One of the biggest problems I've always had with maintaining a blog is the maintaining part. I often find myself too busy, too indifferent, too burned out or, more often than not, too brain dead to care. With the advent of Twitter, things have gotten even worse. I no longer have to sit and think about how to make a worth-while blog post based on a single, tiny string of thought.

Thanks to Twitter, blogging, for the most part, is pretty much dead. Want to share something cool? Well, you no longer have to write a few paragraphs about it, just post the damn thing to Twitter or even Facebook. Thanks to software like TweetDeck, and other push notification services, I can not only post to multiple twitter accounts, I can even update my Facebook status; all with a single click.

Publishing minutea is now easier than ever and you don't have to spend hours making 5 paragraphs from a simple 140 character message.

So, it is because of this small problem I threw together another Posterous integration script. This is basically a cron script I run once a week from the main site which pulls all of my tweets from the previous 7 days, compiles them into a digest and publishes them to my default Posterous blog.

As you can see from my previous entry it's already working. I figure this is a great stop-gap measure which keeps content flowing to the site in between real blog posts.

There are times when I just don't have the time or energy to spend on a decent blog post. As you can see from my archives, good and worthwhile entries are few and far between. Hopefully, this won't always be the case, but at the moment small measures such as this will have to do.

If anyone is interested, I threw this script up on my Github account as a public repository. Feel free to fork it and create something better. It isn't the prettiest block of code I've written, but I prefer the KISS method for something like this.

Here's to Twitter and micro-blogging!

Loading mentions Retweet

I mentioned in a previous post that Ember, a great-looking photo sharing site, was one of the many image hosting services that I wanted to incorporate into Snaggr's growing list of supported providers. Unfortunately, at the time of that posting, their API didn't support image search. I really liked the look of the site and the mix of submissions, so I went ahead and asked them if they planned on adding this functionality into their API. This was about a month or so ago. 

Loading mentions Retweet

A few things, which were of higher priority, came up over the weekend that took up quite a bit of my free time. As a result, some API updates didn't make the weekend's tentative deadline. Not such a huge deal as it's currently the holiday season and bug testing search integration stuff isn't the biggest thing on people's minds at the moment. I think I'll set a more realistic goal to finish documentation and do a closed beta test before the end of the first week of 2010. I have a few days off and won't really have much else to do during the day, so I'll just save most of the work for then.

This should relieve some of the self-imposed stress I've been having about getting the documentation rolled out. I'll just release a few pages a day until everything is sealed up nice and tight. As always, suggestions and critiques are welcome.

Oh, yeah. Speaking of which, I activated Tender Support's email forwarding feature. This should allow you to post suggestions, ask questions and report bugs directly from your email client. I had to upgrade to Tender's "Plus" package, but I think it's a worth-while investment.

As always, be sure to follow us on Twitter and check the site out!

Loading mentions Retweet

So much progress has been made within the past week and a half with Snag.gr it's unbelievable. It's pretty obvious by now that my original timeline was a little too ambitious and I apologize for that, but it's only because I needed this platform to work perfectly. If all goes well, this system will need to process thousands of API calls an hour, so I had to make sure it was designed to handle this kind of load.

To accomplish this, I essentially had to rewrite the entire system from scratch. From an architectural standpoint, this is probably some of my best work. This system is quite flexible and was built for growth. Also, an indirect benefit resulted from all this trouble. I've added so many layers of abstraction to this system that it's basically an API platform. So, in the future, if I want to provide another online service which could benefit from the use of an API, I can easily plug this into the application's existing infrastructure and, shazam, I have a turnkey system ready to use! All that's needed from this point is to create site-specific API service modules which,  thankfully, is an incredibly painless process!

Naturally, a big shout out needs to go to the Kohana team for creating such an easy-to-use framework. It seems as though it was tailor-made for projects which require rapid development cycles. Thanks, guys! If it weren't for you, this endeavor would have taken far longer than it already has.

I'm actually in the middle of writing up all the user documentation on how this platform will work for the public, but I'll be posting new entries for each service in the coming week. Once everything's all nice and written up, I'll be posting it to the documentation area of the main site.

I also thought you should all be made aware of the fact that private testing should, hopefully, begin this upcoming weekend. So, if you're keen, and I hope you are, I'll provide you with temporary API keys!

As always, be sure to follow us on Twitter and check the site out!

Loading mentions Retweet

I've been doing some thinking for the past few days about the initial direction of the Snag.gr project. Originally, I had intended for this project to be a publicly accessible image search engine and, for a short while, that's exactly where it was going. But, recently, I began wondering about how others might want to use something like this. Maybe, there's a use for an API that 3rd-party developers could plug into which provides the very same service as Snag.gr. Then again, maybe not.

Luckily, there was already a "sort-of" API in development for a potential, unannounced, partnership with another image service. With this in mind, I thought it would be an even better move if I built the API functionality first and then built the public search engine on top of it. It doesn't make much sense having 2 completely different systems performing the same functionality within one site. At least this way, the underlying search functionality will be tested and proven enough to provide a reliable starting point; the rest of the site will only be a layer on top of this API.

So, with that said, Snag.gr is to release an image search API. I've been working on some finishing touches and, within the next day or so, the site will have the initial API fully documented and ready for public viewing. The actual API should be ready for private testing within the next week.

When I'm ready, I will post an announcement asking for testing participants. At  this point, they will be given an API key and user account on this site so they can interact with and manage their searches. I'll be honest with everyone, I'm not sure if I want this API to be a 100% public offering or if I want to give full access to a limited selection developers. We'll see in the end, I suppose.

