I'm new to VBA and have been throwing together a small macro application for the Office. We've got about 80 users on essentially identical PC setups, and it will be accessed by all but a few users.
I've been playing around with some automation of accessing web pages using the Web Services references, and I've also loaded the Microsoft Scripting Runtime references into the project. I attempted to run it on a test PC and it complained that there were missing references.
I don't particularly want to go around 80 PCs and manually load the references.
My question, basically, is how should I manage the distribution of this macro-app to 80 odd users so as to ensure that the references will load every time for every user.
Thanks!
For the most part, late binding will solve problems with references in VBA, unless you have some unusual references. Most problems are caused by differences in library versions that can be overcome with late binding. With VBA, it is often recommended that you develop with early binding but release with late binding. The main disadvantage of late binding is changing built-in constants to values (speed is no longer the issue it used to be.)
So:
Dim fs As Object 'Instead of FileSystemObject '
Dim xl As Object 'Instead of Excel.Application '
Set fs=CreateObject("Scripting.FileSystemObject")
Set xl=CreateObject("Excel.Application")
'Value instead of built-in constant '
ForReading=2
Set f = fs.OpenTextFile("c:\testfile.txt", ForReading)
If you have references that your application depends on, that you know are not going to be on the target PCs, then I would strongly recommend you investigate some installer technology.
Using the installer you should be able to install your macro, and install and register all appropriate references / libraries.
There are generally two flavours on windows, Windows Installer based technology and Script based technology.
We use InstallShield for all of our deployment, although there are several options for you to use (there are several discussion on Stack Overflow).
Using windows installer technology, you can build MSI install files, which you are then able to deploy automatically using Group Policy.
Instead of having the documents expose the functionality, make it an add-in for Office (the suite, or the individual apps, your choice). This way, you don't have to deal with references.
Then, just distribute an install package with the add-in which registers the components and registers the add-ins with the appropriate Office apps.
VB6 might be a good idea here, given it's similarity to VBA.
In addition to this answer, which is the bullet-proof solution to solve this kind of issue, but which is quite complex to implement, you can also write some code to be executed when your VBA application starts, checking the 'references' collection of the 'application' object. You can then check (1) if requested files (dll, ocx, tlb) are available on the computer and (2) if reference can be created (application.references.addFromFile ...).
Be careful: object declarations that might be 'reference dependent', such as:
Dim cat as ADOX.catalog
will raise a compilation bug if the reference is not active when the corresponding module is 'compiled'. I then advise you to isolate your 'reference checking procedure' in a startup module (equivalent to an 'autoexec') which deals only with VBA and basic application objects. Check it with your Help Files (Example: in Access, default references that can be used without external references are VBA, Access and DAO).
EDIT:
in case external references depend on other software package and (1) cannot be distributed with a MSI file or (2) can have multiple versions, I think the 'references.addFromFile' is the only solution that can apply. Example:
You have an VBA/Access runtime client
app that needs to refer to Word
(msword.olb file).
For licensing issues, you cannot freely distribute this file with your msi pack
the olb file can be either the 'XP version or a newer one
Our solution is to have 2 tables on the client Access file. One lists all the references that have to be checked or added at startup time (Word will be one of them), and the other one lists all the possible locations of the file (depending if the user has the 'office11' version or a newer one), with a one to many relations between the 2 tables.
So, the best strategy could be a mix between msi packs and management through code:
msi is great for distributing independant dll's or other files that are totally 'embedded' in your app, such as activeX controls (like scanners controls, report or file viewers, etc)
code is the best solution where your app will have to communicate with other applications (word, excel, outlook, etc) that can exist in different versions on your user's machines.
Related
I am developing two addins for Autodesk Inventor (iMateEditor.dll and Addin2.dll) and I want Addin2 to access classes and events from iMateEditor.dll.
I got it to work by following the steps outline in the Microsoft Documents. However, a new problem emerged which is that in order to debug the iMateEditor.dll, I have to remove it from the Global Assembly Cache (GAC). However, once I do that, Addin2 can no longer reference iMateEditor. So my question is, is there another way to have Addin2 reference iMateEditor without having to remove iMateEditor from the GAC to debug it?
Maybe there a way to embed the iMateEditor library into Addin2 whenever I build Addin2?
I have a vb.net application I'm looking to be able to distribute in the near future.
I'm not the original architect and the previous developer referenced a handful of .dll's that are under a GPL license.
All of the software that includes these dll's are freely available online, so my customer can go download and install them if they need that functionality. So I don't have to distribute the DLLs.
