When you view a private file (i.e. a file not in your Public folder), you have the option to “Copy URL to Clipboard”. Normally, one would expect that to be the private URL for your private file. That is, the long URL that requires a DropBox login to view.

Not so in this case! What happens is, a short URL (using DropBox’s own URL shortening service, db.tt) is generated for the file. However, it’s a direct link to download the file. It’s not simply a shortened URL to point to the private URL for the file.

Big security risk? No, not a huge risk, but it does mean that there’s now a chance that your private file will be “brute forced”, or guessed, by someone. The private URL is as good as your password (which should be more than 6 characters! It is, right?!), but this short URL is only as good as the 6 alphanumeric characters.

Something to be aware of (particularly if you’re used to using DropBox outside of the iPhone client), rather than cause for panic!

A further complication is the fact that Google App Engine uses Python version 2.5 (and not the default 2.6 that comes with Snow Leopard). If you just need PIL for Django, you can ignore the 2.5 references (i.e. python2.5 becomes python).

Here’s how I got it working on my machine (the steps are a combination of others’ blog posts, but main credit goes to Michael Richardson‘s post):

1. Download the latest libjpeg .tar.gz (version v7 at time of writing): http://www.ijg.org/files/.
2. Download the latest PIL version for Python 2.5 (v1.17 at time of writing): http://www.pythonware.com/products/pil/
3. Extract libjpeg and do the standard install affair:
export CC="gcc -arch i386"
./configure
make
sudo make install
from within the directory. The first line is the gem that many other blog posts leave out (and it assumes you’re using BASH)
4. Extract the PIL and change in to that directory.
5. [optional] libjpeg will have been installed in /usr/local/lib/, by default, above. Normally this directory will be picked up automatically by the PIL install below, but just for good measure (or if you have problems later in the PIL install), edit the setup.py, look for the line
JPEG_ROOT = None
. Change that line to
JPEG_ROOT=libinclude( "/usr/local" )
and save it.
6. Install PIL by running
sudo python2.5 setup.py install


That was enough to get it running for me. I can start App Engine (I use the Google-provided launcher), and use the Image API.

To remedy the problem, I substituted the call, in 2 places, with the PHP4 equivalent calls. file_put_contents(..) is a shortcut convenience method which is the same as calling fopen(..), fwrite(..) and fclose(..).

As of version 1.1.3 of the plugin the code is under wp-content/plugins/wordpress-mobile-pack/plugins/wpmp_transcoder/ in your WordPress install directory. The 2 occurrences are in the file wpmp_transcoder.php on lines 431 and 448 respectively.

I changed

@file_put_contents($full_location, $data);

.. to ..

$fhout = @fopen($full_location, "w" );
@fwrite($fhout, $data);
@fclose( $fhout );


.. and ..


@file_put_contents("$full_location.meta", "< ?php $"."width='$width';$"."height='$height';$"."type='$type'; ?>");

.. to ..

$fhmeta = @fopen( "$full_location.meta", "w" );
@fwrite( $fhmeta, "< ?php $"."width='$width';$"."height='$height';$"."type='$type'; ?>");
@fclose( $fhmeta );

.. and all was well again.

You could also just not use the “Shrink images” feature to avoid having to mess with any code!

The good news is that there is a workaround! Add *R* and a space to the start of the SMS message (not the number; the message body). While this code will appear in the conversation window, the recipient won’t see it. o2 will strip it out before sending.

When your SMS is delivered, you’ll get a message back which is the raw delivery report. It’ll appear as an SMS in your conversation window.

It’s not an ideal solution, but you do get notification of delivery, at least.

This code also works on Meteor and Tesco Mobile. For a list of codes for networks other than o2 Ireland, see url.ie/2hgg.

ClubW, based in Cork, come to the rescue! It’s clearly an ongoing engagement, and anything in social networking requires time, but the first reaction when one visits MyMunster.com has to have an underbelly of “meh!”.

I’m writing this review wearing a number of hats, which makes it harder. I, of course, strive to be objective. MyMunster.com is clearly aimed at Munster Rugby supporters (check), and users of social networking sites (check). I’ve spent over 8 years authoring collaborative software which has taken a massive swing to the social networking side. That doesn’t make me an expert, by any means, but it does help me identify strong and weak areas.

