ids-devel Mailing List for Image Display System (Page 13)

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Hi, 
Here is the Portuguese (not Brasilian) traslation file for IDS.

Thanks.
Bruno Tavares

>    Personally, I don't use it, however I know of many that do use it, so
>removing it would be a step back for those folks that do enjoy its
>advantages.

I wasn't suggesting that embedded comments be abandoned - just the
futile attempt to use image::info to extract them.

"Anthony A. D. Talltree" wrote:

> I tried that.  No response whatsoever.  Since the API is fundamentally
> flawed, IMHO it should just be abandoned.

    Personally, I don't use it, however I know of many that do use it, so removing it would be a step back for those folks that do enjoy its advantages.  I'll look into it and maybe write a wrapper around its shortcomings as a last resort.  Thanks for bringing this up.

--
H | "Life is the art of drawing without an eraser." - John Gardner
  +--------------------------------------------------------------------
  Ashley M. Kirchner <mailto:as...@pc...>   .   303.442.6410 x130
  Director of Internet Operations / SysAdmin    .     800.441.3873 x130
  Photo Craft Laboratories, Inc.            .     3550 Arapahoe Ave, #6
  http://www.pcraft.com ..... .  .    .       Boulder, CO 80303, U.S.A.

>    Image::Info wasn't written by us, however you can send email to Gisle Aas <g
>is...@Ac...>, the author.  Perhaps you can get them to fix the problems
> with it?

I tried that.  No response whatsoever.  Since the API is fundamentally 
flawed, IMHO it should just be abandoned.

"Anthony A. D. Talltree" wrote:

> I hope that doing *something* about image_info is in there.  Whoever
> implemented it was insane.  It returns a different type for Comment if
> there's one line or multiple.  Sometimes it doesn't see embedded
> comments at all.  I eventually gave up on it and just slapped in

    Image::Info wasn't written by us, however you can send email to Gisle Aas <gi...@Ac...>, the author.  Perhaps you can get them to fix the problems with it?

--
H | "Life is the art of drawing without an eraser." - John Gardner
  +--------------------------------------------------------------------
  Ashley M. Kirchner <mailto:as...@pc...>   .   303.442.6410 x130
  Director of Internet Operations / SysAdmin    .     800.441.3873 x130
  Photo Craft Laboratories, Inc.            .     3550 Arapahoe Ave, #6
  http://www.pcraft.com ..... .  .    .       Boulder, CO 80303, U.S.A.

Paul Delano wrote:

> I got some complaints from people that they couldn't see the images. It appears that the encodeSpecialChars() is too agressive in replacing characters in image names. Specifically, it appears that AOL browsers have trouble with encoded "-" and "_".

    AOL browsers are like really low on my priority list, but nevertheless, it needs to work over a big variety of browsers.  I will incorporate this patch together with all the other ones still pending.  Thank you.

    A call to everyone else on the list: Over the past several months that we have not had any code release, I'm sure many of you may have done your own patches and or modifications to your 0.81 source tree.  I'd like to see what people have been doing, what are the features that people have added to their installations, errors that have been patched up locally and others.  What I'm planning on doing is the following (in this order):

    - Apply the current set of (waiting) patches that have been
      received, and whatever new bug fixes people submit.
    - If there are features that seem to have been incorporated
      several times on custom sites, I will see that they get into
      the current source tree
    - Release another tarball, possibly 0.82 with all the new stuff
    - Implement a code-freeze on 0.8x - this basically means I will
      continue to fix bugs, but will not implement any new features.

    New features will start getting added to the list of the rewrite.  The rewrite itself is currently stagnant.  I need to catch up on Apache::PageKit first before I can dive into John's code (I'm more of a HTML::Mason guy).  So for the time being, let's hold on to 0.8x ...

--
H | "Life is the art of drawing without an eraser." - John Gardner
  +--------------------------------------------------------------------
  Ashley M. Kirchner <mailto:as...@pc...>   .   303.442.6410 x130
  Director of Internet Operations / SysAdmin    .     800.441.3873 x130
  Photo Craft Laboratories, Inc.            .     3550 Arapahoe Ave, #6
  http://www.pcraft.com ..... .  .    .       Boulder, CO 80303, U.S.A.

I came across this bug yesterday. I have a enormous colection of albums,
but i want people to be able to see only some of them in IDS. So, in the
albums directorie i've created symbolic links (im using linux) to the
albums that i wanted available in IDS. So far so good. 
The proble is the search functions, it does not follow symbolic links,
so it is not possible to search comments or filenames.

