I would not have figured out all this if it hadn’t been for someone on the #linode channel with the patience to walk me through the steps.

At any rate, when I moved abettergeek from Dreamhost to Linode, mail quit working in the forums. PHP has a built in mail() function, but it requires some server config to work – apparently, Dreamhost took care of that, while Linode’s default Linux install doesn’t have sendmail installed by default.

This makes sense – Linode basically allows you to select from a number of Linux distros, and they drop a very barebones image onto your VPS. It’s up to you to install the things you need. This is better than, say, installing everything by default and leaving a bunch of ports and vulnerabilities open from the start.

sendmail, however, is a bit pickier than Apache when it comes to making it work properly. These steps are using Ubuntu 8.04 LTS, but they’ll be similar for any Linux distro.

First, you’ll want to actually install sendmail:

apt-get install sendmail

That part was easy…however, the default Ubuntu config did not have my hostname information configured properly. You’ll need to update your hosts file and your hostname file:

sudo nano /etc/hosts

if you have a line that looks like 127.0.0.1 ubuntu, remove it. The top two lines of the file should look like this:

127.0.0.1      localhost
72.14.177.31   abettergeek.com

Then, you need to update your hostname file:

sudo nano /etc/hostname

This needs just one line for your domain

abettergeek.com

Restart your server for the changes to take effect, and you should be able to now send mail using the mail() function in PHP.

My freshman year of college, I knew a kid with a Logitech MX700 cordless mouse. The first time I used it, I knew I had to have one. It was only a few months before I bought the Cordless Elite set, with the Elite keyboard and the MX700 mouse.

I have since discovered that this is the number one best mouse in existence. I’m not a gamer, so I don’t need high sensitivity and hair-trigger response. What I do need on a daily basis are the extra mappable keys that the MX700 (and its siblings) provide.

Logitech canned the MX700 several years ago in favor of the (in my opinion, at least) inferior MX1000. They still, however, produce a corded version – the MX518 gaming mouse. This is what I have in my office at work.

I typically map the second button below the scroll wheel to Ctrl+W, so that I can quickly close tabs and child windows in almost any application. There are, however, two applications I frequently use that do not follow the Ctrl+W convention – Outlook 2003 and SQL Server Management Studio. In Outlook, you need to use Esc to close child windows (messages, appointments, etc.). In SQL Studio, Ctrl+F4 is the only shortcut that will close open tabs in the interface.

Today, on a whim, I Googled to see if there was a way to map each of the mouse’s buttons differently for specific applications. Lo and behold, there is! It turns out that the SetPoint software required for Logitech’s newer mice uses XML for the configuration settings, so someone figured out how to enable lots of additional settings via the aptly-named uberOptions.

This application was ridiculously easy to install and use. I didn’t have to restart, or even manually shutdown SetPoint before installing. Once it was installed, I was able to quickly specify mapped keystrokes for Outlook and SQL Server, and was off and running. Changes take effect immediately – you don’t need to restart an application for its specific mappings to work.

If you have an older MX700-like mouse (the MX500, 700, 510, and 310), you can use an application called LogiGamer, although it looks to be a bit clunkier (and requires the .NET 1.1 framework; it’s old enough that it may not be compatible with newer .NET versions).

If you have a Logitech keyboard, uberOptions will allow you to customize all the keyboard’s extra keys on a per-application basis. I’m just glad that I can finally use the same shortcut button for the same purpose in all my applications!

At work, we use Microsoft SQL Server 2005. The client software requires Windows authentication to connect to a database. However, the credentials are pulled from your logged in Windows account. If you’re not on Purdue’s domain, you can’t connect. While putting my workstation on the domain wasn’t an issue, I prefer using my local user account on my laptop.

Vista supports fast user switching on a domain (XP does not, unfortunately), so I could switch between local and domain to use the SQL client, but running two profiles simultaneously is a bit of a resource hog.

Windows XP has a handy “Run As…” feature. If you right-click on an executable or shortcut, you can click “Run As…” and enter different credentials to load an application. This is especially handy if an admin needs to run something like the Management Console without logging out the user first.

Vista, for some inexplicable reason, has removed this feature. Fortunately, SysInternals came up with a handy little application that brings back the “Run As…” menu. It registers in Explorer’s context menu, and it can be used at the command line to load an application – which means you can create shortcuts specifically to run something as a different user.

It’s small, easy to install, and works great. You can get it here.

So Firefox 3.0 was released recently. I downloaded it today to see if it was worth upgrading from 1.5 (I skipped 2.0 entirely – too many annoyances for me).

The verdict thus far is, as I used to say when I was 15, “big negative on that one”.

I really don’t see much happiness in the future of Firefox. What I see is yet another great piece of open source software going drastically downhill in an attempt to “reach out” to a wider user base – meaning a dumber user base.

There is a delicate balance between making an app so unusable that nobody downloads it (even though you, as the developer, might find it plenty useful) and making an app so dumbed-down that anyone, including a computer-illiterate 90-year-old, can use it. This is particularly a problem with open-source software.

I’ve been a big fan of how Microsoft does things. I’m sure I’ll get blasted by “real geeks” for saying that, but the fact is, Microsoft’s products cater to everyone – from the stupidest of the stupid users to the most advanced power-using geeks out there. While there are certain things that Linux indeed does better, Windows meets all of my needs as a power user – and all of my mother’s needs as someone who only uses the computer for email and the occasional game of Spider Solitaire.

