Sales and Administration

Where's the answer to my sales related query?

Almost certainly on this page or on our purchase page...

You may also be interested in our license page...

How do your license agreements work?

You can find out about licensing on our our license page...

Why doesn't my license key work?

Probably because the license key is not for the product you are using.

You will have been issued a license key via email. Check the product and version referenced in the email and ensure that these match up with the product and version you are using.

It is not uncommon for purchasing departments to get confused between different versions of a product. So you may have asked your purchasing department to buy ABCpdf ASP but they may have purchased ABCpdf .NET instead.

I asked my purchasing department to buy ABCpdf .NET but they've bought ABCpdf ASP instead. What do I do?

If you accidentally purchase the wrong product you can apply for a refund under our 30 day money back guarantee. You may wish to wait for the refund to be issued before placing an order for the correct product or if time is critical you may wish to place another order before applying for your refund.

Alternatively if the price of your two items is the same and you're within the 30 day period we can normally issue you a free electronic upgrade license rather than going through a refund, order cycle.

For an electronic upgrade we need an email from the person who was issued the license stating the order reference number (or license key) and that an electronic upgrade is required. The email needs to state that all records of the license key have been destroyed and that any physical deliverable has or will be destroyed. This assurance needs to be current - not something which will be done in the future. Email addresses can be found here.

For a refund we need a fax on headed notepaper from the person who was issued the license stating the order reference number (or license key) and that a refund is required. The fax needs to state that all records of the license key have been destroyed and that any physical deliverable has or will be destroyed. This assurance needs to be current - not something which will be done in the future. Fax numbers can be found here.

I've lost my license key and receipt. How can I get a copy?

License and receipt emails are typically issued via our automated system at Element5 or cleverbridge. So by searching for emails from Element5 or cleverbridge (rather than websupergoo) you should be able to locate your license key and purchase information.

If you purchased via Element5 and you can't locate the license key or receipt email then you can log onto your account and retrieve your license key that way. There are a variety of logon options available one of which should be suitable.

If you can't get sufficient information to log on to your account you may with to contact Element5 or cleverbridge and ask them if they can help. In general they can provide much more timely help than we can on matters such as these.

I want a free license. How do I get one?

Please read the linkback page carefully and apply via the application forms which you will find there. Please do not email applications.

What support options are open to me?

It is important to us that you feel happy using our software. If you find something that doesn't seem quite right, please tell us and we'll do something about it.

We don't expect you to have any problems with our software and if you have any queries we will do our best to answer them promptly. However if you need an extra level of security we offer a Platinum Level Support program offering a priority channel through to our engineers.

How can I obtain a quote?

The simplest way to obtain a quote is from our cleverbridge reseller. First go to the preferred reseller page and select cleverbridge as the default reseller.

Then go the the purchase page and click on the product you want. Near the top of the page you will see a button called "Request Price Quote".

How do I place a Purchase Order?

If you purchase online using a credit card the whole process is automated. There is no human intervention which means that there is little to go wrong or delay the process. You should have your license within seconds of placing your order.

Other methods such as ordering by cheque or purchase order involve a fair amount of human intervention. They take much longer and there's more to go wrong. You should certainly allow a week for this type of order to get processed.

So... if you can... please order online using a credit card.

If this is something you just can't do...

Go to our purchase page and click on the 'Select Preferred Reseller' link next to the orange icon at the top of the page. Select Cleverbridge as your preferred reseller. Place your order choosing Purchase Order as the payment method.

Please note that you are required to accept our Terms and Conditions of sale.

What is your policy on cracked software?

See our cracked software statement...

I've lost my linkback license. How do I get it reissued?

If you lose your linkback license you will need to reapply under the current rules of the linkback scheme. Providing the link has been maintained during the period you have had the license, we will reissue it.

Please note that maintaining the link is important. Indeed it is the core of the linkback agreement. It is what you agreed when you applied for your original license. If you do not maintain the link at the URL you told us about then you are required to purchase a copy of the software. See the linkback agreement for details.

We regularly crawl your site to check that the linkback is present. When you reapply we will check our records to ensure that the linkback has been maintained throughout the period in which you have had it. If we find that it has vanished at any point then we will ask you to honor the agreement and purchase a license.

So if you know that there have been issues with the maintenance of the link please do not reapply - it will just delay the fact that we will ask you to purchase a license. If you are not sure whether there may have been issues or not, you can use a public service such as Internet Archive to check that the link has been maintained.

General Technical

I'm having problems with Security. What do I do?

To solve security problems you need to understand the complex and somewhat convoluted way in which IIS security works.

If you're on Windows 2003 an excellent overview of security and identities under IIS6 can be found at Microsoft's IIS Insider. If you find other good articles on security please do tell us.

If you're on older versions of IIS (eg IIS5) see "Building Secure ASP.NET Applications". However the essential information is that the ASP.NET process generally runs as the ASPNET user.

For an overview of classic ASP security see Microsoft's "ASP Troubleshooting Tips and Techniques". If you want to run IIS in an isolated process see "Server Reliability Through Process Isolation". Please note that security works slightly differently when running in an isolated process. This is fully covered in the above article.

Your first step should be to run a simple test case as a standalone application (exe) while logged on as Administrator.

If you run code as an application and it works, and you then run the same code from ASP or ASP.NET and it doesn't, then you have a permissions issue. Practically the only difference between these environments is that one set of code runs as you (probably Administrator) and one runs in a restricted permission account (like IUSR_MACHINENAME or ASPNET).

This is not unusual under ASP or ASP.NET which operate a rather strict security policy. In addition patches and lockdowns can modify the security policy in ways that are not publicly documented. The result is that software installations - not just our software - can be complicated.

Unfortunately it's not possible to list all the required permissions. The way that Windows is designed does not make this feasible. There are just too many possibilities and each can impact on the software directly, on a resource which the software uses or on Windows itself. It's practically impossible.

So you now need to track down the permissions issue...

You may have a standard method for doing this. Here is our suggested method.

Start by thinking if the IIS user is likely to have access to the resources required. Often there's an obvious cause.

Enable Auditing. If something fails you want to hear about it. So go to the Security Policy control panel and turn all the auditing on. Try your operation. Check the Event Log. Even if there's nothing there keep an eye on the Event Log. Something may turn up later. Just don't forget to back out the changes after you've finished

If you're trying to access remote resources (a common cause of problems) try making the resources local. If the resource is a file on a share see if the operation works if you put the file on your local machine. If the resource is a URL on the internet try a 'file://' URL on your local machine. Try these operations both from your IIS application and also as an interactive exe application.