I fixed this by changing the line 906 of the index.cgi from:

find (\&searchForFiles, "./albums/");

to 

find ({wanted => &searchForFiles, follow_fast=>1}, "./albums/");

this method makes the File::Find::find to follow the sym links without
performance lost and without no impact to windows servers.

(sorry for my bad english)
Thank You

Bruno Tavares

Hi,

I've got a digital camera and am (as everyone else) looking for software to
generate galleries, both dynamically and statically.  Ids seems to have
most of the features I need.

A complete rewrite is mentioned as a long term goal, what is the status on
this rewrite?  How will it differ from the current implementation?  Is
there a design description, however short, anywhere?  I've done some perl
programming, and would be interested in filling in some parts if there is a
(somewhat) cohorent design thought out.

-- René

On Fri, 2002-03-29 at 19:40, Techwolf wrote:
> This is because Redhat compiles apache with the configure options of 
> --suexec-docroot=/var/www/cgi-bin. I've change it to 
> --suexec-docroot=/home  and rebuild it so I could give users cgi with 
> group/owner id if allowed in httpd.conf, such as IDS. :-)

Thanks for the info - I had assumed it was apache, not RH's build of
Apache.

> Now that I am thinking about it...I may be talked into writing a small 
> bash/shell script that will do it all for fixing the suexec problem for 
> Redhat systems.

Any idea what this would look like?  I've got the problem that at home I
can twiddle configs and move server roots around, but my external host
is out of my hands.  (They're running RH6.2, so for the time being it
shouldn't be an issue)  It would be ice to at least know how to get
around the problem if they ever upgrade.

Thanks,
Tad

nicole wrote:

> Anyone know how to get off this mailing list?? I cant find it on the site.

    Check the headers of each message, as well as the footer banner.

    Header:

    List-Help: <mailto:ids...@li...?subject=help>
    List-Post: <mailto:ids...@li...>
    List-Subscribe: <https://lists.sourceforge.net/lists/listinfo/ids-devel>,
            <mailto:ids...@li...?subject=subscribe>
    List-Id: <ids-devel.lists.sourceforge.net>
    List-Unsubscribe: <https://lists.sourceforge.net/lists/listinfo/ids-devel>,
            <mailto:ids...@li...?subject=unsubscribe>
    List-Archive: <http://www.geocrawler.com/redir-sf.php3?list=ids-devel>

    Footer:

    IDS-devel mailing list
    IDS...@li...
    https://lists.sourceforge.net/lists/listinfo/ids-devel

    (Note about the above URL: you need to scroll all the way to the bottom, enter your email address and click on 'Edit Optinos' - that will then take you to the Unsubscribe option)

--
H | "Life is the art of drawing without an eraser." - John Gardner
  +--------------------------------------------------------------------
  Ashley M. Kirchner <mailto:as...@pc...>   .   303.442.6410 x130
  Director of Internet Operations / SysAdmin    .     800.441.3873 x130
  Photo Craft Laboratories, Inc.            .     3550 Arapahoe Ave, #6
  http://www.pcraft.com ..... .  .    .       Boulder, CO 80303, U.S.A.

Anyone know how to get off this mailing list?? I cant find it on the site.

----- Original Message -----
From: "Techwolf" <Tec...@at...>
To: <ids...@li...>
Sent: Friday, March 29, 2002 5:40 PM
Subject: Re: [Ids-devel] Problems on RH7.2?

> At 01:47 PM 3/28/2002 -0500, Tad Truex wrote:
> >http://test.example.com/ids/  ->  /home/idsuser/www/ids (test was rooted
at
> >/home/idsuser/www)  This failed with permission problems.
> >
> >I reconfigure my server to serve
> >http://ids.example.com/ -> /var/www/html/ids  (now the real server is
> >rooted at
> >/var/www/html/ and the script being suexec'ed is one level below the real
root
> >and at the toplevel of the virutal server).  Everything works fine.
>
> This is because Redhat compiles apache with the configure options of
> --suexec-docroot=/var/www/cgi-bin. I've change it to
> --suexec-docroot=/home  and rebuild it so I could give users cgi with
> group/owner id if allowed in httpd.conf, such as IDS. :-)
>
> If anyone wants details on how to do rpm --rebuilds with different config
> options, send me an e-mail to Tec...@at... or ask here. :-)
>
> If the new IDS that being worked on come out with the feature of multiable
> users, I may be willing to create a rpm package for IDS sence I have
become
> quiet familer with rpm package management.
>
> Now that I am thinking about it...I may be talked into writing a small
> bash/shell script that will do it all for fixing the suexec problem for
> Redhat systems.
>
>
> Techwolf
> Tec...@at...
> http://www.techwolf.net/index.html
> PGP public keys on web site.
>
>
> _______________________________________________
> IDS-devel mailing list
> IDS...@li...
> https://lists.sourceforge.net/lists/listinfo/ids-devel
>

