{5} Assigned, Active Tickets by Owner (Full Description) (131 matches)

List tickets assigned, group by ticket owner. This report demonstrates the use of full-row display.

Results (1 - 100 of 131)

1 2

alexandre (1 match)

Ticket Summary Component Milestone Type Created
#2505 Colaborative Activity paint/drawing-activity (oficina) Opportunity enhancement Jul 26, 2007

Implement colaborative way

andresambrois (1 match)

Ticket Summary Component Milestone Type Created
#8702 Go misses turn if made early in collaboration other-activity 9.1.0-cancelled defect Sep 27, 2008

Go (v.5), running on two laptops with build 8.2-764, has a race condition. These laptops were registered with a school server, and connected using a WiFi AP.

One laptop started the Go activity, and shared with the neighborhood. Another laptop joined the activity. After the opponent joined the activity, but before they had finished changing the resolution of the board (down to 13x13), the first player made a move. This move was never seen by the opponent, causing a deadlock.

bdoin (3 matches)

Ticket Summary Component Milestone Type Created
#6657 gcompris icons in wrong place? gcompris-activities defect Mar 7, 2008

There is a GCompris icon in the activity subdirectory fo each GCompris activity: always the same one: activity-gcompris.svg. There is also a unique svg file in the root directory for each activity. Ideally (1) the icon in the root directory could be swapped in for the common icon so that the activities are distinguishable on the Framw without a rollover. Also, (2) the icons are not "Sugarized" in the sense that they are not grabbing the XO colors; also (3) they often use line elements that are too fine when scaled.

Fixing (1) could be as simple as

cp *.svg activity/activity-gcompris.svg

Fixing (2) is a bit more work. Fixing (3) is a lot of work.

#7138 Updated icons for gcompris activities. gcompris-activities defect May 30, 2008

In preparing Peru's 703-based image, I've cleaned up a number of gcompris activity icons. In particular, chess_computer, scalesboard, and suduko. These icons are attached.

You can use the converter in http://dev.laptop.org/git?p=users/cscott/icon-draw-activity to better clean up the remaining gcompris icons to better match the Sugar icon guidelines. Some icons required additional manual cleanup to ensure that they looked okay when the fill color was transparent (as it is in the sugar frame).

#7862 hanoi_real activity fails to launch gcompris-activities 8.2.0 (was Update.2) defect Aug 7, 2008

The hanoi_real activity downloaded from the main gcompris repository fails to launch from the sugar activity list. It does launch from the terminal activity with sugar-launch. Sugar logs report a problem with rainbow. Seems related to ticket #7239, but I'm not sure if disabling security as stated in that ticket is the best route. Also the TuxPaint activity exhibiting a similar behavior is fixed just by changing the activity.info file. See: http://wiki.laptop.org/go/Tux_Paint#Installing_.28Novices.29

bernie (2 matches)

Ticket Summary Component Milestone Type Created
#1603 XOrg post-1.3: Built-in fonts being removed from the font path on server reset x window system 8.2.0 (was Update.2) defect May 30, 2007

When you run the X server standalone, it quits after initialization because there are no clients left and the server reset clears the builtin font path.

This is a known bug in Xorg's master branch... Nobody seems to be working on it.

Can be worked around by adding -noreset to the serverflags.

#4900 /etc/alsa/asound.state should be in /etc/statetab distro Future Release defect Nov 14, 2007

/sbin/salsa writes directly to /etc/alsa/asound.state, writing through the hardlink and corrupting the pristine copy. In addition, it should be preserved across upgrades. So, /etc/alsa/asound.state should be added to /etc/statetab.

cjb (7 matches)

Ticket Summary Component Milestone Type Created
#4996 Journal doesn't load power manager (OHM) 8.2.0 (was Update.2) defect Nov 17, 2007

build Joyride 299 q2d04

The laptop is not plugged in. Choose reboot. The laptop boots to the point where you get the black border around a grey screen, and then just sits there. Then, if you move the mouse, it shows you the full sugar environment, but the journal has not loaded. I believe this is an issue with ohm interfering with boot up. I say this because I tried this twice, yielding the same result. I then did an rpm -e ohm to turn of ohm. Booting then worked fine, and the journal loaded.

#6134 closing lid when suspended in ebook mode has problems. power manager (OHM) 8.2.0 (was Update.2) defect Jan 21, 2008

Q2D09, joyride-1568.

Should cause the machine to wakeup, the screen get turned off, and then the cpu suspend again (immediately).

Instead, the cpu wakes up for 30 seconds, and then suspends.

When you reopen the lid, the machine does not seem to wake up, (power light does not light), and the screen appears to have been left on.

To reproduce:

  1. start the Read activity
  2. let the machine go idle, until it suspends.
  3. close the lid.
  4. observe the power light comes on one full 30 second timeout.
  5. open the lid
  6. note the power light does not light, and the screen appears to be on.

#8584 need "suppress idle-suspend on external power" checkbox power manager (OHM) 9.1.0-cancelled enhancement Sep 20, 2008

our thinking so far has been that in many deployments, external power is as scarce as battery power, and should be conserved in the same way. this clearly isn't true in all cases -- in areas with plentiful reliable wall power (i.e., most G1G1 markets), or even while powering from a vehicle. ohmd should be willing to suppress idle-suspend if we're plugged in.

idle-suspend interferes with normal laptop operation just enough that it's important to allow the user to suppress it. (as an example, idle-suspend interferes with reliable networking. the times when stable networking is desired probably often coincide with the times that stable power is available.)

#4876 OHM suspends while USB HID devices are attached. power manager (OHM) 9.1.0-cancelled defect Nov 13, 2007

It would be nice if USB-HID devices could also wake up the XO.

[18:30] <cjb> c_scott, pascals: when we are in suspend, the USB host controller is powered down. [18:30] <cjb> There's nothing I can do to keep it powered. What I would do instead is refuse to suspend if a USB device is plugged in. [18:30] <cjb> Or, if a particular class of USB device is plugged in -- ethernet, keyboard/mouse. [18:30] <pascals> cjb: That doesn't seem ideal either. [18:30] <c_scott> cjb: right. but i'd argue that 'refusing to suspect if USB hid device is present' is an update.2 feature. [18:31] <c_scott> cjb: *HID* device. that's the particular class you're interested in. [18:31] <pascals> c_scott: network too, I think. [18:31] <c_scott> cjb: i don't see any problem with powering down disk or network (although I may be surprised by reality) [18:31] <pascals> c_scott: network would mean that Chat would fail interestingly [18:32] <c_scott> maybe i should think about the question in the reverse: are there any external USB devices which it would be *safe* to power-down to prolong battery life? [18:32] <cjb> c_scott: you're running an sshd over the ethernet. the machine suspends through idleness. all the ssh clients lose connection. [18:32] <cjb> c_scott: disks. [18:33] <c_scott> cjb: anyway, this discussion should be in trac. pascals, do you have a trac # yet? [18:33] <cjb> I think it may be in Trac [18:33] <cjb> it's certainly been brought up on devel@.

