Re: [maemo-developers] Application Catalog rewrite [was: Abuse on the ApplicationCatalog page]

2006-04-12 Thread Andrew Flegg
On 4/11/06, Michael Wiktowy [EMAIL PROTECTED] wrote:

 That way, once an app is added to the index page, the information on
 there does not need to be updated each time the detail page is updated.
 Makes it easier to maintain.

Indeed, however one thing which I'd like to keep in mind is my
ApplicationCatalog - RSS feed[1] which, although currently unavailable[2],
is relatively easy to put together with the existing design. Having
to check a page for each application for updates would be more of a pain.

For users not using the RSS feed, being able to look at a single page with
the latest version numbers on is useful, I think.

There's also no point in having a page per application in the Maemo wiki:
there's no file hosting abilities so a separate download/homepage will
be required elsewhere, I'd be happy seeing a short, structured index page
like Jussi suggests.

Perhaps if there is a page per app, it could be a comments or
discussion page like Wikipedia's Talk: prefixes. No official stuff
there about how to use the app, but a way for users to discuss it if
necessary.

 Two main groups of indicies should be Maemo Applications (apps that run
 in the 770 itself) and Maemo Support Applications (apps that run on a
 computer that connects to the 770 which has the primary purpose of
 supporting the 770 in some way [snip])

That sounds sensible.

Cheers,

Andrew

[1] http://www.bleb.org/software/770/#rss
[2] Status of the availability of bleb.org at:
http://aflegg.googlepages.com/tv

--
Andrew Flegg -- mailto:[EMAIL PROTECTED]  |  http://www.bleb.org/
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


Re: [maemo-developers] Application Catalog rewrite

2006-04-12 Thread Jussi Kukkonen
Michael Wiktowy wrote:
 I made a mockup of All Maemo Applications at
 https://maemo.org/maemowiki/ApplicationCatalogMockup . Please comment
 and/or modify the page. 
 
 If there is going to be a more detailed page for each app, I would
 replace the Homepage and Download links with a Details link
 pointing to the more detailed application-specific page ...

But this is the catch: some people want to update the info on their own
web page (and want people to visit their page) and I think that's fair.
I propose only creating application-specific pages when they're needed
(when there's no outside resource) -- this also minimizes the work
needed to keep the wiki up to date.

 ... and a Status indicator to let people know if the app is a WIP,
 Alpha, Beta, Stable release.

I considered this, but I thought it too fine grained and subjective:
Besides, we would already have end-user ready / stable / WIP
indication implicitly with the different pages.

 Two main groups of indicies should be Maemo Applications (apps that
 run in the 770 itself) and Maemo Support Applications (apps that run
 on a computer that connects to the 770 which has the primary purpose of
 supporting the 770 in some way ... be it converting media, browser
 plugins, sync tools)

Support Apps on another page is a very good idea.

-- 
Jussi Kukkonen
[EMAIL PROTECTED]
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


Re: [maemo-developers] Application Catalog rewrite [was: Abuse on the ApplicationCatalog page]

2006-04-12 Thread Jussi Kukkonen
Hi,

Santeri Lindgren wrote:
 As i proposed (might be a lot of typing errors), i propose that the
 ApplicationCatalog -page, as other equivalents to this, Wip and
 Wishlist. Should only be a list, maybe categorized more specifically.
 
 Making the list a really long, isn't very nice when using GPRS to browse
  and search one app you know is in the page but you dont' remember the
 homepage or something.
 
 That way, once an app is added to the index page, the information on
 there does not need to be updated each time the detail page is
 updated. Makes it easier to maintain.
 
 
 Yes, this is a good point, but also, if you make a detailed index page,
 it gets cramped.
 
 If you want some details to the index page, something like:
 
 gategory
   | app name| few centences nothing more
 

I've modified https://maemo.org/maemowiki/ApplicationCatalogMockup into
a more minimal look, would that be an acceptable compromise?

Go ahead and make the table continuous if you want, that would make the
page more compact. I'm not sure it would enhance readability, though.

 Then each of those groups could be subdivided into functional groups
 (Network, PIM, etc.) like they are now.
 
 So, three pages?
 
 index - group - application.
 

Probably not what he meant. Currently they're on the same page, but
under different headers. I think this approach is sensible (taking into
account maintainability, ease of search and machine-readability Andrew
mentioned).

-- 
Jussi Kukkonen
[EMAIL PROTECTED]
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


[maemo-developers] Re: Maemo kit development and remote access

2006-04-12 Thread donars guillaume
Hi again,
I will try to be more precise, last time I wasn't at work.