At 01:47 PM 3/28/2002 -0500, Tad Truex wrote:
>http://test.example.com/ids/  ->  /home/idsuser/www/ids (test was rooted at
>/home/idsuser/www)  This failed with permission problems.
>
>I reconfigure my server to serve
>http://ids.example.com/ -> /var/www/html/ids  (now the real server is 
>rooted at
>/var/www/html/ and the script being suexec'ed is one level below the real root
>and at the toplevel of the virutal server).  Everything works fine.

This is because Redhat compiles apache with the configure options of 
--suexec-docroot=/var/www/cgi-bin. I've change it to 
--suexec-docroot=/home  and rebuild it so I could give users cgi with 
group/owner id if allowed in httpd.conf, such as IDS. :-)

If anyone wants details on how to do rpm --rebuilds with different config 
options, send me an e-mail to Tec...@at... or ask here. :-)

If the new IDS that being worked on come out with the feature of multiable 
users, I may be willing to create a rpm package for IDS sence I have become 
quiet familer with rpm package management.

Now that I am thinking about it...I may be talked into writing a small 
bash/shell script that will do it all for fixing the suexec problem for 
Redhat systems.

Techwolf
Tec...@at...
http://www.techwolf.net/index.html
PGP public keys on web site.

> require script execution from the document root (this would probably be counterproductive to security).
> 

I'm not sure I have completely interpreted the apache page correctly.  I think
you are right that the limitations are not quite so severe for regular cgi
scripts (outside the web directory), but my understanding for things like 
index.cgi, is that they are limited - here's the bit that causes my confusion
(from <http://httpd.apache.org/docs/suexec.html>)

# For  security  and  efficiency  reasons,  all  suexec requests must
#      remain  within  either  a  top-level document root for virtual host
#      requests,  or  one  top-level  personal  document  root for userdir
#      requests.  For  example,  if you have four VirtualHosts configured,
#      you  would need to structure all of your VHosts' document roots off
#      of  one  main Apache document hierarchy to take advantage of suEXEC
#      for VirtualHosts.

I haven't fiddled around with disabling the apache suexec mechanism and 
using the good old +s mode, but I suspect that apache will get all grumpy about
that too...

/Tad

On Thu, Mar 28, 2002 at 01:47:26PM -0500, Tad Truex wrote:

> http://test.example.com/ids/  ->  /home/idsuser/www/ids (test was rooted at
> /home/idsuser/www)  This failed with permission problems.
> 
> I reconfigure my server to serve
> http://ids.example.com/ -> /var/www/html/ids  (now the real server is rooted at
> /var/www/html/ and the script being suexec'ed is one level below the real root
> and at the toplevel of the virutal server).  Everything works fine.

	Remember that the server root is hard-coded into suexec which could explain why it would not
execute scripts in /home/idsuser but could in /var/www .  In addition, I have not found suexec to
require script execution from the document root (this would probably be counterproductive to security).

Jeff

> I'm not sure why Apache has made the directory depth restictions - seems
>somewhat arbitrary to me...

I've never seen such a restriction, but then, I'd never use a
vendor-supplied Apache.  One never knows what goofy args and directories
they compile in.

My understanding of suexec is that all affected cgi's must live under a
single directory, to which suexec permission is granted.  In my case I
gave in and just created a /home/websites, under which I have
www.foo.bar, www.baz.glarch, and which I suexec enabled.  I struggled
with the mess for a while, and this is what finally worked.

It's always seemed weird to me that IDS out of the box wants a wholly
publically-writeable tree to live in.

> 
>     This means those directories aren't writable by the owner of the 
> process.  If you're running suexec, then the IDS process is theoretically i
> owned by that user, which means the directories should also be owned by 
> it.  Otherwise, those directories need to be owned by the web daemon.
> 

This is close to the answer.  In fact, the problem is that Apache's suexec
wrapper places some (in my opinion odd) restrictions on where the script 
can be located.  For a script running on a virtual server in a web a
accessible directroy, the virtual server root must be located no more
than one directory level from the top of the real server's tree.  Furthermore,
the script must be at the top level of the virtual server's root.  In
my case I had

http://test.example.com/ids/  ->  /home/idsuser/www/ids (test was rooted at
/home/idsuser/www)  This failed with permission problems.

