Step 1
Navigate to the Prism installation directory, in Windows this is:
C:\Program Files\Prism\

Open up the extensions folder, then the DOM inspector extension folder:
C:\Program Files\Prism\extensions\inspector@mozilla.org

Now open the “install.rdf” file and find this section of code which can be used (usually) as-is in most extensions pasted in the right place (just use the extension’s preconfigured targets as your guide on where to place it):

<em:targetApplication>
<!– Prism –>
<Description>
<em:id>prism@developer.mozilla.org</em:id>
<em:minVersion>0.4</em:minVersion>
<em:maxVersion>1.0.0.*</em:maxVersion>
</Description>
</em:targetApplication>

Step 2
Open up your extension’s xpi file using 7zip or your favorite archiving tool and extract the install.rdf file.

Open that install.rdf file with a text editor (other than notepad which doesn’t understand the line breaks used by other editors) and find the area where it lists the installation targets and insert your copied text into that or modify one of the install targets with the appropriate information from the DOM inspector extension install.rdf file.

Step 3
Install the extension. This will get tricky due to the way Prism’s developers have modified the profiles and what not. Well, I say that, but it really only gets that tricky if you’re manually installing the extension. Prism developers have seen fit to include the Addons menu item from the Prism menu in the lower right of Prism’s window to help things go more smoothly.

One of the ways to use this new install feature for an extension is globally, which means all webapps will be able to use the extension. This is the least preferred manner due to the fact that you won’t want all extensions available for all webapps in most cases. An example would be having the Better Gmail extension available for Bloglines – don’t think that’s real necessary, do you? So plan accordingly.

Using the second, and more preferred method, you just open your webapp. Open your webapp’s Addons panel, then install by either dragging and dropping the extension onto it, or by clicking the “Install…” button and navigating to the extension and selecting it for installation.

Conclusion
That should be about it. My next Prism post will be about how to make one of two stand alone Firefox/Prism exention webapps – POW local server or Scribefire standalone webapp. Stay tuned, we get to play with chrome!!

Links:

Prism

Any-who, one of our labs has us using mount and umount (did I mention that RHA030 is a basic GNU/Linux skills class with a RedHat flavor?) with a floppy drive. There are a couple of problems with this, however. One, there’s no virtual floppy disk on the virtual machine that’s been configured for class. Two, VMware Workstation 6 has this weird quirk (at least the version we have) that won’t allow you to create floppy images in the img format. It adds flp to the end of whatever you put want to name your file, and then gives this error that the file already exists, which doesn’t make a lot of sense since the Workstation just created the file and should know that.

So, here’s a (mostly) graphical tutorial that allows you to create your own floppy image and use it with VMware Workstation 6.

Step 1. Open VMware Workstation and select the virtual machine you want to add a virtual floppy drive to.

Step 2. Select “Edit Virtual Machine Settings” from that virtual machine’s “Commands” menu.

Step 3. From the “Virtual Machine Settings” dialog’s “Hardware” tab, click the “Add…” button

Step 4. On the “Hardware Wizard” dialog, select the “Floppy Drive” hardware type and then click the “Next” button.

Step 5. On the “Hardware Wizard” dialog, choose the “Create a blank floppy image” radio button and then click the “Next” button.

Step 6. On the “Hardware Wizard” dialog, click the “Bowse” button to choose a location for your floppy image.

Step 7. Type in the name of the image file you want to create. Notice I tried to input an “.img” extension which is one of the types listed in the file type select box. Don’t bother trying that because Workstation will automatically tack on a “.flp” extension regardless of what you input.

Step 8. Check your file’s path and then click “Finish”. What’s supposed to happen is, Workstation creates an image file with the name in the text box.

Step 9. What happens is Workstation adds the “.flp” extension, creates the image file, then gives you this Alert box stating: “Unable to create floppy image. File exists.” Idiotic really. Just click the “OK” button and proceed as follows.

Step 10. You’ll now see that your file name has been appended with the “.flp” extension and that and further “Finish” clicking will merely give you to opportunity to view the Alert box again. But don’t worry, because even though you’re getting error a strange and magical thing has happened. Click “Cancel” to proceed to the next step.

Step 11. Click “OK” on the “Virtual Machine Settings” dialog.

Step 12. Ok, this is less of a step than just a view of the results. Remember that “strange and magical” thing I mentioned? Well, as it turns out Workstation has added the floppy drive to your virtual hardware and created and added the floppy image to your floppy drive just fine even with those errors. Weird, weird, weird, but I’m not really complaining, now I can do my lab!