MyMunster Strengths
The site offers the basic feature of any social networking site; private messaging, public messaging, discussion forums, and areas to share your media. I.e. everything one would expect from a social networking site. We’ll look at those features in a bit more detail later. One twist on social networking that the boys in ClubW bring to the mix is monetisation, and they do that very cleverly!

MyMunster Weaknesses
My biggest gripe would have to be that there is no easy way to consume the information being shared. There are no ways to filter data to suit you. There are no categories of discussion threads, no shared albums, no tagging of content, and when it comes to the news page, it’s all or nothing. The site is lacking in functionality to make it easy for its members to contribute as well as to gain from the site.

MyMunster features:

• Membership.
  To get any value from the site, you must first register, as is pretty much standard these days. I’d expect alternative login methods (e.g. OpenID, Facebook Connect, etc.) on a bigger site, but I’m sure these will come as the site grows. However, I’m still waiting on a confirmation email. This is a necessity for a number of reasons including my own archives (so I know what email address I used), and my own security (so I can report anyone trying to register as me, without my authorisation). ClubW have stuck one gem in the registration, and that’s to encourage users to upload an avatar there and then. Often social networks are devoid of avatars, and this can look poor, but MyMunster have placed appropriate weighting on the value of avatars. You can still skip the step, if you’re not that way inclined, which is, in my opinion, the best way to handle avatars.
• Discussions.
  There is a standard discussion forum available, as you’d expect. It’s a threaded forum, so people can start a topic, or reply to an existing topic. Standard stuff. One big failing of MyMunster.com is to categorise these discussions. It’s a social network based around Munster Rugby, so you’d imagine most discussion will be based on that. However, if you’re to encourage user participation (the measurement of success of any social network), you need to cater for other discussions. MyMunster does this, but lumps all discussions in together. The result is that it’s virtually impossible for a member to focus only on topics, or categories, of discussion. They have to crawl the entire page (and subsequent pages if they are bothered, which research shows they are not!) for a discussion that suits them. MunsterFans.com, the biggest competition to MyMunster, and indeed most discussion forum sites, does do this. Possible categories for MyMunster could include a minimal “Munster Rugby”, “Ireland Rugby”, “General Sport” and “Banter”, and expand over time as patterns of discussion emerges.
• Messaging.
  ClubW have themselves emphasised that this was an important feature, and they’ve revamped MyMunster accordingly. There are three types of messaging available, and all easily accessed from the Messaging page. You can message everyone, in a Twitter-esque form, you can message individuals in a public way, and, finally, you can send private messages to individuals. Given the recognition of the strength of messaging, I’d expect to see further UI tweaks here to make it easier (AJAX username lookup, layout of speech bubbles to put more emphasis on the message rather than its meta data, etc.). It looks like some people are using it in a Twitter kind of way (including the “I’m going for lunch” style of updates), in which case there’s an argument for an auto-refresh of those pages.
• Media Sharing.
  This area, by the site’s own admission, is lacking. They have 2 separate areas for photos and video, although the latter isn’t functional yet on the site. I’m not convinced of the need to have 2 completely separate areas (Flicker, Facebook, and many other media sharing sites don’t differentiate). For now, at least, you can upload photographs and organise them in to albums. The photos are public, and can be seen by other members of MyMunster. What is absent, is the ability to tag photos, which makes it easier to search, and to display multi-member albums like “Heineken Cup Final 2008″. Right now albums are all separated. This space could be vastly improved with the addition of tags. It could also be improved by allowing import (or even just display) of albums/photos from Flickr, pix.ie, Facebook, etc. Similarly, for video, once the feature is implemented.
• News Aggregations.
  Again, nothing new here, but it’s handy to have news specific to the community on display. Again, organisation of the data here is lacking. All the various RSS feeds are dumped in to one big page of data. Information overload. There needs to be a way to filter out Welsh national team injuries from Munster domestic fixtures. I’m not arguing that the RSS feeds should be there, just that they need to be better categorised so that I can manage my information better. I shouldn’t have to read 50 odd headlines (including scrolling) if I’m only interested in the 10 ish Munster-specific ones.
• Blog
  The MyMunster blog seems to be regularly updated with a nice mix of news & opinion on Munster Rugby, with site news & help. I’ve added it to my RSS reader anyway.
• Monetisation.
  This is probably MyMunster’s only unique feature that hasn’t been done to death on other social networking sites. The lads in ClubW have clearly got a grip on the commercial aspect of social networking. I’m sure they’ve experienced this feature as a deal-breaker when negotiating sales. They haven’t just gone for sticking Google Ads on the site and hope that covers hosting and development costs. The system is a credit points based system. It’s a macro currency, as it were. You can top up your credits by purchasing with a credit card, and you get credits when you join to help you get started. You also get credits when you participate (for example I got 5 credits for uploading a photo). So active members are rewarded without the need to splash any cash; a huge incentive. You get to spend your credits on entering competitions (like free tickets, etc.). Plenty of potential here, and easily MyMunster.com’s stand out unique feature.