Currently they are referenced under the "Reference" part of the project file.
My question is, how do I resolve these dll's in a way similar to how the "references" dons it, but at runtime.
My plan is to search the registry for the location of these dll's and reference that location, but given the file location of the .dll, how do I "pull" that code into my project.
Thanks
You may try this
Search for Dll on specified path for dll
Use reflection to load assembly or dll into you code at runtime
Create runtime object from the loaded dll
Call required functionality from the dll
Reflection is the key solution to your problem that you may use to plugin new functionality into your project without distributing the dlls
This is the only solution that works
http://mylifeandsql.com/2018/03/26/replication-readpast-error/
also you can just start your migration with the following command
Sql(#"SET TRANSACTION ISOLATION LEVEL READ COMMITTED");
This will replicate dll changes like adding new column to a replicated table
You will also find that the column is automatically added to replicated articles > columns
No need to create a new snapshot nor set the sync to re-initialization ☺
Thanks
Is it possible to have modules be external to the actual Excel file and call the functions/subs within them externally?
My thinking is if there are multiple Excel files that use the same module, instead of updating each one of those files separately when I make a change, can I just update the one module stored on a server or something?
I have doing something like you describe for years. You can move your VBA code to a VB6 ActiveX dll, organize it into classes, and load that dll as a reference from Excel VBA.
This is a good way to reuse non-workbook specific code. For instance, I have code that queries a mainframe. I like to call it from Excel, but the details of the connection and how data is passed are contained in a dll that I can load from Excel, VB6, Word, .NET, wherever. I have a similar dll for reading data from AutoCAD drawings, one for interfacing with a product DB on a MySQL server, etc.
The code that remains in Excel tends to be simple formatting stuff. Say I return a variant array of strings (technically a COM SAFEARRAY) from some library that I wrote. I would then output it into Excel, maybe do a text-to-columns, and have a list of results returned to the user.
You can also pass and return more complex data structures. The beauty of VB6/COM Automation (and I didn't appreciate this until I learned to do it the harder way in VB.NET or C#) is that the data will flow in and out of your components seamlessly and all the necessary interfaces will be created for you.
The main change to your code will be replacing things like ThisWorkbook or ActiveSheet with explicit parameters like (Byval sht as Excel.Worksheet). These will be caught at compile time in VB6 (since it doesn't know what ThisWorkbook is), so you cannot overlook them; you are forced to pass an explicit reference.
I also notice that my code inside the dll becomes more paranoid if it receives a Worksheet or other Excel object as a parameter. In VBA you might have had more assurance that you were passing a good object since it was private to a given workbook. The dll does not know who is calling it, so I view the passed-in object with more suspicion (check if Nothing, sheet name, formatting clues to ensure I am using what I think I am using).
The only downside I see is that you will have to get a copy of Visual Basic 6.0. I bought mine in 1998. It is no longer available from Microsoft, but surely there is someone out there who will sell it to you. The latest service pack is SP6.
You will also have to become familiar with "regsvr32" and "regsvr32 /u" to deal with the "ActiveX can't create component" errors as you open your workbooks on various computers. I just publish my dlls to a mapped network drive and re-register them on the user's computers whenever there is a significant change. Obviously this is a intranet/single company solution. Publishing the updated versions is much more of a pain the farther you are distributed.
Not sure if this would satisfy your needs, but you could create your common module as an "add-in" and so install it so that all files that you open in the same instance of excel would have access to the add-in code.
It would not be my recommended way of doing it because I would be worried about suitable testing of all the excel files that use it, when you make a change, plus the added complexity of getting users to install your add-in (this may not be an issue for you). I have a "developersToolkit" module I use across 8 different Workbooks, but I import the module into each workbook so its stand alone and I can also test changes for compatibility with each of the 8 workbooks.
When I install an add-in (via the checkbox) what are the rules / mechanics governing how long that add-in stays installed? It seems that it is installed on the application level, meaning that it applies to any workbook in the excel application, until you uncheck it yourself. If this is correct, this means that as long as a user installs the add-in themselves, they should be fine; but they will have to install it that first time (which could be done programmatically).
Regarding VBA references, I imagine these do not need to be 'checked' every time- meaning it is in the scope of the document. But if this is true, why do people recommend using late binding methods when the software is ready to be distributed? Is late binding really just to make it compatible with different versions, but not necessarily to make it so the DLL is 'checked' as a reference? Under this assumption, as long as everyone uses the same version of Excel as the me, would it be safe to forego late binding all together, and just add the references manually?