#5102 Unable to create ~/.gnome2 directory: permission denied pippy-activity Future Release defect Nov 23, 2007

Also says:

Localization of maessages is disabled, using default language. time resolution is 2.320 ns

#10174 usbhid error unplugging a device kernel Opportunity defect Jun 16, 2010

Application crashes when USB device removed (Lego WeDo). Easily reproducible on XO-1 but also on XO-1.5 with latest OS though not as frequently...

[  770.111632] usb 1-3: USB disconnect, address 3
[  770.127433] BUG: unable to handle kernel NULL pointer dereference at 000000a4
[  770.131101] IP: [<d095431e>] :usbhid:hiddev_ioctl+0x25/0x8ef
[  770.138561] *pde = 00000000 
[  770.141495] Oops: 0000 [#1] PREEMPT 
[  770.145141] Modules linked in: usbhid sit tunnel4 usb8xxx libertas ieee80211 ieee80211_crypt pcspkr mousedev ehci_hcd ohci_hcd cs5535_gpio sdhci mmc_core usbcore
[  770.147433] 
[  770.147433] Pid: 2014, comm: squeak Not tainted (2.6.25-20090223.1.olpc.69098d87d56945c #1)
[  770.147433] EIP: 0060:[<d095431e>] EFLAGS: 00210286 CPU: 0
[  770.147433] EIP is at hiddev_ioctl+0x25/0x8ef [usbhid]
[  770.147433] EAX: 00000000 EBX: cda01000 ECX: c018480d EDX: c502b2e0
[  770.147433] ESI: c5220000 EDI: c018480d EBP: bfa4bf98 ESP: c50fcea0
[  770.147433]  DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
[  770.147433] Process squeak (pid: 2014, ti=c50fc000 task=c50e41b0 task.ti=c50fc000)
[  770.147433] Stack: cd5aa240 c04695a8 00000000 c50fcf04 c045f5c8 cd5aa240 cdc60138 ce027824 
[  770.147433]        ce027820 00000001 c045f832 00200246 00000060 bfa4be94 c04dc5ee 00000060 
[  770.147433]        bfa4be94 c50fcf50 c0459cc3 c50fcef0 0000000f 00000000 00000000 0000240b 
[  770.147433] Call Trace:
[  770.147433]  [<c04695a8>] mntput_no_expire+0x11/0x69
[  770.147433]  [<c045f5c8>] path_walk+0x8b/0x93
[  770.147433]  [<c045f832>] do_path_lookup+0x129/0x144
[  770.147433]  [<c04dc5ee>] copy_to_user+0x2f/0x45
[  770.147433]  [<c0459cc3>] cp_new_stat64+0xfc/0x10e
[  770.147433]  [<c04612d8>] vfs_ioctl+0x48/0x5e
[  770.147433]  [<c046153c>] do_vfs_ioctl+0x24e/0x261
[  770.147433]  [<c0461590>] sys_ioctl+0x41/0x59
[  770.147433]  [<c0403782>] sysenter_past_esp+0x5f/0x85
[  770.147433]  =======================
[  770.147433] Code: c0 5b 5e 5f 5d c3 55 57 89 cf 56 53 81 ec ac 00 00 00 8b 5a 78 8b ac 24 c0 00 00 00 8b 93 10 06 00 00 8b 72 20 8b 86 4c 0c 00 00 <8b> 80 a4 00 00 00 89 44 24 24 8b 86 88 0d 00 00 c7 44 24 08 fb 
[  770.147433] EIP: [<d095431e>] hiddev_ioctl+0x25/0x8ef [usbhid] SS:ESP 0068:c50fcea0
[  770.147624] ---[ end trace 5ae1d9f3e0518d20 ]---

#10160 expose the audio codec's "in use" state to user level kernel Future Release enhancement May 12, 2010

for purposes of managing idle-suspend, it would be very convenient to know when the audio device is in use (as opposed to merely "open", status which is already available).

it seems the kernel may already be tracking this, at least for the case of the output device:

<pgf> audio is actually an annoying case.  it turns out it's really easy to tell, from alsa state, whether the input or device is open.
<pgf> unfortunately, all the audio apps open it when they start, and never close it, whether it's in use or not.
<dsd_> the kernel tracks audio codec use and powers it down after 5 secs of inactivity
<dsd_> and this works well, so presumably doesnt face the same problem
<dsd_> i wonder if you can view that state somehow
<cjb> dsd_: yeah, already tried that, doesn't seem to be exposed
<cjb> but would be an easy patch to write, setting and exposing a flag through snd_pcm_playback and snd_pcm_capture.
<cjb> pgf: if you want to file a bug with the interface you'd like, I could work on that
<pgf> oh, that's a good idea.
<pgf> i didn't realize the kernel was already doing that work.
<pgf> so we think the codec is only powered when it's in use (or recently so)?
<dsd_> yeah i think so
<dsd_> but we should also check that its not affected by that same issue with open/close
<pgf> right.
<cjb> pgf: yes, see sound/pci/hda/hda_codec.c; when we stop sending samples, we queue up a power down for CONFIG_SND_HDA_POWER_SAVE_DEFAULT seconds.
<cjb> (=5 for us)

cjl (1 match)

Ticket Summary Component Milestone Type Created
#8043 ship more license translations in olpc-licenses localization Future Release enhancement Aug 19, 2008

Following on from #6928 which says: Let's get the Spanish versions into the release, and then somebody can make a management decision about what other languages to include, and where.

We need to decide which licenses we should ship translations for, and which languages. We will soon be shipping Spanish translations of GPL2, GPL3, LGPL2.1, GFDL alongside the English originals.

Another option is to provide a difference license pack for each deployment, to prevent this growing too large. Or we could have olpc-licenses providing English, and olpc-licenses-es installed on Spanish deployments which provides a few of the translations. And so on...

clash (1 match)

Ticket Summary Component Milestone Type Created
#6115 why dosen't call.xo work other-activity Future Release defect Jan 20, 2008

crazy-chris (1 match)

Ticket Summary Component Milestone Type Created
#6226 xo-get activity fails to start on update.1 build 690 xo-get defect Jan 27, 2008

The xo-get activity, installed by the xo-get.py command-line tool, downloaded on 26 Jan 2008, fails to start on an update.1 build 690 G1G1 XO. I'll attach the xo-get activity log. (BTW, the log viewer isn't displaying the log very well... different issue.)

cscott (1 match)

Ticket Summary Component Milestone Type Created
#23 DNS configuration infrastructure task Jul 26, 2006

How do we find our DNS server? How are caching dns servers on school servers configured to find their DNS servers? Which DNS server to use? Bind, or something more sane? How are laptops named? anycast? mdns lookup? How are these names reflected into world-wide DNS?

