The closest thing I could find was pages & pages of BIND documentation. BIND is great – if you’re, say, actually hosting a proper DNS. But for my purposes, this would have been like using a tractor trailer to give a neighbor a cup of sugar. So, for a while I’ve been looking for a good solution for a cheap and cheerful local DNS — easy to configure and manage. The sole need was to host an internal DNS that could override, but normally be a slave to our public DNS. After a bit of focus I finally found my solution: DNSMasq. On Ubuntu, it’s super easy to install:

apt-get install dnsmasq dnsmasq-base

DNSMasq will read the local /etc/hosts file – or you can add entries to /etc/dnsmasq.conf

Each developer at Shopvolution uses a standard developer image through VirtualBox – each of which has a static IP. On an in-network server, I installed DNSMasq and added entries for each developer’s ubuntu image into it’s /etc/hosts file – ex:

192.168.35.71 chris.example.com

On the internal server, I then added a real DNS server to the /etc/resolv.conf file – then pointed the router’s DNS to the local server. If the in-network DNSMasq server doesn’t know an entry, say, for example.com (which is managed on our public DNS), it forwards the public DNS server’s response. If it does have an entry, it responds accordingly. Our internal DNS doesn’t interfere with the public entry (unless I explicitly override it) and the outside world has no more knowledge of our routing (nor should it care).

Voila – job done! Now everyone can reach developer workstations from their local browser without messy local /etc/hosts changes.

This seems to be a constant obsession of mine: in agile teams, how do you best carry and share task information, sync tasks to subversion (yeah, I haven’t jumped on the GIT bandwagon just yet), and allow programmers to seamlessly submit task information (time spent, feedback questions, commit comments, etc).

In the past I’ve had the individual elements of this perfect equation by running a subversion server, Redmine, whatever text editor developers felt comfortable with (TextWrangler, FDT, Eclipse+Subclipse, Dreamweaver, whatever)  – but that usually requires a lot of jumping around to different windows to gather task info (web browser), update svn (editor/standalone svn client), edit code (editor), commit back in (editor/standalone svn client),  and update task information (again, back to a web browser).

Ideally, a programmer could view their task list, open a task as ‘in progress’, commit/update svn, and mark tasks complete all through their editor. That way programmers would have a completely (or close to it) uninterrupted workflow when addressing issues. That’s the quest for the moment, anyway. Suggestions most welcome!

At Likemind, not only do we design & develop online annual reports, but during the better part of the year, we’re designing and developing large .com sites for pan-European clients. These sites are often 150+ pages and tie into investor relations, careers information, and PR efforts.

Often our clients have a vision of 10-15 geographically dispersed editors and several stages of approval. Web editors are often staffed with their day jobs plus the responsibility of keeping the site fresh. Clients with large, involved IT teams took the requirements from the business and dictated requirements to agencies asking for large, enterprise scale content management systems (CMS) that had the ability to manage workflow (ie, what happens when an editor pushes the “publish” button), integrate SAP HR functions, create custom forms, perform on-the-spot surveys, custom analytics, and allow certain editors limited privileges over certain sections, full permissions over other sections, and no permission to edit yet another area. In all, these were huge development efforts – and products like SDL Tridion, Documentum, and Sharepoint quickly filled RFPs and developers minds for months on end as there were no other real viable systems or strategies for solving the challenges. These systems – from a technology perspective – were *fun* to build. Teams of developers joining together to bang out a scalable site that had all the bells and whistles imaginable.

And what happened? Huge enterprise content management systems were deployed with UIs inherited from the CMS — so complicated all but a handful of editors who performed content edits day-in and day-out knew how to use the system, obliterating the need for system complexity in the first place.

Don’t get me wrong – when users were trained correctly and knowledge about the system was passed from editor to their replacements over the years, the systems themselves performed brilliantly. But just as often, knowledge of the system resided in the handful of editors and stayed there.

Working on sites for the likes of Yara, Shell, and EDF Energy (to name a very few), Likemind has been learning these lessons and altering our expertise and training guidelines. We’ve also learned that after you answer IT’s questions, it’s time to go back to the business drivers and validate and justify cost & requirements.

But also, emerging open source CMS projects like Drupal, Joomla, Magnolia CMS, and up-and-comer Concrete5 have narrowed the gap between smaller content management systems and the heavy hitters. And not only have these projects become viable contenders for the system themselves, but they’ve put tremendous pressure on the larger content management systems to innovate, clean up their UI, and streamline editorial duties forcing a bit of technical evolution. Also, sister systems – like HR and investor feeds have cleaned up their application programming interfaces (APIs) and XML feeds to work with much less effort with the web content management system. This makes incorporating their data and functionality much much easier and eases the need for detailed custom development.

And so while my friends on the design side of Likemind focus on experiences for the website vistor, ever increasingly our technology team is focusing on the experience of content editors. Every week it seems we evaluate a new RFP for an enterprise CMS & remember to put CMS user needs right up with IT needs. Sometimes scrutinizing these requirements and talking to the client team leads us to a much simpler option. An option that meets the multi-editor requirements, permissions based editing, custom workflows, a shorter technical development, and flexible deployment.

