URLMonLibrary and the Internet cache

Hi all,

I've been using URLMonLibrary>>urlDownload:toFile: in an application and it
all worked to my satisfaction on my development environment, but on the
deployed machines I got unpredictable and therefore unreliable results. This
appeared to be caused by URLMonLibrary using the internet cache as the
primary source. Changing the IE internet cache setting from 'Automatic' to
'every visit' seemed to resolve the problems. Now I can't rely on the users
all changing their IE cache setting so I wondered whether there is a way to
enforce URLMonLibrary not to use the cache. In MSDN I didn't find any
reference to the cache with regards to the URL Moniker. I did find ways of
refreshing the cache through the WinInet library, but that didn't have such
a tidy interface as the URL Moniker.

Anybody any other ideas?

Thanks in advance,

Ted

Ted,

> primary source. Changing the IE internet cache setting from 'Automatic' to
> 'every visit' seemed to resolve the problems. Now I can't rely on the
users
> all changing their IE cache setting so I wondered whether there is a way
to
> enforce URLMonLibrary not to use the cache. In MSDN I didn't find any
> reference to the cache with regards to the URL Moniker. I did find ways of
> refreshing the cache through the WinInet library, but that didn't have
such
> a tidy interface as the URL Moniker.

Check Ian's archives - I vaguely recall a similar discussion in the past.
Failing that, try google.  FWIW, in the past three or four problems that
I've solved, MSDN was of no help but google was able to locate the answer
(sadly with a LOT of other stuff).  If you are still stuck, keep in mind
that Squeak contains Smalltalk code for the common Internet protocols.  The
PWS code saved me a lot of work in writing my own web server for debugging.

Good luck!

Bill

--
Wilhelm K. Schwab, Ph.D.
[hidden email]

"Ted Bracht" <[hidden email]> wrote in message
news:[hidden email]...
>
> I've been using URLMonLibrary>>urlDownload:toFile: in an application and
it
> all worked to my satisfaction on my development environment, but on the
> deployed machines I got unpredictable and therefore unreliable results.
This
> appeared to be caused by URLMonLibrary using the internet cache as the
> primary source. Changing the IE internet cache setting from 'Automatic' to
...

This is not what you are looking for, but may be helpful in a pinch.  I have
found that in conditions where a web page does not seem to be refreshing I
can trick it into thinking the page is different by using a slightly
different url.  For example if I add a "?" and a number the url will be
re-queried.  If the url already has parameters then just add a new one with
the number, it will hopefully be ignored.  You could use a random number or
a timestamp.  This is not a great solution, but it might work with a minimum
of effort.

Chris

"Ted Bracht" <[hidden email]> wrote in message
news:[hidden email]...
> Hi all,
>
> I've been using URLMonLibrary>>urlDownload:toFile: in an application and
it
..
> a tidy interface as the URL Moniker.

Hi, I had somehow similar problems with URLPresenter. If it can be of any
help, here is the snippet of code that seems to fix it in the case od
URLPresenter:

>>refreshUrl: url


url ~= previewPresenter view controlDispatch LocationURL

    ifTrue: [previewPresenter view controlDispatch navigate: url ]

    ifFalse: [previewPresenter view controlDispatch Refresh ].


rush
--
http://www.templatetamer.org/

Ted Bracht wrote:

> I've been using URLMonLibrary>>urlDownload:toFile: in an application
> and it all worked to my satisfaction on my development environment,
> but on the deployed machines I got unpredictable and therefore
> unreliable results.

You may want to consider dropping URLMonLibrary -- if all you are doing is
fetching the contents associated with simple HTTP URLs, and don't need to worry
about, say, authorisation or proxies, then doing an HTTP 'GET' with raw TCP/IP
is only a handful of lines of code.  If those conditions do apply (and, of
course, they may not) then URLMonLibrary is over-specced (and under-reliable
;-) for the job.  E.g. #urlDownload:toFile: will fail if the user happens to
have said "work off line" in Outlook Express.

There's code somewhere in the archives, or drop me a line if you'd like me to
dig something out.

    -- chris

Hi all,

>
>
>

Thanks for the suggestions, I'll give them a try.

Ted