Most security issues are related to file access. Occasionally you will get issues related to registry access. However these are infrequent and tend to be associated with COM 0177 errors.

If you know the resource which your code is failing to access...

You may wish to move to a security audit. Just audit security failures for IUSR and the Security Event Log will tell you exactly where the problem is. See "How to Troubleshoot Permissions in IIS" for details. This is the official way to proceed.

However if you are not sure exactly where the failure is occurring...

ProcessMonitor (or FileMon & RegMon for legacy systems) from Windows SysInternals will monitor - every - file and registry access on your machine in real time. It produces a flood of data but all you need to do is to start it up and then make your error occur. Switch it off and then search through with a text editor looking for file access failures.

If you see FAILURE or ACCESS DENIED you should look seriously at why these are occurring.

There should be at most one or two suspicious events amongst them will likely be the culprit. Note that an easy way of analyzing this type of data is to use the 'Auto Filter' menu item under the 'Data' menu in Excel.

When you use these kind of tools - and start to change permissions - please, please, please be careful. If you're not sure - don't do it.

Can I use your products from Cold Fusion?

In general you can use the ASP / COM versions of our products via the object tag. However there are a few things to be aware of.

While we're always happy to help with the use of our products in any environment we are not experts on Cold Fusion.

You cannot use ABCUpload from Cold Fusion because it relies on IIS features which are not available from Cold Fusion.

Cold Fusion does not support default properties so some example ASP code may require modification to explicitly specify those properties. So for example the following ABCpdf ASP code:

theDoc.Rect = "50 50 550 550"

Would require modification to:

theDoc.Rect.String = "50 50 550 550"

Are there any special issues involved in upgrading your products?

Once IIS is using a DLL and related resources it can be difficult to get it to let go of them. The Microsoft Installer (MSI) should take care of this kind of problem but it doesn't always respond in a predictable way.

For this reason it you should restart IIS before installing an upgrade. This procedure is not always necessary but it's a good idea and it is certainly the procedure you should follow if you encounter any installation problems.

We suggest the following procedure to ensure a smooth upgrade path. First restart Windows to release any resource locks that IIS currently has open. Second uninstall the old version of the software. Third install the new version. Four remember to check your serial numbers in the settings application.

Please remember to install new versions firstly on your development machines, then on your staging servers and finally on your live servers. Over 80% of the issues we deal with are the result of setup specific problems (problems with service packs, incorrect security settings etc.) and it is best to catch these on a development rather than a live server.

I'm going to be running on Windows Vista. What do I need to be aware of?

There are two things you should be aware of.

When you install, double click on setup.exe - not on installer.msi.
Launch the Settings application by right clicking it and selecting 'Run as administrator'.

Vista's new security model means that there are subtle but important differences between the way installers are launched.

If you double click on an MSI file (eg installer.msi) the installer will run with standard privileges. This will mean that tasks which require administrative privileges will not run or will run incorrectly. If you click on a setup file (eg setup.exe) the installer will automatically run with administrative privileges. So tasks which require administrative privileges will succeed.

A similar situation can exist when entering license keys. By default Vista will redirect per-machine setting to per-user ones. This means that if you enter a license key it will be maintained on a per-user basis rather than on a machine wide basis.

For situations in which you are running the software as a user this will be fine. For situations in which the license key needs to be accessed by the internet user then it is less than ideal. The solution is to explicitly launch the settings application as Administrator.

I'm going to be running on Windows Vista 64, Windows XP 64, Windows 2003 64 or Windows 2008 64. What do I need to be aware of?

If there is a 64 bit version of the product you should use it. For example if you are using ABCpdf .NET you should be using ABCpdf .NET x64 under 64 bit Operating Systems. If there is no 64 bit version of the product you will need to use the 32 bit version.

At present 64 bit processing and 32 bit processing do not interoperate seamlessly. For example if you are running vbs scripts you should note that there is one 64 bit wscript.exe and one 32 bit wscript.exe. If you run the wrong one then things will not work. Similarly ASP.NET operates in 64 bit mode by default and won't handle 32 bit components in this mode. The ease of integration between the two modes depends largely on whether you are running IIS6 or IIS7.

If you're using a 32 bit version of one of our components in conjunction with IIS7 then all you need to do is set up an application pool to use the 32-bit version of .NET 2.0. Open IIS Manager. Create a new Application Pool and set 'Enable 32-Bit Applications' to true. Create a new Web Application for the section of your site which is going to use the 32-bit component. Assign the new Application Pool to this new Web Application. That's it!

For details of how to switch between modes under IIS6 please see the MSDN Support articles on the subject. Note that you need to set the "Allowed" status of the 32-bit Web service extension every time you switch to 32-bit mode. Also a single ASP Web service extension will show up by default as ASP.NET 2.0.50727 when in 64-bit mode and an additional Web service extension ASP.NET 2.0.50727 (32-bit) will show up when switched to ASP.NET 32-bit mode.

We've found that a good way to make use of 32 bit components on 64 bit machines is to use COM+ / .NET Enterprise Services. That way your main process runs in 64 bit mode and Windows takes care of all the (cross-process) calls to your 32 bit component running in a separate process. There are details of how to create COM+ / .NET Enterprise Services packages in MSDN. For a practical guide you may find our ABCDrawHTML COM+ instructions helpful. For COM components it's the work of five minutes and it works beautifully. For .NET components it's a bit more work.

File Upload - General

Why aren't I getting any files uploaded?

First check the obvious things - your form must be enctype="multipart/form-data". Your file input field must have a name. Check you're submitting to the correct page. Finally try saving a dump of your upload (details in the documentation) and looking at it in Notepad - you will probably be able to see what's going wrong.

Uploading small files works but when I try to upload files larger than about 1 MB I get 'The page cannot be displayed... Cannot find server or DNS Error'.

The most likely cause of this error is that you're accessing your web server via a proxy server that isn't configured or isn't capable of handling large requests.

Your first test should be to try connecting direct rather than via the proxy server. However in some organizations this isn't always practical so here are some other pointers.

To ensure that the problem is due to your configuration rather than ABCUpload you can always intercept the raw data without involving ABCUpload. For details of how to do this under ABCUpload ASP see the topic 'How do I do a raw upload?'.

In addition some tools such as URLScan may block large transfers. The maximum transfer size can be adjusted in the Urlscan.ini file.

Why isn't the ABCUpload progress bar working?

