From http://code.google.com/p/zotonic/issues/detail?id=43

When Zotonic resizes images it performs an additional 'sharpen' to compensate for soft images due
to making it smaller or the jpeg compression induced softness.

This sharpen is also done with png images that are (almost) not resize so are still sharp. Especially
when lossless is also specified.

We might want to disable the sharpen when the image is almost not resized or lossless is specified.

From http://code.google.com/p/zotonic/issues/detail?id=44

I install default zotonic blog, then start it under google chrome 4.0.249.43
page didn't load and was killed.

Actually Opera crashes too on some inside pages.

From http://code.google.com/p/zotonic/issues/detail?id=28

Animated GIF are not animated as they should

From http://code.google.com/p/zotonic/issues/detail?id=41

For performance reasons, query resources should be cached.

In the admin edit you should be able to enable/disable caching, and set the
timeout.

From http://code.google.com/p/zotonic/issues/detail?id=26

The resource #1 is essential to the stability of the system. We should prevent it to be deleted.
Either by a special fk constraint in the database or by application logic (or both).

From http://code.google.com/p/zotonic/issues/detail?id=48

New installation
After loading new media I get errors in log:

=INFO REPORT==== 29-Jan-2010::16:24:56 ===
identify of ""/tmp/zp-cfafou.zptmp[0]"" failed - "identify:
NoDecodeDelegateForThisImageFormat /tmp/zp-cfafou.zptmp' @ constitute.c/ReadImage/531." =ERROR REPORT==== 29-Jan-2010::16:24:56 === webmachine error: path="/image/lib/images/mimeicons/application-octet- stream.png%28187x200%29%28crop%29%2846E0BF9691BD339E2C8C6155B0ED8E34%29.jpg " {throw,"unknown result from 'identify': 'identify: NoDecodeDelegateForThisImageFormat /home/houdini/myErlang/zotonic_test_cms/modules/mod_base/lib/images/mimeic
ons/application-octet-stream.png' @ constitute.c/ReadImage/531.'",
[{resource_file_readonly,ensure_preview,3},
{webmachine_resource,resource_call,3},
{webmachine_resource,do,3},
{webmachine_decision_core,decision,3},
{webmachine_decision_core,handle_request,2},
{webmachine_mochiweb,loop,1},
{mochiweb_http,headers,4},
{proc_lib,init_p_do_apply,3}]}

I see no uploaded image. What is this path? /image/lib/images/mimeicons?

From http://code.google.com/p/zotonic/issues/detail?id=8

In both .zip 0.1.0 and current svn version.
Boxes are amd64 debian sid and i386 debian sid.

web:/home/www/zotonic/zotonic# ./start.sh
Erlang R13B02 (erts-5.7.3) [source] [smp:4:4] [rq:4] [async-threads:0]
[kernel-poll:true]

=PROGRESS REPORT==== 18-Nov-2009::12:37:39 ===
supervisor: {local,sasl_safe_sup}
started: [{pid,<0.35.0>},
{name,alarm_handler},
{mfa,{alarm_handler,start_link,[]}},
{restart_type,permanent},
{shutdown,2000},
{child_type,worker}]

=PROGRESS REPORT==== 18-Nov-2009::12:37:39 ===
supervisor: {local,sasl_safe_sup}
started: [{pid,<0.36.0>},
{name,overload},
{mfa,{overload,start_link,[]}},
{restart_type,permanent},
{shutdown,2000},
{child_type,worker}]

=PROGRESS REPORT==== 18-Nov-2009::12:37:39 ===
supervisor: {local,sasl_sup}
started: [{pid,<0.34.0>},
{name,sasl_safe_sup},
{mfa,
{supervisor,start_link,
[{local,sasl_safe_sup},sasl,safe]}},
{restart_type,permanent},
{shutdown,infinity},
{child_type,supervisor}]

=PROGRESS REPORT==== 18-Nov-2009::12:37:39 ===
supervisor: {local,sasl_sup}
started: [{pid,<0.37.0>},
{name,release_handler},
{mfa,{release_handler,start_link,[]}},
{restart_type,permanent},
{shutdown,2000},
{child_type,worker}]

=PROGRESS REPORT==== 18-Nov-2009::12:37:39 ===
application: sasl
started_at: nonode@nohost
{"init terminating in
do_boot",{undef,[{zotonic,start,[]},{init,start_it,1},{init,start_em,1}]}}

Crash dump was written to: erl_crash.dump
init terminating in do_boot ()

oauth_plaintext.erl uses oauth_uri:calate() which does not return a correctly url encoded signature.

