Monthly Archives: October 2007

Throwing out the Wheel

Filed under Uncategorized

No, I’m not converting my blog here to Craig’s List or anything, but I figured I’d throw out first crack to anyone that reads here.

An Force Feedback wheel is the only way to play driving games IMO. Driving with a freakin’ thumb controller just seems so….dirty.

But I’m giving up on making my Force Feedback Wheel work with Vista.

It’s USB, it’s a HID device. But the drivers just hate Vista. No idea why. I’m guessing that if I knew how to isolate the USB device ID, I could modify the INF files that come with the latest Thrustmaster drivers for Vista and get it functional. I just don’t have the wherewithal, and I’m not a device driver writer by trade, so it’s a tad more research that I want to bite off right now.

Works great in XP, though. Has paddle shifters, strong force feedback, good pedals and a thick wheel with rubber grips.

It’s a Guillemot wheel, and from what I can tell online, that’s actually Thrustmaster under the covers.

Wheel1a Wheel2a

If you’re in the Dallas, Ft Worth area (that’s Texas), let me know if your interested.

I’d let it go for cheap (say 50$ obo).

Are you a Neanderthal SQL Programmer?

Filed under SQL

I generally can appreciate what Joe Celko brings to the table in terms of SQL programming and theory, but a recent article of his had me scratching my head, even considering his sometimes Ivory Tower approaches.

The article is at Intelligent Enterprise here.

What caught my eye was the following paragraph. He’s discussing his own classification system of keys in databases and one of the types is described thusly:

3. An “exposed physical locator” is not based on attributes in the data model and is exposed to the user. There is no reasonable way to predict it or verify it, since it usually comes from the physical state of the hardware at the time of data insertion. The system obtains a value thru some physical process in the storage hardware totally unrelated to the logical data model ” for example, IDENTITY columns or other proprietary, non-relational auto-numbering devices.

Technically, these are not really keys at all, since they are attributes of the physical storage and are not even part of the logical data model. But they are handy for lazy, non-RDBMS programmers who do not want to research or think! This is the worst way to program in SQL

So, to paraphrase, he’s saying that if you use IDENTITY columns in your database schemas, you’re “lazy” and “do not want to research or think”


So how about it, you lazy, slovenly, heathen SQL programmers out there that dare to use Identity columns?

And in case it’s not obvious, I can’t disagree more with Celko on this point<g>

Arcade Ambience

Filed under Arcade, Media, MP3s

If you’re putting together an Arcade cabinet or if you just like the ambience of an old arcade parlor while hacking away at some code, you might get a kick from Andy Hofle’s Arcade Ambience project.


Basically, Andy took recordings of playing dozens of arcade machines, remixed them, adjusting volumes, pans, etc, and ended up with several, very large, mp3 tracks of background ambience that sounds very much like stepping into an old arcade. What’s even better, the tracks are so large (at 70+mb each), they effectively don’t loop, so you don’t get that been there, heard that feeling that’s typical of ambient tracks.

Couple that with an arcade front end that can play an arbitrary mp3 looped as background sounds, and couple that with a recent audio card that supports multiple simultaneous channels, and you get all the ambience of walking into an arcade while playing any emulator, Visual Pinball table, etc.

The USS Millennium Falcon

Filed under Misc, Rants

NASA has a nifty little “picture of the day” feature that you can add to your Google Homepage.

A picture that came up just yesterday shows the shuttle lifting off against a gray sky. Here’s a snippet (from


But the thing that struck me about this is how much our space program is beginning to resemble the Millennium Falcon.


The rusty tanks, the patchy paintwork, using a stapler for repairs. It’s starting to look like the shuttles may have made the Kessel Run a few times themselves.


(Detail from above)

Compare to this 1983 picture from the Wikipedia entry:


And this about 20 months before the first shuttle launch (again from Wikipedia):


Clean and sparkly

Cripes, it’s 2007. Shouldn’t we be flying something like this:



and not this:



Nesting the SQLExpress 2005 Install

Filed under Installations, SQL

I recently had the pleasure of trying to get SQLExpress 2005 to install, on request, during my application’s install, according to various parameters specified by the user during the UI portion of my install. Phew. Yeah, it’s hard enough to say, much less code.

Anyway, the basic problem is that you can’t nest MSI installs. And since an “MSI Install” is considered to be the portion of the install that runs during the EXECUTE sequence of the MSI logic, this means you can’t just execute SQLEXPR32.EXE at some point during the EXECUTE sequence of your install.

