httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From André Warnier>
Subject Re: [users@httpd] Downloading a CSV file
Date Sat, 14 Jun 2008 12:51:28 GMT

I suggest that you first make the following search in Google :
ie +"mime type"
and then read some.
It will tell you what you are up against.

In summary :
The behaviour of various versions and sub-versions of Internet Explorer 
with respect to files received from a HTTP server, is variable, erratic, 
and in most cases conflicts with the specifications of the relevant 
Internet RFCs.
(I am hereby referring to the "Content-type" and "Content-disposition" 

In the practice, it means that if you find one set of HTTP headers that 
works with one IE version/revision, it probably will not do the same 
with other versions/revisions.  In addition, I have found that even with 
identical versions of IE, sometimes client stations will still react 
differently, depending on which (other) program is installed on the 
workstation.  (For example, you may find that depending on whether your 
client has or not MS-Excel installed on his workstation, the behaviour 
of IE versus your CSV files will change).
And since you cannot exactly control which version of IE your client is 
using, nor the programs installed on his workstation, you're in trouble.

The only way that I have found to get a more or less consistent 
behaviour, is to "sniff out" the IE version used by the client (see the 
"User-Agent" request header), and adapt the headers you send in function 
of that.
It's not pretty and no fun.
It will also probably cost you many hours of lost time explaining the 
same over and over to your users and their system administrators, hours 
for which MS will not compensate you.

If anyone else on this list has a different experience and a better 
solution, please let us know.

Note that this topic has alrady been discussed ad nauseam in a multitude 
of lists and web discussions, so adding your own complaint will probably 
just be a drop in the ocean.
But I could not resist warning this new victim.


Octavian Rasnita wrote:
> Hi,
> I've made a program that creates a .csv file and offer it for downloading.
> If a user uses Internet Explorer 6 to download the file and it chooses 
> to right-click then "Save as..." from the context menu, the file is 
> downloaded right.
> But if the user just clicks or hits enter on the link for downloading 
> that file, the user is asked what he wants to do, to save or open the 
> file, and if he choses that he wants to save it, that window remains 
> open and it doesn't ask for the name of the file.
> I've used lwp-request (from ActivePerl) to see which are the HTTP 
> headers received. Please tell me what should I change, add or remove in 
> order to be able to download that file by just clicking on the link.
> Thank you.
> The headers are:
> Cache-Control: no-store, no-cache, must-revalidate, post-check=0, 
> pre-check=0, max-age=0
> Connection: close
> Date: Sat, 14 Jun 2008 12:09:38 GMT
> Pragma: no-cache
> Server: Apache/2.2.8 (Fedora)
> Content-Length: 4358
> Content-Type: text/csv
> Expires: Thu, 01 Jan 1970 00:00:00 GMT
> Client-Date: Sat, 14 Jun 2008 12:13:49 GMT
> Client-Peer:
> Client-Response-Num: 1
> Content-Disposition: attachment;filename="REGS--2008-06-13.csv"
> Set-Cookie: zzzz=6eb405c4774a704757d6455963ec7bfc10810eb9; 
>; path=/
> X-Catalyst: 5.7014
> Octavian
> ---------------------------------------------------------------------
> The official User-To-User support forum of the Apache HTTP Server Project.
> See <URL:> for more info.
> To unsubscribe, e-mail:
>   "   from the digest:
> For additional commands, e-mail:

The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:> for more info.
To unsubscribe, e-mail:
   "   from the digest:
For additional commands, e-mail:

View raw message