So when I run: af-sb-init.sh start in my ssh client, I see in the
xephyr window the environnement trying to run but it disappeared
immediatly.
And I have this error message:

-


 
  
  
 
  
   
   x-SDK_PC: ~]  af-sb-init.sh start
   Sample files present.
   Starting Maemo Launcher: maemo-launcher start failed.
   Starting DBUS system bus
   Starting D-BUS session bus daemon
   Sapwood image server is already running, doing nothing
   Starting Matchbox window manager
   Keyboard is already running, doing nothing
   root window unavailible (maybe another wm is running?)
   Starting MAEMO AF Desktop
   [sbox-SDK_PC:
~]  maemo_af_desktop[16427]: GLIB CRITICAL ** GLib-GObject -
g_object_set: assertion `G_IS_OBJECT (object)' failed
   maemo_af_desktop[16427]: osso_state_close:282: Unable to close state file: Bad file descriptor
   TRACE LOG: status_bar_main: 1 g_new0
   TRACE LOG: status_bar_main: 2 check panel ptr
   TRACE LOG: status_bar_main: 3: init dock
   TRACE LOG: status_bar_main: 4 add prespecified items
   maemo_af_desktop[16427]:
HildonStatusBarItem: Unable to open plugin sound:
/usr/lib/hildon-status-bar/libsound.so: cannot open shared object file:
No such file or directory
   maemo_af_desktop[16427]: HildonStatusBarItem: Unable to open plugin sound
   maemo_af_desktop[16427]: Item not properly loaded.
This function is only called, if a item couldn't initialise itself properly.
   	See hildon_status_bar_item_init/new for more information
   	maemo_af_desktop[16427]: Statusbar item add failed for sound
   
maemo_af_desktop[16427]: HildonStatusBarItem: Unable to open
plugin internet: /usr/lib/hildon-status-bar/libinternet.so: cannot open
shared object file: No such file or directory
   	maemo_af_desktop[16427]: HildonStatusBarItem: Unable to open plugin internet
   	maemo_af_desktop[16427]: Item not properly loaded.
   
This function is
only called, if a item couldn't initialise itself properly.
   	See hildon_status_bar_item_init/new for more information
   	maemo_af_desktop[16427]: Statusbar item add failed for internet
   
maemo_af_desktop[16427]: HildonStatusBarItem: Unable
to open plugin gateway: /usr/lib/hildon-status-bar/libgateway.so:
cannot open shared object file: No such file or directory
   	maemo_af_desktop[16427]: HildonStatusBarItem: Unable to open plugin gateway
   	maemo_af_desktop[16427]: Item not properly loaded.
   
This
function is only called, if a item couldn't initialise itself properly.
   
See
hildon_status_bar_item_init/new for more information
   	maemo_af_desktop[16427]: Statusbar item add failed for gateway
   
maemo_af_desktop[16427]:
HildonStatusBarItem: Unable to open plugin battery:
/usr/lib/hildon-status-bar/libbattery.so: cannot open shared object
file: No such file or directory
   	maemo_af_desktop[16427]: HildonStatusBarItem: Unable to open plugin battery
   	maemo_af_desktop[16427]: Item not properly loaded.
   
This
function is only called, if a item couldn't initialise itself properly.
   
See
hildon_status_bar_item_init/new for more information
   	maemo_af_desktop[16427]: Statusbar item add failed for battery
   
maemo_af_desktop[16427]:
HildonStatusBarItem: Unable to open plugin display:
/usr/lib/hildon-status-bar/libdisplay.so: cannot open shared object
file: No such file or directory
   	maemo_af_desktop[16427]: HildonStatusBarItem: Unable to open plugin display
   	maemo_af_desktop[16427]: Item not properly loaded.
   
This
function is only called, if a item couldn't initialise itself properly.
   
