lynx-dev
[Top][All Lists]
Advanced

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

LYNX-DEV Goto URLs history list bug Re: unanswered queries (2) (fwd)


From: Philip Webb
Subject: LYNX-DEV Goto URLs history list bug Re: unanswered queries (2) (fwd)
Date: Sat, 14 Mar 1998 14:26:40 -0500 (EST)

Randall forgot to copy this to lynx-dev:

Forwarded message:
> From address@hidden Sat Mar 14 14:14 EST 1998
> Message-Id: <address@hidden>
> From: Randall <address@hidden>
> Reply-To: Randall <address@hidden>
> Errors-To: Randall <address@hidden>
> To: address@hidden, Al Gilman <address@hidden>
> Subject: Goto URLs history list bug   Re: unanswered queries (2) 
> In-reply-to: Philip's message of "Fri, 13 Mar 1998 18:55:20 EST."
>              <address@hidden> 
> In-reply-to: Al's message of "Fri, 13 Mar 1998 20:27:32 EST."
>              <address@hidden> 
> Date: Sat, 14 Mar 1998 11:11:11 -0800
> Sender: address@hidden
> Content-Type: text
> Content-Length: 3316
> 
> > From: Philip Webb <address@hidden>
> > upgrade to 2.8 from  www.slcc.edu/lynx/release/  & test that for your bug:
> > if it's still there, tell us again & it may get on the list for 2.8.1
> I don't suppose there is some guest account somewhere, so i could just
> quickly try it out, without having to download/build the thing?
> If you have a 2.8 handy, it's real simple to test out, just do
>   g wig.mit.edu
> or something of the sort once, then
>   g  (up-arrow)
> and see if it says wig.mit.edu or something else.
> 
> I don't see mention of this in the CHANGES since 2.7 file I find at:
> www.slcc.edu/lynx/release2-8/CHANGES
> 
> > if you are capable of C programming, you should try doing your own patch.
> I did try to make a quick patch, but it crashed and burned (crash as in
> seg fault).  But that was a lame approach anyway.  If I am going to attempt
> a patch, i need to discuss with someone familiar with the code about what
> is the correct approach.  This is my idea of how to do it right:
> 
> 
> LYEnsureAbsoluteURL() should return a status value, rather than void.
> [Or maybe not--skip down to note on LYFillLocalFileURL.]
> Check the status before copying temp back to user-input-buffer in
> LYMainLoop.c, in the check-goto-URL section.
> 
> As LYEnsureAbsoluteURL is called only 4 times, by my count with grep,
> this shouldn't be problematic.  If the return value is not needed in the
> other locations, it can be cast to void to avoid compiler warnings; we can
> cover this with a macro if need-be for portability.
> 
> [btw it is good to see that we correctly have "ensure" instead of the 
> so-common
> errors "insure"/"assure"]
> 
> We would also have to deal with LYFillLocalFileURL(), which is called
> just before LYEnsureAbsoluteURL() in the check-goto-URL section.  We
> could either surround the call by an access(2) check, or include this
> access(2) check in LYFillLocalFileURL and have it return a value.
> 
> Actually, now that I think about it again...  Obviously (given that the
> file://localhost... guess is the one that shows up in the history list, and
> that LYFillLocalFileURL is called *before* LYEnsureAbsoluteURL),
> LYEnsureAbsoluteURL does not write into temp unless host-lookup is successful.
> So probably we should just make LYFillLocalFileURL consistent with
> LYEnsureAbsoluteURL in this regard--in other words put an access(2) check
> within LYFillLocalFileURL before writing to the buffer passed in as temp
> (no need for a return value then).
> 
> I count 19 calls to LYFillLocalFileURL (argument against adding return val!).
> Also, access(2) may have to be covered by a macro or function for portability.
> 
> > From: Al Gilman <address@hidden>
> > You are not the first to wish this.  You might want to add this change
> > to the suggestions at       http://www.slcc.edu/lynx/html/todo-list.html
> Thanks for the web address, which I was previously lacking.  I have added
> the suggestion there.  Perhaps you would like to mention this web address
> in the README file.
> 
> Btw, sorry if my note repeated several others on the lynx-dev list; i didn't
> see mention of it when i scanned quite a few headers.
> =randall
> 
> +-------------------------------------------+
> |    ><>        Randall Whitman       <><   |
> |  address@hidden - permanent alias   |
> |    address@hidden - current email     |
> +-------------------------------------------+

-- 
========================,,============================================
SUPPORT     ___________//___,  Philip Webb : address@hidden
ELECTRIC   /] [] [] [] [] []|  Centre for Urban & Community Studies
TRANSIT    `-O----------O---'  University of Toronto

reply via email to

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