Hope this helps anyone still having problems with this.

Project Background

Anyway, the back story for this project is; about 3 years ago, my daughter (around 6 at the time) accidentally spilled a cup of coffee (no she wasn’t drinking it!) onto a brand new Seagate 5GB USB 2.0 Pocket Hard Drive. I had set a cup of coffee down next to the new drive on the coffee table, and needed to go back into the kitchen for something. I specifically warned her to be careful not to bump the coffee or the table because I didn’t want my coffee spilled. Well, that must have triggered some inborn child-instinct that says “be extremely reckless when told to be careful.”

At any rate, the coffee got spilled and I got a chance to get a look at the internals of my new drive while I was drying it out. While doing this, I got to thinking about how small the microdrive was and how it could fit into a lot of different casings. Of course, I was also thinking of how to better protect it from liquid damage and how I had so many – now less useful, smaller capacity – thumb drives laying around.

I thought that if I could get a small enough USB hub that I could fit that inside a piece of PVC or other sealable container and plug the thumb drives into it. With my new drive back together, dry and ready to go, that’s just what I did! After all I had to figure something out to do with all those old thumb drives!

Materials and Build

I got a USB hub and some other pieces and started to work. Of course, all these parts took quite some time to gather. I managed to get a few left over parts from work (sealable caps from network connectors) and ordered a USB-B jack for plugging in a USB-A to USB-B cable into. I could have easily gotten a smaller, mini-USB connector and cable assembly, but my soldering skills are not that hot.

I ordered a 4-port USB hub (to give you some idea of how old this project is, I bought the hub on ThinkGeek and they no longer carry the hub I bought) which had two stacked 2-port USB-A jacks. I also needed some plastic resin to seal the USB jack. I used a 2″ PVC compression coupling as a housing and a couple of 2″ long PVC pipe sections as end pieces.

I removed most of the housing and USB cable from the hub and replaced the USB plug on the cable with the jack I’d gotten. Then I sealed the jack inside the PVC pipe with the plastic resin. Before I did that however, I needed to close the back of the USB jack with a glue stick gun to keep the resin from filling the jack and making it useless. I still ended up getting some inside, but it wasn’t much as I can still insert the USB plug into it.

In order to get all the thumb drives into the tube, I had to strip the housings from them. I used a thin piece of plastic to separate the drives and keep them from performing their own circuit bending operations on any of my data. I had to scrape some tracks into the inside of the compression fitting in order to get the hub’s circuit board in all the way. The other end of the tube was going to hold another USB-A jack, to be used as a kind of feed through USB port, but as it turns out, I didn’t fill the back of that connector correctly and the resin filled the jack. In any event I don’t think the cable connecting the 4th USB port to that jack would have allowed the hub to fit in far enough to fit along with all the drives. So it’s probably better that I just used it for a 4th thumb drive anyway.

The Cable

I also used a piece of conduit as a cable keeper, so that my USB cable doesn’t get tangled with other cables when in transit. I use this trick, albeit with a shorter section of larger diameter conduit, with my laptop cords too. I slit the conduit or take a ¼” strip longitudinally out of the center (depending on rigidity), neatly bundle the cables (folding in half 3 – 4 times), insert into conduit, and viola – instant portable cable minder.

Software

Well all this is well and good, but there’s a problem I didn’t foresee. That would be what a pain in the but this is to unmount all these devices on Windows. With a GNU/Linux system, you simply right a script to unmount the drives all at once (and presumably put that on one of the USB drives), Windows on the other hand, sucks for this kind of functionality. But there is a half-assed solution to the half-assed situation I’ve put myself in. The command:

mountvol X: /d

Where “X” = your thumb drives’ drive letter, works to safely unmount the volume under Windows (thanks and attribution to Joel on Software), but doesn’t solve disconnecting Windows from the device (you still have the little “Safely Remove Hardware” icon on your system tray). So you need the DevCon tool available here. Now you can use the mountvol command and:

devcon remove @usb\*

Conclusion

Hot, no? Actually yes, this set up does get pretty hot after a while. I guess that’s a drawback of having this in a sealed container. Another drawback, I’ll never be able to take this on an airplane… :(

If you’re looking for something to do with all those old USB drives that are getting exponentially smaller with every passing day, then this might be a solution for what to do with your old drives. But I think the next time I do this, I’ll use the left over clear resin and do something a little more slick, like Russell Jones, and encase the whole shebang in a clear plastic block.