See
hildon_status_bar_item_init/new for more information
   	maemo_af_desktop[16427]: Statusbar item add failed for display
   	TRACE LOG: status_bar_main: 5 add user items
   	TRACE LOG: status_bar_main: 6 gtk widget show all
   	TRACE LOG: status_bar_main: 7 if rpc...
   	TRACE LOG: status_bar_main: 8, status bar initialized successfully
   
sapwood-server[16289]:
GLIB WARNING ** Gdk - The gdk_draw_*_image require the drawable
argument to
   	have a specified colormap. All windows have a colormap,
   	however, pixmaps only have colormap by default if they
   	were created with a non-NULL window argument. Otherwise
   	a colormap must be set on them with gdk_drawable_set_colormap
   
maemo_af_desktop[16427]:
GLIB WARNING ** default -
qgn_plat_task_navigation_button_03_normal.png: pixmap[1][1]:
gdk_pixmap_foreign_new(83) failed
   	maemo_af_desktop[16427]: GLIB WARNING ** default - Ai
   
maemo_af_desktop[16427]:
GLIB WARNING ** default -
qgn_plat_task_navigation_button_03_normal.png: pixmask[1][1]: no pixmap
   
sapwood-server[16289]:
GLIB WARNING ** Gdk - The gdk_draw_*_image require the drawable
argument to
   	have a specified colormap. All windows have a colormap,
   	however, pixmaps only have colormap by default if they
   	were 

Re: [maemo-developers] Re: Abuse on the ApplicationCatalog page

2006-04-12 Thread Stephen DeGabrielle
Cool!

You can have 'Ported' and 'Polished' package pages.  Sounds good.

I am confused - shouldn't the 'UNpolished' Applications be in the
[annoyingly named]WIP page?

Stephen


On 4/11/06, Tuomas Kuosmanen [EMAIL PROTECTED] wrote:
 What do you guys think of such a refactoring where the *main*
 application catalog would contain software that has been ported a bit
 further than just wow, it runs! Here is a package to try out!


On 4/11/06, Tuomas Kuosmanen [EMAIL PROTECTED] wrote:
 On Wed, 2006-04-05 at 23:05 +0930, ext Stephen DeGabrielle wrote:
  With that in mind, I think it is probably time that the
  ApplicationCatalog was refactored into multiple pages, as it is a
  128kb monster at the moment and painful to edit.  Of course being so
  large is probably discouraging edits a little so maybe it's not a bad
  thing.
 
  I might refactor it myself. (Probably not - so maybe someone else will.)

 What do you guys think of such a refactoring where the *main*
 application catalog would contain software that has been ported a bit
 further than just wow, it runs! Here is a package to try out!

 There's a lot of interesting stuff going on with porting and making
 things run on the device and maybe providing .deb's is an important
 first step. But there is still work to be done to make it fit the
 Maemo desktop and environment, polishing the packaging to make it fit
 the menus, perhaps adding the dbus service stuff etc.

 I am thinking of a rough outline of a process what it takes to fully
 port an application:

   * basic it works! porting work to make it run, if changes are
 even needed in the code itself
   * packaging for the application installer (where to put it in the
 menus, adding the dbus stuff to make the starting foobar...
 -notifications work etc..
   * hildonizing the UI (menus, toolbars etc)
   * Possibly rethinking the user interface even further to make it
 work better in a tablet environment [*]
   * Ideally working with the original maintainer (if not the same
 person) to get maemo-specific changes integrated upstream,
 ideally maintaining the maemo port in the same code tree as
 the original to reduce forking and making sure the port lives
 with the parent project if possible.

 Did I miss something? Does it make sense to you to have this kind of
 raised bar for the main catalog?

 [*] As you might remember I'm here to help you with the user interface
 part, and in many cases it is not *that much* work to get there. :) And
 well, many times this could also result in good ideas for the parent
 application's user interface.

 I think having such a distinction in the app catalog could be good in
 making us more motivated to making better ports? Comments?

 //Tuomas

 --
 A: No
 Q: Should i quote this on the top?

 ___
 maemo-developers mailing list
 maemo-developers@maemo.org
 https://maemo.org/mailman/listinfo/maemo-developers



--

--

Stephen De Gabrielle
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


Re: [maemo-developers] Zapped wtmp...

2006-04-12 Thread Michael Saunby
I don't know about JFFS2 but back in the old days before Linux yuo 
could be pretty sure that a Unix filesystem didn't waste space storing 
null bytes.   Wtmp and other system files (databases?) take advantage 
of this.


Now you've made me aware of this I've deleted /var/log/wtmp.  I doubt 
it was using as much storage as you suggest but if I'd really been 
thinking I'd have done a df before and after - maybe someone else 
will.


There's no need to link it to anything.  If the file isn't there, then 
logging is turned off - at least that's what should happen.


Michael

- Original Message - 
From: Nils Faerber [EMAIL PROTECTED]

To: maemo-developers@maemo.org
Sent: Wednesday, April 12, 2006 3:34 PM
Subject: [maemo-developers] Zapped wtmp...