Do we need something like CoDNS to improve performance and robustness?

dfarning (1 match)

Ticket Summary Component Milestone Type Created
#8500 add docstrings to alert.py sugar-toolkit Opportunity enhancement Sep 15, 2008

I am doing a test of how round trip modifications work with pydocweb, the new API documenter.

djbclark (4 matches)

Ticket Summary Component Milestone Type Created
#5172 Gitweb fails to html-escape "owner" data that it pulls from a user's GECOS field. infrastructure Future Release defect Nov 28, 2007

If you make a user whose GECOS entry contains an email address (e.g. "Joe Schmoe <joe@…>"), gitweb will fail to escape the email address and will produce invalid XML.

The naive solution of calling esc_html($owner) at appropriate points in the code (around lines 3040 and 3400 of our already-patched gitweb.cgi) fails because the $owner variable contains wide chars and the esc_html function does not deal with wide chars.

#3426 Email problems for rah@laptop.org infrastructure Opportunity defect Sep 13, 2007

Also, my email at laptop.org: I cannot get the emails sent to my rah@… because they have a special protocol to rescue them from my laptop. I need it to be a POP3 type of account. Can you direct me to the person who manages the network so I can put in touch with my IT guy and they can decide how to proceed?

#3394 Add remote console/power support to important colo machines infrastructure 8.2.0 (was Update.2) task Sep 12, 2007

HP Proliant DL 380 (Athlon) Remote Console

  • No SSH keys, only password-based SSH auth

ILO - Integrated Lights Out

#3606 Jabber: jabber.laptop.org -> laptop.org JIDs etc distro Opportunity task Sep 19, 2007


  • danjared | djbclark: shouldn't people officially related to OLPC have @laptop.org JIDs instead of @jabber.laptop.org JIDs?
  • djbclark | danjared: Yes.
  • djbclark | danjared: But as far as I can tell with the current setup that doesn't work. Of course my experiance with the current setup in all of like 10m at the moment. │
  • danjared | djbclark: you'll need to create the appropriate SRV records in DNS │
  • danjared | djbclark: http://jabberd.jabberstudio.org/2/docs/section05.html#5_7 [AKA] http://tinyurl.com/23x845
  • djbclark | I need to go back and read the context around dilinger's comments re: jabber and irc last night. Is/was there actually a plan to move #olpc/#sugar/etc to a Jabber Conference?
  • cjb | No.

Jabber / Trac

  • dilinger | coderanger: where's our jabber/trac integration

AIM Transport

  • Also dilinger wants AIM transport for Jabber

douglas (1 match)

Ticket Summary Component Milestone Type Created
#8610 Theme music for the XS school server Future Release enhancement Sep 23, 2008

The school server needs to use sound to communicate.

For example, it automatically mounts USB drives and attempts various communications, and this can take some time. Rather than have the human guess how long to wait, the school server should play starting and finishing ditties. It could indicate errors, and give musical progress estimates.

Other ideas would be to have the server play an hourly tune that indicated the disk was 95% full, or play a sad song when the network is in poor shape.

These must be played through the PC speaker -- that annoying thing that beeps at you in the console. It can be induced to play melodies through escape codes. For instance, the following plays an A followed by E followed by a higher A:

  echo -en '\033[10;440]\033[11;250]\007' > /dev/tty1
  # sleep because echo is instantaneous.
  sleep 0.3
  echo -en '\033[10;660]\033[11;250]\007' > /dev/tty1
  sleep 0.3
  echo -en '\033[10;880]\033[11;250]\007' > /dev/tty1
  sleep 0.3

  # silently return the beep to something reasonably normal.
  echo -en '\033[10;750]\033[11;300]' > /dev/tty1

So please, add candidate tunes below, either in this notation or with a script that does the conversion.

dsaxena (9 matches)

Ticket Summary Component Milestone Type Created
#7458 Intermittent lockup during WOL suspend/resume stress not assigned 9.1.0-cancelled defect Jul 9, 2008

Email form cjb:

As a heads up, I've seen a hang on resume (or suspend?) with latest testing. It looks like a scheduling problem in libertas -- dmesg overflows such that I can't see what caused the hang, but that seems to be what the rest of it's trying to say.

Currently trying to reproduce and gather some more data. Javier, we might need some help from your end once we figure out what's happening.

#6010 Export EC wakeup event mask to userspace kernel 8.2.0 (was Update.2) defect Jan 14, 2008

We currently have the "suspend" vs. "sleep" distinction made in userspace, which results in many temporary wakeups in order to see whether we're supposed to be woken up fully by an external event.

Instead, the kernel should expose a method for setting the EC wakeup mask to userspace, and OHM should set it appropriately before each type of suspend.

#9054 Speed up USB resume. kernel 9.1.0-cancelled defect Nov 28, 2008

The largest obstacle to resuming aggressively is the ~800ms that USB resume takes. This bug tracks investigation on how that could be sped up; a first area to look at is Arjan van der Ven's recent patchset to parallelize the (currently serial) 100ms delays for USB ports to come up, so that every port is brought up at the same time.

#9128 2595 suspend causes my hub-attached USB devices to stop functioning kernel 9.1.0-cancelled defect Dec 17, 2008

When I remove '/etc/ohm/inhibit-idle-suspend', the XO eventually suspends (power light pulsing very slowly). When I resume by pressing a key on the XO's (green, builtin) keyboard, power is restored to my attached USB devices. However, while my USB keyboard and my USB ethernet adapter function again (both are plugged directly into the XO), my USB mouse and USB CD-ROM (both are plugged into an USB hub) now do not function. [This is a regression. I do not remember which build still allowed USB-hub-attached devices to function following suspend/resume.]

Took 'olpc-log' dump. With /etc/ohm/inhibit-idle-suspend removed, it took nearly an hour for the dump command to complete (normally it takes several minutes). While the 'olpc-log' command was running, the screen on the XO would dim (as it does for suspend) but would immediately brighten again -- the XO's power light remained lit the whole time (it did not go into slow pulsing as it does for suspend).

#6211 patch for support of per packet mesh ttl wireless 8.2.0 (was Update.2) enhancement Jan 25, 2008

We really need to be able to have applications set the TTL field of the mesh frames they generate. Prime example is being able to generate mDNS frames with TTL of 1 so that they don't get retransmitted by every node in dense environments, effectively saturating the spectrum.

The following patch (by Javier) implements that functionality by adding a sockoption in the driver that the application can use to set the TTL for its traffic.

This patch was originally submitted back in July but it was never incorporated because 1) it uses netfilter to track open sockets and 2) the driver needs to register a new sockoption. Since no other option was suggested and since it is crucial for our scalability, I have asked Javier to update the patch so that it cleanly applies to the current driver (after David's rewrite) and I trust it to Jim that it makes it into our kernel.


