I don’t know why it took me so long to do this.  I regularly download files, only to open them up to see that they were created on a *nix computer.  That means it uses CR to symbolize the end of a line rather than CRLF, which is the common end of line on a Windows system.  As part of another project that I am working on, I downloaded a huge archive of spam and regular email.  The only problem was all of the files used CR instead of CRLF.  That, and they didn’t have an extension, so I had to choose each time to open it with Notepad.

So I went into VB.Net, and made a quick and dirty form that recursively looked at each subfolder and converted each file.  As part of the conversion, it was changed to a .txt file, and the original was deleted.

Here is the code to do that:

Private Sub ConvertFile(ByVal fi As FileInfo)
  If fi.Extension = ".txt" Then Return

  Dim r As New FileStream(fi.FullName, FileMode.Open)
  Dim d As New StreamReader(r)
  Dim s As New FileStream(fi.FullName & ".txt", FileMode.Create)
  Dim w As New StreamWriter(s)
  Dim str As String = ""
  Do While Not d.EndOfStream
    w.WriteLine(d.ReadLine)
  Loop
  d.Close()
  r.Close()
  fi.Delete()
  w.Flush()
  w.Close()
  s.Close()
  _fileCount += 1
  lblOutput.Text = "Converted " & _fileCount & " files."
  Application.DoEvents()
End Sub

Hopefully this will help anyone having the same type of frustrations.

I finally just figured out why my database wasn’t updating.  Such a simple matter really, but it wasn’t easy to find through Google so I am posting here, mainly as a future reminder.

I am writing a small app to use as a podcast listener.  Since I don’t own an iPod, I listen to podcasts throughout the day.  Well I was at the point that I wanted to start adding data to the database.  But no matter what I tried, the data wasn’t updated.  It was getting very frustrating.  What I finally figured out was VB was copying the database to the debug directory each time.  A simple setting on the SQL Server Express properties, changing the Copy to output from Always to Copy if newer fixed it.

Hopefully this will save you some time in the future.

I just finished posting the v1.1 of my IP Status gadget.  But you can’t get it yet.  Why?  Because it is “Pending approval.”  When it gets approved, you will notice a couple differences.  One, it now has a shiny blue background.  Two, it no longer requires you to load an asp page on a public website in order to get your public ip address.  It uses a web service from whatismyip.com.  Three, there is no longer a settings page, as it didn’t make sense to have one.  The only thing that I could put into a settings page is maybe the refresh interval, but I am not worried about that at this time.

As soon as it gets approved, I will post the link here.  Maybe now it will get more than 2 stars

Here is a picture of it:

EDIT: It just got approved.  Here is the link: http://gallery.live.com/LiveItemDetail.aspx?li=44a5ddd7-daa1-46cc-a9fa-7204b734f074

Wrap this in HTML, BODY and % tags

Dim UserIPAddress
UserIPAddress = Request.ServerVariables(”HTTP_X_FORWARDED_FOR”)
If UserIPAddress = “” Then
UserIPAddress = Request.ServerVariables(”REMOTE_ADDR”)
End If
response.write(UserIPAddress)


Post this on a webserver that you have access to as something.asp, and use that address in the gadget.

After finally getting it installed (so much for actually reading the documentation), and using it this evening, I have to say I really love Vault.  And really it isn’t specifically Vault, but source control entirely.  I have been experimenting with have a collection that held another collection of a class.  And serializing it out to an XML file.  It was basically MasterCollection, that inherits CollectionBase, and held GroupCollection, which also inherits from CollectionBase.  GroupCollection held Accounts.  The idea was there would only be one MasterCollection, there can be more than one GroupCollection which holds many Accounts.  No problem so far.  Until I tried to add a variable Name to the GroupCollection.  The name never got serialized.  What I got was <ArrayOfArrayOfAccounts></>. 

I ended up trying many different ways to fix this.  I would check the code out of Vault, make changes to it, see if it worked, and if it didn’t, I’d just revert the check out and check it out again.  Sometimes I would check in the code so that I could look at it later.  This allowed me to experiment without having to copy the folder to another place on the hard drive.  I already have multiple copies of previous projects laying around and can’t tell which is the latest/greatest.  With Vault, I put in comments what I changed/modified.  So I can look back at it later.

And something else that I noticed for the first time today.  I use the personalized Google page for my homepage.  I have a tab set up just for the different blogs I look at.  Sometime Google snuck in a new feature.  It now shows me the date and/or time of the posts.  So I can tell by looking at the date (if the date posted is not today) or the time (if the date posted is today) and know if I read it or not.  Each post also has a + that you can click to expand to read the whole post.

I just hope that it shows up correctly for stats when I look at a post using the Google site.

I have mostly finished with the gui portion of the project.  There are a few details to finish up with, namely coming up with the registration code format.  Also, tying it to the website, which isn’t completed yet either.