In the case of OSS, it’s too easy to either make a product really dumbed-down or really complex. This is more or less what happened when Gaim was released as Pidgin several years ago. Since then, Pidgin’s development has gotten drastically more user-unfriendly in an attempt to make it more accessible to people who would otherwise avoid unfamiliar software. In the case of Pidgin, as “features” were added (which translated into actual features being removed, hidden, or dumbed down) and the user base got wind of the changes and requested fixes or reversions, the developers unfortunately were not too interested in listening to their users.

Firefox is a much bigger open-source project compared to Pidgin, which makes me wonder if they’ll be less likely to listen to their users – or more likely because they have such a public presence now. It’s hard to tell, really.

Surprisingly, Firefox 3.0 has made an effort to behave more like Internet Explorer – I tend to believe it should be the other way around. The oversized back button just gets in the way – who even uses the back button anymore? Ctrl+Left is a heck of a lot faster when navigating your page history.

Not only that, but, like IE7, FF 3.0 has condensed the history into a single button. Rather than having arrows next to both back and forward so that you can navigate in either direction in a tab’s page history, it’s all been shoved into a single button. Does this actually increase usability? I tend to think not. When you’re doing a lot of heavy browsing (and lord knows I do that enough), you don’t want to try and work with one list for all your page history – it’s easier when it’s split between what’s behind and what’s ahead in your page history.

The new drop-down URL bar is ridiculously obnoxious. I Googled around a bit to see what others thought, and there’s a pretty clear consensus so far – the new auto search is annoying, frustrating, and at times unusable. While I’m sure there are people who only remember URLs by title, I don’t. I know the actual URL. When I type in “sla” to go to slashdot, I’d rather not have it start throwing back search suggestions and entries from my history or my bookmarks. If I want to find a page in my history, I’ll hit Ctrl+H and search for it in the History sidebar.

I would imagine that this particular problem, given how many people already dislike it, will be fixed soon with an extension.

Then we come to the feature that I was most interested in – memory management. I am a tab addict. On any given day, I’ll generally have between 20 and 40 tabs open in a single Firefox window. On top of that, each tab has enough browsing history that I start to quickly suck up any and all available physical and virtual RAM as Firefox attempts to cache my history in RAM. After a few days of leaving my workstation at Purdue up and running with Firefox open, I’ll look in the task manager to discover that Firefox is using 500MB of RAM (out of 2GB) and another 1.2GB or so of virtual memory. Closing tabs doesn’t fix the problem – the cache is still there. Killing the process entirely and starting over is the only way to free up memory.

I haven’t really put FF 3.0 through the wringer yet on this one. However, in running the same general activities in 1.5 and 3.0 today, both were using almost identical amounts of physical and virtual memory. It’s not a good sign so far, but I’ll have to really do some hardcore browsing to see if 3.0 is actually an improvement over previous releases.

So far, 3.0 has not impressed me. The only feature I’ve seen so far that I liked was the ability to make it remember to always allow SSL certificates with mismatched domains – at work, our development environment’s SSL certificate doesn’t match the server’s URL. It gets annoying to get that “are you sure you want to do this?” popup every time I login (particularly when the login has to expire after an hour). Being able to set it to always authenticate without needing an extension is nice.

Other than that, though, I think I’ll be sticking with 1.5 for awhile longer, much like I’ve been forced to use older versions of Pidgin to compensate for the fact that the developers appear to be writing the application into its grave.

I have friends all over the world. Once in awhile, I make use of the various webcams I own and do a little video chatting. Getting video chat to work well will be for another day – something I learned recently, however, was how to setup a live stream on a webpage. It was surprisingly easy, and works in both Windows Vista and Windows XP.

You need three things : a webcam, a decent high-speed internet connection (the lowest end DSL might not have enough bandwidth to support this well), and Windows Media Encoder, which is available for free from Microsoft. If you’re running Vista, there’s a hotfix you might need (it’s linked on the main WME page), but I have yet to have any problems myself.

You might also want to get a dynamic IP service, like dyndns or no-ip. I used no-ip.org, although they seem to be really big on sending me frequent emails advertising their paid services – I’m not a big fan of free services spamming me about their commerical options. If you don’t use a dynamic IP service with the computer running the webcam, you’ll have to manually update the webpage every time your IP changes.

We’ll walk through how to get setup and going. It’s a bit screenshot heavy, so click the jump to see the rest.

Read the rest of this entry »

I just finished up a review of the Asus eeePC 4g (701) for Julie over at The Gadgeteer (check it out here). Before i finally got around to writing that review, I hadn’t done a whole lot with my eeePC. Since last night, however, I’ve installed Windows and started looking for ways to make my eeePC as functional as possible.

Read the rest of this entry »

Every once in awhile, I like to peruse Sourceforge for new, interesting open-source applications. I was looking for a good screenshot application several months ago, when I was working on writing help documentation on a project at my job. I’ve since found out that the help documentation software, Adobe RoboHelp, includes a very nice screenshot manager called RoboScreenShot. This application, however, is not free, and I’m always interested in finding good open-source or freeware alternatives to commercial software. I have been in need of a small, open-source application that allows me to quickly take and save screenshots of applications, particularly since I’ve started writing these articles.

Enter Greenshot. This is a very small, quick application that allows you to take three different types of screenshots: your entire desktop, a selected window, or a selected region. Taking a cue from the application’s name, selections are highlighted with a translucent green region to show what you’re screenshot will contain. The application is small and fast, and allows you to either copy your screenshots to the clipboard for pasting into another program (for instance, PowerPoint or Word), or save it to your computer in JPEG, BMP, GIF, or PNG format. There are also limited annotation features, but I’ve found that this application is most useful when needing to send a picture of part of your screen to someone quickly and easily.

Read the rest of this entry »