Hi!
I thought I should let you know...

I am not sure if I really found any cure to anything but after the
latest discussions about the size of wtmp I checked mine on the 770 
and

found a more than 200MB file!

After gzip'ping it it shrank to just 5MB because it contains almost 
only
0s. But there must be the information for 200MB stored somewhere in 
the
JFFS2 so I guess that even if it only has 5MB as a comporessed file 
the

impact on a JFFS2 can be much worse.

So I enabled the RD mode, became root and linked /var/log/wtmp to
/dev/null (who cares for this wtmp on the 770 anyway?).

Before I did this I saw more and more weird phenomena on the device,
like crashing browser and spurious power-offs when case closed 
(almost
100% battery, closed device one night, next morning it was powered 
off

and had to be rebooted!).

Now after I zapped wtmp I have the feeling (!) that it behaves much
better again! I can browse pages that before did not work and I hope
that the power-downs are history now too (for the last three days 
they

are but you never know).

Maybe someone else likes to try to this too and report his/her
experience here.

Probably wtmp should be linked to /dev/null in the next software 
release

too.

BTW: With every closure of the case the wtmp grows by about 3kB. You 
can

estimate with your usage how big it will get...

Cheers
 nils faerber

--
kernel concepts  Tel: +49-271-771091-12
Dreisbachstr. 24 Fax: +49-271-771091-19
D-57250 Netphen  Mob: +49-176-21024535
--
___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers



___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


Re: [maemo-developers] Zapped wtmp...

2006-04-12 Thread Santeri Lindgren

Michael Saunby wrote:
I don't know about JFFS2 but back in the old days before Linux yuo could 
be pretty sure that a Unix filesystem didn't waste space storing null 
bytes.   Wtmp and other system files (databases?) take advantage of this.


Now you've made me aware of this I've deleted /var/log/wtmp.  I doubt it 
was using as much storage as you suggest but if I'd really been thinking 
I'd have done a df before and after - maybe someone else will.


There's no need to link it to anything.  If the file isn't there, then 
logging is turned off - at least that's what should happen.


Wrong, this just deletes everything that is stored prior to this. And 
no, what you assume is wrong. Of course the log files should be 
erasable, and if user deletes the log file, then the system should just 
start making a new one.


Creating the log file if it's not there is the right thing. and turning 
off the logging should be done via config file.


And, my wtmp file was 96.5MB, will be posting if the linking helped 
anything. And yes, i have too found that the system is more sluggish 
than in the beginning of use. Even enabling swap-space on the mmc dont 
have any help anymore.


___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


Re: [maemo-developers] Zapped wtmp...

2006-04-12 Thread Michael P. Lococo
 There's no need to link it to anything.  If the file isn't there, then
logging is turned off - at
 least that's what should happen.

 Wrong, this just deletes everything that is stored prior to this. And
no, what you assume is
  wrong. Of course the log files should be erasable, and if user deletes
the log file, then the
 system should just start making a new one.

Without addressing the question of what the behavior should be, man wtmp
on my FC4 system states that logging is disabled if wtmp doesn't exist:

   wtmp is maintained by login(1),  init(1),  and  some  versions  of
getty(1).
   Neither  of  these  programs  creates the file, so if it is removed,
   record-keeping is turned off.

If you've confirmed that the 770 handles things differently, that's useful
information.  Your tone is somewhat excessive, though, especially
considering that Michael Saunby's claim is supported by at least some
related documentation.

Mike




___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers


Re: [maemo-developers] Zapped wtmp...

2006-04-12 Thread Santeri Lindgren

Michael P. Lococo wrote:

There's no need to link it to anything.  If the file isn't there, then

logging is turned off - at

least that's what should happen.

Wrong, this just deletes everything that is stored prior to this. And

no, what you assume is

 wrong. Of course the log files should be erasable, and if user deletes

the log file, then the

system should just start making a new one.


Without addressing the question of what the behavior should be, man wtmp
on my FC4 system states that logging is disabled if wtmp doesn't exist:

   wtmp is maintained by login(1),  init(1),  and  some  versions  of
getty(1).
   Neither  of  these  programs  creates the file, so if it is removed,
   record-keeping is turned off.

If you've confirmed that the 770 handles things differently, that's useful
information.  Your tone is somewhat excessive, though, especially
considering that Michael Saunby's claim is supported by at least some
related documentation.

Mike


Sorry if it feels like that, it wasn't meant to be like that. But, yes, 
the wtmp is created after it is removed.


___
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers