lynx-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

LYNX-DEV http://www.io.com/~sergent/c/bulletin1-version4.txt


From: Klaus Weide
Subject: LYNX-DEV http://www.io.com/~sergent/c/bulletin1-version4.txt
Date: Fri, 11 Jul 1997 15:42:00 -0500 (CDT)

Quoting from the last version of the bulletin1 draft in <URL:
http://www.io.com/~sergent/c/cert-index.html>:


C. One other solution (a source code patch) for this problem, by Klaus
   Weide, can be found at:
        http://www.slcc.edu/lynx/klaus/temp/
   However, this patch should be considered "alpha" quality code, and
   its author is not supporting it at this time.


When I announced that patch, there was no following discussion of the
issues I mentioned (in 
URL: http://www.flora.org/lynx-dev/html/month0597/msg00501.html).
I had no indication that anybody was actually trying the code.

(There was suggested code (from Jonathan Sergent?) for an improved
reimplementation of mkstemp(), but that was a bit beside the point
wheter the patch was a good approach.)

Then Fote started making changes to the fotemods code which more
or less addressed the issue, in a different way (not based on my
patch and not based on mkstemp()).  Ok, somebody was working on the
thing, in code that ctually got used and didn't chang Lynx's behavior
in subtle ways.  So the development code just took Fote's changes.

Major differences in my approach:

- Use mkstemp() [could use similar function] for atomic secure
  naming+opening of file.
- Never reuse a temporary filename one the file has been used and
  closed.

If my approach should be for some reason (sufficiently) better
than Fote's current aproach, I could put it in the development code.
But I am not convinced that that is the case, and don't understand
all the different implications.

  Klaus



;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;

reply via email to

[Prev in Thread] Current Thread [Next in Thread]