#7981 EC mask setting is inefficient kernel 8.2.0 (was Update.2) defect Aug 15, 2008

From rsmith:

So looking my EC logs from the newer joyrides I see that there are like 6 repeat calls for setting up the mask. This is very non-optimal. Without the fastpath EC code you are talking about 100ms or so of EC command time with fastpath its still going to be 25 or 30ms.

Just seems silly to me. Why can we not also have a control that echos a single hex value? Just like the SCSI debugging code.


(Note that this isn't so urgent, because it only affects entering "sleep", and not leaving it. It's still a waste, though.

#9329 olpc-kernel-rpm target doesn't build debuginfo kernel defect Jun 10, 2009

#2886 some files shouldn't be compressed by jffs2 kernel enhancement Aug 17, 2007

Storing a PDF of 35MB takes 25s. to be written to NAND, processor bound.

If this is caused by JFFS2's on-the-fly compression, marking those files as already compressed using chattr or xattr would make these operations much faster.

Also, I presume reading perf would also be improved.

#7393 Kernel, initramfs, and wireless firmware Refresh release-contract 8.2.0 (was Update.2) task Jun 30, 2008

Steps to release:

  1. Nominate a package. (Deepak, Scott)
  1. Include in a release-stream build. (Michael)
  1. Run the best regression test we've got. (Joe, public)
  1. Security audit.
  1. Iterate.

(NB: Please annotate this release contract with all relevant tickets and wiki pages.)

dsd (6 matches)

Ticket Summary Component Milestone Type Created
#11641 yum upgrade fails on kernel rpm due to conflict over /boot/olpc.fth kernel Future Release defect Feb 16, 2012

"file /boot/olpc.fth from install of kernel-<...> conflicts"

Given our current kernel rpm setup, kernels will conflict over /boot/olpc.fth . Using "rpm -Uvh new-kernel.rpm" literally upgrades the rpm package; yum however will install new kernel rpms in parallel, assuming that the rpms have no files in common.

We only hit this when we change olpc.fth, because rpm has the strange property that two packages can own the same file, and only conflict over it if the file contents differ. We seldom change olpc.fth .

One workaround is to run "rpm -U /var/cache/yum/<repo>/packages/new-kernel.rpm", then re-run yum update.

Yum upgrades are not supported at this stage, so this is very low pri.

#11647 1.75 external SD card unmounts & remounts on resume kernel Future Release defect Feb 22, 2012

Observed on XO 1.75 B1 with os28 in GNOME desktop.

The mounted SD card (with ext4 filesystem) icon dissapears for a moment, and then reappears and opens the Nautilus browser whenever the system resumes. If the Nautilus window is left open, it closes and reopens on resume.

#11898 xo-1 wlan association LED blinks after resume from blank idle suspend not assigned Future Release defect May 28, 2012

install os12, remove /etc/powerd/flags/inhibit-suspend.

associate with an access point (mine was WPA2).

let the xo-1 suspend, then dim, then blank the screen. observe that the wlan association LED is still lit.

wake with the touchpad or keypress. observe that the wlan LED goes off, then comes back on a bit later.

this is repeatable for me. i realize idle suspend isn't enabled by default, but this feels like a regression to me.

#11968 Possible skb leak in asix/usbnet kernel Future Release defect Jun 22, 2012

I saw page allocation failures upgrading a 512 MB XO-1.75 from 11.3.1 to 12.1.0 os15 while olpc-update was running on 11.3.1.

These occurred when an asix USB ethernet adapter was used with such frequency and were logged against so many things that the XO was unable to upgrade without a reboot using wifi after the asix adapter was removed.

After a reboot the XO was upgraded successfully to 12.1.0 via wifi & the libertas card, which is much slower (only 100-200 KB/sec reliable bandwidth in our area).

Psmem.py did not show unusual memory usage by any userspace process. I have not retried the upgrade with the asix adapter since.

Jun 21 18:08:26 xo-74-34-46 NetworkManager[735]: <info> Activation (eth1) Stage 5 of 5 (IP Configure Commit) complete.
Jun 21 18:08:27 xo-74-34-46 avahi-daemon[753]: Registering new address record for 2002:ad0c:4b0c:1:21c:49ff:fe01:6c0 on eth1.*.
Jun 21 18:08:27 xo-74-34-46 avahi-daemon[753]: Withdrawing address record for fe80::21c:49ff:fe01:6c0 on eth1.
Jun 21 18:15:51 xo-74-34-46 kernel: imable:4504kB kernel_stack:1312kB pagetables:1516kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all
_unreclaimable? no
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.276938] lowmem_reserve[]: 0 0
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.286502] Normal: 3170*4kB 2782*8kB 0*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB 
0*8192kB 0*16384kB 0*32768kB = 34936kB
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.306098] 78186 total pagecache pages
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.315583] 0 pages in swap cache
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.324851] Swap cache stats: add 0, delete 0, find 0/0
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.334032] Free swap  = 0kB
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.342970] Total swap = 0kB
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.368843] 131072 pages of RAM
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.377385] 9078 free pages
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.385717] 2541 reserved pages
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.393908] 4085 slab pages
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.402014] 51335 pages shared
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.409978] 0 pages swap cached
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.417740] asix 1-1.4:1.0: eth1: kevent 2 may have been dropped
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.425512] rsync: page allocation failure: order:0, mode:0x20
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.433168] [<c00369ac>] (unwind_backtrace+0x0/0x120) from [<c0091534>] (warn_alloc_failed+0xe0/0x104)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.448274] [<c0091534>] (warn_alloc_failed+0xe0/0x104) from [<c00940bc>] (__alloc_pages_nodemask+0x528/0x580)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.463394] [<c00940bc>] (__alloc_pages_nodemask+0x528/0x580) from [<c00b891c>] (cache_alloc_refill+0x2d8/0x5a
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.478567] [<c00b891c>] (cache_alloc_refill+0x2d8/0x5a4) from [<c00b8c70>] (__kmalloc+0x88/0xd8)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.486395] [<c00b8c70>] (__kmalloc+0x88/0xd8) from [<c02b6bd0>] (__alloc_skb+0x48/0xd8)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.494201] [<c02b6bd0>] (__alloc_skb+0x48/0xd8) from [<bf0b4090>] (rx_submit+0x24/0x328 [usbnet])
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.509708] [<bf0b4090>] (rx_submit+0x24/0x328 [usbnet]) from [<c0211318>] (usb_hcd_giveback_urb+0xa0/0xec)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.525774] [<c0211318>] (usb_hcd_giveback_urb+0xa0/0xec) from [<c0222328>] (ehci_urb_done+0xc4/0xe0)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.542501] [<c0222328>] (ehci_urb_done+0xc4/0xe0) from [<c0224980>] (qh_completions+0xb4/0x458)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.551198] [<c0224980>] (qh_completions+0xb4/0x458) from [<c02258e0>] (ehci_work+0xb8/0x970)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.559970] [<c02258e0>] (ehci_work+0xb8/0x970) from [<c0228d5c>] (ehci_irq+0x308/0x370)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.568771] [<c0228d5c>] (ehci_irq+0x308/0x370) from [<c0210898>] (usb_hcd_irq+0x34/0x88)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.577558] [<c0210898>] (usb_hcd_irq+0x34/0x88) from [<c008888c>] (handle_irq_event_percpu+0x30/0x174)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.594774] [<c008888c>] (handle_irq_event_percpu+0x30/0x174) from [<c0088a28>] (handle_irq_event+0x58/0x80)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.611999] [<c0088a28>] (handle_irq_event+0x58/0x80) from [<c008a8a0>] (handle_level_irq+0xc4/0x114)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.629315] [<c008a8a0>] (handle_level_irq+0xc4/0x114) from [<c00883dc>] (generic_handle_irq+0x20/0x30)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.646841] [<c00883dc>] (generic_handle_irq+0x20/0x30) from [<c002b060>] (asm_do_IRQ+0x60/0x84)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.655912] [<c002b060>] (asm_do_IRQ+0x60/0x84) from [<c0030e18>] (__irq_usr+0x58/0xa0)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.665076] Exception stack(0xd41bffb0 to 0xd41bfff8)
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.674246] ffa0:                                     00077e68 00000011 00000000 00087cf3
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.683629] ffc0: 00081ae0 00000000 00000001 00087cef 00000000 000815b0 000824b0 00083c5b
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.693140] ffe0: 00008257 beff90e0 0004ba60 00049dac 60000010 ffffffff
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.702688] Mem-info:
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.712191] Normal per-cpu:
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.721710] CPU    0: hi:  186, btch:  31 usd:  30
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.731358] active_anon:15717 inactive_anon:2161 isolated_anon:0
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.731367]  active_file:16307 inactive_file:59507 isolated_file:0
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.731376]  unevictable:8 dirty:5794 writeback:0 unstable:0
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.731385]  free:8734 slab_reclaimable:2959 slab_unreclaimable:1126
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.731394]  mapped:7644 shmem:2372 pagetables:379 bounce:0
Jun 21 18:15:51 xo-74-34-46 kernel: [ 1162.780471] Normal free:34936kB min:2884kB low:3604kB high:4324kB active_anon:62868kB inactive_anon:8644kB active_file:65228kB inactive_file:238028kB unevictable:32kB isolated(anon):0kB isolated(file):0kB present:520192kB mlocked:32kB dirty:23176kB writeback:0kB mapped:30576kB shmem:9488kB slab_reclaimable:11836kB slab_unreclaimable:4504kB kernel_stack:1312kB pagetables:1516kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no

