Subject:
|
Re: Inventory Helper app
|
Newsgroups:
|
lugnet.db.inv
|
Date:
|
Fri, 6 Apr 2001 13:44:57 GMT
|
Viewed:
|
635 times
|
| |
|
|
In my ongoing quest to keep an archive of our discussion,
and with Dave's permission, I'm posting his response to the group.
Regards,
Hakan
On Fri, Apr 06, 2001 at 06:26:10AM -0500, Dave L. wrote:
> I like the tab-delimited format just fine, as output. But it's not
> terribly machine readable when we inventory authors keep changing our
> minds about what header fields to include or whether quantity should be
> listed before color, or all the other variations we've seen! I was
> thinking of some standard, portable baseline that *could* be assembled by
> hand but could also more easily be compiled by an applet. Then the data
> could be output/exported in whatever form we like, including the
> tried-and-true tab-delimited style.
on the other hand, an XML (or any other tag system) format is not easy
to manually input... you'd be forcing people to use a program to make
inventories, instead of using any text editor or spreadsheet program.
> In its simplest form, the idea of a helper app could be just a simple
> macro or script that analyzes a hand assembled tab delimited inventory to
> make sure all the tabs are in the right places, etc. so that it would be
> machine readable at Dan's end (or wherever the master database ultimately
> resides).
I already have a validation script, that does many checks - including part number/description match, color validation and format validation. That script is very simple, just because the file format is so basic.
> My pseudo-markup tag language is not XML compliant. It was just a draft
> of an idea that I used while making the app. It has worked relatively
> well, but does have some limitations. With smart folk like you to help,
> making something out of the XML spec could be achieved. The main thing I
> like about using a text-based file format is that it could be included
> inline in HTML pages and just commented out so that normal browsers would
> not pick it up. Any pages that were known to include it, though, could be
> run through the helper app to display and work with the inventory. I used
> angle-bracket tags because they are easier for me to read in a text-editor
> than some other markup language styles (like RTF for instance).
Inlining the raw data in html is an interesting idea... but you can still
do it with the tab-based format - just comment it out just as easily,
perhaps with a leading and ending marker, for your html-to-inventory
script.
> Actually, I think it's impossible to just post to the mailing list version
> of the newsgroup. At the top of each digest edition I get is this bit:
> "To reply to a message within this digest, visit the Message Archive URL
> shown at the top of that message, then compose your reply using the web
> interface to the newsgroups. Do not reply directly to this digest
> message..."
>
> If there is some other way that I'm not aware of, someone will have to
> tell me what it is. So, you certainly have my blessing to quote me and
> forward my comments to LUGNET.
most of my use of lugnet used to be via the mailing list only - if you
set your delivery options to "as soon as messages appear", you don't get
the digest, and you can reply to each message as you go along. You may
or may not need to go through the "mail setup" on lugnet, but I can
understand why Todd would want to make sure people agree with the TOS...
---
> Hmmm. Perhaps I wasn't as clear as I should have been! I did not mean to
> imply that there was anything wrong with the tab-delimited format that you
> guys use. In fact that is ideally what I had imagined as the *output*
> format of any final inventories, and is what I have used in the past when
> I did inventories by hand. What I meant to say is that if someone submits
> a tab-delimited inventory to you, don't you have to manually input their
> data into your master database? Or if you have some scripts to do the
> conversion for you, doesn't it screw things up if there is a space when
> there should have been a tab, or some field that you don't use has been
> added? That's what I meant by automatic parsing of the info: stuff
> submitted by other people being entered into your big database.
actually, like I mentioned my script does the validation, and can deal
with the basic variations of inventories - mainly if there is a "bag"
column or not. also it deals with files that have converted the tabs
to spaces, so if I get an inventory from the web interface, it's easily
converted. Other than those 2, I havn't really seen many errors.
> The other benefit from using a helper like this is speed. In the dozens
> of inventories I have completed since I wrote my little app I have noted a
> marked decrease in the time it takes to complete one.
compared to what? doing it all by hand can be slow, esp when you need
to look up numbers for pieces... but almost everyone using windows
has excel installed, and there's a very simple spreadsheet I wrote that
will do those lookups for you. I'm working on a linux equivilant now,
so I won't have to boot to windows to inventory... We've found that we
can inventory as fast as it takes to sort and count out the pieces.
I don't mean to be harsh, and I do think writing apps to help creating
inventories is a great idea - the easier it is to inventory, the more
data we all get! I'm just not sure that a new file format is needed...
:)
--
Dan Boger
dan@peeron.com
--
"Any sufficiently advanced bug is indistinguishable from a feature"
|
|
Message has 1 Reply:
3 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