I tracked the issue down to here:

On line 87 of oauth_uri.erl:

Value = io_lib:format("%~2.1.0s", [erlang:integer_to_list(C, 16)]),

I changed to this:

Value = io_lib:format("%~s", [erlang:integer_to_list(C, 16)]),

From http://code.google.com/p/zotonic/issues/detail?id=23

I don't know why this is done this way, but shouldn't the 'default' site
just be another site?

From http://code.google.com/p/zotonic/issues/detail?id=11

This is a problem with code:lib_dir/2.

We have to change the "home" directory lookups in some other method, as this is too confusing.

From http://code.google.com/p/zotonic/issues/detail?id=1

In admin section, there is no action "Delete" in list of pages, etc. Only
view/edit, so it take too many clicks to delete each page. Maybe it would be
better to put delete link in list and even make checkboxes and "Delete
selected" link.

From http://code.google.com/p/zotonic/issues/detail?id=6

If I make a page which has a custom page_url, like /contact; and I go to
that URL, the resource_page_path controller redirects me to something like
/page/1232/contact.

Instead, I would like to stay on that page; calling the right resource for
the page that has been resolved.

This solution would use some kind of recursion into the webmachine handling
process.

From http://code.google.com/p/zotonic/issues/detail?id=35

The management of API keys is now a very hidden feature which is probably
broken.

It should be moved to the admin backend, since we don't support logged in
users through the frontend anyway for now.

From http://code.google.com/p/zotonic/issues/detail?id=31

When a zotonic site runs behind a proxy, it usually runs on something like
{host, "localhost:8000"}

Unfortunately, this is the base URL it will use for internal redirects. It should
probably use something like a m.config.site.url setting from the admin
backend, if it's available, and fall back to default value if it isn't.

From http://code.google.com/p/zotonic/issues/detail?id=33

I've seen a couple great and beautiful implementations of Zotonic. (you know who you are).
It would be awesome have a great looking, feature show casing default website. Still the blog
format, but more visual and feature rich, perhaps with the geo-stuff in there and as default /
theme: Amsterdam with our own footage.

From http://code.google.com/p/zotonic/issues/detail?id=36

What steps will reproduce the problem?

1. Open "categories"
2. Drag-and-drop some categories around, changing categories' parents etc.
3. Everything becomes slow, because in the background, a proces is
   recalculating all category numbers for all things in the database.

Solutions:

• do not brute-force change all resources in the database, but only those
  resources which are affected by the operation
• Create a "save" button which will do all drag-and-drop stages in one
  change. After clicking this button, a dialog pops up to give the user a
  clear warning that this slowness will happen.

From http://code.google.com/p/zotonic/issues/detail?id=34

Searches could be saved as rsc resources as well, as a kind of "dynamic
collection".

The page of such a rsc would look like the search result. In the admin
editor of that rsc, you would be able to edit the query terms and the
sorting for it.

From http://code.google.com/p/zotonic/issues/detail?id=16

What steps will reproduce the problem?

{% lib
    "css/blueprint/screen.css"
 %}

What is the expected output? What do you see instead?

Should probably generate something along the lines of
<link href="/lib/css/blueprint__screen~59958273600.css" type="text/css" media="screen"
rel="stylesheet" />

Right now generates
<link href="/lib/css/blueprint/screen~59958273600.css" type="text/css" media="screen"
rel="stylesheet" />

Such a file doesn't exist and causes an error in zotonic itself

What version of the product are you using? On what operating system?

r84d4e97594 on MacOS X 10.6.2

From http://code.google.com/p/zotonic/issues/detail?id=24

What steps will reproduce the problem?

1. Create a news item and save it
2. Create a news item by duplicating the first one
3. Repeat ad infinitum
4. Modify a page template to include prev and next links

What is the expected output? What do you see instead?

The search should easily move between items, whereas it is usually stuck in
displaying just a couple of items

What version of the product are you using? On what operating system?

Zotonic 0.2.0

Please provide any additional information below.