#11877 device permissions for USB devices should be more permissive distro Future Release enhancement May 21, 2012

we often need to add non-root access for USB devices -- typically devices used for Lego or other robotics applications.

we should add a udev rule to give rw permissions to everyone, based on some useful criteria: perhaps if no kernel driver has claimed the device, or perhaps simply for all devices that aren't of certain classes (e.g., everything but storage devices).

#11763 gcc segfaults a lot on XO-1.75 12.1.0 distro Future Release defect Apr 6, 2012

Compiling ffmpeg on XO-1.75 with 12.1.0 os6 ends in failure. From a clean install, run the following commands:

sudo yum install -y --nogpgcheck yum-utils gcc make

wget -c http://download1.rpmfusion.org/free/fedora/development/17/SRPMS/ffmpeg-0.8.9-1.fc17.src.rpm

rpm -i ffmpeg-0.8.9-1.fc17.src.rpm

tar -xjf ~/rpmbuild/SOURCES/ffmpeg-0.8.9.tar.bz2
cd ffmpeg-0.8.9

after approximately 20 minutes, it ends with:

libavcodec/ansi.c:162:17: note: ‘width’ was declared here
CC	libavcodec/apedec.o
libavcodec/apedec.c: In function ‘entropy_decode.constprop.1’:
libavcodec/apedec.c:470:1: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugzilla.redhat.com/bugzilla> for instructions.
The bug is not reproducible, so it is likely a hardware or OS problem.
make: *** [libavcodec/apedec.o] Error 1

I managed to capture the gcc command line:

gcc -I. -I/home/olpc/ffmpeg-0.8.9 -D_ISOC99_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_POSIX_C_SOURCE=200112 -D_XOPEN_SOURCE=600 -DHAVE_AV_CONFIG_H -std=c99 -fomit-frame-pointer -marm -pthread -g -Wdeclaration-after-statement -Wall -Wno-parentheses -Wno-switch -Wdisabled-optimization -Wpointer-arith -Wredundant-decls -Wno-pointer-sign -Wcast-qual -Wwrite-strings -Wtype-limits -Wundef -Wmissing-prototypes -Wno-pointer-to-int-cast -O3 -fno-math-errno -fno-signed-zeros -fno-tree-vectorize -Werror=implicit-function-declaration -Werror=missing-prototypes -MMD -MF libavcodec/apedec.d -MT libavcodec/apedec.o -c -o libavcodec/apedec.o libavcodec/apedec.c

It doesn't always fail but happened on 3 different laptops. If it succeeds, make fails with the same error on a different file. I haven't successfully compiled ffmpeg.

Adding swap doesn't seem to make a difference.

dwmw2 (6 matches)

Ticket Summary Component Milestone Type Created
#5024 Libertas BUG() - Command Queue locking kernel defect Nov 19, 2007


[    1.200565] usb 1-1: can't restore configuration #1 (error=-32)
[    1.200582] logical disconnect on port 1

and then an oops, as attached, and wireless didn't come back. C2 machine, q2d04, update1-635.

#190 JFFS2 power failure handling kernel Future Release defect Oct 18, 2006

A pathological pattern of power cycles can cause JFFS2 to entirely run out of space to perform garbage collection. For example, hook up an automatic power switch triggered by the NAND controller and make it cut power in the middle of each page write, unless it's the first page in an eraseblock.

Thus, each erase block will have only a _single_ valid node and thus can't be deleted -- but the rest of the space in each block will be wasted by failed write attempts. And then you'll have no free space left into which you can garbage collect, and you see nasty messages like ‘Argh. No free space left for GC’

There are ways to deal with this -- you can reserve an extra eraseblock for "panic" garbage collection, where you don't GC normally, but instead you just _copy_ nodes intact from the least-used eraseblock, just as we copy REF_PRISTINE nodes. If power is interrupted during this 'panic GC', we can erase the target block and try again. If we manage to write the whole thing, then we can erase the _source_ block and it becomes our panic block for the future, while we still have the tail end of the block we've just written which is available to allow normal GC to proceed.

