Cadifra UML Editor[^] is a nice lightweight UML diagramming tool. Cadifra diagrams can be OLE embedded in Office docs. The XML output is simple and clear -- it only took me a half a day to write a code generator from the output of Cadifra's FSM mode. Regards, Doug

Hi Paul, You don't want to be buying cables from a AV retail store. They slap a huge markup on cables, knowing that consumers consider cables to be a commodity item. Most customer are totally focused on the price of the TV, and don't pay close attention to the price of the cable that is casually added to the deal at the end. For example, at Tiger Direct (in the US) you can get an HDMI-to-DVI-D cable for US$20[^]. I expect there are similar prices available in Ireland. Cheers, Doug

I really like Perforce -- it is hands down the best config management software I have used. Client and server both run on pretty much any platform you can name (Windows, UNIX/Linux, Mac, etc.), and the Visual Studio integration is excellent. For a small development team (1 to 2 devs) Perforce is free, and they also offer free licenses to open source projects. Cheers, Doug

Hi Håkan, 1) Based on your criteria I would consider either Vancouver or Ottawa. Ottawa can get cold in the winter (below -10C would be normal in January and February), but on the other hand if you like the outdoors there are lots of winter sports. Vancouver is rainy all winter (like never see the sun for four months), but comparitively warm. In the summer both are great. Vancouver has the ocean and mountains and warm weather, Ottawa has hot weather and lakes and rivers and hills. Ottawa is fairly close to Montreal and Toronto, and a reasonable drive away from interesting American cities like Boston and New York. Vancouver is close to Seattle. Vancouver is a big city, so you might drive a lot, depending on where you live. Ottawa is a smaller city (< 1M), so it is easy to bike to work or downtown or into the countryside. Vancouver housing is very expensive (by Canadian standards), Ottawa's is reasonable. 2) Both Vancouver and Ottawa have good high tech communities. Ottawa's is large due to the telecom and government sectors. Healthcare workers are in demand in Canada, but as someone else noted you need to be careful the professional qualifications will be portable. My impression is that European degrees are recognized, but the specifics vary from province to province, since health care is a provincial responsibility. 3) Wages vary a lot depending on industry and specific company. A senior developer could be anywhere from C$80k to C$120k. I think the salary for an occupational therapist would be around the $50k range. 4) I'm not sure about the immigration paperwork, being a native myself. Canada gives preference to skilled workers. Have a look at this government web site for lots of good info: http://www.cic.gc.ca/english/skilled/index.html I hope this helps. Good luck! Regards, Doug

My little company develops Java under the latest version of Eclipse. It is pretty great, as is VS 2005, but Eclipse excels in different ways. If you are looking for a clone of VS for Java development you are probably going to be disappointed. I am going to try and re-pitch Eclipse to you :-) I would contrast Eclipse and VS by saying that VS emphasizes polish and solidity, while Eclipse emphasizes feature richness and extensability. I find that VS has very clean, obvious work flows, while Eclipse is more opaque but has loads of little features and knobs, some of which are a kind of odd. To pick a couple of specific examples: First, the VS C# code formatting and coloring is still painfully basic -- my 12-yr old emacs config is still far more capable. But the Eclipse for Java is the first IDE I have found whose code formatting and coloring capabilities exceed emacs'. As a second example of the IDE contrast, the VS project management and external source code control plugin system are all-around excellent, while these functions under Eclipse are kind of wierd and clunky. In these two examples, clearly Eclipse team cares deeply about the the joy of beautiful code, while the VS team cares deeply about developer producivity. As for alternative IDEs, I tried Netbeans, but IMO it really doesn't stack up to Eclipse, except in Swing GUI layout. IntelliJ IDEA is highly rated, but it costs $, so for a hobby it may not meet your needs. Really for Java you are looking at Eclipse. Regards, Doug

Sure, you *can* use proprietary drivers on Linux, but within the Linux community as a whole there is a lot of antipathy to the idea. Strictly speaking the kernel license doesn't accept loading of non-free loadable modules (e.g. drivers) into the kernel. Of course there is nothing stopping you from doing this, but Linux folks frown on this. The problem is, as pointed out by an earlier poster, that the above is an untenable position, and the Linux folks are being driven by dogma rather than common sense. It does not seem to be possible to write a *good* open source driver for bleeding edge video chipsets, because the chips churn quickly and the chip programming APIs are proprietary and not publicly available. My Ubuntu box has an ancient NVidia GeForce2 video card, and even for that old card the proprietary 'nvidia' driver is far better in several ways than the mediocre OSS 'nv' driver. And to get the "good" non-free driver I had to reconfigure the package manager to download the driver and then manually edit my /etc/xorg.conf to install it. For the critical non-technical majority this is a non-starter. Speaking of Ubuntu, currently the most popular of the non-commercial distros, it does not support Java out of the box, due again to OSS dogma. Don't get me wrong, I think Ubuntu is great, and Ubuntu and Canonical have done great things for GNU/Linux. But not supporting Java out of the box is just silly. For me, a professional programmer, this is merely irritating and inconvenient, as I know how to modify the 'apt' repository config so the package manager can find Java and install it, or I know how to download and properly install Java manually from Sun's site. But again, for the critical non-technical majority this is a non-starter. My stock broker's trading platform is a Java app. If I was non-technical I wouldn't know how to install Java, I wouldn't be able to run one of my daily-use sw packages, and I would therefore reject Ubuntu as a desktop platform. Modern GNU/Linux distros are very good. If you have not tried GNU/Linux for a few years (or ever) then you would probably be shocked at how good it is. IMO it is 80% there. But 80% is simply not good enough for the mass market, and that final 20% will require a culture change. The lack of broad Linux community support for proprietary hardware drivers, proprietary codecs, proprietary DRM (e.g. Apple's FairPlay), and proprietary software (e.g. Java VM) is a deal breaker for most people. Cheers, Doug

