Re: Problems updating global preferences

--===============70348000722009418==
Content-Type: multipart/signed; micalg=pgp-sha1;
	protocol="application/pgp-signature"; boundary="wxQIpZ+KqQj4wYAN"
Content-Disposition: inline


--wxQIpZ+KqQj4wYAN
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Fri Sep 26, 2003 at 01:11:33PM -0700, Scott Brown wrote:

> The plot thickens...
>=20
> I haven't been able to try 0.2.4pre yet... will likely do so shortly,=20
> although a diff between the two admin.php files (1.28 and 1.30) doesn't=
=20
> show me anything that would affect the issue first reported yesterday!
>=20
> I have found some other quirks, though, not quite sure what they mean. I'=
d=20
> much rather run this on a *nix box, but here I am...
>=20
> 1) d_DoDiff
>=20
> This is probably a quirk on my end, but my mail server didn't like mixed=
=20
> case on Cc:, so I had to change it to CC:.

This shouldn't be an issue for any MTA, I would suspect (changing it to CC).
I can make this change.

> 1.5) newuser.php
>=20
> Similar to the above, probably a quirk unique to my setup, but I had to=
=20
> change line 117 so that it did not include $rname. My mail server (I=20
> *think*; at least I didn't see where the code was doing it, except around=
=20
> $email. PHP might be, but chasing that would be a pain!) would add the=20
> carats "<>" around the entire thing, like <User Name <username-9IKiO1iGCm/[email protected]=
>>=20
> and throw a syntax error. So my newuser.php just sends $email to @mail (n=
o=20
> carats). I still need to change it where $go =3D=3D "newpass" (line 16).

Ok, your MTA is really really wierd.  That's a standard way of writing a To
line.  I mean, I can ditch the real name part (that was more of a cosmetic
thing than it being functional).

> Those two might be good for a "getting Anthill running on Windows" FAQ or=
=20
> something. Or ArgoSoft Mail Server Pro. It looks like it was designed wit=
h=20
> Sendmail in mind?

It was designed for a normal MTA.  =3D)  Really, it works with postfix,
sendmail, qmail, exim...  I'm not sure about some of these windows MTAs...
seem kinda non-standard to me.

> 2) The thing with admin.php (previous thread)
>=20
> Strangely, if I repeatedly refresh the page after a POST, I can=20
> *occasionally* get the setting change to take (very unpredictable), and g=
et=20
> the success message at the top of the page. I have been over and over the=
=20
> code, and can't for the life of me see what is going on. Even when I=20
> successfully get a change to "take", I have to refresh the page again to=
=20
> see the font change. None of the elements, or any combination of elements=
,=20
> alleviate the issue. I don't *think* it is a client issue. If someone=20
> wanted to check it out, I can email them off-list with the URL and a=20
> username and password. It's strictly in a (more or less) sandboxed area f=
or=20
> evaluation at the moment, so I'm not sweating it. It would certainly help=
=20
> me to determine whether or not this is a client issue, or a server issue.

Please, I'd like to be able to see what's going on.

> The screwy thing is, every other form works fine. I even reviewed my=20
> settings in my php.ini to make sure safe mode was off (it is, and was). A=
s=20
> a note, variables_order is set to "EGPCS". Does this matter? gpc_order is=
=20
> also commented out (deprecated) does this matter?

Should work with safe_mode on as well.  I don't think variables_order should
make a difference... I've never muddled with that and don't think it should
matter.

> The only thing I can think of is that for some reason, this statement:
>=20
> if ($go =3D=3D "change")
>=20
> Doesn't resolve properly, yet on every other page, it seems to do just=20
> fine! Maybe IE6 thinks "change" is a reserved word or something, and isn'=
t=20
> posting it properly. I might try changing it (pardon the pun) to somethin=
g=20
> else. The reason I say that it must be this statement is because I would=
=20
> expect that if it were resolving, but a database error occurred, I would=
=20
> still get the success message at the top of the page, but it would fail t=
o=20
> do the update. No errors are thrown, and whenever I see the message, the=
=20
> update was successful.

I don't think it can be a reserved word.  The variable name is go, the value
is change.  Unless go is a reserved word, but that would mean the other
forms would be broken as well.

> Needless to say, I am stumped!

You and me both.  That's pretty odd.

I have to admit, when it comes to running under windows, I'm pretty
clueless.  I do my development work on OS X and run it on various Linux
boxes, so when windows problems crop up, I more or less rely on others to
debug.

> Another thing I should probably add is that I copied my 0.2.1 db over to=
=20
> the new install of 0.2.3; I don't see any changes to its structure, or to=
=20
> admin.php between 0.2.1 and 0.2.3, though. Also, I am able to reproduce t=
he=20
> same problem in the 0.2.1 install... and since I can sometimes get it to=
=20
> take the changes by refreshing the page, I don't think that's it.

There were some minor changes.  In the etc/ directory is a 0.2.2-0.2.3.sql
file; you need to apply that to your database.  That's not going to solve
this problem, I don't think, but you should do it.