A while back, Chris Bray, a friend back in NYC posted: Another installment of “What you sold, what you built, and what they needed”. A perfect graphic representation for technologists to keep in mind.

But moving on – lesson learned for me: even after umpteen years living in large metro areas, I still need to do a better job of keeping my belongings in sight.

The positives:
First off, the laptop had been on a time machine backup for a few months – but unfortunately the backup had filled up & I had taken to just hitting cancel. Not smart. But I also keep source code checked into a subversion server & email is all server based, so there won’t be a huge problem getting back up & running w/ new hardware.

The negatives:
A few things are just a pain: first, my visa to stay in the UK was in the backpack. Dealing with that will take weeks off my life. My prescription sunglasses as you might know if you’ve ever purchased these sorts of things before, are priceless to me & useless to someone else. So I have a bit of an incentive to get these things back – and thought the stolen laptop might inadvertently come to my aid.

I have dropbox & xmarks on the computer – programs that automatically hit a server and do things (like sync a file directory & sync my bookmarks). Now chances are the thief will just drop a fresh install on to get around my login password & wipe the computer clean before selling it on. But – if – and a big if – they actually manage to get my password or figure out how to login any other way those two programs will hit & I’ll be notified of my new friend’s IP address, hopefully getting me one step closer. A long shot to be fair.

But to the point – there’s an open source project that I’ve (too late) realized would have helped me that much more: Prey is a lightweight application that will help you track and find your laptop if it ever gets stolen. It works in all operating systems and not only is it Open Source but also completely free. We like. It is now going on all my computers & recommending that my office install on all portables as well.

Also, I should have registered mobile belongings here at the UK’s property registrar which may have helped stop the person selling on the laptop on the open market.

Finally, I’m thinking of several ideas which might electronically tie items to me. Say, if my bag had a RFID chip & talked to my phone, I could program it to go off if my bag went out of range….though peripherals are in development and look half baked at the moment, it’s rumored that the next iPhone will have reader built in…here’s hoping…

Though it’s been said that Wave is email’s Segway, Google has a pattern of seeming to bite off more than they can chew – only to succeed in the end. Gmail was, at the time of release, yet another web mail client with seemingly too much hype. But only a few years on, Gmail’s killer feature of unlimited size of your inbox has not only transformed their users’ behavior but now unlimited storage is ubiquitous among web mail providers. Wave is in fact a protocol, not just a Google hosted service and Google is planning on releasing the source code for their implementation, hoping that there will be as many Wave servers running as there are email servers today. With the vision clearly set on completely transforming online conversations, I’m not underestimating their prowess in revolutionizing a seemingly ingrained technology.

However, Wave is pushing boundaries and making bets users will change several engrained behaviors. While developers have been using versioning systems for years, Mom & Dad (unless software engineers) will grope to understand the power of tracking file versions (‘playback’ in wave). Also, users need to switch from thinking linearly about online conversations (email, chat, forums, even twitter) to an asynchronous model – which again is a basic human behavior that will have significant barriers to change.

Unlike Search, Gmail, Google Voice, Calendar, and Docs, in order for Wave to be useful, let alone truly powerful, everyone in your conversation will need a Wave account – either on Google’s servers or elsewhere. Here, Google is counting on two curves of adoption – first in Google Wave, then in the ubiquity of everyone having a Wave account somewhere in the federation of Wave servers. The first will likely have to be a success before the second curve starts its uptrend. At this point with a limited user base and a slow invite process, it’ll be some time until Google Wave hits it’s inflection point.

If you want a Google Wave invite comment & I’ll put your email into the queue.

If you’re really lazy, just hit this:

There you go! Happy Blogger Search!

Incidentally, here’s a tip for all you wireless roamers: when you’re on the road, especially in a public wireless network, it’s best to be really careful about surfing so as to keep your belongings secured at all points (”Did you pack your own bag?”, “Was your bag with you at all times?”, “Have you let every person in the airport know your email password by sending it unencrypted over our unsecured wireless network?”) . I can never remember to be that careful, so here’s what I do – be ready to be geeked.

I setup a secure ssh tunnel and proxy my web traffic through that. I happen to have a few servers that I can run this through, but you can always use your broadband connected home mac (or PC with ssh-server installed). At the command line, I type:

ssh -D 9999 evan@freshkills.xyz.com -p 22

Then in Firefox, I go to my network preferences (Firefox, Preferences, Advanced, Network, Settings) and select “Socks Host” and type 127.0.0.1  with 9999 as the port. This routes all network traffic in Firefox through freshkills.xyz.com over ssh (secure shell).  Fun stuff, I know.

Oh – by the way – check out http://www.no-ip.com/ if you want to setup a free url that points to your home machine at all times. And remember to update your router to handle ssh traffic!

I wrote about Photosynth a while back – a Microsoft owned application that can take photos from a stream and mash them together. So, a source like Flickr would be ideal to contribute to something like …oh, I don’t know, a world collaboration to map the earth? Yahoo owns Flickr. Microflickrhoo could conceivably take a significant chunk out Google’s maps and who knows what else? Maybe that’s where they came up w/ their valuation – throw everything Yahoo away except for a few properties Microsoft doesn’t have…say..Alibaba? My gears are turning on this..