If you have ASP script debugging enabled the progress bar will not work. When debugging is enabled the script engine locks IIS down to one page at a time per session. This means that the progress bar cannot update. So if you want the progress bar to work you need to disable debugging.

A similar situation can apply to ASP.NET as the Trace module can interfere with the progress bar. This should not be a problem for deployment since you should not have tracing enabled on a live server but it sometimes causes confusion when testing on development servers.

Most proxy servers will recognize that the progress bar page is changing frequently and update it as required. However some proxy servers insist on serving the same out-of-date progress bar again and again rather than updating it as requested.

Some proxy servers cache the entire request/upload before releasing it to the server in one fell swoop. So as far as IIS is concerned there's a long wait and then it gets all the data in one go. The progress bar will sit refreshing over and over again but will read zero until the entire upload is transmitted at which point it will disappear.

How do I ensure that my Mac upload is sent in MacBinary format?

You must be running Internet Explorer on the Mac for your upload to be encoded in MacBinary format. You can change the MacBinary settings using the IE Preferences. In the 'File Helpers' section there's a list of file types. If a particular file type is selected as 'Binary Data' and 'Macintosh File' then uploads of that type will always be in MacBinary format.

Why doesn't the ABCUpload progress bar work with Netscape Navigator 6.1?

There is a bug in the latest version of the Netscape JavaScript interpreter that can result in the incorrect submission of multipart forms when using JavaScript methods. This is easily demonstrated using the following self-contained page.

If you select a file and then click on 'Upload' the file input box will be cleared. In a real world environment this results in the form being submitted without any file.

In practice it is possible to make this kind of form work. For example if you delete the font tags in the above example it will work correctly.

How can I allow users to select multiple files into one file upload box?

Unfortunately it is impossible to do this kind of multi-selection upload unless you use client side code (eg Java, ActiveX or .NET).

Many people on the internet cannot use this kind of code. For example people in large companies often cannot use this kind of code because IT departments tend to disable client side code for security reasons.

An additional benefit of using pure server-side code is that you get security for free. You can do your uploads using an HTTPS link and both you and your clients can be certain that their data transfer is secure.

If this feature is essential you can use Flash 8 (which is a client side control) to provide the file upload capabilities on the client side. ABCUpload will process the data identically on the server side because Flash sends the data in exactly the same way that a browser does. For details see the Flash site.

ABCUpload .NET

I'm having problems with ...?

Please, please, please start by installing the example site and trying that out. When we get support requests saying that ... isn't working our first step is to establish if this is due to a configuration issue or a code issue. To establish whether this is a configuration issue we'll be asking you if the example site is exhibiting the same problem. We can skip this step if you can say yes or no.

You can find the example site under the ABCUpload .NET menu item. All you have to do is create a new virtual directory in IIS and then copy the files into the folder.

The one other change you're likely to have to make is to adjust permissions on the 'write' folder within the example web site to allow the ASP.NET user full access. Otherwise it won't be able to save your uploads.

Then open your browser and try it out. That's it!

Really and honestly this is the fastest way of resolving problems.

But could you take a guess at what's going wrong?

Well many of the issues we deal with are a result of incorrect configuration. Make sure that your web.config file contains the same entries as our example site web.config file.

Some issues are a result of misunderstanding. Server side events work by submitting your form to the server. If there is a file upload field in your form and your client has selected a file for upload then this will require that the entire file is sent to the server before the event can occur. This is just the way that ASP.NET works.

What's the maximum size for an upload?

Assuming you're using GigUpload you will be able to cope with multiple ongoing uploads each up to 2GB in size. IIS does not allow uploads larger than 2GB.

Why is my upload being cached in memory? Isn't GigUpload supposed to avoid this?

It is not practical for GigUpload to intercept every request. ABCUpload .NET will only intercept requests if the Progress Bar is enabled, GigUpload is enabled and there is an Upload ID present in the URL. Additionally when you save uploaded files you should use the ABCUpload .NET classes rather than the standard .NET ones.

The progress upload example provided with ABCUpload .NET fulfills these conditions. However the other example pages do not use Upload IDs and so GigUpload will not intercept uploads made from these pages.

If you've been following the Codebehind Progress Bar Example please ensure that you've set the enableLargeUploads attribute in the web.config.

Why is my Event Logged filled with messages System.Web threw exception 'Thread was being aborted.'?

This is the error you will get if your upload times out (see the documentation for how to alter the timeout). The reason this error is in your Application Event Log is because you have Error Logging turned on in your web.config file.

Is there a file download method I can use which won't max out ASP.NET'?

The key here is to stream your file in chunks and to wait until each chunk has been picked up by the client before streaming the next one. Something along the following lines:

Dim rdStream As New FileStream(oFile.FullName, FileMode.Open, FileAccess.Read)
Dim lReadLength As Long = 1024000
Dim iPos As Long = 0
Dim iPosEnd As Long = rdStream.Length
Dim bReader As New BinaryReader(rdStream)

Response.ClearContent()
Response.Buffer = True
Response.ContentType = "application/octet-stream"
Response.AddHeader("Content-Length", oFile.Length)
If mbView Then
  Response.AddHeader("Content-Disposition", "inline; filename=""" & oDAFile.FileName & """")
Else
  Response.AddHeader("Content-Disposition", "attachment; filename=""" & oDAFile.FileName & """")
End If

iPos = 0
Do While iPos < iPosEnd
  If iPosEnd < iPos + lReadLength Then
    lReadLength = iPosEnd - iPos
  End If
  Response.BinaryWrite(bReader.ReadBytes(lReadLength))
  Response.Flush()
  iPos += lReadLength
Loop

rdStream.Close()
rdStream = Nothing

ABCUpload ASP

Why am I getting "Upload size is greater than the maximum allowed" or "The page cannot be displayed... Cannot find server or DNS Error"?

If your uploads are greater than 8 MB you need to adjust the XForm.MaxUploadSize property. See the ABCUpload documentation for more detail.

Why does ABCUpload tell me 'Access is denied. - (0x00000005)' when I try and save a file?

For safety ABCUpload will not, by default, overwrite existing files. If you try to do so you will receive an access denied error. If you wish to overwrite files you should set the XForm.Overwrite property to true in your script.

Less commonly it may be that the IUSR internet user does not have sufficient permissions to write to the location you have specified. In this case you should alter the permissions on the write destination to allow access.

Why does ABCUpload keep telling me that the 'Upload stopped unexpectedly.'?