> 3) Style sheet and BODY statement conflicting?
>=20
> I did notice that the style sheet over-rides the settings, anyway (mostly=
=20
> playing with the font sizes above), if I am set up correctly (initially, =
I=20
> don't think I was, and it was not finding the style sheet). My initial=20
> setting had 'webroot' set to the actual www web root, which is a director=
y=20
> above the location of the Anthill install. I changed webroot to the webro=
ot=20
> including path info, and thought I had blown something out (hehe), becaus=
e=20
> suddenly, everything was darkcyan. It's a VirtualHost now ;-), lol. I=20
> caught it when I fixed the emails, and the link was wrong.
>=20
> I believe this is expected behavior in IE, that style sheets supercede BO=
DY=20
> tags, but I could be wrong. They don't supercede FONT tag containers,=20
> methinks. Gotta install Mozilla to see... Although, since they are=20
> conflicting concepts, maybe a way to turn styles on and off, which, when=
=20
> on, disables the font options on the admin screen (and removes them from=
=20
> the body tag) in a future version?

All of this is going to change in 0.3.  I'm going to move to more
customizable template format and css will be used to control almost all
design aspects (ie. instead of changing stuff in the admin page, you'll have
to change the css).  I may make a "css designer" type control so you can
generate css changes from within Anthill and not have to modify the css by
hand.  But that likely won't be happening for quite a while.

> Maybe this is already on the change list for 0.2.4pre. If so, forgive me!=
 I=20
> seem to remember it (the changelog) mentioning something somewhere on the=
=20
> styles. I'm not beta testing, but evaluating the production version, for=
=20
> use. So, I am skeptical of using a pre-release for testing.

0.2.4pre is pretty much 0.2.4 final without some fixes I've committed
recently.  Anthill is pretty stable no matter how you cut it, and the added
features to 0.2.4 have been pretty small or subtle.  Using 0.2.4pre should
be fine.  I would have releaed it by now except that I've been waiting for
language translations (the spanish, italian, and french translations aren't
up to date and I'm hoping someone steps forward to update them).

> 4) Just a question...
>=20
> In the email(s) when changes are made to a bug report, I am getting some=
=20
> weird text that includes path info in it:
>=20
> --- 2	2003-09-25 17:17:34.000000000 -0700
> +++ C:/OpenSA/Apache/htdocs/cc/ICS/shadow\2.t562.tmp	1969-12-31=20
> 16:00:00.000000000 -0800
> @@ -1,43 +0,0 @@
>=20
> I looked at the .tmp file directly, and it isn't in there, and the=20
> date/time above is just wacky. Is this expected behavior with the windiff=
 I=20
> downloaded from the main site? Can I get rid of it? I don't see where the=
=20
> code is adding it right off...

Hmmmm...  not sure.  I don't use windiff here.  The date is definitely
wrong, I don't know what that is from.  It also shouldn't be putting the
full path to the temporary file like that, but I suspect that's due to my
regular expression.  It should look like:

--- 2
+++ 2.tmp

Can you tell me what the value of $_CONF['shadowdir'] is?  Do you use
forward slashes or backslashes?  Can you try applying this patch to
functions.inc.php and see if the full path for the temporary file
disappears?  It really shouldn't be there.


--- functions.inc.php.org	2003-09-26 16:38:59.000000000 -0600
+++ functions.inc.php	2003-09-26 16:39:57.000000000 -0600
@@ -662,9 +662,8 @@
     }
=20
     // create the diff
-    $stmpfile =3D ereg_replace($_CONF['shadowdir'] . "/","",$tmpfile);
     $antdiff  =3D tempnam($_CONF['tmpdir'],"ant-diff.");
-    $cmd      =3D sprintf("%s -ruN %s/$bugnum %s/$stmpfile",$_CONF['diffto=
ol'],$_CONF['shadowdir'],$_CONF['shadowdir']);
+    $cmd      =3D sprintf("%s -ruN %s/$bugnum %s",$_CONF['difftool'],$tmpf=
ile,$_CONF['shadowdir']);
     $output   =3D exec($cmd,$diffarray,$return);
     if ($return =3D=3D 1) { // there are differences
       $x =3D implode($diffarray,"\n");

> I hope I am not overstepping myself here, just wanted to throw out some=
=20
> things I unearthed. The real big issue here is the admin.php not working.=
 I=20
> can manually go into the db through an administrative tool for MySQL I am=
=20
> using and make changes, but I'd really like to nail this down.

You and me both.  While I'm not developing for windows and never initially
meant for it to run on windows, I know a fair amount of people are using it.
Which is what kinda wierds me out here... this is the first I've heard about
this kind of problem.

> On an aside, the fix for the query screen in IE is super! Much better (I=
=20
> started out yesterday with 0.2.1).

=3D)

--=20
MandrakeSoft Security; http://www.mandrakesecure.net/
Online Security Resource Book; http://linsec.ca/
"lynx -source http://linsec.ca/vdanen.asc | gpg --import"
{FE6F2AFD : 88D8 0D23 8D4B 3407 5BD7  66F9 2043 D0E5 FE6F 2AFD}


--wxQIpZ+KqQj4wYAN
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (Darwin)

iD8DBQE/dMDVIEPQ5f5vKv0RAt78AJ90ajIwrNyPt+ph9b2M1G+5lkdWcwCfc3Cl
XkGwt7V68TL7rJ+8uAawOSU=
=piKT
-----END PGP SIGNATURE-----

--wxQIpZ+KqQj4wYAN--


--===============70348000722009418==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Anthill-users mailing list
Anthill-users-+mQDA36h6dWw5LPnMra/[email protected]
http://vmlinuz.ca/mailman/listinfo/anthill-users

--===============70348000722009418==--