• a file archiver – my favorite is 7-zip, but you can just as easily use any archiving program that is compatible with zip files
• a text plain text editor

For more advanced webapps, a more advanced context highlighting text editor can be helpful for coding custom style sheets or adding javascript customizations. And don’t forget icons! I use a combination of Paint.Net and @icon sushi (which covers everything but MacOS, which is because I can’t seem to find software for Linux or Windows to accommodate icons for it – recommendations anyone?).

The No “.webapp” WebApp

There are several ways to open webpages with Prism or WebRunner. The first and easiest is to install Prism, then create a link to it on your desktop. Then change that link adding the option “-uri” and the url of the page you want to open. Here are two examples:

Windows:
c:\your\path\to_prism\prism.exe -uri http://web_url_of_site c:\your\path\to_prism\prism.exe -uri c:\location\of\local_file

GNU/Linux:
/your/path/to_prism/prism -uri http://web_url_of_site /your/path/to_prism/prism -uri /location/of/local_file

This method does not require a webapp file, it just launches the website or file you specify. This means, no icons are required, but you also don’t get to add any custom themes or js files. This would be a really good method to use with applications like Wiki on a Stick (woas).

WebApp Bundles

The second method is also pretty straight forward (until you make it complex by adding more functionality). For the most basic webApp bundle you only need to create one file; the webapp.ini. Here’s the content from the example webapp.ini on the Prism wiki:

[Parameters] id=unique-app-id@unique-author-id.whatever uri=http://[the-url-what-you-want-to-connect-to]/ status=yes location=no sidebar=no navigation=no

The “id” parameter is required to be unique. I’ve been using my website’s url (as most others have done as well) preceded by the name of the site, but any unique identifier will work. The “uri” parameter is required and will accept both url of the target website as well as local file paths. Everything after that appears to be optional and the values show are the defaults. These remaining parameters are options to tell Prism whether or not to show certain GUI elements (“status” = status bar;”location” = address bar;”sidebar” = sidebar;”navigation” = well, quite frankly, I have no idea – maybe navigation buttons that haven’t been implemented yet?).

The next step after saving your changes to this document, is to use your archiving program to zip up this file creating, as an example, “your_webapp.zip”. Replace the “zip” extension with “webapp” and you’re ready to roll! Now either double click the webapp file or you can launch the webapp from the command line thusly:

Windows:
c:\your\path\to_prism\prism.exe -webapp your_webapp_bundle

GNU/Linux:
/your/path/to_prism/prism -webapp your_webapp_bundle

For more advanced webapps you can add icons, scripting and style sheet support (see the Prism page for more details).

Happy coding!

Prism
Prism/WebApp Bundle
Prism/Scripting
Prism/Styling
7-Zip
Paint.Net
@icon sushi
Wiki on a Stick

Update!!
Had to add a comment to the replacement code below in order for this to work.

080404 Update!!
Yet again, code changes for new versions and what not.

Looks like Mozilla has added WebRunner to it’s Mozilla Labs line up and changed the name to Prism. From what they say, it’s going to be a cross between an open version of Adobe Air and Microsoft Silverlight and a localized lancher for web applications. Which, by my definition, is what utility computing is supposed to be like. Prism is currently for Windows only (they’d better hurry up with a Linux version), but it runs the same webapp files and also has added style sheet support – among other things – so you can style your webapps your way.

Anyway, I’ve been kind of discouraged by the lack of extension availability for WebRunner (and XulRunner for that matter). You’ll find in another of my posts a webapp for accessing the WebRunner extensions panel. Using this webapp, you can add new extensions (by installing or dropping them onto the panel), but at this point, adding extensions will fail because the version checking mechanism for WebRunner/XulRunner. To get around that, you need to change the extension’s compatibility information. These instructions will not work for every extension and your results will vary…

I recommend you find a smaller extension, download it and make the following changes:

1. Open the extension package (you can open it with 7-zip, ALzip, Winzip or any other archive program that works with zip files.
2. Extract and open the install.rdf with a text editor, I’d recommend Wordpad if your using windows.
3. Locate and change the following or something that looks like the following (this will vary by extension, but look for the Firefox description section – and you may find that you have to remove all references to all other Mozilla programs that the extension is compatible with for this to work):

