Pat Gunn (dachte) wrote,
Pat Gunn
dachte

Tapping the Wire

I've griped a lot about servlet development, mostly in person to peoplearound work, but also a bit on here. In particular, I'm writing stuffthat uses a lot of forms, and it's really hard to debug, as I'm beingkept outside of the 'black box' that I'm developing, and need to inferwhat's happening from the results (in other words, no debuggers, noprintf (that go to a sensible place anyhow), and a lot of other stuffthat's just not good for development). However, I've developed twomethods to deal. The first is that, as most of our form submission happensthrough a http POST, the user gets redirected after the jsp that catchesthe POST finishes. So, I typically comment out the forward URL, andinstead serialize whatever printing I want to happen through that.No spaces or underscores, and it sucks for handling a lot of printing,but it works. I might have code sprinkled through my jsps and servletssaying things like forwardurl += "value_of_parameter=..."It's ugly, but it's easy to set up. At some point, I'll look into betterways to do this. The second thing I've found, which I actually grabbed forhome programming stuff 2 days ago, is a tool called LiveHTTPHeaders.It's *REALLY* good for figuring out what's going on with your CGIs and forms.I was just debugging something for about 15 minutes, trying to find why alistbox's values were landing in the POST form as a null, when I rememberedthe tool, grabbed it, and found that I was capitalizing the form value inthe submit, and the POST was looking for it in lower case. So, yeah, it's areally great Mozilla plug-in.

While I'm boosting, let me say that google is a really cool company. I'veblocked microsoft's bots from my site (no, IE users, I'm not blocking you,I'm blocking the crawlers that populate Microsoft's search engine). I'vemade a few more tweaks, and google is now crawling my site, including myentire blog. Huzzah. There's a few things I could do to make it easieryet for google to find the whole thing -- currently to go to an arbitrarypage in the archives, which is what I really want indexed, google hasto go to the initial index and step forward, something that their bot isreluctant to do to any depth. A page that directly links to each archiveindex will fix that, and would be pretty simple to write.

On the downside, when I returned from shopping yesterday, a blackout+surge hadvisited my apartment, resetting all the clocks and turning off all mycomputers. Alas, although everyone else booted back up without damage,kryten's hard disk is now toast. Without the system that acted as thegateway to the internet, I'm without access at home.. almost. I can stillget internet access from one of the neighbors' wireless APs, for my laptop,but that's outgoing only. I'll probably start checking the spare IDE harddisks I have to see if any of them are good and could have the latestOpenBSD installed onto them.. although if I do that, I'll probably wait forthe coming release. Either that, or perhaps I'll put Redhat Severn ontoit.. hmm.. If I had the money, this would be a good time to cancel my verizonDSL and switch to telerama, but I don't want to drop the money for theDSLmodem right now.

Sweden voted nay on joining the EU. Phooey.The Israeli cabinet is talking of forcibly relocating Arafat, and oneguy, Ehud Olmert, wants him assassinated. Just like Zeevi, it would bea good thing if Olmert got his just desserts, and were to be killed himself.

Sometimes I wonder what Wally does all day when I'm not around.

Tags: tech
Subscribe
  • Post a new comment

    Error

    Anonymous comments are disabled in this journal

    default userpic

    Your reply will be screened

    Your IP address will be recorded 

  • 0 comments