I'm in the process of taking over an Access application that was developed
by someone else. Not my favorite work, but the client is a friend. The front
end is Access and the back end is SQL 2000. When I looked at the existing
application, I discovered that it fails to compile. There appear to be a
number of forms and reports where the code references non-existent controls.
There may be other issues also, but I got tired of commenting out invalid
code and quit trying to get it to compile. I suggested the client contact
the original programmer, who is still working on this project, and ask why
the application doesn't compile. The programmer questioned my knowledge,
that I would suggest such a thing as requiring successful compilation.
Apparently he relies on the incremental compilation and is satisfied if the
parts being used work.

I've been doing software development for many years. I've worked with Access
since version 1. In all that time I've never delivered a non-compiling
application to a client, in Access or any other tool. I sent the client a
few quotes from "Code Complete", by Steve McConnell, indicating that
successful compilation was one of the minimal tests for a successful daily
build.

I'm just wondering if maybe I'm missing something here. Does anyone here
deliver applications that don't compile? On purpose? And defend the
practice?
Paul Shapiro

"Paul Shapiro" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...


>
> I'm just wondering if maybe I'm missing something here. Does anyone here
> deliver applications that don't compile? On purpose? And defend the
> practice?
> Paul Shapiro

As a pracince, I would NEVER deploy an apocation that does not compile. In
fact, I alwasys deployed a mde.

However, we don't live in a perfect world. For example, while I just said I
would "never" deploy a mdb, and *always* use a mde for each workstation, it
not the end of world. We don't have to put someone in jail if they don't
always use a mde.

However, I do think that a "basic" standard for deployment is that the
application should at least compile. So, if you asking is your reaction
"fair", or normal...sure, I agree. A bit of aghast here is needed. I think
deploying an application that don't compile is a sign of laziness, or simply
a lack of professional standards here....

I also think the same about deploying a split application, and again, it
hard to phantom an un-split application..even in single user environments...


--
Albert D. Kallal (Access MVP)
Edmonton, Alberta Canada
(E-Mail Removed)

Per Paul Shapiro:
> I suggested the client contact
>the original programmer, who is still working on this project, and ask why
>the application doesn't compile. The programmer questioned my knowledge,
>that I would suggest such a thing as requiring successful compilation.
>Apparently he relies on the incremental compilation and is satisfied if the
>parts being used work.
>
>I've been doing software development for many years. I've worked with Access
>since version 1. In all that time I've never delivered a non-compiling
>application to a client, in Access or any other tool. I sent the client a
>few quotes from "Code Complete", by Steve McConnell, indicating that
>successful compilation was one of the minimal tests for a successful daily
>build.
>
>I'm just wondering if maybe I'm missing something here. Does anyone here
>deliver applications that don't compile? On purpose? And defend the
>practice?

I can't imagine putting something into production that doesn't
compile.

Also, it seems to imply that he doesn't do regular /Decompiles
and Compact/Repairs while developing.

I think the guy's way, way, waaaaaaaaay out in left field -
especially in defending the practice.
--
PeteCresswell

"Paul Shapiro" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...

> I'm just wondering if maybe I'm missing something here. Does anyone here
> deliver applications that don't compile? On purpose? And defend the
> practice?

No defense that I can think of. I've been developing Access databases for
just shy of 15 years, and every one of my applications compile. Incremental
compilation only worked on Access 97 and earlier databases, so maybe he is
still working on old versions.
--
Arvin Meyer, MCP, MVP
http://www.datastrat.com
http://www.mvps.org/access
http://www.accessmvp.com

As far as I can tell the on-demand compilation is still working in Access
2003, with an Access 2002-3 format db. Some of the forms with code in them
open without any compilation error, but the application as a whole fails to
compile. From the online help in A2003: "Compiling on Demand: It's a good
idea to explicitly compile the modules in your project by using the commands
described above, but it's not necessary. Microsoft Access compiles a module
before running a procedure in it if the module hasn't already been
compiled."
Paul Shapiro

"Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> "Paul Shapiro" <(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...
>
>> I'm just wondering if maybe I'm missing something here. Does anyone here
>> deliver applications that don't compile? On purpose? And defend the
>> practice?
>
> No defense that I can think of. I've been developing Access databases for
> just shy of 15 years, and every one of my applications compile.
> Incremental compilation only worked on Access 97 and earlier databases, so
> maybe he is still working on old versions.
> --
> Arvin Meyer, MCP, MVP
> http://www.datastrat.com
> http://www.mvps.org/access
> http://www.accessmvp.com

But on-demand compilation in Access 2000 and newer attempts to compile all
code in the database, whereas in Access 97, you could compile a single
module. In other words, there's no way the developer could be doing
on-demand compilation without having the errors appear.