Enhancements Suggestions
The ClubW lads obviously have a good head start in that they have a functional social networking site with basic features, and have the backing of a huge commercial entity (worth many, many MyMunster credits in advertising!). My constructive criticism can be best summarised as follows:

• Categorise data. Make it easier for your members to digest the information other members share. Some will be interested in what everyone says, most will not. Most will have a limited amount of time to digest information from the site. Use categories in the discussion forums. Categorise the RSS feeds “Domestic, Province, National, Other Rugby”. Yes, it might take more work to manually parse the RSS feeds, but promote some active members of the community to monitor/categorise the feeds and give ‘em credits!
• UI tweaks. A very personal thing for the finer detail, but there are areas that could do with enhancements that have long been acknowledged as requirements on other sites, rather than personal tweaks. Things like auto-refreshing the live updates messaging page, using a Adobe Flash-based file uploaded (to show progress, handle timeouts, support multiple uploads, support zip file uploads, etc.). There are areas of the site that are in need of a usability review. I won’t dwell on the point, just to note that it’s in need of some effort.
• Data Sharing. Right now it’s not too easy to share data between members. Yes, there is messaging, but that is limited. I’d suggest the support of groups “The crew of the Munster Red Renault 19 Ferry Trip to Cardiff ’06″, or at least add multiple recipient support (so I can message Dave, Joe and Tom, and then can easily reply to all). For data/video sharing, there really is need for more than just uploading to individual albums. I’d suggest adding tagging support would be the easiest way to create shared group albums, e.g. “#cardiff08″. It would also make a nice tag cloud feature as a good starting point to offer casual browsers. Tagging support would fit nicely in with the public messaging as well.
• Data Privacy. One cause for concern when I uploaded a picture was that I wasn’t sure what rights, if any, I was giving up. Google, Facebook, etc., have all been in high-profile media criticisms over what happens to user content uploaded. You should have a clear policy on this. I had a quick, and I emphasise quick, glance over the Terms and Privacy pages, but couldn’t see anything obvious. If you’re not claiming the right to royalty-free use of my content, then brag about that, rather than say nothing at all.
• Leverage other social media.
  Encourage users of other established networks to share their content in to yours, or indeed to sign in to yours without the necessity for registration. There may be value in registration, but there are ways to make this easier for members of other networks (e.g. OpenID or Facebook Connect), which you should leverage. Similarly, leverage the established methods of sharing the data between those networks. “Import album from Flickr” or “Create Flickr-based album” along side “Create a new photo album” and similar options. Indeed, offer an export to Facebook/Flickr from your site, so that you are the originator of the data; an added incentive for a member to use your site. Twitter is an obvious area for integration as well. Basically, there are other sites that do messaging, discussion, media sharing and that do them better than you have done, however harsh that may seem. My suggestion is to tap in to that.

Summary
MyMunster tries to be the jack of all trades, master of none. If a corporation is looking to implement their own social network, with monetisation built right in, then ClubW can definitely deliver. However, to build a social network and make it grow, then MyMunster is lacking in too many areas to make it easy for this to happen. They don’t harness the power of external resources (APIs, other social networks, etc.), meaning it’s a very isolated network. Or, in other words, a very expensive network to build. This has been offset by monetisation, but it’s unlikely that’ll be enough, long term.

