02 Dec 2004
I should prefix this by stating that I don’t really know an awful lot about Sxip. The concept is interesting but I have yet to have the opportunity to pla
y with their system. However, they’ve been in the news recently so I’ll be reading over their materials on sxip.org, m
ight even try out their java development kits.
I think they look pretty cool. I’ve been meaning to listen to a talk by Dick Hardt on www.itconversations.com, but just haven’t had time to do it yet.
Its interesting to see how they’re building their market. Obviously they have an understanding of how open-source works (and how to make a business out of
it) and are already building substantial grass root support. That kind of an approach works very well with open source and they could really take off.
They are targetting the building blocks that people use for a lot of commercial applications (SDKs for php, perl, java, python and recent support in Drupal
and Bitflux). By attending all the notable conferences they are targetting and getting the attention of the medium-senior level application developers,
the kind of people that are active in technology decision making.
Microsoft has been trying to build a useful single-sign on network with their Passport service but it’s fairly platform specific. I’m a Java guy so it’s p
retty useless to me. Sun has their liberty alliance but I have no idea how that’s going (have not really had a need to follow it to date). Sxip is trying
to be as open as possible by providing SDKs and support for many different platforms. I’m not sure how mature their spec is but personal identity managem
ent on the internet will only become more important and evolve over the short-medium term. It’s something I’ll keep in the back of my mind, and it’s alway
s nice to see another local company doing well.
27 Nov 2004
My hovercraf
t is full of eels: Assistant Orange Peelers
The author raises a very good point. I don’t believe you can be successful (or perhaps, its more difficult to be successful) if you break software develop
ment down into a series of specializations. Although a developer may have more training and intimate knowledge in a specific area of code, it is not effic
ient to let him/her focus solely on that area.
A good developer needs to have the ability (ie: knowledge and tools available) to diagnose and debug a problem regardless of where it may lie and what tech
nological cross-sections it might touch.
When I look for a developer, I look for someone with good cross-sectional exposure to technologies and methadologies. These are the types of people that a
ble to easily adapt to different environments and share their experiences while contributing to valuable discussion. What I term a vertically focused deve
loper (someone who is a specialist in a single area) could potentially be very set in his ways. Design/implementation designs should be based on a holisti
c view of the entire system and not on someones personal perferences (although the experiences of a domain specialist are valued).
This leads to my last point, the over-reliance on specialists to make far-reaching decisions.
Hypothetical Situation:
Company: We know we’re doing something wrong and its going to bite us later, but at that point we’re going to bring in a high-priced outside senior perso
n to magically fix the problem for us. For now we’ll look at ways to “optimize” the code.
Maybe I’m not optimistic enough, but I don’t see the point of waiting for someone senior to come in and point out the mistakes (and magically fix them) wh
en you could start taking iterative steps towards reaching a suitable goal right now. I’ve been in situations where I’ve been brought in as a so-called ex
pert to provide man power on a dead-end project. It’s difficult to fight past momentum and really make the right (and necessary) kind of change. So, on a
whole, all developers need to be agile and prepared to work collectively on all aspects of a code base. It’s not possible to have intimate knowledge of t
he entire code base, but it should be possible to have a birds-eye view of the different components and perhaps their most important points of interaction.
The code is owned by the collective and although you may not be a specific expert in a given technology, that’s no excuse not to raise a concern or quest
ion a course of action.
Management should encourage their developers to cross-train and become self-sufficient, I believe a cross-sectional approach to development far outweighs
a specialist vertical approach.
24 Nov 2004
The software company I work for, GenoLogics Life Sciences Software, is looking for a few qualified software develo
pers.
We definately do J2EE and would encourage all qualified applicants willing to relocate to the very livable Vancouver Island, British Columbia. We’re a gro
wing (and fairly technologically progressive) software company with a strong focus on biotech management software.
A full listing of jobs is available here. I would be curious to see what kind of applicants a
blog posting attracts so if you are thinking about applying, feel free to drop me a line or mention where you saw the posting on your application.