IE can't cache .js files above 50K

[Manish]

I observed that IE is not caching the .js file if its size
is 50K or greater. If I cut down the size to say 45K, it
caches it fine !!!

Is this a known bug or feature ? Is there an IE setting I
need to do tweak for this to work ?

Thank you,

Manish.

 

[Robert Aldwinckle]

Do you have a sample link?

 

[Guest]

Robert:
I work with Manish
I sent an email to (e-mail address removed) a JSP and couple of .JS files beacuse our IP addresses are not exposed outside our network. The mail got bounced back saying that your mail box is full. Do you have any email id where I can send the files.

Thanks
Mahesh
----- Robert Aldwinckle wrote: -----

Do you have a sample link?

 

[Robert Aldwinckle]

Do you have a sample link?

I was assuming from the description that this was a general problem
with large .js files. If I can find a publicly available large .js file (>50K)
do you think that you would have the symptom with it?

If the problem only occurs with certain large files then I would suspect
the content as much as I would suspect their largeness.

Also relevant will be what headers your server is sending with its response.
For example, we recently observed that when some Apache servers
send a Vary: header that IE seems to avoid caching it even though
the programmer is doing everything he can to otherwise mark it cacheable.
Can you trace the response of the suspect file and see if there is a clue
that way?

I sent an email to (e-mail address removed) a JSP and couple of .JS files

I'm not set up to run a personal web server so I doubt if I should be the
one to be getting your files anyway but FWIW my webmail Inbox is
most likely to be available at about 10:00, 18:00, and 00:00 EST.
Of course the most recent E-mail exploit isn't helping me provide a place
for receiving legitimate messages. ;(


Robert
---

Robert:
I work with Manish
I sent an email to (e-mail address removed) a JSP and couple of .JS files beacuse our IP addresses are not exposed outside our

network. The mail got bounced back saying that your mail box is full. Do you have any email id where I can send the files.

 

[Guest]

Any large .js file (> 50K), that file would not be cached. So I guess it has nothing to do with the content. As both the files one large .js file (>50 K) and other .js file of 47 KB are included in the same JSP, the header information for both should be same, If I understand your point rightly. Could you give me a pointer or two I can try.

Thanks
Mahesh Mutham

----- Robert Aldwinckle wrote: -----

I was assuming from the description that this was a general problem
with large .js files. If I can find a publicly available large .js file (>50K)
do you think that you would have the symptom with it?

If the problem only occurs with certain large files then I would suspect
the content as much as I would suspect their largeness.

Also relevant will be what headers your server is sending with its response.
For example, we recently observed that when some Apache servers
send a Vary: header that IE seems to avoid caching it even though
the programmer is doing everything he can to otherwise mark it cacheable.
Can you trace the response of the suspect file and see if there is a clue
that way?

I sent an email to (e-mail address removed) a JSP and couple of .JS files

I'm not set up to run a personal web server so I doubt if I should be the
one to be getting your files anyway but FWIW my webmail Inbox is
most likely to be available at about 10:00, 18:00, and 00:00 EST.
Of course the most recent E-mail exploit isn't helping me provide a place
for receiving legitimate messages. ;(


Robert
---

I work with Manish
I sent an email to (e-mail address removed) a JSP and couple of .JS files beacuse our IP addresses are not exposed outside our

network. The mail got bounced back saying that your mail box is full. Do you have any email id where I can send the files.

 

[Robert Aldwinckle]

Could you give me a pointer or two I can try.

The first step would be getting an HTTP tracer.
Do you have one?

Does Netscape have a trace? You could try the assumption that
whatever it gets would be what IE would get (not necessarily valid
because of the User-Agent and because of whatever differences
that each browser would induce the server to make by what they send
with each request.)

Ideally I think IE would provide one but it doesn't so we have to improvise.
Solutions range from DIY proxies to real proxies to real packet sniffers.
Ethereal is one of the latter but I have never had any success using it
transparently. I think that ProxOmitron might have a trace feature but
I'm not sure about that; I've never used it. I use a developer's private
tool which works as a proxy; so it's not transparent either but at least
it doesn't have the performance hit the Ethereal seems to have and
in any case is easier to use. If you have a MS SMS you may have
their netmon package which operates similar to Ethereal.

Hmm... there may be another option which I have previously been
unaware of.

<TITLE>310875 - Description of the Network Monitor Capture Utility</TITLE>

This implies that you could *capture* the trace data using netcap
(e.g. from the XP Support Tools). I think that even if you don't
have netmon to interpret the trace that you still might be able to use
Ethereal to do it for you. The only reason you might want to do this
instead of using Ethereal to capture as well is if you find that netcap
does it more transparently than Ethereal.

The article doesn't make very clear how we might filter our capture
to only HTTP packets. If it could do that it would help a lot towards
making the capture phase more transparent.

It should not be this difficult just to see the underlying HTTP packets
but I have been looking for a long time and haven't found a really simple,
reliable, publicly available tool.


Good luck

Robert
---

Any large .js file (> 50K), that file would not be cached. So I guess it has nothing to do with the content. As both the files one

large .js file (>50 K) and other .js file of 47 KB are included in the same JSP, the header information for both should be same, If
I understand your point rightly. Could you give me a pointer or two I can try.