stebalien / list-it Goto Github PK

On Mar 1, 11:01 am, Danny <[email protected]> wrote:
Can't seem to do this from within list.it options dialog box (Ubuntu
GNU/Linux, Firefox 3.0.6)

Using the mozillla keyconfig add-on this is
possible:http://forums.mozillazine.org/viewtopic.php?t=72994

Using for F1 now, but that won't work across keyboards (think happy
hacking) so thinking of remapping Caps Lock :)

However note that after doing this, while F1 does work, the list.it
options dialog box still shows the old 2-key combination, which
actually no longer works (firefox was restarted in there so the bug is
persistant).

-Danny

Actually typed too soon here. After restart list.it actually reverted
to using the old 2-key combination :-(

What steps will reproduce the problem?
1. Write a note and save it.
2. Click on the note, change it in some way.
3. Close the sidebar with the hotkey.
4. Re-open the sidebar, changes you made to the note are not saved.

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

> Expected: Note with changes survives closing/reopening sidebar.
> What I see: Changes not saved when closing/reopening sidebar.

Please use labels and text to provide additional information.

Exception "Node Not Found" occurs when notes are double-clicked 

feature request from user

"One thing, though, that I wish it could do (or maybe it does and I
haven't found out) is automatically append the URL of the page one is
viewing while adding the note.  This, of course, is not always
desired, so maybe the URL would be appended to the note if the user
pressed ALT+Enter instead of SHIFT+Enter."

What steps will reproduce the problem?
1. Use a skin
2. Open list.it
3. Everything fails

What is the expected output? What do you see instead?
The background is white, should be gray. Probably plenty of other bad color
problems too.

We just need to try installing skins and making changes until we find out
what's causing it. I don't know much about how skins work, but I'm guessing
it's just a css conflict.

What steps will reproduce the problem?

0. open listit
1. start a new negative search (e.g., "-hello")
2. save it.
3. try engaging the search

desired effect is not achieved (it searches for the string "-hello" instead

*all versions*

code for displaying detected semantics duplicates entire code blocks from
scroll.js. This makes things brittle.

we also need a better way of turning off/on manuscript illumination

What steps will reproduce the problem?
1. Open bookmarks sidebar (Ctrl-b)
2. Open list-it with hotkey (Shift-Ctrl-f)
3. Close list-it with hotkey (Shift-Ctrl-f)

What is the expected output? What do you see instead?
Bookmarks sidebar status is retained after closing list-it

What version of the product are you using? On what operating system?
list-it 0.4.9.9 on Ubuntu Linux 10.04.2

Please provide any additional information below.

Problem with latest 0.4.9.1 version of list-it and Firefox 4.0b4:

If you launch list-it you will notice no research bar! It's gone. Where'd it 
go?  Showing and hiding the bar from the config page doens't help. It's just 
gone.

I'll look into this when I can.

FEATURE request from a clever user:

why not support anonymous sync without signup for those users who can't be 
bothered to create an account - could still provide backup facilities, such as 
by using their unique client id

AND if they wanted to sync between computers, they would have to sign up

that way we can protect them against data loss

feedback from clintster

I'd love the idea of creating a new note on one computer running
listit, and when I open up my browser on my other computer listit
informs me that new notes have been added from another machine by
popping a little tooltip/notification/changes icon indicating this

David reported to me that the 'X' is difficult to target (and indeed it is!)

It seems that Firefox is being SO clever that when you want to click the X,
you have to actually click on the non-alpha masked pixels.  As a result
it's really really easy to miss.

We should make clicks within N pixels of the center of the X all cause the
X to trigger. 

While a potential danger is that the "X" will become _too_ easy to click, I
do not think this will be the case. 

unsure that this is listit related, but it seems to frequently haunt me..
no visible effects. (ff3.5b4)

Error while parsing contractID: QueryInterface - TypeError:
contractArray[1] is undefined
firefox-bin(281,0xa06ae720) malloc: *** error for object 0x5d64:
Non-aligned pointer being freed
*** set a breakpoint in malloc_error_break to debug

Error while parsing contractID: QueryInterface - TypeError:
contractArray[1] is undefined

Getting service...
Debugger() was called!

users would like to be able to:
    * print selected notes to a file;
    * print selected notes on separate pages; or,
    * highlight individual notes on the pages of notes about to be printed (the print menu for my printer denies me access to the pages).

Demand for supporting spell check in note creation fields

I think we can just use this
https://developer.mozilla.org/en/Using_spell_checking_in_XUL

The dustbin seems to have broken.

Steps to reproduce:

1. open list-it
2. click dustbin.

Versions affected: latest SVN.
Bug level: Release-blocking (Critical)

Currently we use preferences to store synchronize usernames and passwords.
 Using Firefox's nsIPasswordManager to do this would probably be better.

