I run a lot of builds. I run the vast majority of those builds directly within Eclipse taking advantage of its integration with Ant and some of those builds are large. More accurately, some of those builds are verbose, whether it’s because I’m explicitly documenting what’s happening for later debugging or whether there’s a lot of output from a given task (svn export is a perfect example) that’s being displayed. Sometimes there’s so much output that the console’s buffer discards a portion of that output and prevents me from debugging effectively if a problem arises.

By default, Eclipse limits the console output buffer to 80,000 characters, but this can be changed easily enough. In my case, I like to remove the limit all together:

1. From the menu bar, go to Window > Preferences > Run/Debug > Console.
2. Uncheck the box labeled Limit console output.

With that limitation removed, I can access the unabridged version of my build output. This setting is not specific to Ant, of course. It will remove the buffer size limit for any action that directs output to the console.

Having been on vacation, I just read that Google is planning to release their own browser, Google Chrome, into the already crowded market. Although I’m a huge fan of Google’s existing products, I have to admit to some mixed emotions about this one.

On the one hand, Chrome offers multiple processes. Each tab would be anchored by its own process so no single tab would be able to crash the entire browser application. The benefits of this are obvious and huge in today’s world of increasingly complex, resource consuming web applications.

On the other hand, it’s yet another browser.

On the one hand, Chrome is based on the WebKit engine, the engine that powers Safari. WebKit is a very capable engine that, judging from my limited experience with Safari, seems very fast. It’s also open and standards-compliant.

On the other hand, it’s yet another browser.

On the one hand, Chrome includes some significant Javascript innovations including isolation through a virtual machine, performance and precision garbage collection. This, of course, in addition to the multi-process capability mentioned above.

On the other hand, it’s yet another browser.

On the one hand, it borrows some of the best of what others have to offer. Tabs (of course) – now as the primary metaphor rather than the secondary, an intelligent address bar, etc.

On the other hand, it’s yet another browser.

If their online book is to be believed (and I have no reason to think it shouldn’t), they’ll be introducing some very welcome innovations on both the back- and front-ends, but at the same time it’s another browser in a market that’s already too crowded and, more accurately, too diverse. I can’t decide whether I’m excited about the innovations or dreading the additional testing I may have to do as a developer of web applications.

After making the whole-hog switch to Linux at work, I’ve been gradually reassembling the pieces of my development environment in the order of need (not priority). Some of these pieces are easy. Web server? Easy. IDE? Easy. ERD tool? Not so easy. I didn’t really have any idea what ERD tools, specifically for use in database design, were available for Linux.

In Windows development environments, I tend to use a Win-centric product called DBDesigner 4. I’m familiar with Dia and have even used it on Windows, but I’ve never particularly enjoyed that experience so I was hoping to find something better. After digging around and asking a few questions, what I found is that there’s a something of a dearth of these tools available for Linux. After a day or so of searching and waiting for answers to my questions, it occurred to me that maybe I could just install DBDesigner 4 via Wine.

I’d heard that the functionality of Windows apps in Linux can be pretty spotty using Wine, but I’d lose nothing by trying. Besides, even on Windows, DBDesigner has more than a few quirks (though nothing serious or even seriously annoying). To my surprise, it worked beautifully. Once I installed Wine and downloaded the DBDesigner installer, I was able to run the installer executable via Wine without a hitch. What I’d heard about spotty functionality turned out to be true – at least with respect to this app which, it should be pointed out, is not listed as a supported application in the Wine database. Nonetheless, after roughly a week of regular use, I haven’t run across anything serious and I’d call the experience a good one. Certainly better than having no ERD tool at all.

If anyone is looking for a decent ERD tool for Linux, you could do far worse than to install DBDesigner 4 via Wine. The project has been rolled into the MySQL Workbench effort, but the legacy installer is still available on the fabFORCE downloads page.

A few days after installing DBDesigner, I was told about a product called SQL Developer that also looks pretty nice. It appears to be a Java application that will, of course, run natively on Linux. It seems to be worth a look.

This weekend I had one of those experiences that exemplifies the reason that I switched to Mac rather than Linux nearly two years ago. I spent the better part of this weekend trying to get my girlfriend’s laptop to dual boot into Linux because she’s been having all sorts of inexplicable issues with Windows. After hours and hours (and hours and hours) of driver hell…no mas. I’m done.

The laptop is an old one of mine, but it’s hardly ancient. It’s a Dell D600 Latitude – about four years old, I’d say. The specs are still pretty decent too, except that the wireless and ethernet cards are provided by Broadcom, a company that appears to be somewhat (in)famous for not offering Linux drivers. The result: I couldn’t get online – wired or wireless. That put me at a distinct disadvantage with respect to getting anything else done, so I tried to make it work.

Continue Reading »

Having grown tired of the wrapping post titles that resulted from the ridiculously constraining, three column layout that I was using, I made a quick and dirty change to my Chyrp theme to allow for more space. The modified layout also allowed me to increase the font size which I hope will assist with readability. This end product (some might argue by-product) of this modification has its own issues and violates more than a few of my own aesthetic preferences, but it addresses the most egregious need and I’ll get to the rest when I update my entire product code base to the latest version.

I tell you this because I care. And for absolutely no other reason.