Mailing List Archive: Stuff for 0.8.3...

Stuff for 0.8.3...

Jul 25, 1995, 9:07 AM

Post #1 of 9 (598 views)

What follows is what's in my current sandbox. The only seriously
controversial item that I'm aware of is the no2slash business (the
directory_walk routine now finds PATH_INFO first, then makes a copy of
the filename and runs no2slash on that). I want this fixed, at least
if the old code did it "right", and if I understand them correctly,
then Rob and Andrew (who originally reported it) want it as well, so I
count three in favor, which is why I fixed it.

If no one objects too strenuously, I'll put it on hyperreal later in
the day.

rst

*) Edited distribution README to *briefly* summarize installation
procedures, and give a pointer to the INSTALL file in the src/
directory.

*) Upgraded imagemap script in cgi-bin to 1.8 version from more
recent NCSA distributions.

*) Bug fix to previous bug fix --- if .htaccess file and <Directory>
exist for the same directory, use both and don't segfault. [Reported
by David Robinson]

*) Proper makefile dependencies [David Robinson]

*) Note (re)starts in error log --- reported by Rob Hartill.

*) Only call no2slash() after get_path_info() has been done, to
preserve multiple slashes in the PATH_INFO [.NCSA compatibility,
reported by Andrew Wilson, though this one is probably a real bug]

*) Fixed mod_imap.c --- relative paths with base_uri referer don't
dump core when Referer is not supplied. [Randy Terbush]

*) Lightly edited sample config files to refer people to our documentation
instead of NCSA's, and to list Rob McCool as *original* author (also
deleted his old, and no doubt non-functional email address). Would be
nice to have examples of new features...

Re: Stuff for 0.8.3... [ In reply to ]

Jul 25, 1995, 3:33 PM

Post #2 of 9 (596 views)

> *) Only call no2slash() after get_path_info() has been done, to
> preserve multiple slashes in the PATH_INFO [.NCSA compatibility,
> reported by Andrew Wilson, though this one is probably a real bug]

I don't see why a web server should ever merge multiple slashes, even
in the unix path. If it doesn't like the path, it should reject the
request, rather than munging it.

I can't imagine anyone relying on this behaviour, namely that
http://server.com/dir/path.html
and
http://server.com/dir//path.html

give the same result, whereas I have seen at least one complaint in the
newsgroup that allowing the latter form means that relative links cannot be
guaranteed to work.

Apache should fault the latter, and I think this is best done in
directory_walk().

David.

Re: Stuff for 0.8.3... [ In reply to ]

randy at zyzzyva

Jul 25, 1995, 6:49 PM

Post #3 of 9 (599 views)

WhoooHoo 30+ messages today! Like the old days! :-)

> I can't imagine anyone relying on this behaviour, namely that
> http://server.com/dir/path.html
> and
> http://server.com/dir//path.html
>
> give the same result, whereas I have seen at least one complaint in the
> newsgroup that allowing the latter form means that relative links cannot be
> guaranteed to work.
>
> Apache should fault the latter, and I think this is best done in
> directory_walk().
>
> David.

I agree with David's take on this. I think it is wrong to
allow this abuse of the filesystem. This is just plain wrong.

Re: Stuff for 0.8.3... [ In reply to ]

akosut at nueva

Jul 25, 1995, 10:06 PM

Post #4 of 9 (590 views)

On Tue, 25 Jul 1995, Robert S. Thau wrote:

> *) Upgraded imagemap script in cgi-bin to 1.8 version from more
> recent NCSA distributions.

Hi. Me again. One note: in order to make this work, you need to remove
the util.o references from the imagemap: line in the cgi-src Makefile.
Version 1.8 has the required util functions built-in to the source.

> *) Lightly edited sample config files to refer people to our documentation
> instead of NCSA's, and to list Rob McCool as *original* author (also
> deleted his old, and no doubt non-functional email address). Would be
> nice to have examples of new features...