This is normally what you’d want to do because then, that action would happen regardless of whether your installation was running silently (ie no UI) or not.

Ok, that’s off the table. Hey, Microsoft! Believe it or not, nesting installs is something real world apps have to do in order to get all those nifty libraries of yours out there onto client machines! Sigh.

Fortunately, in my case (and in many others I’m guessing), not being able to run the SQL Express installer during a silent install is not a deal breaker. More than likely, if my installation is being rolled out silently, whoever’s rolling it out is using some sort of automated rollout tool to do it, and it’s likely they’ll just automate the rollout of SQLExpress as well.

But, what about when a user is simply trying to run my installation to get the product on their machine. In this situation, you want a no fuss, no muss process that has a user up and running with a little decision-making as possible. The last thing I want a user being prompted for is the instance name to use for the SQL Server that’s about to get installed. Sheesh.

Well, I can’t nest MSIs, technically, but I can execute SQLEXPR32.EXE from the UI sequence of the install.

Aha! Success! Except that, as well documented by any number of posts online, it doesn’t work consistently. Often, you’ll get an MSI error 110 indicating that SQL can’t access a file (usually sqlncli.msi) that it needs to complete the install.

            ******* CommandLine: **********
MSI (s) (18:00) [12:59:55:555]: Machine policy value 'DisableUserInstalls' is 0
MSI (s) (18:00) [12:59:55:575]: Note: 1: 1309 2: 5 3: C:\23661fc1e2705da3b45f5b05\setup\sqlncli.msi 
MSI (s) (18:00) [12:59:55:595]: MainEngineThread is returning 110
The system cannot open the device or file specified.

Trouble is, the file is there, exactly where the log says it isn’t.

The problem is clearly an issue with the SQLExpress installer. I can say that pretty confidently because you can nest all of SQL’s prerequisite installs just fine, including:

  • MSXML6.MSI (The XML 6.0 libraries)
  • SQLCNLI.MSI (SQL Native Client)
  • SQLServer2005_XMO.msi (SQL Server management objects installer)
  • And even VSTOR.EXE (the Visual Studio Tools for Office)

Just not the SQLEXPR32.EXE itself.

One possible solution that I found mentioned online was to extract all the SQLEXPR32.EXE files, then actually include them directly in your installation and at the end of your installation (in the UI sequence) launch the SETUP.EXE that starts the SQLExpress installation.

So I extracted them all:


Ouch. And I mean owwwwwwwwwchaaaaaa. 400+ files scattered over dozens of directories. The InstallShield Component Wizard choked trying to add them all as components and there was simply no way I was going to manually set up all that felgercarb in my install. Yeah, I went there<g>

Then it struck me. The extraction process (that /x command line arg) isn’t actually an msi install, so I should be able to nest it just fine. And if that’s the case, then I could extract all the files from SQLEXPR.EXE dynamically after installing the exe, and then fire up SQL’s setup.exe directly, just as I’d tried launching the SQLEXPR32.exe before (but that failed with the 110 error).

So, first, make sure you install the SQLEXPR32.EXE along with the rest of your application’s files.

Then, set up a Custom Action in your MSI project, in the EXECUTE sequence, that extracts all the files from SQLEXPR32.EXE to some folder (usually inside your INSTALLDIR somewhere):


(note the use of /q; that will keep the extraction process quiet, including suppressing a rather bothersome “Extraction process done” message box when it finishes, Also note that this adds substantially to your app footprint so you may need to accommodate that in available space calculations).

Next, create another Custom Action to execute the SQL Express SETUP.EXE that was extracted in the previous step.


Set this CA to execute towards the end of the UI sequence (because this is an MSI setup and this action can’t be started from the EXECUTE sequence of your installation, no getting around that one).

I set a command line of /qb to use the basic install ui for SQL Express, but you might also want to include other command line parameters, like:


or any of the other options you can specify. More info about the command line options for SQLEXPR32 are pretty easy to find online.

And finally, to keep your installation from looking like it just hung during the extraction part above, be sure to author a row in the ACTIONTEXT table for that custom action, and give it some text, like “Extracting SQL Server files…”. The text will show up over the progress bar that normally displays during an InstallShield-authored installation.