The reason for this is that duplicating an item also duplicates its
publication date. Therefore the search used by next and previous queries
always returns the same set of items regardless of the current item.

Attached is a possible solution to that (it also takes into account the
modification date of an item)

From http://code.google.com/p/zotonic/issues/detail?id=7

Webmachine 1.5 has been released.

It has support for hosts, though not the depth of the support we might need. The most limiting
factor is that wm only accepts a single hostname per dispatch rule.

There is also no implementation of the "redirect to primary host name" feature we have added.

We have to evaluate if and how we can use the 1.5 version.

From http://code.google.com/p/zotonic/issues/detail?id=19

http://code.google.com/p/zotonic/source/browse/src/dbdrivers/postgresql/epgsql_pool/pgsql_pool.erl

Maybe a typo, but:

%% Return the current time in seconds, used for timeouts.
now_secs() ->
{M,S,_M} = erlang:now(),
M_1000 + S. %%%%% <--------- should be M_1000000 + S.

:)

From http://code.google.com/p/zotonic/issues/detail?id=27

Imagemagick has an auto-orient option which can be used to automatically
orient the image.

Pictures like the one attached should be properly oriented (currently
they're not).

From http://code.google.com/p/zotonic/issues/detail?id=13

"visit website" is static link, so it will always redirect to "127.0.0.1:8000" no
matter where is your website.

From http://code.google.com/p/zotonic/issues/detail?id=22

mod_twitter uses the 'mod_twitter' atom as name of the gen_server. This
does not work as it only allows one mod_twitter per zotonic instance,
instead of one per site.

From http://code.google.com/p/zotonic/issues/detail?id=40

I would like Zotonic to have this feature:

A button in the TinyMCE toolbar to create an internal link to another
zotonic page on this site.

The dialog window would have an autocomplete interface so you can easily
find a thing to link to.

From http://code.google.com/p/zotonic/issues/detail?id=30

What steps will reproduce the problem?

1. Create a page, save it
2. Duplicate it

What is the expected output? What do you see instead?

The new page should get a resource_uri.

What version of the product are you using? On what operating system?

Version 0.2, r914f08e03f9383806675c55bda6650e88a598013

Please provide any additional information below.

Only found out that this is happening by examining my the atom feed is
invalid. It uses resource_id to generate ids for feed items, and quite a few
of those are empty

From http://code.google.com/p/zotonic/issues/detail?id=18

It would be nice if mod_development has a "debug mode" button / state that
will reduce the amount of caching that is being done, to speed up development.

From http://code.google.com/p/zotonic/issues/detail?id=9

Line 82 of scomp_base_pager.erl

Now: "\n</ul"
Must be: "\n</ul>"

From http://code.google.com/p/zotonic/issues/detail?id=37

Create an updating mechanism which allows fundamental database changes to
happen between Zotonic releases.

From http://code.google.com/p/zotonic/issues/detail?id=3

What steps will reproduce the problem?

1. Make a directory 'services' under priv/sites/default
2. Make a file service_default_test.erl with appropriate content
3. Try to reach http://127.0.0.1:8000/api/default/test

What is the expected output? What do you see instead?

I should get the output from the API call in JSON, but instead I get a 404.

From http://code.google.com/p/zotonic/issues/detail?id=46

There is a little bit untidy page makeup in the admin interface. Please,
take a look to the attached screenshot.

From http://code.google.com/p/zotonic/issues/detail?id=47

What is modernizr currently used for? Do we really need it?

On Chrome on Linux, modernizr lets the browser crash.

From http://code.google.com/p/zotonic/issues/detail?id=38

It would be nice to create a zipfile with downloadable documentation.

Basically this would be a mirror of the zotonic.com website, especially
everything under /documentation/.

From http://code.google.com/p/zotonic/issues/detail?id=5

What steps will reproduce the problem?

1. Upload a non-image file, like a .tgz or a .txt, using the "add media"
   button in the admin edit page.

What is the expected output? What do you see instead?

The extension of the file will be changed to .bin. This causes browsers not
to display the file correctly.

From http://code.google.com/p/zotonic/issues/detail?id=4

At the moment we use a fixed db connection pool size of 10 connections.
Most web sites don't need that many connections, only on a surge so many connections are needed.

We should make the pool size dynamic. Growing from 2 to a maximum depending on the amount
of connections needed.

