Now I found why under some sizes it worked. It was because in the config
file php.ini, the variable output_buffering was set to "4096". When the
buffer reached that size the rest of the cookie cannot be stored and failed
:)

well, that's what I suppose reading the documentation of this option:

; Output buffering allows you to send header lines (including cookies) even
; after you send body content, at the price of slowing PHP's output layer a
; bit.  You can enable output buffering during runtime by calling the output
; buffering functions.  You can also enable output buffering for all files
by
; setting this directive to On.  If you wish to limit the size of the buffer
; to a certain size - you can use a maximum number of bytes instead of 'On',
as
; a value for this directive (e.g., output_buffering=4096).

Thank you all :)

2006/9/19, German Piqu� <extremdur2@[...].com> :
> 
>   I found the problem... and the solution
> 
>  To test it, i set a group of messages in the code of the class cookie when
>  the variable "debug" is active. That was leading to all the errors. It's
>  like the script was generating html code and after that it tries to modify
>  the cookies.
> 
>  I set the debug variable to false and only show the cookies' contents
>  after the writing to it, and it works with all the strings length I tested.
> 
>  BTW, how many bytes a caracter in a string consume? In C a char is equal
>  to one byte, but I don't know if in php is the same. I want to know to limit
>  in the class the cookies' size.
> 
>  Thanks to all :)
> 
>   2006/9/19, German Piqu� <extremdur2@[...].com>:
>  >
>  >
>  >
>  > ---------- Forwarded message ----------
>  > From: German Piqu� < extremdur2@[...].com>
>  > Date: 19-sep-2006 10:38
>  > Subject: Re: [PHP-WIN] cookies time? - now it doesn't write
>  > To: "Frank M. Kromann" < frank@[...].info>
>  >
>  >  Now works... Oo
>  >
>  > But i have another problem.
>  >
>  > My script must check a DB and inform of the changes that have been
>  > produced since the last update. I tried two different ways to do it:
>  >
>  > 1) Use a file in the server to store the diferences
>  >
>  > 2) Use cookies to store the info.
>  >
>  > The first solution has the problem that the update will take place for
>  > all the users that access the script. I think that would be risky for the
>  > integrity of the file or the info retrieved in a case of to much accesses...
>  >
>  >
>  > The second option i think that is better. But i encounter problems.
>  > Depending on the content, the setcookie returns an error. The idea of the
>  > code is that first search if the cookie exists. If yes, shows it. otherwise,
>  > no. After that, when the access to the DB is made, it stores the results in
>  > the cookie to be reused again. I tried to avoid the 4 KBs limit problem by
>  > creating a class that controls the lenght of the string and if it surpases
>  > the limit, creates another cookie, with the increment of a counter attatchet
>  > to the base name of the cookie.
>  >
>  > It's curious that if I put the string manually, it can be larger than
>  > one string created through the union of one array's cells.
>  >
>  > Could it be a memory problem, like the class is taking all of it leaving
>  > nothing to the cookies? I think that this sentence has no sense but i don't
>  > know what it could be.
>  >
>  > I tried to destroy the cookies after reading them, too. But when I try
>  > to re-create it later in order to insert the new info causes an error, and
>  > in the next refresh, there are no cookies and I can write to the cookies!!
>  > Then, in the next refresh, again I can read the cookie, but I can't write to
>  > it. In the next, there are no cookis and I can write again... Why? Why if
>  > the cookie is destroyed I can write it but if it's create I can't rewrite
>  > it?
>  >
>  > That's the cause that leade me to think that it's a memory problem.
>  >
>  > Well... any ideas?
>  >
>  >
>  >
>  > 2006/9/19, German Piqu� <extremdur2@[...].com>:
>  > >
>  > > So, it's another problem. -_- I'll give it a try.
>  > >
>  > > 2006/9/18, Frank M. Kromann <frank@[...].info>:
>  > >
>  > > > both time() and mktime() works fine on Windows.
>  > > >
>  > > > - Frank
>  > > >
>  > > > > Well, reading the documentation about cookies i saw that the
>  > > > function
>  > > > time()
>  > > > > works using UNIX epochs. If I try to use this function (or mktime)
>  > > > in
>  > > > > windows, the php shows a white page, like there's an error
>  > > > somewhere.
>  > > > >
>  > > > > It's a problem with windows machine's or there's another solution
>  > > > to
>  > > > this?
>  > > > >
>  > > > > Meanwhile, I put no time option (so cookie expires at the end of
>  > > > the
>  > > > > session) but i would like to try to set up a timeout.
>  > > > >
>  > > > > I don't demand absolute solutions, I only want a clue to work on.
>  > > > >
>  > > > > Thank you :)
>  > > > >
>  > > >
>  > > >
>  > > >
>  > > >
>  > >
>  > >
>  >
>  >
>  >
> 
>