--
Doug Steele, Microsoft Access MVP
http://I.Am/DougSteele
(no e-mails, please!)


"Paul Shapiro" <(E-Mail Removed)> wrote in message
news:e6YzdM$(E-Mail Removed)...
> As far as I can tell the on-demand compilation is still working in Access
> 2003, with an Access 2002-3 format db. Some of the forms with code in them
> open without any compilation error, but the application as a whole fails
> to compile. From the online help in A2003: "Compiling on Demand: It's a
> good idea to explicitly compile the modules in your project by using the
> commands described above, but it's not necessary. Microsoft Access
> compiles a module before running a procedure in it if the module hasn't
> already been compiled."
> Paul Shapiro
>
> "Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...
>> "Paul Shapiro" <(E-Mail Removed)> wrote in message
>> news:(E-Mail Removed)...
>>
>>> I'm just wondering if maybe I'm missing something here. Does anyone here
>>> deliver applications that don't compile? On purpose? And defend the
>>> practice?
>>
>> No defense that I can think of. I've been developing Access databases for
>> just shy of 15 years, and every one of my applications compile.
>> Incremental compilation only worked on Access 97 and earlier databases,
>> so maybe he is still working on old versions.
>> --
>> Arvin Meyer, MCP, MVP
>> http://www.datastrat.com
>> http://www.mvps.org/access
>> http://www.accessmvp.com
>
>

On demand compilation is not quite the same as incremental compilation.
Access 97 and earlier could compile just the module you were working on. The
choices from the Debug menu Item were: Compile Loaded Modules, Compile All
Modules, and Compile and Save All Modules. Now the choice is simply: Compile
[ProjectName]
--
Arvin Meyer, MCP, MVP
http://www.datastrat.com
http://www.mvps.org/access
http://www.accessmvp.com

"Paul Shapiro" <(E-Mail Removed)> wrote in message
news:e6YzdM$(E-Mail Removed)...
> As far as I can tell the on-demand compilation is still working in Access
> 2003, with an Access 2002-3 format db. Some of the forms with code in them
> open without any compilation error, but the application as a whole fails
> to compile. From the online help in A2003: "Compiling on Demand: It's a
> good idea to explicitly compile the modules in your project by using the
> commands described above, but it's not necessary. Microsoft Access
> compiles a module before running a procedure in it if the module hasn't
> already been compiled."
> Paul Shapiro
>
> "Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...
>> "Paul Shapiro" <(E-Mail Removed)> wrote in message
>> news:(E-Mail Removed)...
>>
>>> I'm just wondering if maybe I'm missing something here. Does anyone here
>>> deliver applications that don't compile? On purpose? And defend the
>>> practice?
>>
>> No defense that I can think of. I've been developing Access databases for
>> just shy of 15 years, and every one of my applications compile.
>> Incremental compilation only worked on Access 97 and earlier databases,
>> so maybe he is still working on old versions.
>> --
>> Arvin Meyer, MCP, MVP
>> http://www.datastrat.com
>> http://www.mvps.org/access
>> http://www.accessmvp.com
>
>

I understand now. But I think just the Compile menu items changed. I
remember there used to be the option for Compile Loaded Modules, and now
that option is gone. But if you don't use the Compile Project menu item,
Access still only compiles each module when you use it. In this case the
developer must never use the Compile command, since it fails. He just lets
Access compile a module as it's loaded. As long as you don't load the broken
ones, you don't see the error.
Paul Shapiro

"Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
news:OuEt9f$(E-Mail Removed)...
> On demand compilation is not quite the same as incremental compilation.
> Access 97 and earlier could compile just the module you were working on.
> The choices from the Debug menu Item were: Compile Loaded Modules, Compile
> All Modules, and Compile and Save All Modules. Now the choice is simply:
> Compile [ProjectName]
> --
> Arvin Meyer, MCP, MVP
> http://www.datastrat.com
> http://www.mvps.org/access
> http://www.accessmvp.com
>
> "Paul Shapiro" <(E-Mail Removed)> wrote in message
> news:e6YzdM$(E-Mail Removed)...
>> As far as I can tell the on-demand compilation is still working in Access
>> 2003, with an Access 2002-3 format db. Some of the forms with code in
>> them open without any compilation error, but the application as a whole
>> fails to compile. From the online help in A2003: "Compiling on Demand:
>> It's a good idea to explicitly compile the modules in your project by
>> using the commands described above, but it's not necessary. Microsoft
>> Access compiles a module before running a procedure in it if the module
>> hasn't already been compiled."
>> Paul Shapiro
>>
>> "Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
>> news:(E-Mail Removed)...
>>> "Paul Shapiro" <(E-Mail Removed)> wrote in message
>>> news:(E-Mail Removed)...
>>>
>>>> I'm just wondering if maybe I'm missing something here. Does anyone
>>>> here deliver applications that don't compile? On purpose? And defend
>>>> the practice?
>>>
>>> No defense that I can think of. I've been developing Access databases
>>> for just shy of 15 years, and every one of my applications compile.
>>> Incremental compilation only worked on Access 97 and earlier databases,
>>> so maybe he is still working on old versions.
>>> --
>>> Arvin Meyer, MCP, MVP
>>> http://www.datastrat.com
>>> http://www.mvps.org/access
>>> http://www.accessmvp.com

Correct. However, broken modules still can cause problems even if never
loaded. Those with broken references can and usually will break common VBA
functions. Uncompiled/broken modules cannot be imported, nor can an MDE be
made. That also means that the versions cannot be upgraded.
--
Arvin Meyer, MCP, MVP
http://www.datastrat.com
http://www.mvps.org/access
http://www.accessmvp.com

"Paul Shapiro" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
>I understand now. But I think just the Compile menu items changed. I
>remember there used to be the option for Compile Loaded Modules, and now
>that option is gone. But if you don't use the Compile Project menu item,
>Access still only compiles each module when you use it. In this case the
>developer must never use the Compile command, since it fails. He just lets
>Access compile a module as it's loaded. As long as you don't load the
>broken ones, you don't see the error.
> Paul Shapiro
>
> "Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
> news:OuEt9f$(E-Mail Removed)...
>> On demand compilation is not quite the same as incremental compilation.
>> Access 97 and earlier could compile just the module you were working on.
>> The choices from the Debug menu Item were: Compile Loaded Modules,
>> Compile All Modules, and Compile and Save All Modules. Now the choice is
>> simply: Compile [ProjectName]
>> --
>> Arvin Meyer, MCP, MVP
>> http://www.datastrat.com
>> http://www.mvps.org/access
>> http://www.accessmvp.com
>>
>> "Paul Shapiro" <(E-Mail Removed)> wrote in message
>> news:e6YzdM$(E-Mail Removed)...
>>> As far as I can tell the on-demand compilation is still working in
>>> Access 2003, with an Access 2002-3 format db. Some of the forms with
>>> code in them open without any compilation error, but the application as
>>> a whole fails to compile. From the online help in A2003: "Compiling on
>>> Demand: It's a good idea to explicitly compile the modules in your
>>> project by using the commands described above, but it's not necessary.
>>> Microsoft Access compiles a module before running a procedure in it if
>>> the module hasn't already been compiled."
>>> Paul Shapiro
>>>
>>> "Arvin Meyer [MVP]" <(E-Mail Removed)> wrote in message
>>> news:(E-Mail Removed)...
>>>> "Paul Shapiro" <(E-Mail Removed)> wrote in message
>>>> news:(E-Mail Removed)...
>>>>
>>>>> I'm just wondering if maybe I'm missing something here. Does anyone
>>>>> here deliver applications that don't compile? On purpose? And defend
>>>>> the practice?
>>>>
>>>> No defense that I can think of. I've been developing Access databases
>>>> for just shy of 15 years, and every one of my applications compile.
>>>> Incremental compilation only worked on Access 97 and earlier databases,
>>>> so maybe he is still working on old versions.
>>>> --
>>>> Arvin Meyer, MCP, MVP
>>>> http://www.datastrat.com
>>>> http://www.mvps.org/access
>>>> http://www.accessmvp.com
>
>

In news:(E-Mail Removed),
Paul Shapiro <(E-Mail Removed)> wrote:
> I'm just wondering if maybe I'm missing something here. Does anyone
> here deliver applications that don't compile? On purpose? And defend
> the practice?

No. No. No.

--
Dirk Goldgar, MS Access MVP
www.datagnostics.com

(please reply to the newsgroup)