The most likely cause of this is that there's a reference to the built in form object somewhere in your code. If the form object is referenced it grabs all the uploaded data which means that ABCUpload can't get it. This sometimes occurs when an include file is changed so that it references the form object.

If the error occurs during a large upload the cause is typically that the client has clicked on stop on their browser window or that they've dismissed the browser window or that there's been a network problem which has broken the connection. If the error occurs immediately on form submission you should look at the following explanation.

ABCUpload reports an error at the point it needs to obtain data. Typically the point at which it first has to gather the data is when you reference a field or you ask for information (such as the field count) that can only be derived from the upload data. The essential cause is that IIS isn't providing this data.

You can completely remove ABCUpload from this equation and you should find that the problem remains. Comment out your form upload code and try inserting something like:

theLen = Request.TotalBytes
theData = Request.BinaryRead(theLen)
Response.Write "Upload Size = " & theLen
Response.Write "Data Size = " & UBound(theData)

You should find that either the upload size or the data size is substantially wrong (they should both be appx equal to the size of your uploaded files) or that an error gets thrown.

When uploading very large files I keep getting 'Upload stopped unexpectedly.'.

Our standard tests run up to about 100 MB. We've had clients uploading files of over 500 MB. However with files this big you can run into unusual problems so here are some points to check.

Make sure you're using ABCUpload Version 4.6 or later. Make sure you've got enough disk space to save your files. Check the SessionTimeout and ScriptTimeout properties as mentioned in the 'Script Settings' portion of the documentation.

Ongoing uploads are stored on disk in the temp folder (you can change this location if you want). This means that your system drive has to have enough space. If you have five simultaneous 100 MB uploads you'll need minimum 500 MB free disk space.

If you have file system quotas then check the file system quota for the IIS user.

Your client browser has to have disk space free. Some browsers save the encoded form on the system disk before sending it through to the web server. This cache can be as much as twice as large as the uploaded files. If your client does not have sufficient space free the results can be unpredictable.

There is a theoretical limit to the number of large uploads available. Because you're working in a 2 GB address space the total upload size (for all ongoing uploads) must not exceed this value. Of course some other things will take up space as well so the total will probably be somewhat less than 2 GB. This is a limit that is only likely to come into play for large numbers of very large uploads.

If this doesn't help please mail us with an exact description of the problem and the system on which you're running. Also check the Application Event Log.

How do I do a raw upload?

You can perform a raw upload using the following code rather than by creating an XForm object. This code does reads the raw data without processing it and without using any external components.

theLen = Request.TotalBytes
theData = Request.BinaryRead(theLen)
Response.Write "Upload Size = " & theLen

If the problem appears to be due to ABCUpload try saving a dump of the upload using the XForm Save method and then send it through to us for analysis. You can save a dump using the following code.

Set xf = Server.CreateObject("ABCUpload4.XForm")
xf.Save "mydump.txt" ' saves in your virtual directory

My upload just hangs. What gives?

Under some circumstances if you have an ASP with more than 32k of static HTML and you locate your XForm script after the 32nd kilobyte of data then the upload can hang. We suspect that this is because IIS starts the HTTP response before the client has finished the request and this can confuse browsers.

My NT 4 web server occasionally hangs during large uploads.

See the Microsoft article "ASP Request.BinaryRead Method Causes 100 Percent CPU Utilization". Please note that all upload components use this method to obtain files (even if it is hidden from you) so all are vulnerable to this NT 4 bug.

ABCpdf

Can ABCpdf convert my Microsoft Office documents to PDF?

Yes. ABCpdf .NET Version 6 can do this. See the Doc.Read method in the documentation for details.

Can I use ABCpdf from Visual Basic?

Yes. All the example code will be exactly the same but you'll need to replace ASP specific code like Server.CreateObject with generic code like CreateObject.

Why does the trial version ask me for a license key?

ABCpdf is a component not an application. The PDFSettings application is simply for entering a license key when you purchase one.

You don't need a license straight away - the component will work for 30 days without any intervention. All you need to do is look at the documentation and start writing code.

If your web application is reporting that your license has expired...

Log on as Administrator and open up the PDFSettings application and see what it says. It will probably tell you that you have a trial license - this is OK.

If it says your license is OK then restart and try your code again. If you still get the same message then you have a permissions issue.

ABCpdf needs to be able to read the license key from the registry. If your web application is unable to read this key it will not work. You need to enable read access to the ABCpdf registry settings at:

HKEY_LOCAL_MACHINE\SOFTWARE\WebSupergoo\ABCpdfNET

HKEY_LOCAL_MACHINE\SOFTWARE\WebSupergoo\ABCpdfCOM

You can adjust registry permissions using regedt32. Please take the relevant precautions when modifying the registry - if you're not sure what these are then please don't perform this type of operation.

Can I import an existing PDF document and then manipulate it on the fly?

ABCpdf will allow you to open, modify and save PDF documents.

ABCpdf will allow you to draw on top of PDF documents or add or delete pages or modify document data. However because of the way that PDF documents are structured it's unlikely that you'll be able to reflow existing content.

So if there are empty spaces which you can draw your entries into that will work great. Indeed you might want to draw a white box over existing content and then draw on that.

However you shouldn't expect to be able to edit and re-flow text that is already in your PDF.

Why doesn't AddImage work with my GIF images?

You can pass AddImage a path to a JPEG, TIFF or EMF file. This inserts the raw compressed data into your PDF document. Other image types (e.g. GIF) cannot be inserted direct - they must be drawn into an Image object and then the Image object can be added to the PDF document. There are examples in the ABCpdf documentation.

My code works on my dev server but fails when I deploy. What gives?

The vast majority of configuration issues are related to security and permissions.

Dev servers often operate a more relaxed security policy than staging or live servers and so this kind of problem often becomes apparent only during deployment.

Your first step should be to run ABCpdf interactively while logged on as Administrator. This will allow you to eliminate permissions as a source of the problems.

If you are running under .NET you will need to create and compile a simple VB or C# application to perform the some basic operations. You can download some sample projects here.

If you are running under ASP you can create a VBS script to mimic the effect of your ASP. Copy the following text into a text file and then rename it mytest.vbs. Double click to run.

Set d = CreateObject("ABCpdf6.Doc")
d.FontSize = 96
d.AddText "Hello World!"
d.Save "mydoc.pdf"
MsgBox "Finished"

If the issue appears to be permission related you can track down the cause of the problem using the security troubleshooting procedure detailed on this page.

Alternatively it can be possible to completely bypass security problems by running your code in a Component Package under .NET Enterprise Services or COM+. This has a number of advantages.