In the process of testing, I determined that my way that I was going to store data will not work.  So I deleted that section and started over.  But after I deleted the section and saved it, I thought it would be nice to be able to see how I did it before.  This led again to getting Vault set up and running.  Well, I am working through the dependencies now.  First, I had to install IIS (I am using WinXP in a Virtual PC to run Vault).  Once that was done, it then told me that I needed to install the .Net 1.1 framework.  So I go to Windows Update, which lists 67 critical updates.  This is essentially a clean OS install.  Adding .Net 1.1 framework made it 68 updates.  Right now it is on #8.

And on it goes…

My older son (who is 8 years old) has recently begun to like country music.  Or specifically one band, Rascal Flatts.  This started because they did a song on the Cars movie soundtrack.  He received their new CD for Christmas this year.

Now I don’t mind him listening to country music.  I like country music myself.  And he could be listening to a lot worse!

But now that he is starting to pay attention to music and radio stations, I learned about the sorry state of affairs in FM programming.  I usually only listen to XM radio.  I have the boom box in my office at work and the car kit in my vehicle.  The only time that I listen to anything else is I turn on the local AM news/talk station to hear traffic reports, usually only after I am in a traffic jam.

There used to be one station that I listened to on FM.  It’s format was 80’s music.  But they changed a while ago to 80’s – today, and I don’t particularly care for today’s pop music.

Now the local stations have done a format dance again.  The oldies station is now country music (making it two country stations in the area), and the former 80’s station is now oldies (which is now defined as 60s-80s).

It has been tried in the past to have two country stations in the area.  It hasn’t worked in the past, but that doesn’t keep radio stations from trying.  I just wish the established country station would be the one to change format and dump their morning show.  And this is mainly because their morning show has a guy pretending to be an old woman.  That just drives me nuts.

As far as I know there are only two FM stations in the area that have kept the same format for years.  One is the aforementioned country station, the other is the classic rock station.  The one-time 80’s station was a gospel station when I moved to this area.

Enough ranting.  Back to regularly scheduled programming

Or how I spent this evening.

My one goal for this evening was recovering files from a hard drive.  The hard drive had lost its partition information.  The program that I regularly use for this purpose (RecoverMyFiles) was not working well.  It gave me a divide by zero error when trying to do a Fast Format Recover.  The Long Format Recover took hours (around 12 hours if I remember right) and was not able to recover much of the files.  It could not rebuild the folder structure and a lot of the filenames had @s in the names.

So I go searching google for lost partition hard drive or something like that.  I downloaded a couple programs to try out.  One had a nice looking UI, but the button images did not make any sense at all.  And that one wanted to phone home when exiting.  ZoneAlarm caught it doing that, so all it guaranteed was that I was going to uninstall it.  Somehow the clueless wonders that wrote the program didn’t manage to get it listed in Add/Remove programs.  Luckily they had a uninstall link in the start menu programs folder.  Both of the programs that I downloaded had big readability problems.  Poor english, using custom boxes for “tips” that did not have scroll bars, so you could only read about half of the tip.

Another program thought it was so good, that there wasn’t any reason to offer a trial version.  I paid for that, realized it would not work, and then had to open a support ticket to get a refund.  If I do not get a reply back by tomorrow evening, I will have to charge it back, which means instead of a $25 sale, they get $0 in sales and $25-30 in bank charges.  Maybe that will make them rethink not having a trial version.

Well I finally got the files recovered.  I used a DOS program called SpinRite to fix the hard drive, then a program called GetDataBack for FAT to recover the files.  A whole evening wasted!

And I still don’t know why I get that divide by zero error in RecoverMyFiles.  I am waiting for a response from support about the bug I sent them.  They did give me a debug version so they could capture more information.

This is the first post in this blog.

I plan on using this blog to flesh out ideas that I have regarding development, sales, and marketing my software product.

Right now I think that I will begin working on a database application that I will be able to use at my day job.  It would have entries for each piece of equipment in my office, and allow me to trace paths from one piece of equipment to another.  Sortof like if you had a room of servers, switches, cat5 patch panels, etc.  My app will have an entry for a switch, and then each port on the switch will be traced to where it plugs into. 

I believe that this will help a lot in untangling the web of wiring, panels, and equipment that I have to deal with.  Yes, my current employer has a huge database that has all this information in it.  The problem with that database is it is huge (containing every panel in 70+ offices), and hard to find the information that I am looking for.  The other problem is that each time my office was purchased by another company (2 times in the last 2 years), the migration from one database platform to another is difficult.  As a result, most of the information that I can put my hands on is in Excel spreadsheets.

For one office, the huge database platforms are too much.  I want something that is easily updatable and searchable.  I will be able to put a customer name on a path, along with what the ID is now, what it was last year, and the original order that installed it.

Anyway, that is where I am at now.  More info to come.