by ·0 Comments

Just got a new Twitter follower. Wasn’t someone I knew, so I went out to check out who they were. CEO of a tech company who has a large number of folks who they follow. My suspicion is that this guy noticing my Twitter feed was due more to me following someone else that they follow than anything else, but might be related to OWF or to rugby or to open-source or to who knows what else. Twitter is slightly nondiscriminate.

So, this guy’s tweet feed. Typical executive feed: links to articles they find interesting, which, actually, is far more useful than to me as a Twitter follower than some of the folks who mostly use it to hold visible conversations. Seeing a ‘thanks!’ tweet or ‘I thought the same!’ isn’t enough to make me go check out the full thread to see if there was something useful there. Seeing an interesting article cross-referenced: usually more useful.

2nd tweet down the list:
These High-Tech Underwear Keep Your Farts From Smelling | Co.Exist | ideas + impact http://buff.ly/16u2aON #finally

This poor guy, who I’ve never met / will likely never meet, will forever be linked in my brain as the guy who advertised to the world that he’s interested in underwear that covers up his farts… And THAT’s why you need to be careful what you link to. And now I’m sorely tempted to link to it, just to pass along that ‘hee hee’ moment to someone else.

by ·0 Comments

Working on a simple application to let a group have a set of forms in which they collect data. Something like a simplistic form of WuFoo which lets the group update their forms and save the data to a backing data store.

First level idea: use Google Forms. Forms are very readily editable, even by non-technical folk, data’s saved to a backing spreadsheet, and you can use Google’s tools for analysis.

Implemented a first pass solution:

Advantages:
– data validation, either through their selections, through regular expressions, or (not yet tested) scripts
– really easy to share the form’s editing and responses with a set of collaborators
– very easy to add new elements
– the control set (textboxes, checkboxes, etc) are non-sophisticated but OK for my needs

Disadvantages:
– No great way to handle theming. You can pick from a set of stock themes, and there’s one sitethat seems to have found a way to handle it, but definitely not something that seems like it’s embedded in the core capability. Since I expect the group to need to adjust this ongoing, this is probably a non-starter
– No great way to link records.. e.g., if I have the same user fill out multiple things, no good way to connect them through.

OK… committed to taking a look at a first-level solution, and this doesn’t seem to be the right one. Next I’ll look at a hosted open-source solution.

by ·0 Comments

Today I’ve:
– presented half-working stuff at a retrospective and then spent the rest of the day moving it forward
– figured out that one reason my widget wasn’t working was because it was being served as http, rather than https, which caused mixed-content headaches
– configured Apache to deal with https. Worked much more nicely on Linux than it did on Windows. Meaning, I was completely hosed on Windows.
– got PHP working with PDO and MySQL. No, PHP 5.3 does NOT automatically work with PDO. Though yum install nicely gets things working, once you figure out the right set
– compiled node and npm on our Linux instance, as it sure seemed nicer to pull what we need from Git directly to the box with a known build environment, rather than build it and push it up via ftp. Note: compiling node requires things like gcc and make – haven’t seen those since C++ in college. Glad everything just worked, once the right path revealed.
– reviewed a briefing deck for a meeting tomorrow… unclear what these guys are asking for, but they’ve definitely got a pitch in mind
– signed timesheets, updated mine
– talked through demo possibilities with another team

More stuff, too, including futzing with git feature branches, investigating how to create separate user accounts on the ec2 instance so that a git clone wouldn’t expose my password in history, and a few other things I’m sure I’ve forgotten…

And I thought I’d only spend 10 hours “managing” this project.. Uh, maybe that’s right. Except that most everything above is _doing_ rather than managing. That said, a healthy mix keeps me happy. Better bounds on the work day would make me even happier. Too bad I don’t give up until a problem’s solved.

by ·0 Comments

Success: getting an Amazon EC2 instance up with your software running on it. Able to take a screen capture to show work actually complete before shutting down the instance to avoid running up unnecessary charge.

Despair: testing stop / restart and then realizing that your public DNS has changed, which means your site can’t be bookmarked, its certs no longer work, and you have to update a properties file with a new host name.

Hope: the box itself still has your stuff on it, you can log in with the same keys, and there’s this thing called an Elastic IP available for a charge from Amazon.

Elation: stop / restart works! New user ids granted through IAM, console renamed to give something useful, PEM sent out for access, and JIRA ticket marked as resolved.

Contentment: Now time for bed, in time to wake up and think of the other three things you wanted to accomplish in the evening with Fringe playing in the background courtesy of NetFlix.