We should add a button in the dustbin to allow users to permanently delete
all of their notes from the dustbin.

Firefox wouldn't start when I had wireless turned off. I got this message
in the terminal (the end is probably the interesting part). It was begin to
startup, but then crash. I think before the window even started to open.

-------> listit component startup() <---------
Enabling context.
Scheduler init........
CMStore init {"filename":"listit-lifelog.sqlite","debug":false,}
CREATE TABLE IF NOT EXISTS entities(id TEXT, type TEXT, data TEXT);
entity store initialized : [object Object]
CMStore init {"filename":"listit-lifelog.sqlite","debug":false,}
CREATE TABLE IF NOT EXISTS event(type TEXT, start NUMERIC, end NUMERIC,
entityid TEXT, entitytype TEXT, data TEXT);
CREATE INDEX IF NOT EXISTS startend_index ON event ( start, end);
CMStore init {"filename":"listit-lifelog.sqlite","debug":false,}
CREATE TABLE IF NOT EXISTS entity_aliases(alias TEXT, entityid TEXT,
entitytype TEXT);
Setting commit interval to 30000
CMS is [object Object]
WatchMe.initialize
... WatchMe starting ...
watchme initit .params is [object Object]
watchmecontexprovider initialized
Initializing Oil Location services
nsOilLocation::Init() url_base=http://oil.nrcc.noklab.com:8080/api
db_filename=/tmp/oil.db
Localizer using server URL http://oil.nrcc.noklab.com:8080/api
enabled contextmaiden scheduler.[object Object]
Enabling analysis features
plumdate is [object Object]
plumdate.init---->
::initializing plumdate::[object Object]
Running initOverlays
OilOverlay constructor for document [object XULDocument]
sqlite3_open_v2 returns 14
localizer.cpp:236 rc=0 locdb=0x1a9ce108 sql=create table if not exists
version (         id integer PRIMARY KEY AUTOINCREMENT,       name
varchar(17) UNIQUE NOT NULL,       version integer,        modified
datetime default CURRENT_TIMESTAMP);
oilworker ClientMac=00:17:f2:e6:78:fa
dyld: lazy symbol binding failed: Symbol not found: _fputs$UNIX2003
  Referenced from:
/Users/Tortilla/list-it/src/firefox/listit/components/nsOilLocation.dylib
  Expected in: /usr/lib/libSystem.B.dylib

dyld: Symbol not found: _fputs$UNIX2003
  Referenced from:
/Users/Tortilla/list-it/src/firefox/listit/components/nsOilLocation.dylib
  Expected in: /usr/lib/libSystem.B.dylib

If you clear your notebook completely and start list-it, the creation of
new notes fails with the error

TypeError:before is null

in the error console

Opening list.it 0.4.0b5 is slow, causing script timeout errors on some
computers and taking an unacceptably long time on others.

I think this is caused by insertNote doing redundant things (e.g.,
re-applying .sortability()) for batch inserts in the DOM.

This is very important, critical to releasing 0.4.0

We've got some requests for a video tutorial on how typical people use
list.it to manage their to-dos.

This is on our to-do list for 0.4.0 release!

Note orderings are not in the server schema, and therefore are not
replicated across computers during a sync.

This is a major problem, because users will think that notes are missing.

Many users are not using FF 3

(An old issue that was never explicitly added here i believe G said he
could address)

The researchbar should be made to expand automatically to accommodate more
entries than can fit on one line.

Ideally it should start with height=0 (so that the extra space isn't
wasted) and expand to accommodate the first line, etc.

Support for skins! Like Ubiquity.

I think we could benefit from having a limited number of backgrounds that
we know look good to give to our users, but also let the creative ones
create their own.

Not sure how to do this yet; will take a look.

Firefox starts very quickly while list.it is installed and Disabled, but
when enabled the start-up time is about 5 seconds.  During this time and
for a second or so afterwards CPU usage is at maximum - much higher and for
longer than when list.it is disabled.

I'm Win32 (XP SP3), Firefox 3.6, List.it 0.4.8.3.3

report from team leader

"I ran into a fatal flaw caused by the hidden window---when I "closed"
firefox, I was not closing the hidden window, so firefox thought I was
only closing a window, so it forgot all my open tabs (which it remembers
if you close the last window)."

Firefox notifies me of update to version 0.4.8.8 and then to 0.4.8-8 and
then to 0.4.8.8 amd then 0.4.8-8 and so on ad infinitum.

This seems to be related to the change in r2128

I'm running Firefox 3.6.

I am really not sure why this is happening but for some users it seems that
their Firefox is configured to give a user agent corresponding to Firefox
2.  This tricks our script and makes us mistakenly tell people that they
are not running FF3.