<!--Firefox-->
<RDF:Description RDF:about="rdf:#$djS7s"
  em:id="{ec8030f7-c20a-464f-9b0e-13a3a9e97384}"
  em:minVersion="0.8"
  em:maxVersion="2.0.0.*" />

To:

<!--Webrunner-->
  <RDF:Description RDF:about="webrunner"
    em:id="webrunner@webapps.org"
    em:minVersion="0.1"
    em:maxVersion="10.0" />

or (again it will depend on which way the developer implemented the Firefox description):

<em:targetApplication>
  <!--Webrunner-->
  <Description>
    <em:id>webrunner@webapps.org</em:id>
    <em:minVersion>0.1</em:minVersion>
    <em:maxVersion>10.0</em:maxVersion>
  </Description>
</em:targetApplication>

If you’re using Prism, change to the following:

<RDF:Description RDF:about="prism"
  em:id="prism@developer.mozilla.org"
  em:minVersion="0.1"
  em:maxVersion="10.0" />

or

<em:targetApplication>
  <Description>
    <!--Prism-->
    <em:id>prism@developer.mozilla.org</em:id>
    <em:minVersion>0.1</em:minVersion>
    <em:maxVersion>10.0</em:maxVersion>
  </Description>
</em:targetApplication>

4. Then save the rdf file and add it back into the xpi archive.
5. You should be ready to install the extension using the Extensions.webapp

I found this parts of this bit of magic on Alex Sirota’s website along with the tip to check in the applicaiton.ini file for any XulRunner application’s ID. If you run into errors, then something is probably malformed in the rdf file so check that first. I won’t be much help in trouble shooting, and I’m not about to offer support for other peoples extensions on an unsupported application, but if you have problems, post a comment and I’ll give editing the rdf a go myself.

I’m no real programmer, I mean, I know enough to be dangerous. I’ve got a good handle on html, css, javascript, asp etc. In the past I’ve played with the source code for Portable Firefox from PortableApps.com trying to get XulRunner portable. I didn’t have much luck. Ran into too many code errors to resolve myself. What I needed was a way to make the program portable without programming anything new.

X-Launcher

What I was able to find was X-Launcher. X-Launcher is a nifty little app that allows you to launch just about any application portably – that is, without leaving much of a trace on the host computer. I’ve been able to use X-Launcher without too much effort (hey, just my style :).

You can get X-Launcher from winPenPack.com (downloads section) along with a host of other open source software all made portable using X-Launcher or other means. WinPenPack is an Italian website, but you can get English on most pages by choosing it from the language menu. Nice site and a lot of good software!
X-Launcher itself is an easy to use program, you simply configure the X-launcher.ini file for it and you’re ready to use it. One of the things that really sets X-Launcher apart is the fact that you can have multiple launchers in the same directory. Just rename the X-launcher.ini file and the X-launcher.exe to any name you wish (just as long as both have the same file name) and you can put a second set of exe/ini files with a different set of names in the same directory together. Why would you do this? Well, for one, this allows you to run multiple versions of a program with wildly different configurations. Another thing this buys you (and more relevant to this discussion) is the ability to launch different WebRunner apps portably.

Normally if you launch a program you can pass it options through command line switches. If you’re using another program to launch the application you want to run, then you don’t usually have the option of passing switches to the second program from the first. With X-Launcher, these switches are built into the configuration file.

The default configuration file that comes with X-Launcher is loaded with information on feature settings, but it’s also a bit overwhelming. Fortunately, if you run X-launcher.exe with no like-named ini file, it will create it’s own, scaled down X-launcher.ini. This configuration file is a heck of a lot easier to read!

There are several application variables that can be defined, the most important one being the application name. This will eventually point to folder names and the application itself. Here’s the contents of a X-launcher.ini for a webapp with included explanations for most settings:

wikipedia.ini

[Setup]
AppName=WebRunner
;Application name, this will also be used below as the
;name of some of the folders.

[FileToRun]
PathToExe=$AppName$\$AppName$.exe
;This is the name and path to the file you're trying to
;launch.
WorkingDir=%HOME%
;This is the name of the directory which will be used
;by the program as it's working directory.
Parameters= -webapp wikipedia@luckydisasters.com
;List of any parameters to pass to the application, in
;this case we're telling WebRunner to load the webapp
;"wikipedia@luckydisasters.com"

[Environment]
HOME=$HOME$\$AppName$
;This folder will be used to save any user specific file
;information - like the Mozilla profiles folder.
USERPROFILE=%HOME%
;As you can see I've used the same folders for all the
;programs files, this will make it a lot easier later if we
;want to see what the program is doing or make
;changes.
PATH=%PATH%;@ScriptDir@
;This adds the @ScriptDir@ variable to the system
;path used by the application

[Functions]
DirCreate=%HOME%
;if this folder doesn't exist when you start the
;program, it will created. You can place other
;functions here as well if you need them.

