Sent:  Mon 3/6/00 11:44 PM

Quick Reference:

1.    What’s New? 
2.    This update on the web (nicely formatted).

Hey Everybody,

First of all I’d like to welcome a whole bunch of retailers that responded to my e-mail.  I apologized in advance for the “spam”, but I hope that it was worthwhile and they will stick with me until I’m finally published. I figure we’re all in this together and I hope that the final product is enjoyable and possibly educational.

I’ve gotten some questions regarding the status of the project.  I’ve been bad about updating that section on the site, because not a whole lot has changed since the holidays.  I’m really waiting on Jason to finish up his pencils.  He’s been trying to get them done while changing jobs and mostly putting in a few hours each night.  Although I would think that would equate to at least a page a week, but he puts in a ton of detail so it takes a lot longer.  I know I’m able to spend about an hour or two each night to work on miscellaneous stuff now that my son is going to bed around 8:00 or 9:00 each night.

Photoshop 5.5
This program rules!.  I had been stuck using Photoshop 3.0, but I was okay with it only because I went to elementary school with one of the guys on the team, David Corboy.  I’m not sure if he’s still at Adobe, but wherever he is I’m sure he’s doing quality work.  I couldn’t believe I was okay with having to make about 20 different back-ups of each image I worked on.  This version supplies a history window that is awesome for correcting mistakes or just going through “what if” scenarios.  The text manipulation is a million times better than 3.0.  I don’t have to accompany each file with an associated Word file that contains the text that I want to put into Photoshop.  In the old version you couldn’t edit the text; you had to erase and do it over and hopefully you put it in it’s own layer.  Now version 5.5 does that for you.  “It’s about time too.”  I think this will allow me to letter the rest of the issues with “relative ease”.  “Huh!”  Sorry about that… DK 64 on the brain.  

Promotional Cards
I’ve only gotten two responses from retailers saying they want a few cards to put in hold slots (you know who you are, thanks guys!).  First I need to know if you would like some, and second I need to know how many you need.  I’ll get some out to you as my supplies last.

The Ashcan
I’m getting quotes right now from North Suburban (here in Denver) and Brenner (in Texas).  I know I need to do the entire first issue for the retailers so they can determine the quality of the book.  My guess is that it will be a gray scale version of what you see out on the site.  I’ll have to put in a pinup or two to round things out two a multiple of 4.  That’s 32 pages if you do the math and you don’t count the cover.

Much of this update will be geared towards promotion and mailing lists.  If you don’t want to learn how to do mailing lists, etc. then I’ll see you in Update #12.

Promotion Using “The Master List”
Again, I’d like to thank Robert Scott for pointing me to The Master List, which is maintained now by Mark Adams.  I still don’t know whether to thank Rob or to kill him.  I’ve definitely putting in some late nights trying to get The Master List into a workable form.

Once I got the website address from Rob I checked it out.  WOW!  There’s a ton of data out there!  I kind of got my hopes up, because I thought there would be a spreadsheet or database that I could use to produce some mailing labels quickly.  I contacted the web-master and Mark replied with “I’ve been meaning to do that.”  I was completely blown away that he did all of this BY HAND.  I commended him on his efforts and his ability to deal with the tedium.  Mark sent me an e-mail list of about 400 retailers.  Unfortunately that was in HTML format and wasn’t easily “mail-merge-able”.  I knew I had a few hours of coding (programming) to do before I’d have a workable database where I could start generating mailing labels.

I figured I’d start with the e-mail list and then work my way to the address data hidden on the website.  Granted, the data isn’t really hidden; it’s all there, but locked away in HTML pages.  I already had a program that searched an e-mail file and pulled out e-mail addresses.  You know that spam that you get with thousands of e-mail addresses in it.  Don’t you want to tell these people to get a clue.  Yeah, well I did that a few times.  I got a few “thank yous” and I got many “how did you get my e-mail address?”.  Microsoft or Disney isn’t giving away money.  Someone is, however, collecting e-mail addresses and I’m sure they’ve written programs like mine to pull out e-mail addresses once that message circulates back around to somebody that cares.  Anyway, I already had code that scanned files and I modified it to look for certain HTML tags within the file Mark sent me.  Eventually I was able to pull out the Store Name, City, State/Province, and E-Mail address from the file and write it out to a “|” delimited file.  Some of the store names needed to be conditioned a little, because of the way the “&” is represented in HTML (“&”).  

I was then able to take that file and import it into Microsoft Access 2000.  I created a couple of new fields for that table such as “DateSent”, “DateReplied”, “DateFollowedUp”, “IsValid”, etc.  I then borrowed some code that I wrote for my comic database.  It was code that sent out e-mail based on a list of criteria.  I coded up my initial introductory e-mail and then hit the “spam” button.  I felt bad doing it, because I hate getting spam, but I felt that this would be information worth having.  Also since I got the e-mail addresses from The Master List so I felt exonerated, because that’s what the list is for. 

