johnbarrytech.com

I’d just like to throw out the suggestion for anyone stuck in Windows world to look into AutoHotkey.
I’ve combined it with python to create some on-the-fly scripts that autopopulate test-plans at work.

But it’s the little things that make it useful immediately. I have a Lenovo ThinkPad at my job. Great little laptop, but the F1 key is where the escape key is on a regular keyboard.
This is a huge problem in vi; my muscle memory goes right to the location and hits “help” instead of changing modes almost constantly.
That’s where AutoHotkey comes in – I remap F1 to Escape when I’m in a gVim window.
I don’t even notice and if I really needed to, I can always :help

The other nice snippet is this one:
Copy/Paste

It flips copy/paste around – Ctrl + V now pastes sans formatting; if you want formatting you Ctrl + Shift + V
Unbelievably useful.

Haven’t updated in a while since I’m at a new job. Everything I’ve been building is too specific to share (trade secrets, business process-related) so I haven’t.
Hopefully I get some generic stuff in the future.

Long/short:
I have a Very Good Idea that nobody’s doing yet.
I ran it by someone in the problem domain, and he agrees not only is it a Very Good Idea, but yes, he hasn’t heard of anyone doing it.

We begin fleshing out an application to fill said need.

We’re at the point where we’ve only invested time and energy into the design phase, but that’s about to change. Brainstorm lists have transformed into user stories, and user stories are starting to have documents that look an awful lot like functional specifications.
I’m not really into extensive requirements gathering if I don’t know the problem domain very well – I’d rather develop via tracer bullets and a few prototypes of basic functionality to make sure we’re going about it right.

The frustrating part of this idea:
I don’t know how to make any money off of it.
It’s a chicken/egg issue – if there’s a ton of users then there’s a resource for people to pay for. If there’s no users, there’s nothing worth paying for.
But in order to attract users, you have to give it away.

So, what is the right business model?

• Premium Membership, a la Monster/Dice/Ladders?
• Advertisements that pay via clicks?
• Give product away but sell support? (same chicken/egg trap applies here)
• Something brilliant I haven’t thought of?

This thought’s occurred to me – maybe the problem isn’t solved because there’s no money in solving it.

Probably not giving out enough information to get help, so here’s the nutshell:
We want to be a document harvester for certain classes of documents that relate to specific standards.
The information is already captured for other purposes; we’re trying to archive it in a format that allows querying/modification/usage of said documents.
Sort of like a data library. There’s more to it that’s specific to the problem-domain that differentiates the product, but that is the nutshell.

Like any library, if it’s three people’s book collections it doesn’t give us a lot of utility.
When it’s 5000 (and we’ve got a dewey decimal system of some sort) we start to see some power.
Hive-mind, infinite retention of knowledge. . . that sort of thing.

It almost feels crass to ask, “how do I get mine if I invent this?”
But a guy’s gotta eat.
So how do I get mine if I invent this?

(just an aside – I think this blog gets 20 visitors a month. It’s funny to query the audience when aware of its size; it makes it an act akin to prayer. “Help me, oh illustrious few who could mentor me in this strange journey!”)

I feel guilty about saying I hate TOAD in the last post after spending the last 3 months with it open.
In fact, it’s the only application that’s open on every Virtual Window on my work machine.
I hate editing in TOAD. I like its autoformatting, love its schema browser and filtering on data grids, and love love love the fact you can read the script for tables, views, etc.

Vim’s too powerful to ignore.

In a perfect world I could run vim as my editor in TOAD; I haven’t figured out how to do that yet.

I don’t think I’ve extolled the virtues of vim, but it’s my favorite text editor by far. Whenever I’m in something that’s NOT vim, I can feel the pain.
What should take 3-4 keystrokes involves mousing and an excessive amount. Auto-indent, appropriate tab stops, shift widths, all that stuff is touch and go depending on the IDE and the time you want to take to configure it.
With vim, I sit in a .vimrc, set everything up the way I want it, and never deal with trivialities like that again.

Lately I’ve been stuck in PL/SQL-land – a chunk of new development in Oracle Time and Labor and some Pricing procedures I’ve been trying to optimize. I’ve been using TOAD, but I actually hate TOAD with most of my black little heart. Its Find functionality is spotty at best, and the amount of keystrokes/navigation is a pain in the butt even when you’re using hotkeys and bookmarks.
The newer version’s code folding is nice, but clutters the screen to a great extent.

