Note: CogDogBlog has a new WordPress powered home at http://cogdogblog.com/. All entries from this version have been moved there, so as a guide dog service try finding this article in its new home by title search.
No, Willie Nelson did not pen this song, but I think Alan Cooper may have been an influence ;-) And this old song has been eating a lot of my time and productivity this week.
But here is the moral first- Programmers are brilliant, intelligent people in their areas of expertise, but for the most part (ducking to avoid the lobbed rotten vegetables), they should never be allowed to lead or do web design for use by any normal humans (e.g. potential site visitors).
Last September, one of our offices faculty recruitment internship programs was shifted from an internal focused program to one that would be open nationally, one part being a redesign we did form the old site (HTML tables + CSS) to a new XHTML, 21st century design. But that was the cosmetic change.
The team behind this wanted to have a 100% online application and review system, something I am rather in favor of given all of the paper we still print and push around our system. However, we were smack in the middle of updates and development of our successful online Learning Grants application system and with a technical staff of 1.5 here (and I am the "1") it was not feasible to create something brand new for a January 15, 2004 launch date.
It was then suggested and offered that since the application needs were similar to our hiring application (this later turned out to be wrong), that it would be "easy" for their programmer to de-deploy our online job application for this use.
I was leery as I had seen some of the application screens and on hiring committees suffered through deciphering the electronic applications (I highly recommend anyone who wants to apply for a job at Maricopa to request and submit a printed application!). But we had no choice.
The worry for us was that this application would use software we did not use (Colf Fusion) and would be on a a server we lacked access to- we would be at the total whim of someone in our technology department.
Then we started seeing some of the screens as this person began developing the online forms. Oh sweet lord of HTML, what did we get into! This person may have had come across the use of web style sheets perhaps in a copy of MS Word, but it was a mishmosh of FONT tags with CSS classes applied, and everything was structured in HTML tables, and displayed the kind of way people used tables when NetScape introduced the table tag in the mid 1990s- boxes and boxes and boxes....
At least they included a suggestion we use on all of our web pages- a link to a standard feedback form that automatically captures the URL of the page it was sent from and some technical details from the HTTP request (browser version, operating system) as well as the person's comments. On our sites, this is emailed to me. On this application system they decided to ignore the browser details (so we have no idea if their problem was platform related), and rather then generating the email with the writer's email address in the "Reply-to" field, it has my email address. So if I need to rely to the person that sent feedback, I have to copy/paste it from the body of the message- now this is minor, but for a few weeks I was addressing 5-10 messages per day.
It took a few iterations of requests to do some simple things from the review end of the system, such as a linkt o generate a list of all applicants, to output search results in last name alpha order, so be able to swtch the status of the application.
And then this programmer got shifted to another project, so we were left at the mercy of requesting (and cc:ing 3 more people up the chain of command) for fixes to problems.
But looks weren't everything, and as we worked to makes sure the directions were clear and the system worked, we were ready in time for the system to launch January 15. A major part of the application was the necessary uploading of 3 different Word documents- a statement of interest, a course completion form, and a resume- the first two we set up standard formatted Word docs as "forms" where the applicant merely had to type info in, tab from field to field etc...
For the first two weeks there was no check that all 3 of the documents were submitted. At least 1/3 of the applicants neglected to insert there name in the field labeled "Applicant Name". A handful submitted Word Perfect documents, despite the numerous directions.
Anyhow, some 200 people managed to actually survive this application form. Now I have some 475 .doc files that needed to be converted to .pdf. I did find a nifty free Mac OSX app that does this via drag and drop, with the curious name of Strangelove which does a decent job. Here is a peek of it- you drag file icons in, it converts to PDF in about 1 second, and then you drag the PDF out. If I had bothered to learn AppleScript way back when, I might have sped up this process:
Later I found out that it did not format the 2 Word items that were set up as Protected Word forms, so I am resorting to opening in Word and Printing to PDF... And speaking of which, why the *$%# when you Print to PDF in OSX does it not append a .PDF extension? And why does MS Word, after the OSX 10.3 update, insist on naming each printed PDF "Formatting Palette: if that palette is in view??
Our web site designs from this site do not go through usabbility testing and are do involve people with skills in graphics, web-standards, information design, usability, but hey, they sure know C++ and Oracle.
Update A few hours later (I had to take a break from the monotony, to go out side and pull some weeds) I am about 80% done processing the supplemental docs to this application system. What is really bunring me is that I have seen sometimes 5.7.10 even iterations of a document from an applicant. The programmer who built this system decided to create each uploaded word doc as a file name like 86787555.doc, likely some sort of date/time stamp, but the only tie to who's document this is is via a database table. Since there three distinct documents each applicant should have, if I were designing this systemt he files would have been named according to the database id (unique record number) for each persons, so rather than the person at record number 127 having a spray of documents named 8879032.doc, 854345.doc, 8943453.doc, I would have named them, app127_file1.doc, app127_file2.doc, app127_file3.doc. This would reduce the number of *%$#ing files I am dealing with now because if person #127 uploaded a new version of his/her resume, it would overwrite the one sitting there. More so, all I would need is a list of applicant names and database IDs and I could easily scan a directory to find their files.
But what do I know- I am not a programmer.
Back to the grind....blogged February 29, 2004 05:09 PM :: category [ web bad dog ]