[Options]
;These settings are other X-launcher application
;options, such as showing a splash screen and
;what not, feel free to make changes, but these
;should be fine as is
DeleteTemp=true
MultipleIstances=true
FixAppData=false
RunWait=true
ShowSplash=true
WriteLog=false

[SplashScreen]
;This section allows you to add a custom splash
;screen and sets the length of time the splash
;screen is shown. Leave the Image and Title blank
;for X-launcher defaults. The image path (as well
;as all paths in this ini file) are relative to the
;current location of the ini/exe files for the
;X-launcher app.
Image=
Title=
TimeOut=1500

Download the Wikipiedia X-launcher files here:
Wikipedia X-launcher

7-Steps

What I’ve done to use the above webapp portably is as follows (assume that the USB drive/portable device is drive letter “X”):

1. Make sure you have WebRunner and the Wikipedia.webapp installed.
2. Create a folder under “X:\PortableApps\” (or where ever you store your portable applicaitons) called “X-WebRunner” (“X:\PortableApps\X-WebRunner”)
3. Copy the “C:\Program Files\WebRunner” folder to “X:\PortableApps\X-WebRunner” (“X:\PortableApps\X-WebRunner\WebRunner”)
4. Copy the wikipedia.exe and the wikipedia.ini files into “X:\PortableApps\X-Webrunner”
5. Run the Webrunner.exe, this will create the user profile you’ll need to copy your webapp into
6. Now copy your webapp folder from your profile folder on your hard drive usually located about here:
   “C:\Documents and Settings\YOURUSERNAME\Application Data\WebRunner\Profiles\RANDOMSTRING.default\webapps\WEBAPPNAME”
   to your new profile folder on your portable device:
   “X:\PortableApps\Application Data\WebRunner\Profiles\NEWRANDOMSTRING.default\webapps\WEBAPPNAME“
7. Last step, start the wikipedia.exe file again, this will start the webapp that you’ve copied.

My “X-WebRunner” install:

Now you can start moving your webapps over and making X-launcher.exe and X-launcher.ini files for each one, the only line in the ini file that needs to be changed for each is “Parameters“, which gets changed to the webapp folder name for each webapp. So basically all you have to do is copy the original wikipedia.ini file, rename it and change the Parameters setting.

Next I’ll have to configure Plain Old Webserver USB (POWUSB) to run with X-launcher. Then I can carry around a nice little XUL Webserver!

Links:

winPenPack
X-Launcher download
My Wikipedia X-launcher configuration
My Wikipedia.webapp Bundle

copy access\*.* access.log

This copies all the files in the access directory into the file “access.log”.

How To: Redirect Errors From Standard Out to Nowhere (*nux)

Here’s how to get rid of those annoying error messages you really don’t care about when running a script (everybody’s got their own excuse for not wanting to see this type of thing, even though it can be handy now and again):

user@computer:~$ [command] [options] 2>&1

Stick “2>&1” at the end of your command string, that about covers it!!

How To: Launch Firefox With a Specific Chrome Interface (Windows)

Ever install a nifty little Mozilla Firefox extension and want to run it as a standalone program? Here’s how:

C:\Program Files\Mozilla Firefox>firefox.exe -chrome [chrome path]

Add the “-chrome” option to the firefox command, then add the chrome path after that. Here’s an example of how to launch the bookmarks manager as a standalone program:

C:\Program Files\Mozilla Firefox&gt;firefox.exe <strong>-chrome chrome://browser/content/bookmarks/bookmarksManager.xul</strong>

Enjoy!!

1. You need three things to better promote your website. A meta tag to summarize your site contents, a sitemap.xml file to direct search engines on where to look for content you want them to find, and a robots.txt file to tell search engines what not to look at. Keep in mind that, from a security standpoint, any one of these things can be used against you by someone gathering information about your site (reconnaissance is the first step to any action). So balance your desire for traffic with a healthy dose of paranoia (see paranoia is healthy).

Meta tags should include search phrases and text, but don’t go and overload the tag with content. Keep it as lean as possible while still covering as much as you can. The more focused your meta tag, the better your page rank will turn out.