I do not know what plugin/feature users are using to get their FFs to act
like FF 2's but it would be nice to be able to correctly detect
compatibility with their browser.

Very often when entering notes, you have notes for a project or for
specific action... You want to remove from the search such words or
specific project (overloading the search result). My idea was to use a "-"
prefix to remove a word from the search. But this doesn't work in version
0.4.0r1.

pasting a url into list.it like:
http://www.nytimes.com/2009/05/28/technology/personaltech/28basics.html

gets caught by the url recognizer and the date recognizer, which ends up
producing this:
<a semdate__w0zz2xm39dlc="" semantics_date="" href="<span class="
onclick="openLink(event);">http://www.nytimes.com/2009/05/28/technology/personal
tech/28basics.html"><span
class="semantics_date semdate__w0zz2xm39dlc" style="color: rgb(255, 255,
255);">http://www.nytimes.com/2009/05/28/technology/personaltech/28basics.html</
span></a>

which becomes unclickable.

If you want to copy/paste text from your notes, you have to now select to
_Edit_ it and then copy stuff out.

Add a "handle" so that drag and drop is restricted to just the left hand
side of the note.

See http://docs.jquery.com/UI/Sortable -> .handle property

Something is wrong with the way the background of the researchbar is
rendered in FF3.5

We should investigate as to whether this is our bug or Mozilla's.

What steps will reproduce the problem?
1. engaged filter "todo"
2. added a note
3. engaged filter "foo"
4. added a note
5. crash! "m" is not defined

I'd love some way to keep a note on top, for instance, the todo list that
I'm currently working on.  Way easier than re-dragging it to the top :)

steps to reproduce:

0. open listit
1. perform a search with a new tag e.g., "test"
2. add it to the research bar (+)
3. toggle it on 
4. delete "test" while it's toggled on
5. it's still on!!

Synchronization still sometimes hangs the client for >200ms.  If we offset
sync to a web worker (to be introduced in FF3.5) we can probably avoid this
stuttering effect and make people happier.

Many people want to be able to easily print their notes.
We should support this.

Many people have requested the ability to drag notes to re-arrange their
order.  We are currently working on designs to make this feasible.

Expected roadmap: 0.5 

Currently all communications between client and server are encrypted (via
SSL) but notes are not stored in an encrypted form in either client (sqlite
store) nor server. 

The proposal is to encrypt/decrypt note contents on the fly so that it is
stored no where on disk in unencrypted form.  

This is an important enhancement because many people keep sensitive
information in their notes. 

We have had this on our TODO list for months, but it has been postponed due
to manpower shortage.

What steps will reproduce the problem?
1. open list-it
2. close/kill firefox
3. open/launch firefox
4. attempt to use open/close hotkey

What is the expected output? What do you see instead?
Should respond on the first try. But it only responds on the second try.

Please use labels and text to provide additional information.

See
http://groups.google.com/group/list-it/browse_thread/thread/b4d8edc066d0ff2c

Too many re-search buttons overflow the research bar.

Maybe gvargas could look into this for me?

Input bar should support Shift-Enter or Enter according to the user's
preferences.  Also I think this means that the quick input bar could grow
on demand to support multi-line input.

User suggestion:

Opening and closing list.it insufficiently discoverable.  Right now the
current mechanisms (View -> Sidebar -> list.it) and hotkeys (which are
completely invisible) do not achieve our goal of making it easy to open and
close.

Suggested solution:
Make the little note icon open the listit sidebar, and right-click to open
the note input pane.

Alernatives?

What steps will reproduce the problem?
1. Using beta xpi
2. Restarting Firefox
3. (Then it doesn't work)

What is the expected output? What do you see instead?
I expect to see list.it working!  Instead, it's this blank sidebar :(

Please use labels and text to provide additional information.

Given the lack of support for any operations dealing with hidden chrome
windows, our hack to make sure the singleton listit components [model and
database routines] don't get crushed by first window destruction have
proven futile.

I think that our only hope at this point is to convert the core listit bits
to a js XPCOM component.  This would involve severing all ties to objects
grounded in the window.* frame, including

setTimeout/setInterval/clearTimeout/clearInterval
XMLHttpRequest
netscape.*
Components ? [presumably we can get ahold of this another way]
alert (we dont use this anyway)

IF anyone felt brave about geting a start on this that would be great

Max

UPDATE this bug list, dudes.

It's way obsolete.

Max

From pmman in the google group:

Please consider adding an "auto-save" feature which will automatically
save a new note if the user forgets to hit 'save'. I lost an important
note yesterday when my PC crashed before I hit 'save'.


gv : Good idea, we should look into an easy way to keep this from
happening. It's easy to leave a note open for a long time without realizing
that it hasn't been saved and then lose it because of an unrelated firefox
crash.

when you delete a re-search, it doesn't reset list-it's search mode