It allows your code to run in a completely different process from your web code thus isolating the two.
It allows you complete control over the user which your code runs as without impacting the security of your web code. All you do is change the user using the Component Services Administration Tool.
Because your web code and your ABCpdf code are operating in different address spaces there is more free space in each process.
You can tell your component package to shut itself down when it's not being used.

If you are using ABCpdf .NET you'll need to write a .NET Managed Component wrapper round your code. The PDFEnterpriseServices example project under the ABCpdf .NET menu item shows you how to do this for converting HTML into PDF. If you're using ABCpdf ASP the process is even easier - you may find our ABCDrawHTML COM+ instructions helpful.

My code is throwing errors like 'Unable to render HTML' or 'Unable to read file' or is producing a blank output.

First read the previous section carefully.

Firewalls and proxy servers can present particular problems for URL rendering because they may require some kind of logon. Your IIS user will most likely not have a logon. You need to allow ABCpdf to logon automatically.

Windows Authentication can produce a challenge that your IIS user may not be responding correctly to. You will find details of how to manage Authentication in the documentation.

Windows 2003 Server defaults to an Internet Explorer Security policy which may interfere with HTML rendering. You may have to modify or disable the policy to allow access to the pages you want to render. If you wish to remove the security component you can do so using the "Windows Components" section of the "Add & Remove Programs" control panel.

If this doesn't help...

First work through some simple sanity checks and common causes of this type of issue.

Do ensure you can browse to the appropriate location using IE while logged on as Administrator. Make sure IE doesn't throw up any blocks or warnings. ABCpdf requires a clear route through to the page.
If you're rendering a URL on your localhost then ensure that the IIS user has read access to this location.
Lack of access of access to the windows/temp directory is a frequent cause of this type of issue. Ensure that your IIS user (eg ASP.NET or IUSR_MACHINENAME) has read/write permission and also (if it fits with your security policy) the ability to create folders. You may also find that ASP.NET is being denied access to its own preferences (eg C:\Documents and Settings\MACHINENAME\ASPNET or C:\Documents and Settings\NetworkService).
It is our experience that it is surprisingly common for web servers not to be able to resolve URLs to themselves. This means that when testing using an IP address your code will work but when you try and render a web page on the local server using a URL you may get issues.
If all this looks fine and you're still seeing problems we strongly suggest putting in ten minutes with ProcessMonitor. In most cases it will flag an obvious cause. We suggest you do not apply ProcessMonitor filters as problems may not always be quite where you expect them to be. You will find details on how to use ProcessMonitor on the "I'm having problems with Security." section of this page.

If this doesn't help...

You need to establish a test case to establish the boundaries of the problem.

If you are using ABCpdf .NET please download the .NET base test case. If you are using ABCpdf ASP please download the COM base test case.
Log on as Administrator. Because the code in the test cases is run as the logged on user you are bypassing any permissions issues which may exist in ASP.

You first need to run the base test cases as-is. These will default to a simple render of a file based URL. If you see an issue then restart and try again. If you still see an error then ensure your machine is fully patched and up to date and then try again. If the error still occurs then please contact us describing exactly what you've done and what error you're seeing.
Assuming the base test case works then you need to try a different URL. Try one one the web (eg http://www.google.com) but do first ensure that you can browse to that URL using IE on the same machine as you're running ABCpdf on. You should see the same content in IE as you do in your output PDF. If not please let us please contact us describing exactly what you've done and sending us the URL and the incorrect output PDF.
Assuming your web based URL works then try the URL you're using in your code - the actual page you want to render. See what kind of output you get.

If the output is correct then copy the exact same code to an ASP page and try it. Any difference between the operation of the code under ASP and in the test cases will be due to permissions issues. See this support page for how to resolve permissions issues.

If the output is incorrect then browse to that URL and save the web page using IE. Then copy the saved web page to the same web site and try rendering that (static) page using ABCpdf. If the output from the static page is correct then your dynamic page is not returning the output you think it is. If the output from the static page is incorrect then send us through the saved page and the output PDF and tell us what's wrong with the PDF.

Please work through these instructions before contacting us. If you can tell us what you've done and what results you've had this will save us all a lot of time.

My code is throwing errors like "'Unable to render HTML. Page load timed out. Unable to load page".

The typical reason a timeout is reported is because a page takes too long to load. ABCpdf expects a page to finish loading within 15 seconds. If it doesn't it will time out. So... hardwire your URL into the code so you're sure it's correct. Open the same URL in IE. How long is your page taking to load when you view it in IE?

Assuming it takes longer than 15 seconds you need to adjust the timeout - see the HtmlOptions.Timeout property.

Don't make the timeout too high - keep it realistic. If you set a HtmlOptions.Timeout value of ten minutes and a HtmlOptions.RetryCount of 5 then you are telling ABCpdf that it can spend up to an hour trying to get the page. If this happens, to all intents and purposes it will look as if the process has hung.

Occasionally timeouts may be reported because ABCpdf does not have permission to access a page. However this is unusual. If you suspect that this may be the case you will need to run through the blank page troubleshooting instructions in the previous section.

I'm getting a "Windows error 193. %1 is not a valid Win32 application" error.

This is a permissions issue related to the creation of MSHTML objects. Please start by checking the "I'm having problems with Security." section of this page.

Sometimes this permissions issue can have a base cause in an old or inconsistent installation of MSHTML. If this is the case then updating IE and restarting may well resolve the issue.

As with all permissions changes take great care. It is vital you ensure you know how to back these changes out should you decide that they are inappropriate. If you are unsure then please do not make any changes.

Open the Component Services control panel. Right click on the 'My Computer' icon to get up generic COM/DCOM permissions. Check the Launch and Access permissions. If your IIS user (typically ASPNET or IUSR_MACHINENAME) does not have Launch and Access permissions it may be denied permissions to create or use MSHTML objects.

The issue may be related to the fact that your IIS user doesn't have "log in a batch job" permissions. Try assigning the permission to that user.

Restart your computer and try your code again.

Why doesn't my font work?

ABCpdf will allow you to use any valid TrueType or Type 1 (PostScript) font.

ABCpdf will not embed fonts which do not explicitly permit it. This is a measure in place to protect you from inadvertently infringing copyright. However you can override this behavior and force ABCpdf to embed these fonts. See the Doc.EmbedFont method and the Doc.HtmlOptions.FontProtection property in the documentation for details.

ABCpdf maintains a font cache. This means that for ABCpdf to pick up on a newly installed font you will need to restart any processes that are using ABCpdf.