This relies on being able to recover properly after the power failure -- we have to:

  1. Know that we're in panic mode, which is easy enough because there are no free eraseblocks.
  2. Find the block which we were copying into, which is also relatively easy because it contains only nodes which are _duplicates_ of other nodes on the medium, with precisely the same version number. Currently we silently drop one of the duplicates, but we can change that to keep a record so that we can find blocks which are entirely made up of such nodes.

This also helps us deal with the uncertainty around garbage collection, where it was never mathematically proven that we were reserving enough space to allow GC to proceed under all circumstances. The ‘panic mode’ should allow us to avoid getting stuck, as has occasionally been observed to happen.

#3032 JFFS2 performance on content bundles. kernel 8.2.0 (was Update.2) defect Aug 26, 2007

We need to study JFFS2 and overall system performance for content bundles.

Some questions are: should we have large files with lots of content, or multiple small files (which depends on performance of large directories vs. performance of large files). Should we compress the whole file and implement the chattr ioctl interface we've spoken about for so long to tell JFFS2 not to bother? etc.

We need to fix whatever performance issues exist, and give coherent guidelines to those providing content.

Some samples of potential content are at...






[02:17] <Arael5> I could stuff a lot more content into XO if I could use file.html.gz and iso image for storing thousands of small files.

[02:17] <Arael5> I have tried to open 250kB html file on XO and it took 30 seconds to open it on build 406.

[02:17] <Arael5> 100kB html file loaded in 7 seconds.

[02:23] <Arael5> The 1mb file is here http://arael.shtooka.net/fi/html/1mb.html

[02:24] <Arael5> If XO could open it in less than 10 seconds I will not have to use many small files in my bundles.




#5723 Actiontec Router Issues...no Gmail/Google/Wiki wireless Future Release defect Dec 27, 2007


Just received my XO and I've been having difficulty connecting to the web with the browser.

I'm running an actiontec router...had issues even having it show up in the neighborhood view until I switched over from channel 9 to channel 6 per the recommendation of a friend. This effectively made the actiontec router show up in the neighborhood view.

Browser worked intermittantly.

Today I go to log on again and I can get certain websites: http://www.burgerking.com; http://www.mcdonalds.com; http://www.wcnx.org; http://craigslist.org. I can ping google.com and get packets back, 109ms. But if I type google.com or http://www.google.com or http://mail.google.com into the browser it runs incessantly and then I get a network timeout error.

Here are the specs on the WAN/LAN etc from my router, just incase:


WAN Status: Conencted; PPPoE; IP Address:; Subnet Mask; Gateway; DNS 1:; DNS 2:

Lan Status: IP Address:; Net Mask DHCP On.

SSID Broadcast: Enable; 802.11b/g mode: 802.11g only

mac authentication: accept all clients. no exceptions.

wireless on; ESSID: GriffithAp; Channel: 6; Security: Off

Both my ubuntu machine and my olpc show up in the ARP table and the active user list as DHCP connected and IP assigned ( for the OLPC; for ubuntu).

Any ideas as to why this is behaving this way??

#7037 Wireless not working after idle-suspend wireless defect May 19, 2008

MP-machine, joyride-1949, connected to a D-Link DWL730AP w/ WPA, loses connection after idle-suspend.

(05:30:02 PM) bertf: cjb: mine loses connectivity after suspending ... but that's not what you mean?
(05:32:10 PM) cjb: bertf: after sleepipng, you mean?
(05:32:22 PM) cjb: (lid close or power press)
(05:32:32 PM) bertf: cjb: after idling
(05:32:38 PM) cjb: bertf: that doesn't happen here
(05:32:41 PM) cjb: bertf: which build?
(05:32:51 PM) cjb: (how long idling?)
(05:33:07 PM) bertf: cjb: latest joyride, mp machine, about 30 secs
(05:33:15 PM) cjb: huh
(05:33:21 PM) cjb: I think that's what I'm running too, and it works
(05:33:40 PM) cjb: bertf: which joyride exactly?
(05:33:48 PM) cjb: bertf: what happens to the connection?  does ifconfig still have an IP?
(05:34:15 PM) bertf: cjb: 1949. ifdown/up fixes it
(05:34:31 PM) cjb: bertf: I'm curious whether you have an IP after this happens or not
(05:34:52 PM) bertf: cjb: yes, ifconfig reports the right ip
(05:34:57 PM) cjb: bertf: that's not good
(05:35:28 PM) cjb: dwmw2: do you know what the status of the libertas that's in latest joyride is?  Bert seems to be losing connectivity after suspend.
(05:35:42 PM) cjb: bertf: I'll try on my MP, I'm a few builds back, thanks.
(05:36:22 PM) dwmw2: cjb: not offhand. Is that still the stable kernel?
(05:36:23 PM) bertf: cjb: I'm not sure this is a recent problem. I usually had the idle-suspend disabled, but upgrade recently
(05:36:39 PM) cjb: dwmw2: dunno.
(05:36:51 PM) cjb: bertf: but it works fine here.
(05:36:58 PM) dwmw2: would be interesting to see libertas debug output when it fails.
(05:37:02 PM) cjb: bertf: this is a wireless connection via NetworkManager?
(05:37:09 PM) dwmw2: didn't we still have some outstanding concern about suspend/resume sometimes failing?
(05:37:27 PM) cjb: dwmw2: not that I know of
(05:37:37 PM) bertf: cjb: yes, to my access point. 
(05:37:42 PM) cjb: (we have an outstanding concern about RTC resumes being possibly flaky, so we don't do that anymore.)
(05:37:53 PM) bertf: cjb: the wireless led is off
(05:38:08 PM) cjb: bertf: I wonder if dmesg says anything about it
(05:38:24 PM) cjb: bertf: would you mind filing a bug, assign to dwmw2, attach dmesg and look at getting libertas_debug output too?
(05:38:34 PM) cjb: I'll try to reproduce, but since I've never seen this before I'm not sure I'm going to be able to
(05:39:05 PM) bertf: cjb: will do ...

#7340 mesh interface doesn't send association events wireless defect Jun 24, 2008


When setting an ESSID on the mesh interface no association event is ever signalled. Which means NM can't know when the interface is ready to handle traffic.

Eben (3 matches)

Ticket Summary Component Milestone Type Created
#163 Background activities interface-design Future Release task Oct 10, 2006

Bert Freudenberg: Is there a general policy of background activity? I guess the chat, for example, would continue to receive messages even when not at top. An animation should be stopped if nobody sees it. But what about a longer-running simulation? Or some time-lapse data measuring from the analogue input?

Marco Pesenti Gritti: We should probably have a policy for this in our (planned) activity guidelines

#352 Need design for integrated chat ("mesh" chat) interface-design Opportunity task Nov 8, 2006

The icons are too small (e.g. the alpha and italic and up and down arrows, and the smiley). The "Send" button gets truncated to "Sen".