After more than a few years in any position one becomes comfortable -- too comfortable -- and the rate of learning drastically slows. The people I have seen stay many years in a position start to stagnate, and they begin to lose the ability to cope with new technologies and situations. Working in tech is like living on the African savannah -- movement means survival, and if you stay still you're dead (professionally). The sad thing is it is often in an employer's interest to keep people in one place: the company gets an expert in a narrow area, and the employee finds it harder and harder to leave. Good for the employer, bad for the employee. Don't fall into this trap. It is not that the employer is intentionally bad, it is just that companies put their best interests ahead of yours. In your first position out of school all of the above is doubly applicable. In your first position you will think that how your organization does things is the only way to do them, and you don't have the breadth of experience deeply understand that every organization does some things well and others not so well. Early in a career you need breadth of knowledge: learn diverse technologies, different ways of solving problems, different ways of organization people, different processes. Expposure to different ways of operating is the only way for you to decide what you believe in. Realistically the only way to develop this is to work in several companies early on.

My background: I have worked at a multinational telecom company, a couple of well funded startups of 100+ employees, back to the telecom company for a few years, and now I'm at a bootstrap startup. If I were you I would skip the big company and go to the startup. The major reason is you get to do what you love, and that is worth a lot. Telecom is certainly interesting (that's where I grew up), and was a hot area for about 10 years, but it has entered a period of stagnation. Telecom today is where the defence industry was in the early 90's -- lots of downsizing and consolidation, and not a lot of movement in the state of the art. The routers being designed today are not significantly different than those designed 5 years ago. There are a lot more interesting areas of technology to work on. Life moves a lot quicker at a small company. You will learn more about the profession of software development at a small company in 1 year than you will at a big company in 3 years. There's a reason that big companies are referred to as the Dilbert Sector. Big company politics are fun the hear about second hand, but not so fun to have to put up with. The big company I worked at had plenty of sharp engineers working there. But so did all the small companies I worked at. The difference was that at the big company we were often hamstrung with heavyweight processes and comporate inertia. At the small companies there was more freedom to take chances and achieve interesting things -- and also to fail. The best reasons to work at a big company are the relative stability, the good benefits, and the low risk -- the company generally protects you from doing things that will harm it and by extension yourself. But if you have some fire in your belly, are confident in your abilities, and are tolerant to more risk, then a good small company will offer you a more satisfying career path. As far as the CV goes, in my experience as a hiring manager, and also as an engineer being hired, there was no inherent advantage to having worked at a big company versus a small one. It was what you *did* at the previous jobs that mattered. And in almost all cases you will do a lot more at a small company. I have interviewed and met many big company people who had incredibly narrow work experience. I met few small company people about whom I could say that. Finally, hiring managers at small companies usually look very suspiciously at candidates that spent

My last organization used Clearcase, and it was brutal. It is almost like Clearcase has complexity for complexity's sake. After having used Perforce for four years, moving to Clearcase was constant annoyance and pain. I have changed jobs (for other reasons) and am back with Perforce, and the pain is gone. Here are some other devs' opinions on Clearcase. And Clearcase is expensive. Not just by dev seat, but also in support. The companies I know who use Clearcase always have at least one full-time Clearcase admin per product, if not two admins. Factor that into the cost! In contrast, Perforce just works, and there is no need for a full time admin; Perforce might require couple of hours a week of some dev's time, and that's it. Perforce is solid and easy to use, and in my experience devs rarely if ever need admin assistance. That is a huge cost saving. So who buys Clearcase? Middle managers at big companies do, because they get sold a bill of goods by slick/highly paid IBM sales people. And big companies like to buy from big companies. The mentality: nobody ever got fired for buy Clearcase (sadly). In summary, please do yourself and your organization a favor: avoid Clearcase, and take a long look at Perforce.

Check out Perforce. - Excellent in its own right. I have used it at 3 of my last 4 companies for a wide variety of team sizes, product types, and platforms. Once you use Perforce it will be painful to use anything else. - Rock solid stability - Good Visual Studio integration - Cross-platform support (Windows, Linux, etc.) - Reasonably priced for commercial development -- free for 1-2 devs, otherwise ~$1k/dev.

As stated earlier, CD-Rs have a surprisingly short shelf life, from 2 to 5 years. I believe this is because the info is burned into an ink layer, unlike commercial CDs which are pressed into a silver layer. On the CD-Rs the ink degrades over time. Here are two storage options for your photos: - Buy a second hard drive and be diligent about backups. Of course HDs crash too, so make sure your data is always in two places. My USB backup HD crashed recently... - Subscribe to an online storage service. The storage costs are typically pretty expensive per GB though. Cheers, Doug