MyMunster are on a winner with the monetisation, and have the very valuable backing of a huge brand. However, the ability to grow a social network requires more work from the boys at ClubW in functionality. ClubW and MyMunster; off to a good start, but more needed.

Having used a few similar site, most notably FreshBooks.com, I’m drawn back to 1time because of it’s simplicity. I was able to set up my own details and a client’s details in a matter of minutes (versus at least half an hour with other similar services).

As a self-employed web application developer, 1time is ideal in that it allows me to track time down to granular levels and report on that time, by category, by client, or by a range of criteria. My favourite feature of 1time is its built in client reporting. I can simply give my client a (reasonably secret) URL where they can get reports on my time tracking (related to their projects, only). They can drill down and see where my time is being spent. I don’t have to worry about sending ongoing timesheets.

1time stands out from the rest, because of its ease of use. There are a lot of features in 1time that I don’t use, but on exploring those, I can see that they’re all as easy to configure, use, and report on, as the rest of the application that I’m more used to using.

The only feature missing that would really suit me, is a clock facility (where you start tracking your time, with a mouse click, and come back and stop it when you’re finished, automatically tracking the time spent). I’ve been informed that this is on the way. Great! 1time have never received money from me, but they were quick to assure me that a feature I want (again, for free) was on the way. I admire that!

I’d certainly recommend my peers to give 1time a shot; it costs you nothing. It can automate a lot of time tracking, at let you categorise it as appropriate for your client. Your clients will thank you!

I’ve blogged about Toddle in the past, as part of the Tuesday Push, as that was the only site I’d used that was being pushed. I aim to continue with that principle (only review/talk up something I’ve used). However, from now on, I’ll actually go and use the site/product so that I can review it.

First up, I’ll be posting about 1Time, a nifty time-tracking website. After that, I’ll aim to keep up with the Tuesday Push!

You can read more about the Tuesday Push, or sign up to have your site/company pushed.

@donncha and @micheleneylon tweeted about the stats of the blogging technology behind this year’s Irish Blog Awards, organised by Damien Mulley. It got me thinking about the stats of the software used by all those short-listed in the blog awards. Here they are!

Blog Software

First up are the overall stats. A resounding win for WordPress (both hosted, and self-installed). WordPress is by far the more popular choice, with Google Blogger/BlogSpot a distant 2nd. Between the 2, they’ve pretty much wrapped up Irish blogging (that are award-winning, anyway).

Blog Hosting

Next, just over one third of bloggers use free blog hosting services offered by Google’s BlogSpot/Blogger and WordPress.com, with the majority preferring to install the software on their own servers, or hosting accounts:

Other stats

Other, bordering on silly, stats include:

• 65% of short-listed WordPress installations are out of date, not including WordPress.com hosted blogs, obviously! 85% of Winners are out of date.
• 15% of short-listed WordPress installations hide the version tag in the headers (that WordPress ask that you leave intact). 8% of winners hide it.
• There are no 2 same versions of MovableType installed amongst the short-listed
• The only category where all the blogs are of the same type is Best Newcomer – Sponsored by Teamworkpm.net, and all of those blogs are out of date (with 1 having hidden version info). So the newcomers are choosing WordPress, but not keeping it up to date.
• Best Music Blog – Sponsored by DownloadMusic.ie have the best record for up-to-date WordPress versions (a WordPress.com hosted blog is automatically up-to-date), whilst Best Technology Blog/Blogger – Sponsored by Bitbuzz and Best Newcomer – Sponsored by Teamworkpm.net have the worst record.
• Only 15% of the short-listed are on .ie domains. 20% of winners are on .ie.

Installed Blog Hosting

Finally, Blacknight are the big winners. They’re the most popular host, by far, of self-installed blogs, amongst the short-listed. They’re also joint-favourite hosts for all of the short-listed, including free blog hosting services and self-installed blogs.

Take out free blog hosting services (like WordPress.com and Blogger/BlogSpot), and Blacknight are clear leaders:

Summary

WordPress and Blacknight are the winners, in technology terms, of the Irish Blog Awards ’09.

Note: the sample size is 100. There were 19 categories plus 1 overall winner (source: IWA Finalists Blog Post).

The raw data is in a Google Spreadsheet here.

Any other metrics required?