#1795 Design object chooser palettes based on the journal interface-design Future Release task Jun 22, 2007

This dialog/palette will allow activities authors implement a way for inserting in their activities entries from the journal.

ebf (2 matches)

Ticket Summary Component Milestone Type Created
#907 Provide a way to run "unsugarized" applications under sugar... distro Opportunity defect Feb 20, 2007

That may require multiple windows open at once...

While this is much less desirable than having applications sugarized, we need some way to run unmodified applications that require multiple windows open simultaneously: an example, just to name an application that does this, not that we expect to run it, is the gimp. We need applications to be able to run, even if they don't meet the needs of some kids due to UI complexity.

Ion, for example, provides a way to do this, from talking with Mako. It may be Matchbox needs to do this as well.

#16 voice messaging infrastructure task Jul 26, 2006

Since many kids (and family members) won't have continuous contact to the internet, despite our best attempts, having non-realtime voice messaging of some sort would be a great help to many people.

Ideally, the UI for this operate well int he sugar environment, and there are many possible transports for the message. Automatic opportunistic behavior would be very desirable (if you have a message bound for someone, and that person becomes available, taking the opportunity to deliver the message immediately would be very good).

Note that conventional mail systems are not typically very good at this: they queue mail, and only run queues on a periodic basis. This might be quite different, more like "buddy pounce" on IM systems, where a message is forwarded whenever the recipient appears.

Of course, delivery via a school server is another option; a message might arrive on one laptop in the morning and get home to the recipient on another laptop that evening.

edsiper (2 matches)

Ticket Summary Component Milestone Type Created
#3158 Powertop should be a pane on the developer's console. developer-console 8.2.0 (was Update.2) defect Sep 4, 2007


We need to make performance testing easy...

#4859 Mesh View: Selecting an AP/mesh icon does not dismiss its palette sugar 9.1.0-cancelled enhancement Nov 13, 2007

If the user want to connect to some AP or Mesh network, he has to click the right icon but the palette of this icon is never hided and that make it feel that nothing happens. As the palette never come's hided the user try to click the palette title :/ .

For buddy icons, all actions are in the palette and for ap/mesh network over the icon, I think this must be fixed.

Here's a patch attached in order to close the palette when the user click the AP, but would be better to handle the 2 ways to connect: clicking the icon and closing the palette and add a 'connect' menu item in the palette.

erikb (2 matches)

Ticket Summary Component Milestone Type Created
#5622 Record doesn't save EXIF data record-activity 8.2.0 (was Update.2) enhancement Dec 21, 2007

The Record activity doesn't record EXIF data identifying the XO as the creating entity, nor any exposure info etc.

This would be useful for many advanced photography applications, including stitching and stop-motion.

This would also be helpful for tagging purposes when files are uploaded to sites like flickr or gallery2.

#4209 Record'ed items should inherit Record's current title record-activity 8.2.0 (was Update.2) enhancement Oct 12, 2007

616 q2c28 B4. When you change the title/name/label/'thing in the top left of the Activity tab', only the Journal entry for the activity itself is affected. Media documents created by the activity still get their default title. They should instead incorporate the activity title.

So if the Record title is 'bird near the bus', the resulting photos should also mention that, without requiring the hand relabeling of n images.

erikg (3 matches)

Ticket Summary Component Milestone Type Created
#7258 possible XO security issue psybnc security defect Jun 11, 2008

An XO was apparently hacked and an instance of psybnc was apparently installed.

Investigate and document how this happened.

Determine what can be done to mitigate the risk.

Erik suggests setting up a "honeypot" XO.

We could run tripwire on the XO and have a "man-in-the-middle" to run tcpdump and use ACID and SNORT to analyse the traffic.

#808 Rotate mapping of arrow keys onto rocker/buttons when rotating screen x window system 8.2.0 (was Update.2) task Jan 24, 2007

When the display is rotated, the default arrow key mappings to the rocker buttons should rotate as well, so that someone holding the laptop in tablet mode can navigate naturally.

#6605 Screen rotates clockwise, while rotation button shows counter clockwise arrows sugar defect Mar 1, 2008

As said in the summary, the screen rotates clockwise when the rotation button is pressed. However, the rotation button shows a counter clockwise rotation. I think that the screen should be rotated counter clockwise, so that it matches with the graphic on the button.

erikos (1 match)

Ticket Summary Component Milestone Type Created
#6461 increase-mcast-rate needs adjustments distro defect Feb 14, 2008

Two comments on the "/etc/NetworkManager/dispatcher.d/increase-mcast-rate" script

1 - The rates passed to iwpriv must be doubled. The way it is the rates are being fixed at 18Mbps.

2 - Also, using msh0 or eth0 makes no difference. We don't need both iwpriv commands. The command does not affect infra-structure traffic, either way.

So, for clarity sake, I believe what is intended is:

iwpriv msh0 mesh_set_bcastr 72

Or, if the intended rate is 11Mbps:

iwpriv msh0 mesh_set_bcastr 22

ffm (1 match)

Ticket Summary Component Milestone Type Created
#8924 Write sugar-control-panel manpage sugar 9.1.0-cancelled task Nov 2, 2008

garycmartin (1 match)

Ticket Summary Component Milestone Type Created
#9812 Bridge 2 doesn't quit not assigned defect Dec 4, 2009
  • noodling about in bridge
  • press x on top left of keyboard
  • we're not entirely sure, but we think bridge dissappered and was replaced with the "starting" screen with throbbing bridge in the middle (we can't remember if it was actually throbbing).
  • try resume from the frame menu
  • we're not sure when the bridge frame menu became doubled up, but it has two stops and two resumes
  • bridge no longer starts or stops or works
  • according to ps (attached) bridge doesn't appear to be running anymore
  • bridge doesn't show up in the list below the XO in the activity wheel
  • bridge /does/ show up in the frame menu (and still has two resumes and two stops)
  • other activities work
  • attached are three bridge logs with timestamps contemporaneous to this error.
  • lacking any other diagnostic tools, we restarted the XO to clear this, advice on further diagnosis would be greatly appreciated

hhardy (5 matches)

Ticket Summary Component Milestone Type Created
#7364 Need some limitation to readership for security-related issues not assigned defect Jun 26, 2008

Is there a trac feature we can use to limit access to bugs related to security issues until the fix is found and released?

#6406 Set up fileserver to host 1 + 1 TB of materials infrastructure defect Feb 11, 2008

We could use a fileserver with some disk redundancy with at least 1TB of usable space, to mirror large collections.

Separately, I have a ~1TB disk image of an eGranary available to schools that wish to use it, which should ideally be mirrored and kept in our colo. Because it takes hours to make a raw disk image (and significantly longer to copy files via rsync), it is useful to reserve a dedicated pair of disks for this.

#6409 install semantic mediawiki extension website enhancement Feb 11, 2008

The OLPC Wiki is going to stay a size that is reasonable for adding semantic-forms and semantic mediawiki. This will turn templates into forms, track structured data more clearly, and provide a related data-enhanced search.