The new columns in the table were helpful in keeping things organized when I received the fatal “SysAdmin” replies.  When I got those, I knew the address was bad and I marked them as being not valid.  Some replies were valid, but they didn’t want to be on the list, but most replies were filled with words of praise and willingness to help out.  I was pretty bowled over by the responses.

           Stores with + Response:                   61
           Stores with – Response:                     4
           e-mail bounced:                               39
           No response:                                 320
           Total e-mail addresses:                 426

 I know I’m missing two addresses somewhere, but I’m not sure where.

I let that message simmer for about a week, before I sent out the follow up.  I knew MS Word could do an e-mail-merge so I thought I’d give that a shot.  I had tried it before, but it was with the old version of Word.  As the responses were coming in I entered their addresses or phone numbers into the database.  I knew I didn’t want to use Outlook to manage my Contacts, because attempting that from home would be extremely slow over a 28.8 modem.  I extracted the data I needed from Access and gave it to Word as my mail-merge data source.  I was able to use Word to type up a nice follow-up, instead of trying to code up a response and send it through Access.  The mail merge worked like a charm and it was fast too, with no failures (not like my attempt in Access, which could only send about 25 at a time before it completely bailed and I had to restart).  Thanks Microsoft!

I haven’t figured out how to take data from an Access database and write it into the Outlook Contact database, but that will be next on my agenda.  I hope to have a full-blown Outlook database here really shortly.

Scraping-Up Is Hard To Do?
NAH!  Once you’ve got the code in place, the rest is just tedious.  The next step was to get the data from the site into an Access database so I could use it for mailing labels.  I had written a program a while back that will drive another program using a set of commands.  You can use the old “SendKeys” command in MS Visual Basic and send keystrokes to a program as if you were typing them in yourself.  I had made the program read a file of commands and then those keystrokes would be sent to a program such as Lview (a public domain image editor).  I could tell Lview, through this set of commands, to read in a set of files and re-format them and then save them out again.  This is great when you’ve got a huge set of scans that you might have made at 150 dpi and you want to shrink the images so they’re the same size, but a different resolution, say 75 dpi.  I have also been able to use this program to navigate the web using the MS Explorer.

I thought I would give that a shot.  Sure enough, The Master List website can be navigated completely with the keyboard.  With a couple of well-placed down-arrows, backspaces, and <ALT>-F-A sequences I could save the entire site to my hard drive.  I had to tweak the duration between navigation, but I didn’t care how long it took because I could just walk away and about 50 meg later the site was there (maybe 50 meg is an exaggeration, but there were quite a few images too).  I didn’t really need the images so I trashed them.

Once all the pages were downloaded and saved, I was able to start the parsing process.  I modified the code, which parsed the e-mail file (above) only slightly, and I had data pretty quickly.  There were almost 88 files to parse so I modified the program again to read in all the files in my download directory.  When all was said and done, I had “|” delimited file which was ready to be brought into Access.  The file I created wasn’t perfect and had some blank rows in it, but I wrote a few queries in Access to take care of those.  Now that the data is in a workable form, I’ve got 8,310 addresses at my disposal.  Unfortunately not all of them were complete and many of them are only card shops.  I wrote a few more queries to pick out the names that have “comic” or “comix” in them.  I assumed the names with the word “card” in them, but not “comic” were card-only shops.  I needed to narrow my search down and I can sweep the rest later.

Hard Data
So that’s pretty much it as far as data acquisition goes.  Don’t ask me for a copy of any of the data.  You’ll want to go through Mark @ The Master List.  I’ll be passing everything I do over to him.  Once I’ve cleaned up the data, I plan on normalizing the database and I’ll add some columns that Mark has asked for, so he can keep track of the contacts for each of the shops.  I’m still not sure if I’ll make the source available, because I’m guessing that Mark won’t want other people to do what I did.  He could end up doing what SwitchBoard did and hide their zip codes under another mouse click.  Talk about tedious!  But if you can mimic the way a person navigates the web, then the data can be acquired all in due time.  Thanks a lot Mark for maintaining that site.  I know a little recognition goes a long way…

Thanks for all your support,
_______________________________
Parker D. Smart
President - Spilled Milk, Corp.
psmart@SpilledMilk.com      http://SpilledMilk.com 

The standard removal policy if you should ever need it.  Just reply to this message with the word remove in the subject and you'll be deleted from the mailing list.  You might make it in ALL CAPS or else I might not take you seriously J.