2. The more your site is linked to, the higher your page rank for certain terms. Usually this type of rank is context sensitive, i.e. where the link appears weighs toward your page is rank. If you post to forums which are related to your site, I’d recommend adding a link to your site in your signature or in your user profile, or even hosting content on your site and posting links to it in these forums. When a search engine crawls that site, it will pick up those links and help increase your page rank.

3. Submitting your page to search engines does help. As a matter of fact, Google offers quite a few good resources for webmasters. Webmaster Tools, Webmaster Central and Webmaster Central Blog can help considerably. Also, if your trying to submit you page to Google, the link isn’t always the easiest thing to find, so here’s the link. Google’s Webmaster Tools is a good way to both verify your sitemap.xml and robots.txt. It also allows you to view other statistics about how Google is crawling and caching your site.

Here are some other resources you might find helpful:

Web Server Administrator’s Guide to the Robots Exclusion Protocol
sitemaps.org
.htaccess Tutorial
Webmaster Help Center – How do I add my feed to the search results for Google’s personalized homepage or Google Reader?
WEBMASTER GUIDELINES DIRECTLY FROM GOOGLE
11 Ways to Find New RSS Subscribers for Your Blog

[cpg_imagefixthumb:145]A few months ago I picked up a two pack of Great Stuff expanding foam insulation. I needed the first can when I bought it so the other can sat on a shelf in our storeroom until today. I’d decided to finally fill in the gaps around the window in our bathroom. So I got the can, shook it up and was about to put the applicator/handle on when I noticed that it wasn’t assembled. It’s supposed to come assembled. In other words – it was broken, damn – now what. I didn’t want to run out to the store to get another can just for another applicator.

That leaves just one option, trying to fix it myself. I tried gluing it using the only adhesive I had. Crazy-super-extra-hold-tight-glue-stuff (you know, the generic stuff), it didn’t work, I held it together for several minutes and it fell apart as soon as I removed pressure. So either the glue’s bad (it is an old tube) or whatever plastic they used for the applicator is some sort of polymer that resists chemical adhesion (well I guess it would be, after all Great Stuff doesn’t exactly stick to the applicator and it does a great job at sticking to my skin).

Well now what… Time to improvise! I looked around for a properly sized tube of some sort, preferably with a handy bend in it. Then I remembered, we have bendy straws in the kitchen! Woohoo – what can’t bendy straws do!

[cpg_imagefixthumb:144][cpg_imagefixthumb:143]I checked the size of the straw against the size of the threaded tube coming out of the can. The straw was just a little too small, That’s easily fixed I used a knife and a pair of scissors as augers to stretch out the end of the straw so that I could fit it on the can. Unfortunately this process twisted the straw a few times and made the straw too weak to press down onto the can’s tube without bending the straw (making the situation worse) or squirting some stuff out. I could have tried another straw (after all I had a whole pack to work with), but something told me I’d be going through quite a few straws before I got one that worked. Or worse yet, I wouldn’t get one that worked.

Hmmm… What to do? Actually, my wife had the solution. She handed me a green tube, which was very stiff and just a hundredth of an inch or so larger than the straw. My wife being sometimes the helpful sort, but more often than not the tortuous sort (she knows this, and seems to be just fine with it) told me with a smile that this tube was from a tampon.

Yikes… Well – I guess that works too…

[cpg_imagefixthumb:142]So I slid the green tube over the straw. Then, I carefully slid the straw onto the the can’s threaded tube. Carefully because Great Stuff is, well, great stuff, but it takes forever to clean off your skin and I really didn’t want it spraying out of the can while I was trying to put something that doesn’t quite fit onto it. [cpg_imagefixthumb:141]I then slid the green tube down the straw and over the can’s tube. This gave me an extra tight fit and enough rigidity to use the straw as the applicator. I then went to work on the window. I pressed on the can’s tube to start the flow, it hissed and the stuff started to flow! The tampon tube/straw combo worked! Thanks to the tampon tube it didn’t pop off right away and [cpg_imagefixthumb:146]the straw didn’t bend when I put pressure on it. The straw did pop off once when the can was nearly empty, but this was because I was pressing on the can’s tube where the straw was fitted to it, and I pushed it off – I was able to put it back on and continue.

So there you have it. If you loose the applicator or it breaks, you can always grab a tampon applicator and go.. to… town… Ahem, yeah.