Alternatively you can pass the path to your font file to the AddFont or EmbedFont method. This will automatically load the font file. Do not move the font file after doing this - ABCpdf relies on fonts staying in place.

ABCpdf will automatically find any font in the Fonts folder. However applications such as Adobe Type Manager may place the font files in a different folder. If this is the case then ABCpdf won't be able to see them.

Sometimes permissions are placed on individual font files which may restrict access from restricted permission accounts such as ASP or ASP.NET.

Occasionally TrueType fonts are corrupt or non- standard. This can cause problems for ABCpdf (which will refuse to recognize them) or Acrobat (which will refuse to use a font embedded in the PDF). However this type of problem is relatively infrequent and tends to be restricted to unusual fonts such as bar-codes.

If you hit a problem you think is related to a corrupt or nonstandard font please mail us the font and we'll see what we can suggest.

How do I automatically print my PDF documents?

You can use ABCpdf to render your PDF. However while ABCpdf will manage the rasterization you will need to manage the printer control. The ABCpdfView example project (included with the ABCpdf .NET install) shows how to do this and provides starter code for printing using ABCpdf.

Alternatively to script Acrobat to print a document you can run it silently from the command line. Use the following syntax.

AcroRd32.exe /t path printer driver port

The path is the path to the PDF document. The printer is the name of the printer. The driver is the name of the driver - as it appears in the 'Driver Used' box. The port is the printer port (this cannot contain backslash characters).

You can also script Acrobat to print via Automation. You need to use the AVDoc.PrintPages... suite. More information can be found on the Adobe web site.

How do insert bar codes into my PDF documents?

There are lots of free and commercial bar code fonts which will allow you to do this quickly and easily.

ABCpdf will allow you to insert barcode TrueType fonts into your PDF and thus produce barcode output. Please ensure you embed the barcode font and please ensure you have permission to do this.

We suggest starting at the barcode-1 site for your code 39, code 93, code 128, DataMatrix, Interleaved 2 of 5 & 2 of 5, OCR-A & OCR-B, PDF417, Postnet and UPC/EAN barcode fonts.

Look for TrueType fonts which are embeddable into PDFs for use by ABCpdf.

How do I change the PDF Page Size?

See the Doc.MediaBox property as detailed in the documentation.

Why don't my documents measure up when I print them?

In every case we've come across the issue has not been with the PDF output produced by ABCpdf - it's with the way that Acrobat and Windows does the printing.

Normally you can solve these issues by by adjusting the PDF and by ensuring that the print options on Acrobat are correct.

Check the print dialog. Make sure you don't have the 'Shrink oversized pages to paper size' or 'Expand small pages to paper size' or 'Auto-rotate and center pages' options checked.

Are you printing on a postscript printer - if not you may not get exact output.

Try using AddGrid to add a background grid to your image. This may show how the PDF is being changed as it moves from screen to paper.

You may find it helpful to specify a CropBox for your document (this may affect the way that Acrobat prints pages). To insert a CropBox for the current page use:

theDoc.SetInfo(theDoc.Page, "/CropBox:Rect", "30 30 562 732")
' standard mediabox inset by 30 points
' you will probably need to adjust this rect

If you can't see a solution send us a simple example of a PDF which prints correctly and one which doesn't. If you can do this we should be able to find what it is that is different between the two. We can then tell you how to achieve this using ABCpdf.

Can I convert my PDFs to JPEG/TIFF/GIF using ABCpdf?

Yes - see the Doc.Rendering property as detailed in the documentation. You need an ABCpdf Professional License for PDF rendering.

Simply write code like this (this is ASP but the .NET code would be very similar):

Set doc = CreateObject("ABCpdf6.Doc")
doc.Read "c:\myfile.pdf"
doc.PageNumber = 2
doc.Rendering.Save "c:\mypreview.jpg", ""

Why am I having problems displaying my PDFs in Explorer?

The combination of IE and Acrobat is not always trouble free - particularly under https or using older versions of IE. It's difficult to know exactly where the problem lies because there is an interaction of the Operating System, IE and Acrobat. Any of these can be the cause. IE 5.5 can be particularly troublesome.

Sometimes IE gets 'stuck' on a particular content type and insists on displaying your PDF as HTML. In this case you will see random text starting with %PDF. Often this can happen if you send PDF data to a window which was previously displaying HTML.

Sometimes server-side debugging results in extra data being inserted into the content stream. While this may not matter for HTML it will corrupt binary documents like PDF.

Sometimes your code may inadvertently insert extra data into the content stream. Again this will corrupt the PDF.

HTML Compression is currently something which Acrobat is rather unhappy about. If your server provides compressed output to IE then Acrobat may simply display a blank page. You should know whether you've enabled IIS compression or not - it's off by default. If you're really not sure then use a utility like IEWatch to check the content headers being returned. If you see gzip then you're seeing compressed content. If you're using IIS6 compression, this can be switched off on a file by file basis by altering settings in the Metabase.

For testing purposes you may wish to change the content-disposition from 'inline' to 'attachment'. This will allow you to download the data rather view it in your browser. You can then check the downloaded document using Acrobat or a hex editor. All PDFs should start with "%PDF" and should end with "%%EOF".

Alternatively if the problem is that PDFs seem to be cached you may wish to check the 'Enable Content Expiration' checkbox you will find under the Web Site Properties.

We would suggest two steps:

Your first step should be to eliminate ABCpdf as the cause. Why not save the PDF to disk at the same time as sending it to the client? That way you can establish that the PDF is fine. If you want to take it further you could then read the PDF data from disk and stream it direct to the client.
The example site streams PDF data direct to the client. So install the example site into a new virtual directory and establish if the same issue exists for the example site. If it works then it's simply a matter of moving your current code base and the example site code base towards each other until you find the cause of the issue.

If this doesn't help please do get in contact with us describing exactly what you've done.

Why are my styles and images not working when I use AddImageHtml?

HTML does not exist within a file and so it does not have a location.

External stylesheets and images are often referenced via relative URLs. Because the HTML has no location it is impossible to resolve these relative reference.

So you need to provide your stylesheet and image links as absolute references. Or you may be able to use the HTML BASE element to specify an appropriate base location. Or you can save your HTML to file in an appropriate location and then use AddImageUrl.

Can I put ABCpdf .NET on my shared / hosted site?

In general - yes.

You need to follow the instructions in the Manual Installation section of the documentation. Any DLLs will need to be placed in your bin directory and the license (or trial license) key will need to be incorporated into your code.