Let me know if this works for you, or if I’ve missed some esoteric combination in InstallShield that prevents it from working.

As always, YMMV.

How NOT to set up an issue tracking email service

Filed under Rants

Here’s a snippet from a recent Microsoft email I received when I opened a ticket with them through the MSDN support channel.

******* The following is an email for a support case from Microsoft Corp.
******* DO NOT REPLY TO THIS MESSAGE--your email will not be added to
******* the case if you do.  Instead, FORWARD your response to the
******* email address COMPMAIL@MICROSOFT.COM and place your text after
******* the keyword 'MESSAGE:'.  Also, delete all other text above
******* and below the keywords 'CASE_ID_NUM: SRnnn' and 'MESSAGE:'
******* to ensure proper delivery of your email.  Thank you.

Now, I’m pretty technical, but any support email conversation that you have to read several times over in order to even reply to is, uh, shall I say, unfriendly. 

And the wording, “DO NOT REPLY TO THIS MESSAGE…”, and “Instead, FORWARD…”. Wow, the antithesis of friendly.

What’s worse, when I did exactly as was indicated, the email failed to go through with my attachment. I ended up, get this, having to send a reply to the very FROM address that the snippet above indicated NOT to reply to!

Now, I’m not pointing fingers at the techs themselves here; I doubt the guy that clicked the button to send this email was actually responsible for the content of this header.

But seriously, someone at Microsoft needs to take a lesson or three in email etiquette.

The Keybowl

Filed under Hardware

I’ve played around with some strange input devices, including the Felix (now called the Altra MicroPoint), the Kinesis Keyboard, and a mouse with a trackball for a scroll wheel among the more notables.

But the Keybowl has to be the most, well, interestingly named of the bunch so far.

What is it?, well, here


You basically put your hands on the two “domes” and slide them around. The left hand is like an 8 way “shift” key, and they right hand selects a specific character from a set of characters depending on the position of the left hand dome.

For someone with limited hand mobility, I could certainly see some benefits, but it does look a bit, er, strange.

They should wrap it in knit.

Sorry, I couldn’t resist.

True Or False

Filed under Code Garage, VB Feng Shui

It seems like such a simple task; check a variable for whether it contains a true value or a false value.

No problem, right? Well, if you’re dealing with numbers, maybe, but when input comes from config files or databases, the truth <ahem> may not be so obviously out there.

As part of my code garage, I thought I’d post two functions I’ve used for ages to do just that; convert a variable, virtually any variable, to a boolean result.

They are especially useful in configuration handling, where you might want to support multiple values that might mean “true” or “false”, like on/off, yes/no, etc. I’ve also found that they can make code clearer by specifying the “assumed” default value if the variable can’t be concretely identified one way or another (is “Bob Thomas” true or false?).

Public Function IsTrue(VarToTest As Variant, Optional ByVal Default As Boolean = True) As Boolean
   '---- Resolve an input variable to a boolean
   '     but convert common "true"/"false" phrases as well
   '     Also, this allows for an easy way to indicate a
   '     "default" value in cases of an undetermined (ie blank)
   '     value
   Dim t$
   Dim s

   Select Case VarType(VarToTest)
      Case vbArray
         Err.Raise 5, "IsTrue", "Can't test an array for true"
      Case vbObject
         IsTrue = ObjPtr(VarToTest) <> 0
      Case vbString
         If Len(VarToTest) = 0 Then
            '---- is true assumes blanks are true
            IsTrue = Default
            On Error Resume Next
            '---- strip to first space
            t$ = Trim$(Left$(VarToTest, 15))
            s = InStr(t$, " ")
            If s > 0 Then t$ = Left$(t$, s - 1)
            '---- clean out any tabs
            t$ = Replace(t$, Chr$(9), vbNullString)
            '---- accept some synonyms (any other good ones?)
            If InStr(1, t$, "YES", vbTextCompare) = 1 Then
               IsTrue = True
            ElseIf InStr(1, t$, "NO", vbTextCompare) = 1 Then
               IsTrue = False
            ElseIf InStr(1, t$, "ON", vbTextCompare) = 1 Then
               IsTrue = True
            ElseIf InStr(1, t$, "OFF", vbTextCompare) = 1 Then
               IsTrue = False
            ElseIf InStr(1, t$, "TRUE", vbTextCompare) = 1 Then
               IsTrue = True
            ElseIf InStr(1, t$, "FALSE", vbTextCompare) = 1 Then
               IsTrue = False
               IsTrue = CBool(VarToTest)
               If Err Then
                  IsTrue = Val(VarToTest) <> 0
               End If
               On Error GoTo 0
            End If
         End If
      Case Else
         If IsEmpty(VarToTest) Then
            IsTrue = Default
            On Error Resume Next
            IsTrue = CBool(VarToTest)
            If Err Then
               IsTrue = Default
            End If
            On Error GoTo 0
         End If
   End Select
End Function