The biggest pain, though, is just knowing it’s so damn inefficient when compared to a real text editor.
I’ve had 2-3 times in the last week where I realized a set of code should be sitting in an IF/ELSE construct. In vim it’s typing IF blah THEN, tapping j, then >>, then . for every line I need indented. Takes about 3 seconds tops. In TOAD I have to navigate, tab, then go to the beginning of the next line, hit tab, etc. ad nauseam.

So, naturally I want to use vim. I don’t want to be alt-tabbing and copy/pasting into TOAD to compile, however; I want to just compile in vim with :make.
Lo and behold:
The Answer To My Problem

This is awesome. I got it working today after an hour or so of fiddling with it. I would consider this my greatest accomplishment this week.
My boss would be enraged to hear it, I’m sure.
“You have deadlines!”

Yeah, I do.
And now that I’m vimming it up for PL/SQL, I’ll meet them easily.

Upgraded to Jaunty Jackalope this week.
Had a minor issue with my rig. See, my PC currently does not return from sleep ever ever ever with an internet connection. This is, I suspect, a hardware issue. It happens both in Vista and in Linux.

The last time I was logged into the Ubuntu side of the world I’d been troubleshooting it in a variety of ways. Shutting off IPV6 (some known issues w/Firefox in Ubuntu), looking through the init.d files for any clues, scouring logfiles and the internet trying to get a direct hit, googling my hardware setup to see if anyone else has run into it. . . nothing.
One of the things I was doing was attempting to create a reproduction at will for the bug, which involved tweaking some power settings to test for it. I promptly forgot about tweaking these settings until . . .

11 minutes into my upgrade to Jaunty Jackelope, my power management kicks in and sleeps.
Which causes the internet connection to die inexplicably.
Which means when I log into Ubuntu again the next morning, I see a failed install and 982 packages to update.

Ubuntu is only nominally Linux at this point, however, and rather than forcing me to research and fix hundreds of potential errors, it just gracefully recovered by telling it to Upgrade again.

---

One of the more interesting things Ubuntu does is allow a lot of alternatives to live in the same space without a significant amount of effort.
E.g. Java – in most linux installs, living with multiple java homes (say for Java 5 and Java 6) will cause you a few headaches scripting opportunities to switch these up.
With Ubuntu, however, there’s some handy dandy commands that do this for you:

sudo update-java-alternatives -l

and

sudo update-alternatives --config java

This takes out most of the pain and suffering of swapping between java and javac.
More info is here but it seems to need an update, as it’s a few versions behind. (we’re in java 6 with Jaunty, for example)
I still can’t get stupid Maven + Hibernate to work on this godforsaken development environment.
Currently hunting down what the heck po2xml is and why I don’t have it. The Good Times Are Killing Me. . .

I’ve installed Apache, installed ruby on rails.
Ruby on Rails works fine on the weBRICK server, but it’s not working on the apache server, just giving me a 500 Internal Server Error.
Better check the logs.
ls -ltr ~/www/myApp/log/

Lo and behold, production server’s the most recently modified!
Check its log, see stuff like:

Status: 500 Internal Server Error
Could not find table ’sessions’

This has gotten me three times using the Ubuntu Community Documentation. It’s in there, but it’s not formatted as an instruction.
Relevant text:

murb: I had to add RailsEnv development as well to get around the ‘no route found to match “/rails/info/properties” with {:method=>:get}’ warning… (can someone elaborate on why?) apparently this is because /info/properties is buggy and no longer supported: http://www.ruby-forum.com/topic/161924

(emphasis mine)

I added RailsEnv development to my httpd.conf, bounced apache, and everything started firing away smooth as silk.

Should have been “The sky is falling!”
It does pay to keep your passwords separate, distinct, and difficult and use a tool to manage them.

If you’re using twitter, there’s a fair likelihood your password was compromised. It’s on Twitter’s Main Site.

They’re recommending changing your Twitter password, which means that anything using a combination of the same Username/Pass or your email password is the same, I’d change it immediately.Tech Crunch has more.

New design.

Ruby coming soon.

Automatic saving functionality – make a change, it’s saved.
It’s not so bad with revision control; if the changes don’t work out the way you want them, revert and you’re done.
But what if your workflow is:

1. Open image you plan on editing
2. Make changes you want
3. Either Save As… or Save the current picture. Either way, until the edits are done you’re committed to neither choice

With the “automatic save” functionality, once (2) is done, your original is toast. To preserve originals, I’m forced to change my workflow, creating a copy before any work is done.

Save a recovery copy automatically, but let me choose Save or Save As… – I’m smarter than you because I’m actually editing the photo myself. I probably know what I want to do with it better than the computer.