Most installations of ABCpdf .NET on shared servers are seamless. However you need to be aware that you are a guest on the server and your host may have locked down permissions in ways which will make your life difficult. If this occurs there is typically little that either you or us can do about it.

This is why we recommend deployment on a dedicated server. Dedicated servers are cheaper than you might think and the level of control they allow is only one of the very significant advantages they afford. It is one of the best decisions you can make.

The essential thing to do is to deploy early and discover any issues before they become major problems.

Why are my pages being cached when I use AddImageUrl and stream the resultant PDF to the browser?

There are two possible causes.

The first is that the URL is in some way being cached. Just because you've disabled the ABCpdf cache doesn't mean that it's not being cached somewhere else. See the HTML / CSS Rendering section of the documentation for details.

The second is that the PDF itself is being cached. Most commonly this kind of thing can happen if you're streaming the PDF direct to the browser and you have certain IIS settings (eg Expire Content) disabled.

Your first step should be to narrow down the problem. Why not save the PDF to disk at the same time as sending it to the client? That way you can establish whether the PDF itself is being cached or whether the URL is being cached (resulting in the same PDF being created again and again).

If the URL is being cached you will need to look at the dummy URL method. ABCpdf is not doing the caching but something else is and you need to work around it.

If the PDF is being cached you will need to look at your IIS settings. Again ABCpdf is not doing the caching (and indeed it cannot cache the PDF in this way) it is something which is happening either in IIS/ASP or on an intervening proxy server or on the client.

In most cases the dummy URL method (detailed in the documentation) is the fastest route to a solution.

How do I use ABCpdf from WebMatrix?

The way in which Assemblies are linked from a particular development environment is very much dependent on the development environment. So for best practice you should consult the documentation for your chosen environment.

However you might like to use a register tag to link to the assembly and import the namespace:

<%@ Register TagPrefix="pdf" Namespace="WebSupergoo.ABCpdf6"
Assembly="ABCpdf, Version=6.0.0.5, Culture=neutral, PublicKeyToken=a7a0b3f5184f2169" %>

Obviously you'll need to modify the version and possibly the key to match the assembly you're using.

Why is Acrobat telling me my fields are malformed after I use ABCpdf to delete them?

If you create forms using Adobe Designer (which comes with Acrobat 7) it will make use of some new field references which were not used in version 6.

If fields are deleted and the field references are left intact then Acrobat may complain that the fields are invalid.

To solve the problem delete the AcroForm after deleting the fields. For example:

theDoc.SetInfo theDoc.Root, "/AcroForm:Del", ""

Alternatively if you are simply setting field values then you may be able to achieve what you want by deleting the XFA elements:

theDoc.SetInfo theDoc.Root, "/AcroForm/XFA:Del", ""

Why am I finding it so difficult to work with Adobe Designer forms? Why are my Acrobat 7 forms working so differently from my Acrobat 6 ones?

Adobe Designer may be installed as part of Adobe Acrobat but it is a fundamentally different application which simply happens to use PDF as an output format.

In more detail...

You can use Acrobat to edit forms using the Advanced Editing tools available under the Tools menu. However if you choose items from the Forms menu or toolbar then you will probably find that you end up editing your form in Adobe Designer rather than Acrobat.

Adobe Designer uses PDF as an output medium. However the way that Designer operates means that forms created by Designer are fundamentally different from forms created by Acrobat.

For example an Acrobat created form typically contains a background and then a set of fields. The fields operate separately from the background.

Adobe Designer created forms do not make this distinction. They use a separate data store to specify the fields. The PDF content is merely the visible rendition of this field specification. The underlying field specification is made up of chunks of XML embedded in the PDF.

Because Designer documents are PDF documents you can add content to them using standard ABCpdf methods of adding PDF content. However if you then open them then in Designer the content will most likely be deleted as Designer recreates the PDF appearance using the separate field specification.

Equally because the PDF output is merely the visible rendition of a separate field specification the fields and background may be tied to each other. So you might use ABCpdf to delete a field and find that the border has been left behind.

If you want to modify forms you will generally find it easiest to work with Acrobat created forms than Designer created ones.

I'm having problems with OpenOffice.org when I try to convert documents to PDF. How do I troubleshoot these?

It is not uncommon to come across permissions issues when running OpenOffice.org from ASP.NET.

We offer a three step guide to help you through these. Follow the instructions precisely and you should have no problem.

Step 1. Validate your existing solution

Download and install OpenOffice.org. Open some documents to check that it's working correctly. You may wish to disable the OpenOffice.org Quickstarter to ensure that the OpenOffice.org processes are shut down after use rather than kept alive in the background. This may be important if you're going to be changing the user at a later stage.

Build and run the doc2pdf project. If it fails then this would imply a generic configuration issue. Check your OpenOffice installation. Let us know if you can't see a cause.

If it succeeds (as invariably it will) it will imply a permissions issue under ASP.NET.

So assuming it succeeds we need to eliminate the permissions issue. The easiest way to do this is to bypass it using .NET Enterprise Services.

There is an example project ready for you included with your ABCpdf installation. You can find the PDFEnterpriseServices project in the Projects folder under the ABCpdf menu item. All you need to do is build and run it.

Let's assume you're logged on locally as Administrator. All you should need to do is open the solution. Then set right click on the web site project and select it as the startup project. Then run - just press F5.

Note for Vista users with User Access Protection (UAP) enabled. UAP means that Vista switches between user and administrator mode as requested. Visual Studio may not always make appropriate requests during the build process and this can lead to access failures or generic errors. We have found it significantly easier to disable UAP temporarily rather than work through these errors.

Step 2. Deploy to your destination server

Once you've got the .NET Enterprise Services project working on your local machine you may want to deploy it to a remote server.

You should note the following points before deployment:

For server versions of Windows (eg Windows 2003 Server) it is important that the latest updates from Microsoft are installed.
It is important that ABCpdf .NET has been installed by running the installer on your server. This is necessary to ensure that ABCpdf is available via the GAC.
It is important that the version of ABCpdf .NET you compile your project against is exactly the same as the version you install on your remote server. If the versions are different, .NET will refuse to load the assembly.

Once you've built the .NET Enterprise Services DLL on your local machine you can copy the project to the remote machine and register the DLL from the command line while logged on locally as Administrator. Simply make the following call from the command line,

regsvcs.exe PdfEnterpriseServices.dll

After you've registered the DLL you should set up a new virtual directory and point it to the web site. Then browse to the site.

