Making things easier for myself
Posted by Jim Jagielski on Monday, May 21. 2007 in Programming
I find myself using external "port kits" more and more often... Instead of keeping track of, and hand-compiling, various external packages that I use and rely on, I'm using MacPorts (formerly DarwinPorts). As previously posted, I'm also finding myself using Locomotive to handle my RoR environment, again, instead of worrying about the versioning and dependencies myself. It's not laziness, just efficiency. By not "being bothered" by some aspects, I can focus on the things that are more important. Plus, I alternate from a G5 desktop system to a Intel MacBook semi-regularly, and using Locomotive ensures parity between these development systems for me.
FWIW, I hardly ever use Fink anymore... It used to be my main external source, but I like MacPorts better.
More...
Personal break Boss?
Posted by Jim Jagielski on Tuesday, May 15. 2007 in Programming
Between what I do for fun and what I do for work, I deal with a lot of different languages and frameworks. I find myself switching gears quite often, sometimes several times a day, from one programming environment to another. So it should come as no surprise that even though I consider myself "proficient" using all these variants, I have not committed to memory all the different library/class/method calls, parameters lists, etc. So I find myself occasionally pulling out a "pocket reference" (or man page) as needed to refresh my recall.
Yet, from what I hear about the normal working practices in several companies (including large companies as well as those that really should know better) as well as standard interview "techniques" this classifies me as a low grade moron. Sure, I guess if you use just one language and just one framework then yeah, I guess, you should know them down cold. But who is so specialized nowadays? And how does simply rote memory tests mean anything regarding how well you design and implement code?
Of course, one rationale for such "characterization" is that it increases productivity; that having designers/programmers/coders "constantly" needing to look stuff up (assuming, of course, their IDE doesn't do this for them anyway) wastes time. Well, it's certainly been my experience that programming is not just simply typing. You spend time thinking things through, designing things in your head, with others, testing and refining, profiling and reworking. These hardcopy lookups are hardly in any way an impact on schedule. If you don't understand that, you aren't running a software development house; you're running a sweatshop. And Boy! I bet the quality of that code leaves much to be desired.
More...
No AC EU 2007 for me
Posted by Jim Jagielski on Friday, May 4. 2007 in Programming
What with the Hawaii vacation trip, and the inevitable work-load that results whenever you take time off, I was unable to attend Apache EU this year, something I'm still quite bummed about. However, I've already set aside time for the US show, plus I even "submitted" 3 presentation proposals as well (it's in quotes because every time I tried to submit a proposal via the website, it wouldn't let me... I ended up just sending my proposals in via Email). It appears that the conference was a nice success.
In the meantime, in addition to the "real" work, I managed to find some time to attend to a lot of things that I've been meaning to do. For example: finally removing the dependency on the (non APR-ised) ftp_glob() function in mod_ftp; releasing the 1st beta of the 1.2.0 version of Telaen; voting on various backports for Apache httpd 2.0 and 2.2; trying out Locomotive for ROR development on OS X instead of creating the runtime environment by hand; and using that setup to create an online gradebook to make tallying up the grades at the MSB summer camp easier (and less time consuming)...
And to add to all that, I even found time to badly sprain my ankle yesterday! I could have done without that.
More...
Work and Fun
Posted by Jim Jagielski on Thursday, April 26. 2007 in Programming
So I "just" got back from our mini-vacation/company-trip to Lana'i (counting travel time, it was a week), and I'm slowing slogging through the pile-up of Email and other stuff that I on purpose didn't check while away... OK, that's a lie; I *did* check my email, but only with a rough cut "This can wait..." approach. But I did force myself to stay away from the laptop and enjoy myself and relax.
The funny thing is that one thing I do during my free time, and part of how I relax is that I hack on code... So even though I would force myself "not to do work" I would still find myself opening up the laptop and code. Once, while doing this, Eileen looked over and said "You're not working, are you?" "No, of course not," I said. Then she smiled and said "How can you tell?"
You know it's pretty cool when you enjoy your work. It's even better when what you do for work is also what you do for fun. This is pretty common for many of us open source people. We're a lucky breed.
More...
Quiet before the storm?
Posted by Jim Jagielski on Thursday, March 29. 2007 in Programming
Yesterday, Filip Hanik and I co-presented a webinar that we thought would have created more controversy than it has (not that creating controversy was the goal). The webinar was about interfacing Apache httpd and Tomcat efficiently and reliably. From a support point of view, we've seen countless issues related to AJP, and with the recent advances within httpd and Tomcat, the summary of the presentation was simply to drop AJP, stay with HTTP all the way and use Apache httpd -> mod_proxy -> Tomcat HTTP connector (basically, treat httpd as a "standard" reverse proxy and Tomcat as a backend HTTP server). Fronting Tomcat with Apache httpd still makes a LOT of sense (Mladen Turk's ApacheCon presentation looks to be very interesting and, although not likely to support the conclusion that AJP causes more harm than help, appears to support the general topic that an Apache httpd -> Tomcat implementation is still the best layout by far) but maybe AJP's time has come... HTTP itself has certainly shown itself to be pretty good enough to create scalable infrastructure around snicker
Anyway, I really think that Filip and I are just enjoying the quiet before the "storm"... Before Apache 2.2 and the APR/NIO Tomcat connectors, such a concept would have been dismissed out of hand. Today, it has some teeth.
More...
Syndicate This Blog