Using late binding is better because you can't guarantee that every system will have the same references available. So you should only select references that you can guarantee will be on every machine and distribute the DLL files for the rest.
As far as scope, the public subs/functions/variables in your add-in will be available to anything in the application as long as it is installed and active. The references in the add-in will be available only to the routines in the add-in.
Add-ins are installed to the Application-level. You can fine-tune how an Add-in may be exposed (or not) to various Workbook(s) using Ribbon XML if needed.
as long as a user installs the add in themselves, they should be fine; but they will have to install it that first time
Yes, they'll have to install it.
Regarding VBA references, I imagine these do not need to be 'checked' every time- meaning it is in the scope of the document.
Yes, version control. It also saves you the hassle of having to try and programmatically add references. This can be done by path (which requires knowing the OS, version, etc. ) or the GUID (I've never actually been able to do this successfully). Both of these would then require error-trapping (what if the path doesn't exist or is otherwise inaccessible? etc.). So just use late binding.
While developing using early binding is helpful because of the intellisense, from a user perspective, there is generally no observable difference in how they might perform although EB is arguably faster, the difference is usually negligible for most applications. Related, if while using EB you rely on the New keyword to instantiate objects, I believe that the CreateObject function which you would use with LB is actually faster. This is probably not noticeable, though.
can I forego the late binding all together, and be safe by just adding the references manually?
I would simply do the development with EB, and then modify the code to late-bound objects before compiling the Add-in.
I've had a very difficult time implementing an Add-in in Excel.
I've got buttons on a worksheet which reference code in the Add-in, and I've also got code in the worksheet which references functions in the Add-in. What was really throwing me off, was that I could inactivate the Add-in (options, add-ins, Manage, uncheck) and the code continued to access the add-in. This was particularly frustrating during development and debug of the Add-In, because what I wanted to do, was to inactivate the Add-In (during development) and have my application worksheet reference the open Add-In .xlsm development file. What took me a while to figure out, was that due to the code "References" in the application worksheet, it was still calling to the code in the .xlam file, not the .xlsm.
Once I figured this out, things when smoother, but now, each time I make a change and want to test, I need to close my application, Inactivate the Add-in, Close Excel, Save my .xlsm as .xlam in the default Excel Add-Ins directory (which BTW requires administrator rights), Open excel, Activate the Add-in. [Close Excel], Open my workbook application. A throughly exhausting process. Perhaps I could skip the last [Close Excel] step if the add-in becomes active immediately when checked.
And then dealing with users on Win10 and Win7 really complicates things. Different path to default Add-in folder. Users have to change references path, etc. Very ugly. Very Microsoft.
I have a vb6 program that needs to use MSOLAP80.dll to display its pivot tables properly. But because MSOLAP90.dll has some compatibility issues with this I cannot use MSOLAP90.dll and still have the pivot tables display.
I have registered MSOLAP90.dll and then registered MSOLAP80.dll again and everything seems to be fine. I however don't know if both are actually registered or if MSOLAP80.dll is the only one registered, because I have no reference point as to what is new in MSOLAP90.dll. Is it possible that both are registered and the program is just using MSOLAP80.dll and if there are programs that need MSOLAP90.dll then it will know to use that one?
I guess I am just confused about how registering DLL's work and if it is possible to have both of these registered at the same time. Can somebody help with an explanation?
If you want to know for sure which one is registered, you can:
Look at the References dialogue for a type library which matches your DLLs' paths.
Open up RegEdit, and search for MSOLAP80.DLL or MSOLAP90.DLL (uncheck "Match whole string only").
If you find references for both DLLs, then you are safe because you can bind to a specific version. If you find a reference to the wrong DLL, then unregister the wrong one, and register the right one.
COM originall only allowed one version of a set of CLSIDs (which uniquely identify classes), IIDs (which uniquely identify classes' interfaces) at any one time. It is possible to have more than one reference to a LIBID (which identifies a type library - a resource embedded in the DLLL) but they have to have different versions.
From Windows XP onwards, it became possible to do side by side DLL access in which an executable can access a specific version of a DLL, overriding the value in the registry. You need to embed or have a .manifest file in the same folder as the EXE file.
Unfortunately, the documentation for this seems to have disappeared from MSDN, and is only referred to by a couple of Knowledge Base articles:
http://support.microsoft.com/kb/828629
http://support.microsoft.com/kb/843524