If you later need to update the package you must first delete it using the Component Services Administration tool and then re-register the DLL otherwise your changes will not be picked up.

The Identity of the Component Package should be left as the Interactive User. Indeed you should be logged on locally as the Interactive User. By all means change the user later - but please do not be tempted to do this before you have got the package working.

So why do we say this? The problem with changing the user is that each user has to - on first run - accept the OpenOffice.org EULA. Running as the interactive user you can see this and accept it yourself. If you change to someone who may not have access to the desktop you may never see it and you'll end up with a timeout. It's important to start with a simple scenario so that you know what you're dealing with.

Step 3. Change the Component Package user

The Interactive User is fine as long as you are logged on locally. However this may not always be the case for web servers. So you will most likely want to assign a user to the package.

When you change the user it should be to one that has accepted the OpenOffice.org EULA. If the user has not accepted the EULA then ABCpdf will time out. In this case you will probably want to copy over the EULA acceptance documentation to the relevant user.

Let's assume you entered appropriate values for the EULA and auto-update settings when you originally ran OpenOffice.org. These are stored in your personal settings on the computer. For example:

C:\Documents and Settings\Michael\Application Data\OpenOffice.org2

So all you need to do is copy these settings to the Component Package user. On Windows 2003 you will probably need to copy the OpenOffice.org2 folder to:

C:\Documents and Settings\Default User\Application Data\

Note that there is a lock file at the top level of the OpenOffice.org2 folder which you may need to delete after copying the settings.

On Windows XP the setup may be slightly different and depending on whether you've set up the Component Package as a Service or not you may find the user settings at one of the following locations:

C:\WINDOWS\system32\config\systemprofile\Application Data\
C:\Documents and Settings\NetworkService\Application Data\

If you're unsure about the location assigned to a particular user you can use Process Monitor from Microsoft/SysInternals (mentioned on this support page) to find out.

ImageGlue

I've installed ImageGlue 6.1 ASP. When I open IGSettings I get an error "Run-time error 339 Component 'mscomctl.ocx' or one of its dependencies not correctly registered: a file is missing or invalid.". What do I do?

IGSettings in ImageGlue 6.1 ASP (but not in ImageGlue .NET) uses various Visual Basic controls which normally come pre-installed.

We don't install them ourselves because they're extremely common and overwriting existing versions can be just the kind of thing you don't want. We try and keep web server installations as light as possible so you know what's going on.

Microsoft discontinued free support for Visual Basic 6 in April 2005. They are slowly withdrawing VB6 support from all aspects of Windows. You may well find that on fresh installs of recent releases of Windows that the standard VB6 controls are not available.

If these controls have not come installed on your system then you can get them at the Microsoft web site. However in the long run you should be transitioning to ImageGlue .NET.

The example web works but my code doesn't.

If you are having problems getting ImageGlue to do what you want we may be able to help. Write to us but please provide a concrete example as well as the code you are using. Most problems are a result of misunderstanding and without both of these elements we can find it difficult to work out what's going wrong.

How do I return an image to the browser without writing a file out?

You can use Response.BinaryWrite to send raw data to the client. For example the following ASP code will display a red image. You can use similar code under .NET.

<% @Language="VBScript" EnableSessionState=FALSE %>

<%
Set ca = Server.CreateObject("ImageGlue6.Canvas")
ca.Color = "red"
ca.Width = 100
ca.Height = 200
Response.ContentType = "image/png"
Response.BinaryWrite ca.GetAs("dummy.png","")
%>

I installed ImageGlue ASP on my Windows 2000 Server and I keep getting error 'Server object, ASP 0177 (0x8007045A)'.

This error is often caused by overly strict security. Please see Microsoft Q175804. IIS needs (at minimum read) access to the registry and to the Windows directory where the QuickTime DLLs are installed - and of course read access to the directory where ImageGlue is installed.

This kind of error is not unusual on Windows 2000 Advanced Server. You need to change the IIS Anonymous user to some user that has appropriate access privileges. In fact you may well find that if you simply switch your IIS virtual directory properties to 'High - Isolated Process' the problem will just solve itself.

This error may also occur if another process has the ImageGlue DLL locked. For example if you try to access ImageGlue from a VBS script while IIS has the DLL locked you will get the same error from your VBS script.

Less frequently this error may occur if the DLL is not registered at all or if it has been unregistered. If you believe this is the case you should register the DLL manually - see below. After registering you should restart IIS to ensure the new settings are picked up.

How do I manually register ImageGlue ASP?

To register the ActiveX DLL for ImageGlue ASP (COM) you should first locate the DLL. Perform a search and look for ImageGlue6.DLL. To register you should:

Open up a DOS window
Navigate to the location where the DLL is located
Type regsvr32 ImageGlue5.DLL
Press return
You should receive a confirmation message that ImageGlue has been registered
The Application Event Log should contain an information message to confirm that ImageGlue has been registered

Note that this procedure is only applicable to ImageGlue ASP - .NET components do not need to be registered and indeed cannot be registered in this way.

When I try to use ImageGlue ASP from VB I get an error 'Variable uses an Automation type not supported in Visual Basic'.

The current version of VB doesn't understand the new neutral threading model. It tries to pre-validate the threading model and gets all confused. A simple work around is to use Late Binding and CreateObject rather than New. For example,

Dim MyCanvas As Object
Set MyCanvas = CreateObject("ImageGlue6.Canvas")

... rather than,

Dim MyCanvas As ImageGlue.Canvas
Set MyCanvas = New ImageGlue.Canvas

Why do I get these error messages? "A QuickTime error occurred indicating that the file was not found (-43)." "JPEG Error Message - Invalid file name."

Both error messages indicate that a file could not be written. This can be caused by an invalid file path. It may be that your IUSR user does not have permission to read/write to this directory or that he doesn't have permissions to overwrite or delete a file of the same name that already exists in this directory.

My text isn't being anti-aliased when I draw text at very small font sizes. Why?

ImageGlue uses standard NT anti-aliasing requests (the ANTIALIASED_QUALITY flag). Normally this works fine but (and MSDN is a little vague on this) MS state that it may not work at either 'very small' (about 10 point) or 'very large' font sizes. It also only seems to work correctly in 16 or 32 bit color depth.

The height limitation is hardwired into the True Type Font itself in what is known as the GASP table (Grid-fitting and Scan-conversion Procedure). If you remove the GASP table using a utility such as TTFGASP the font will anti-alias at all sizes. However the size limitation is there for a reason and some characters may disappear completely if they are anti-aliased at very small sizes.