When connections are not used over time we should close them again.

From http://code.google.com/p/zotonic/issues/detail?id=10

Can't start zotonic, and can't exacly say why, previously everything was fine.
Try to do this in debian, beside my MacOS, same thing. This is terminal
output right after starting ./start.sh.

From http://code.google.com/p/zotonic/issues/detail?id=29

Currently we have

{% all include name %} to include all templates with name "name" and {% include name %} to include the first "name" found

However, I might want to include just _html_body from mod_seo_google

Something like:

{% include "_html_body" seo_google %}

From http://code.google.com/p/zotonic/issues/detail?id=14

What steps will reproduce the problem?

1. Choose a new item from the search box
2. Drop it on an existing menu item to add it as a submenu

The second-level item does not show up in the menu editor, but it is
visible in the html source of the page.

Inspection with firebug tells that the second level tree is hidden by the
ul ul {display:none}
declaration in zp_admin.css on line 677

I was not digging deeper to see if maybe a 'show' event is missing, but at
least i can't see javascript errors either. Commenting out the
display:none fixes the problem, but possibly breaks something in the
unlinker section where it's declared...

What version of the product are you using? On what operating system?

• latest hg update as of 21.11 (hg identify says 92911c945e88+ tip)
• FF3.5 / Opera 10 / Ubuntu 9.04

Please provide any additional information below.

Meanwhile I'm continuing having fun with zotonic ...

/David

From http://code.google.com/p/zotonic/issues/detail?id=12

It's not very clear what is default data in zotonic and what is data, that could
be deleted. I mean when you come to Pages, toy see a lot of data, and most of
you don't need to remove, or change. It's kind of confusing. I think default
required pages should be separated by style somehow from what is new data,
like news, some new posts, etc.

From http://code.google.com/p/zotonic/issues/detail?id=45

Steps to reproduce:

1. goto http://zotonic.com/admin/
2. insert 'blabla' as login and 'barfoo' as password
3. click 'logon' button

The output is:

500 error
Click for error information.

[{webmachine_resource,do,3},
{webmachine_decision_core,decision,3},
{webmachine_decision_core,handle_request,2},
{webmachine_mochiweb,loop,1},
{mochiweb_http,headers,4},
{proc_lib,init_p_do_apply,3}]

I use Zotonic both 0.3 and 0.4-dev versions. OS: Debian GNU/Linux 'Lenny' 5.x.

From http://code.google.com/p/zotonic/issues/detail?id=2

When I run ./start.sh i get CRASH:

=CRASH REPORT==== 13-Nov-2009::13:46:54 ===
crasher:
initial call: default:init/1
pid: <0.109.0>
registered_name: []
exception exit: {bad_return_value,{error,eacces}}
in function gen_server:init_it/6
ancestors: ['z_module_sup$default',<0.75.0>,z_sites_sup,zotonic_sup,
<0.55.0>]
messages: []
links: [<0.108.0>]
dictionary: []
trap_exit: true
status: running
heap_size: 1597
stack_size: 24
reductions: 1757
neighbours:

From http://code.google.com/p/zotonic/issues/detail?id=39

Hi,

I'm trying to create a module in one of my sites but Zotonic doesn't seem to know to start it
when the server starts. It lives here:

zotonic/priv/sites/default/modules/mod_example.erl

After compiling and running start.sh, I run this in the interpreter:

1> gen_server:call({global, mod_example}, foo).
** exception exit: {noproc,{gen_server,call,[mod_example,foo]}}
in function gen_server:call/2

But if I start the gen_server myself, it works:

2> mod_example:start_link([{context, default}]).
{ok,<0.223.0>}
3> gen_server:call({global, mod_example}, foo).
bar

I've attached mod_example.erl. It's pretty much the bare minimum example described in the
documentation, except start_link is global.

Do I have to do something else to add mod_example to the supervisor tree somehow or is this a
bug?

(It exhibits the same behavior on both 0.2 and trunk (286:8a34b4d359c0))

From http://code.google.com/p/zotonic/issues/detail?id=20

now_msec() ->
{M,S,Micro} = erlang:now(),
M_1000000 + S_1000 + Micro div 1000.