Public Function IsFalse(VarToTest As Variant, Optional ByVal Default As Boolean = False) As Boolean
   '---- basically, the inverse of IsTrue above
   '     mainly for convenience

   IsFalse = Not IsTrue(VarToTest, Default)
End Function

Hey, CBOOL is undoubtedly faster, but:

   If IsFalse(SettingValue) Then 
      '---- handle the negative condition here
   End If

just seems so much clearer.

The optional Default argument allow you to specify what value to return if the value to test can’t be resolved satisfactorily one way or the other. Practically, this allows you to easily specify whether a blank value or a “non-boolean” value should translate to true or false. This is especially important when you’re reading config options from a file or the registry where the option may not exist at all.

It allows to do something like so: 

   If IsTrue(SettingValue, False) Then 
      '---- handle positive conditions here, but if the SettingValue is blank, we default to False
   End If

Simple, but handy.

Video Game Idea (or has this already been done)

Filed under Uncategorized

I was chatting today when the new game Portal from the Half Life guys came up (thanks, Ralf!).

Now, what does this have to do with Visual Basic? Well, nothing. But development ideas can come from the strangest sources so it’s good to keep your horizons open.

At any rate, I’d read about something similar in Wired several months ago, but apparently the concept got picked up by the Half Life team and fleshed out fully. (disclaimer: I have no idea whether I’ve got the backstory straight or not, that’s just how I remember reading it)

Essentially, you wield a interdimensional portal generator, but just think of it as a gizmo that you fire once at a flat surface to open one side of a tunnel, and fire it at another surface to open the other end. Then you step through one end and end up at the other, complete with all the inertia you had going in. You can see the portals in this screenshot:


(image from the wikipedia entry)

Pretty slick idea and it opens up a ton of puzzle possibilities in a game.

But it reminded me of a concept I had years ago, but to date haven’t seen it implemented, but then, maybe I live in a bubble.

Basically, I see the game as a typical first person shooter, land based (like Quake, not like Descent), but the trick is that all the walls, and the ceiling exhibit gravity, either consistently or in varying amounts and not just a few, but all of them. Hence, you can walk not only on the “ground” but on the walls and ceiling, which effectively means there are no walls or ceilings. If you ever read Orson Scott Card’s Ender’s Game, the battle training sequences described there would give a pretty good picture.

Surely, it’s been done before, but does anyone know by whom?

Visual Pinball. Is it real or is it…

Filed under Arcade

Take one Bally pinball plunger off EBay (7.99$)


Hack together a plate and attach a roller microswitch to it (4.95$ at Happ, but 1.98$ at Fry’s)


Position things such that the switch is pressed when the plunger is at rest, but when you pull back on the plunger, the switch releases, like so:


(it’s a vertical picture, normally the plunger sits horizontally)

Then wire the NC (normally closed) connection to an Ultimarc IPAC as just another pushbutton.

Finish up with a couple of Competition pushbuttons mounted to the side of your desk (or some other convenient place).

Now, load up Visual Pinball, grab a few table files (the Black Hole table is especially nice), crank the volume way up, and line up some quarters on the lip of your monitor, just for some ambience.


All (ok, most) of the arcade goodness of the early eighties in a fraction of the space. This nifty thing is, wired up this way, when you pull back on the plunger, the switch is released, which, because of wiring it to the NC connection, has the effect of pressing and holding that button, until you release the plunger. This exactly models the Visual Pinball concept of pressing the Enter key and holding down on it longer to “pull farther back” on the plunger. Granted, it’s not as exact as, say, a cog driven mouse wheel hack, but it seems to be pretty accurate so far and it was a heck of a lot simpler.

Next? Hook a tilt plumb bob


to several contacts to simulate a full-on tilt mechanism (Visual Pinball supports tilt-left and tilt-right keys, not sure about other pinball sims).