a v2.0 plugin for MT is forthcoming. Due to this I have now returned to the regular MT comment-system.

However I still wanted users to get neat avatars next to their comments, and the easiest way I found to do this was to add support for the Gravatar user-pic service and Favicons to my comment-listings. Adding them was very straightforward as there are ready made Movable Type plugins for both, except of course that neither of the Gravatar-plugins worked. *Sigh*
After some research it turned out that the Gravatar-supplied plugin is outdated and also only supports Movable Type versions pre-3.0, and the MT supplied plugin only supports MT 4.0 or later. Combine this gap in supported versions with Murphy's Law, and it is just obvious that I still had to be using MT 3.3. I did upgrade to MT 4.x once, but it caused so many problems for me so didn't keep it for long, and attempting that upgrade again was not an option right now as I just don't have the time to get it working properly.

That leaves the option of hacking the plugin! Yay! With no plugin-documentation to be found and me never having toyed with MT-plugins before, the task naturally stumped me a bit at first. However I figured out soon enough how to get it working, so for anyone else having the same problem I hereby present you with the updated Movable Type 3.3 Gravatar plugin! Enjoy :-)

And now all that remains is for my readers to get their own Gravatars. Go fetch!

Posted by Svein-Magnus Sørensen at 10:33

Bookmark/Share Email this Reddit Digg! Stumble! Technorati Facebook Linkedin Ma.gnolia Google

Tags: performance, troubleshooting, windows November 21, 2008 Fixes for slow browsing in Windows Explorer

Folder iconAfter a recent defrag of my harddrive it suddenly became excruciatingly slow to browse 'My Computer' and other folders with Windows Explorer (explorer.exe) in Windows XP. Sometimes just opening a regular folder with a few files in it would take more than 5 minutes, as well as having explorer.exe hang and be not responding. As none of my other applications were noticeably affected it really had me stumped as to why a regular defrag would cause such a slowdown for folder-browsing, until I discovered that browsing was still near instantaneous when using other file-explorers like Total Commander. Relieved that it wasn't a problem with my drive but likely just a bug in Windows Explorer I set out to find a fix that would bring it back to its old self, but that was easier said than done, and I spent several days searching and experimenting with various fixes.

Apparently many people are having problems with Windows Explorer being slow, and for a host of different reasons too. Most commonly I found the obvious suggestions to run windows update, antivirus, antispyware, defrag and chkdisk, as this will commonly fix the performance and many common issues that crop up on computers that are not kept and maintained by professionals. The next step is to improve performance by adjusting the Folder Options in Explorer. Good tips here are to disable the automatic search for network folders and printers, as well as using simple folder view and to not cache thumbnails. But it didn't make any difference and explorer was still just as slow afterwards. Then I downloaded and ran CCleaner to do a full systems checkup and registry cleaning, and I also removed all recent network paths from my "Network Neighbourhoood" as these things also appear to cause many slowdowns too. Still none of these suggestions made any difference. A thread at Google Answers hinted to NeroVision Express as a possible culprit, but I didn't have that installed of course.

Finally I stumbled across a Techspot thread describing how you by logging in as a different administrator user and deleting the folder "c:\documents and settings\<username>\local settings\application data\microsoft\windows", can fix the problem of extremely slow browsing in Explorer, and this actually worked!
It seems that something had corrupted that part of my windows user profile, but deleting the folder and inducing an automatic recreation of it on the next login fixed all of the problems I was experiencing. Another good thing to come out of this is that all the other tweaks has made the rest of my computer faster too :-)

Posted by Svein-Magnus Sørensen at 13:48 | Comments (3)

Bookmark/Share Email this Reddit Digg! Stumble! Technorati Facebook Linkedin Ma.gnolia Google

Design and content by Svein-Magnus Sørensen. All rights reserved ©2002-2022

Slideshow from my life.