Amen Tony.

I don't know how many times I get a call from a user saying for "some"
reason my XLS/XLSX parser doesn't work with the "Excel" file they have. What
the user really has is a CSV/TSV file with a XLS file extension that some
programmer decided to name the file. This is because EXCEL doesn't bring up
the CSV/TSV layout when the file has a .xls extension.  However, my engine
expects a true binary xls or x-zip xlsx envelope when the file has the
xls/xlsx extension. Now the engine reads the first 20-50 bytes to determine
if  it's really an Excel file before attempting to parse it.

It's NOT only U2/Pick programmers who do this little "trick" on the users. I
have well known insurance company that does this, and it annoys the hell out
of me.

I've created excel files with PERL and the spreadsheet::write module when I
worked on a HPUX/UD server and windows server for the past 12 years.

-----Original Message-----
[] On Behalf Of Tony Gravagno
Sent: Monday, February 06, 2012 10:06 PM
Subject: Re: [U2] Building an Excel File

Not responding to any particular quote here, just the CSV topic
in general.

Respected colleagues, CSV is not Excel. If you have an end-user
that asks for Excel and you give them a CSV you're just
perpetuating the myth that Pick is a dinosaur. They will gladly
spend tens of thousands of dollars to replace your application
with something that creates real Excel (and PDF) despite the fact
that such things can be attained at low cost or no cost right
now. Trust me, I've seen it happen.

This dove-tails with the reasons why people get 20 people to
support Oracle when they can have 3 working on Pick.  The reason
is that the Oracle people say "yes", and give them pretty
reports, when their Pick guys say "no", and give them plain text
in columns and rows and call it "Excel".

Please don't let that happen to you.  Be sure you are properly
responding to end-user requests. Just ask them what they do with
the documents after you generate them. If they really just want
raw data, OK. But if they go on to tell you how many days it
takes to reformat the data, assemble the multiple CSVs into a
single workbook, etc, then you have found a great deal of room
for improvement. Yeah, I've been there too.

Off the soapbox, thanks.

U2-Users mailing list

U2-Users mailing list

Reply via email to