I reconfigure my server to serve
http://ids.example.com/ -> /var/www/html/ids  (now the real server is rooted at
/var/www/html/ and the script being suexec'ed is one level below the real root
and at the toplevel of the virutal server).  Everything works fine.

I am sure there is some other way to fix this, but it seems to work for now.

My guess is that the suexec mechanism changed sometime after RedHat 6.2 
(which was the last time this sort of thing worked for me.)  I'm not sure why
Apache has made the directory depth restictions - seems somewhat arbitrary to me...
Oh well.

Thakns for the input,
Tad

Tad Truex wrote:

> Could not write a description (album-data/Sample Album/album_image.txt)
> Couldn't create temporary file: (Permission denied) at /idsShared.pm line 226.
> Couldn't unlink album-data/Sample Album/album_image.txt.21979-206 .
> No such file or directory at /idsShared.pm line 233.

    This means those directories aren't writable by the owner of the process.  If you're running suexec, then the IDS process is theoretically owned by that user, which means the directories should also be owned by it.  Otherwise, those directories need to be owned by the web daemon.

--
H | "Life is the art of drawing without an eraser." - John Gardner
  +--------------------------------------------------------------------
  Ashley M. Kirchner <mailto:as...@pc...>   .   303.442.6410 x130
  Director of Internet Operations / SysAdmin    .     800.441.3873 x130
  Photo Craft Laboratories, Inc.            .     3550 Arapahoe Ave, #6
  http://www.pcraft.com ..... .  .    .       Boulder, CO 80303, U.S.A.

I'm not sure this is the correct place for this, as I'm pretty sure it is a 
RedHat specific issue, but I'm having trouble with directory permissions on 
a stock RH7.2 installation.  I get the following errors in my error log
when running version 0.81.

Could not write a description (album-data/Sample Album/album_image.txt)
Couldn't create temporary file: (Permission denied) at /idsShared.pm line 226.
Couldn't unlink album-data/Sample Album/album_image.txt.21979-206 .
No such file or directory at /idsShared.pm line 233.

My guess is that this has something to do with the suexec mechanism but I'm
not entirely sure.  Any thoughts?

Thanks,
Tad

I'm not enough of a Perl hacker to fix this, but IDS has at least two
problems with handling parentheses. One is in the name of a comment
poster. A poster used ":)" in his name and it caused the "Show recent
comments" function to completely break. Try it.

Another (bigger) problem is filenames with parentheses. You cannot rename
them through the IDS admin interface. Try it.

Hopefully someone can come up with the fix for this (as well as confirm my
suspicions)... Seems to be a pretty big problem.

Thanks!

Moshe

-- 
Moshe Jacobson
http://runslinux.net
moshe at runslinux dot net
404-206-4060 -=- AIM: Jehsom

Ok, lets start with the bug

On the 'Recent Comments' page, if the person's name has () brackets in it, 
such as Erin (Maggie), then the comment line is left blank.  This does not 
occur on the picture pages, just on 'Recent Comments'.

Secondly, is there any way to make it so that it would allow users to see 
multpile pages of 'Recent Comments' such as in groups of 10 or 20 so they 
can scroll back through added comments without stopping to check each picture?

Other than that let me just say I love this software!

Doug

Hi All,
	I've just started using this and its seems to be pretty close to 
what I was looking for. I have (at least) two suggestions/questions.
	I tried using the rotate function, however, it wants to modify the 
original image. I'd rather it made a copy in the image-cache directory 
rather than overwrite teh original (I have it setup so that the web server 
can't actually write to the original directory). I looked at the Rotate 
method in the admin/index.cgi file and it looks simple to change it there, 
however, would it be easy to get the ids/index.cgi to pick up that version 
instead of the unrotated original?
	The second thing is that it doesn't seem to display the EXIF 
information from my pictures. You can browse these at;

	http://www.sgowdy.org/cgi-bin/ids/index.cgi

if you want to have a look at some of the pictures (I noticed such a 
request for a similar problem in the archives).

						regards,

						Stephen.

-- 
 /------------------------------------+-------------------------\
|Stephen J. Gowdy                     | SLAC, MailStop 17,       |
|http://www.slac.stanford.edu/~gowdy/ | 2575 Sand Hill Road,     |
|                                     | Menlo Park CA 94025, USA |
|EMail: go...@sl...       | Tel: +1 650 926 3144     |
 \------------------------------------+-------------------------/

I figured it out.  Image::Info's image_info has bizarre behavior:

Textual comments found in the file.  The value is a reference to an
array if there are multiple comments found.