Since this was my idea, I decided to put my foot where my mouth was and
edit the conf files to put in examples of new features. The revised files
are at http://www.nueva.pvt.k12.ca.us/~akosut/conf.tar, and contain
examples of the following directives:

AddLanguage
BindAddress
CacheNegotiatedDocs
VirtualHost
XBitHack

I also updated the DirectoryIndex entry, and added commented-out AddType
entries for httpd/send-as-is (BTW, isn't this really just message/rfc822,
in disguise?) and application/x-httpd-imap. Plus I re-edited the copy at
the top of the files to make a bit more sense.

Also, I added some AddLanguage entries. I looked at RFC 1766, which is the
document that describes the Content-Language header, and it referenced a
ftp://ftp.isi.edu/in-notes/iana/assignments/languages/, which supposedly
defines languages. However, this directory does not appear to exist. So,
not having ISO 639 (which RFC 1766 also references) handy, I used the
examples given in the RFC itself to define a couple of AddLanguage
lines. Hopefully I did that correctly.

Anyhow, think that's it. Hope someone likes my new configuration files.
Or not. Whatever.

Thanks!

--/ Alexei Kosut <akosut@nueva.pvt.k12.ca.us> /--------/ Lefler on IRC
----------------------------/ <http://www.nueva.pvt.k12.ca.us/~akosut/>
The viewpoints expressed above are entirely false, and in no way
represent Alexei Kosut nor any other person or entity. /--------------

Re: Stuff for 0.8.3... [ In reply to ]

brian at organic

Jul 26, 1995, 4:08 AM

Post #5 of 9 (592 views)

On Tue, 25 Jul 1995, Alexei Kosut wrote:
> Since this was my idea, I decided to put my foot where my mouth was and
> edit the conf files to put in examples of new features. The revised files
> are at http://www.nueva.pvt.k12.ca.us/~akosut/conf.tar

+1. Who's doing the build?

Brian

--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
brian@organic.com brian@hyperreal.com http://www.[hyperreal,organic].com/

Re: Stuff for 0.8.3... [ In reply to ]

randy at zyzzyva

Jul 26, 1995, 5:55 AM

Post #6 of 9 (600 views)

> Also I'd recommend using a dot in
> AddLanguage en .en
> which makes it clearer which is the language and which is the suffix.

The most recent MIME RFC seems to indicate that the addition of a dot
when specifying types is invalid. The dot is assumed. I agree that
it does read a bit better, buuuut.... Standards are Standards. I think
we have been trying to clean the dots out of config files.

Re: Stuff for 0.8.3... [ In reply to ]

Jul 26, 1995, 9:25 AM

Post #7 of 9 (594 views)

The most recent MIME RFC seems to indicate that the addition of a dot
when specifying types is invalid. The dot is assumed. I agree that
it does read a bit better, buuuut.... Standards are Standards. I think
we have been trying to clean the dots out of config files.

The dot goes on the file extension, not on the language tag (which is
not a MIME type in any event) --- and it helps to make clear which is
which, which will probably save people a bit of trouble when they want
to make the two different (as in the much-talked-over Polish case).

At any rate, I don't see how the MIME RFCs apply to notations for
filename extensions.

rst

Re: Stuff for 0.8.3... [ In reply to ]

Jul 26, 1995, 9:27 AM

Post #8 of 9 (591 views)

+1. Who's doing the build?

Well, I already built a 0.8.3; I could build another one with Alexei's
changes to the conf files fairly easily this evening...

rst

Re: Stuff for 0.8.3... [ In reply to ]

guillaum at clipper

Jul 26, 1995, 2:27 PM

Post #9 of 9 (592 views)

First, a +1 on freezing and releasing something soon.

About the config files :

You'll also want a line with LanguagePriority (please adapt my English
if needed) in srm.conf-dist :

# LanguagePriority allows you to give precedence to some languages
# in case of a tie during content negotiation.
# Just list the languages in decreasing order of preference.

LanguagePriority en fr de

Also I'd recommend using a dot in
AddLanguage en .en
which makes it clearer which is the language and which is the suffix.

Finally, maybe we could uncomment the lines
#AddEncoding x-compress Z
#AddEncoding x-gzip gz
because this has become really widspread now, and probably should be
in by default.

Florent

--
Florent.Guillaume@ens.fr