%%% should be: M*1000000000 + S*1000 + Micro div 1000.

in:
zotonic/source/browse/src/support/z_utils.erl

From http://code.google.com/p/zotonic/issues/detail?id=17

While doing a load test we got the following error message. It happens only in max 3-4% of the
requests, and without load it doesn't seem to happen.

I've send a mail to Will Glozer if he can have a look at it as well.

=ERROR REPORT==== 8-Dec-2009::11:49:49 ===
webmachine error: path="/toekenning/50226/vprojazzlive-in-het-bimhuis"
[{erlang,iolist_to_binary,
[{exit,
{{undef,
[{pgsql_connection,querying,
[{parse,[],
"select r.id from rsc r where r.visible_for = 0 and r.is_published and
r.publication_start <= now() and r.publication_end >= now() and r.pivot_category_nr = 44 order
by r.id offset $1 limit $2",
[]},
{<0.21787.0>,#Ref<0.0.3.217355>},
{state,<0.25169.0>,#Port<0.81081>,
[{<<"client_encoding">>,<<"UTF8">>},
{<<"DateStyle">>,<<"ISO, MDY">>},
{<<"integer_datetimes">>,<<"on">>},
{<<"IntervalStyle">>,<<"postgres">>},
{<<"is_superuser">>,<<"on">>},
{<<"server_encoding">>,<<"UTF8">>},
{<<"server_version">>,<<"8.4.1">>},
{<<"session_authorization">>,<<"postgres">>},
{<<"standard_conforming_strings">>,<<"off">>},
{<<"TimeZone">>,<<"Europe/Amsterdam">>}],
true,"mediafonds",undefined,
{<0.21840.0>,#Ref<0.0.3.217310>},
{55342,950732435},
{statement,[],
[{column,<<"id">>,int4,4,-1,1}],
[int8,int8]},
73}]},
{gen_fsm,handle_msg,7},
{proc_lib,init_p_do_apply,3}]},
{gen_fsm,sync_send_event,
[<0.25167.0>,
{parse,[],
"select r.id from rsc r where r.visible_for = 0 and r.is_published and r.publication_start
<= now() and r.publication_end >= now() and r.pivot_category_nr = 44 order by r.id offset $1
limit $2",
[]}]}},
[{gen_fsm,sync_send_event,2},
{pgsql,equery,3},
{z_db,q,3},
{z_search,search_result,3},
{m_search,search,2},
{m_search,m_find_value,3},

{template_mediafonds__users_pmaas_dev_erlang_zotonic_priv_sites_mediafonds_templates_toeg
ekend_toekenning_tpl,
render2,2},

{template_mediafonds__users_pmaas_dev_erlang_zotonic_priv_sites_mediafonds_templates_toeg
ekend_toekenning_tpl,
render,2}]}]},
{webmachine_decision_core,encode_body,1},
{webmachine_decision_core,decision,1},
{webmachine_decision_core,handle_request,2},
{mochiweb_http,headers,4},
{proc_lib,init_p_do_apply,3}]

From http://code.google.com/p/zotonic/issues/detail?id=21

It would be nice to have the name of the current dispatch rule in the template.

If you are using the same template on different dispatch rules, this is
really handy, for instance for the pager.

From http://code.google.com/p/zotonic/issues/detail?id=15

Context windows in site admin did not show scrollbars when resized.

From http://code.google.com/p/zotonic/issues/detail?id=32

There is a rudimentary content importer outside of the repository. It would be awesome to have an
import script for one dimensional content (one model, no relations, just text fields).

Ultimate goal would be perhaps connecting with (for example) to the API of Wordpress to import:

• Articles (including relations to:)
• Tags
• Categories
• Media
• Comments
• Pages

When the ''host' setting in the config contains dots or other "strange" characters then the template compiler will fail with quite non-understandable error messages.

We should add some pre-flight/sanity checks to the config parser and fail with clear messages.

From http://code.google.com/p/zotonic/issues/detail?id=25

It would be nice to be able te create multi-level indented lists in the WYSIWYG
editor

From http://code.google.com/p/zotonic/issues/detail?id=42

It would be nice to have better reporting for the following aspects of a
zotonic system.

This dashboard functionality would encompass the following:

• crashes: which gen_servers have crashed and when (also: live feedback
  when a gen_server crashes using a broadcast message!)
• hits / page requests / stats (integration with google analytics)
• current connections (users)
• an error log per gen_server with crash reports