Who the heck dreamed up the idea of not always returning the same data
type?  AUGUGGUGUGHGHHHH!!!!!!

Here's a workaround for that design flaw. I have no idea how it managed
to work in the past.

*** index.cgi	Fri Oct  5 16:15:28 2001
--- /new/index.cgi	Thu Jan  3 22:26:32 2002
***************
*** 686,697 ****
  	$imageNameTrimmed =~ s/\.([^.]+)\Z//;
  	$fileExtension = $1;

! 	my $embeddedComments;

  	if ($displayImageData eq 'y') {
  		my($picinfo) = image_info("albums/$albumtodisplay/$imagetodisplay");
  		if (defined($picinfo->{'Comment'})) {
  			$embeddedComments = join("<br />\n", $picinfo->{'Comment'});
  	  	} else {
  	  		#Set the embeddedComments string to null, so we don't have to
keep checking
  			$embeddedComments='<i>'.$localization{'image-noComments'}.'</i>';
--- 686,702 ----
  	$imageNameTrimmed =~ s/\.([^.]+)\Z//;
  	$fileExtension = $1;

! 	my $embeddedComments, $ref, $t;

  	if ($displayImageData eq 'y') {
  		my($picinfo) = image_info("albums/$albumtodisplay/$imagetodisplay");
  		if (defined($picinfo->{'Comment'})) {
+ 		    if ($picinfo->{'Comment'}  =~ /^ARRAY/) {
+ 			$t = $picinfo->{'Comment'};
+ 			$embeddedComments = join("<br />\n", @{$t});
+ 		    } else {
  			$embeddedComments = join("<br />\n", $picinfo->{'Comment'});
+ 		    }
  	  	} else {
  	  		#Set the embeddedComments string to null, so we don't have to
keep checking
  			$embeddedComments='<i>'.$localization{'image-noComments'}.'</i>';

For some reason when viewing an image with an embedded comment, I now get

 ARRAY(0x888ae0) 

instead of the comment.  Ideas?

IDS 0.81 is now available:
http://ids.sourceforge.net/

Changes:
*	Slovak, Swedish, and Japanese (partial) localizations
*	works with ImageMagick 5.40
*	popup menus work both with and without JavaScript
*	file upload should now work on Windows servers
*	more tolerance for corrupt images and non-standard data tags within 
images
*	revised IIS configuration script

Thanks to all of those who contributed.

John

nicole wrote:

> this is a stupid question really.. but when linking to .htpasswd from
> .htaccess what is the form supposed to be? relative or absolute?
> i cant get it to accept my password when i hit the first page (/admin)

    Absolute.  Always.

    My .htaccess file looks like this:

AuthUserFile /full/path/to/hidden/htpasswd
AuthGroupFile /dev/null
AuthName "Gallery Administration"
AuthType Basic

<Limit GET>
require user admin
</Limit>

--
H | "Life is the art of drawing without an eraser." - John Gardner
  +--------------------------------------------------------------------
  Ashley M. Kirchner <mailto:as...@pc...>   .   303.442.6410 x130
  Director of Internet Operations / SysAdmin    .     800.441.3873 x130
  Photo Craft Laboratories, Inc.            .     3550 Arapahoe Ave, #6
  http://www.pcraft.com ..... .  .    .       Boulder, CO 80303, U.S.A.

2001	Jan (2)	Feb (48)	Mar (16)	Apr (14)	May (42)	Jun (36)	Jul (57)	Aug (13)	Sep (2)	Oct (23)	Nov (17)	Dec (3)
2002	Jan (5)	Feb	Mar (10)	Apr (1)	May (6)	Jun	Jul (2)	Aug (1)	Sep (12)	Oct (26)	Nov (11)	Dec (37)
2003	Jan (11)	Feb (27)	Mar (5)	Apr (21)	May (11)	Jun (38)	Jul (8)	Aug (29)	Sep (2)	Oct	Nov (7)	Dec (5)
2004	Jan (5)	Feb (2)	Mar (19)	Apr	May	Jun	Jul	Aug	Sep (3)	Oct	Nov (1)	Dec
2005	Jan	Feb (3)	Mar (3)	Apr (4)	May	Jun	Jul (1)	Aug	Sep	Oct	Nov	Dec
2006	Jan	Feb	Mar	Apr	May	Jun (2)	Jul (2)	Aug (1)	Sep	Oct	Nov	Dec (8)

ids-devel Mailing List for Image Display System (Page 13)

ids-devel — IDS developers use this list