You will also notice a few updates to the site. It's much more responsive now and you will see I've added a few new sections in preparation for the actual launch of the service. The original microsite was just a placeholder which pulled the latest tweets and blog entries for the service. The updated version uses the Kohana framework and contains all the brains of the upcoming API!

More to come, so in the mean time, be sure to follow us on Twitter and check the site out!

Loading mentions Retweet

I spent a few hours today throwing together a microsite for Snag.gr. Nothing too special, but it's a decent placeholder until the new functional parts of the search engine are ready to go. It just lists the latest entries from the Snag.gr Twitter account and Posterous blog at the moment, but I might add a Flickr photostream down the road for interface updates and whatnot.

We'll see what happens, but in the mean time, go ahead and bookmark the site!

Loading mentions Retweet

Looks like 101Domain finally submitted my request to fully register Snag.gr and it finished propagating last night. Nothing up at the moment but a simple splash page, but I plan on throwing up a microsite this weekend that will pull the latest blog entries and twitter updates for the service. I have some big things planned for this, so, hopefully soon, I'll be able to throw up the new version and start building a roadmap.

I went ahead and setup a Twitter account for the site. So, if you're interested, go ahead and follow @SnaggrTweets. I'll be posting more updates on this and other upcoming services very soon.

Stay tuned!

Loading mentions Retweet

I realize you all could probably care less about play-by-play logo concept updates, but I appreciate any feedback I can get. I just took my previous draft and made a few updates to it. I think it's turning out quite well and I am, for the most part, happy with it. I've included the first draft for reference.

   

Click here to download:

Snag.gr_Logo_Draft_2_tags_Deve.zip (9 KB)

I'm a bit torn on a few things, though. At first glance, you can tell that some bits have a drop shadow while others don't. Everything has a glossy finish to it, but while it's a very hard edge effect on most of the text and shapes, the blue dot has a curved one. It's a bit inconsistent, but we'll see what happens in the end.

Who knows, maybe I'll just end up hiring this guy again and be done with it.

Tell me what you all think!

Loading mentions Retweet

I think I made a great decision the other day when I purchased a new domain name; snag.gr. I figured it would be the perfect choice to replace the name of my poorly-titled pet project; Imaggregator.

For those of you who don't know what Imaggregator is, think of it as an image search aggregator (hence the name). Say, for instance, you're looking for a nice stock photo of a Mustang. What do you do? Well, you search for one on the web. But, where do you search? Flickr, Google, Bing, Picasa or Photobucket? In most cases, you'll end up searching multiple sites individually until you find a suitable image with the appropriate license. Not that big of a deal unless you find yourself doing this multiple times a day. Searching all these individual sites can be a serious inconvenience.

I figured people shouldn't have to search in all these places. So, with a bit of thought and research, I came up with a way to pull results from all these sites with a single search. It's much easier just typing in what you want and then have the ability to browse the results from multiple sites on one page.

So, what's the problem? The problem lies with the name of the service. In it's most basic form, Imaggregator is an image search aggregator; it pulls search results from multiple sources and aggregates them into a single page. The name makes sense, but you have to admit it's a pretty poor choice. It's too long, difficult to spell and say, not very attractive and a bit too nerdy. When I refer to the service by name, people usually give me that "what the hell were you thinking" look.

"It's a brilliant idea, but the name is fucking terrible.", is the usual verbal response. So, I set out in search of an appropriate replacement. The new name needed to be short, catchy, fun and, most importantly, inexpensive. I mean, let's face it, the project isn't making any money at the moment, so I don't really feel like dropping $1,200 USD a year (ahem, swi.pr) on a domain name. After a few hours of brain storming with mind maps, a thesaurus and the help of Doma.nr, I came up with a brilliant name: snag.gr!

It's a play on the word "snagger", as in to "snag" or "take" something. In this case, at least for the first phase of the site, it snags images from multiple sources. I think this is a perfect replacement for the service. Lucky, too, as it isn't even finished yet which means I won't have to worry about search engine indexing issues and what not.

It will take a few days for the registrar to complete the process and I'm stoked you can register Greek domain names without any registration restrictions. Got it at a very fair price from 101domain.com for $79 USD for 2 years, too, so I can't complain. Personally, I think it's a great investment as I plan on spending quite a bit of time on this service. It's a great idea and I believe people will find it to be useful.

I also went ahead and threw together a little logo for the site. I wanted to keep it simple and I think I pulled it off. Maybe later on down the road I will hire a real designer to give it, and the rest of the site, some flair. I guess we'll see what happens!

Comments and critique are always welcome!

Loading mentions Retweet

Yeah, yeah, I know it's been a while, but I've been working on a super-secret project for the past few months. After which, I'll find the time to finish Great British Walks. Speaking of which, I should probably shoot myself in the face because it's been in development in some form or another for the past few years. Not sure what my problem is as I usually finish my contracted projects post haste. Honestly, Pete is the most patient guy in the universe... and I suck.

ANYWAYS... A lot of people have been asking about MyTopix lately and where they can download a copy. I'm assuming the memo didn't go out to everyone, so I'll just say it again. MyTopix is no longer supported by me in any shape or form. It is an open source project now and can be forked by anyone who wants to keep the project alive.

For those of you who are interested, you can ALWAYS find the LATEST build tucked away in its Google Code repository.

Seriously, it's right there. Check it out, create a fork and publish your own distro. If you people like it so much, go ahead and keep it alive and you'll definitely have my full backing.

Loading mentions Retweet