We should install both extensions (after updating to mwiki 1.11).

#6045 remove projects/jokemachine from git infrastructure 8.2.0 (was Update.2) task Jan 16, 2008

Joke Machine is hosted under mamamedia/joke-machine.
Please remove the projects/jokemachine entry because it leads to confusion.

#6665 Wiki Google search *.laptop.org instead or in addition to wiki.laptop.org infrastructure enhancement Mar 9, 2008

That Google search on the left hand-side of the Wiki goes to a Google page where you can ( ) Search the Web and (*) Search wiki.laptop.org (default).

I think it would be handy if instead wiki.laptop.org that thing would search *.laptop.org (e.g. including dev.laptop.org; this Trac!).

Or maybe, for larger public who is less interested in Trac, keep the current default for wiki.laptop.org, but add another third option for *.laptop.org (if that's possible with Google).

homunq (1 match)

Ticket Summary Component Milestone Type Created
#1070 Classic-style debugger develop-activity Future Release task Mar 15, 2007

Develop needs a classic debugger with Next, Step, Traceback, and inspection features.

humitos (10 matches)

Ticket Summary Component Milestone Type Created
#6099 browser's Spell Check does not offer alternatives browse-activity Future Release defect Jan 19, 2008

In Browse, go to a web site and enter mis-spelled words in any textarea (like this page).

The Gecko engine's spell checker underlines "badword" in red, but there's no access to corrections.

I don't know if this is simply lack of a right-button context menu in Browse, or a design issue. Apparently Write has the same issue (bug 5394).

Obviously, spell check should work identically in the Write, Browse, anything else using XULRunner, and any other program with text input.

#7401 youtube still too hard to get working browse-activity defect Jul 1, 2008

Playing videos from YouTube (And google video) has long been one of the most frequent requests of XO users... it still takes too many steps to make this work.

These videosites don't work by default with the latest Gnash, and it's still too hard to update Browse to make it so : it requires opening finding the right how-to page or FAQ, opening a Terminal, and downloading & applying the Flash RPM.

While sites such as dailymotion are providing ogg videos for us, we should make it easier to enable the most commonly used sites (and should probably have a list against which to test how well Browse works by default).

#7663 Images should be optional in browse browse-activity Future Release defect Jul 26, 2008

End result : Allow the user to disable images in browse.

Use case : Alice is on a 56k modem, sharing it with 60 other children. Nonwithstanding the use of SOCKS (which only works best when more than one child gets the same content), alice should be able to disable images, flash, etc, to speed up her browser on slow connections.

#7684 Browse should get user agent from /etc/olpc-release. browse-activity 9.1.0-cancelled defect Jul 28, 2008

Browse currently reports its user-agent as 'OLPC/Update.1'. This should be updated for the 8.2 release, and in fact it should be getting this string from either /etc/olpc-release, /boot/olpc_build, or /etc/redhat-release (in order of preference).

#8389 copy/paste between terminal/browse not working browse-activity 9.1.0-cancelled defect Sep 9, 2008

used the copy icon (in the frame) to copy text from Terminal to try and view a url in Browse. the paste icon is in the edit tab in browse, but the url entry field is in the "browse" tab. i couldn't get paste to work.

using a two-button paste did work. however, that resulted in an extra "garbage" character (a newline or carriage return?) in the url field. (firefox did not have this issue.)

(build 8.2-759)

#9068 Browse should tag downoads with source-url metadata browse-activity 9.1.0-cancelled defect Dec 4, 2008

When downloading a file, Browse should add the source-url as journal metadata.

#10091 Browse fails to correctly render fonts with large point size browse-activity defect Mar 28, 2010

In an HTML page, setting text to a large point size via CSS results in garbled text when viewed in the Browse activity.

See what happened to the words: "am" and "and" when font-size is set to 120pt here:


#10514 Scrolling using the arrow keys or dpad in Browse browse-activity Future Release defect Dec 6, 2010
  • open a pdf in Browse
  • try to scroll with the arrow keys or the dpad
  • the scrolling behavior is inverted (up scrolls down, down scrolls up)

This is because by default the Spinbutton has the focus. And 'up' means here increment the page number (going down) and 'down' means here decrement the page number (going up).

I think the interface is a bit misleading. I think we should use the same interface as Read to be consistent.

#11306 Browse PDF plugin doesn't show shared bookmark pdf browse-activity Future Release defect Oct 5, 2011

As pointed out by Simon:

In 11.2.0, when sharing a bookmark which is actually a PDF file embedded in the browser, other users who join the activity can click the bookmark and see the PDF embedded in their browser too.

In Browse-129.1 (for 11.3.0) the users joining the activity can click the bookmark but then the PDF is downloaded instead of shown in the browser. This is a regression.

#8097 Alt-Esc exit from Browse while in Full Screen leaves persistent Full Screen icon. browse-activity 9.1.0-cancelled defect Aug 22, 2008

In joyride-2301, I was using Browse-95 in Full Screen view mode, then exited using alt-esc shortcut from keyboard.

This leaves a very persistent "Full Screen" icon in the upper right corner, that just won't go away.

See Screenshot attached.

jcardona (1 match)

Ticket Summary Component Milestone Type Created
#7825 can't complete WPA handshake with D-Link WBR-2310 wireless 8.2.0 (was Update.2) defect Aug 5, 2008

Testing a D-Link WBR-2310 in WPA mode. Authentication completes, association too, but WPA PTK/GTK handshake fails.

There are 4 handshake messages as usual (supplicant, authenticator, supplicant, authenticator) at which point the PTK is installed. It then enters group handshake mode but wpa_supplicant does not see the next EAPOL.

I ran a system in monitor mode next to it and confirmed that the AP did send the 5th EAPOL. It did not retransmit it, suggesting that it also received acknowledgement (my monitor does not show acks). So the message is being lost somewhere.

Interestingly, it is timing-sensitive too. If I run wpa_supplicant in verbose mode on the XO console, the extra delay imposed by printing a lot of messages results in WPA PTK+GTK completing. Odd.

jirwin (1 match)

Ticket Summary Component Milestone Type Created
#648 Sugarize "puzzles" collection. sugar Opportunity enhancement Jan 2, 2007

http://www.chiark.greenend.org.uk/~sgtatham/puzzles/ is a fun collection of lightweight GTK2 puzzles written in C, which we could distribute. We'd just need a few changes:

  • a sugar activity for choosing between the puzzles
  • conversion of the drop-down menu items to icons, since we don't want to have to translate the options, and many kids couldn't read them even if we did.

We likely won't have time to do this ourselves for a good while; if anyone's looking for a project to introduce themselves to sugar, this could be a fun one.

Looking further, there is a huge collection of games at http://www.happypenguin.org/... we should take some time to sort through them for good ones.


jnettlet (4 matches)

Ticket Summary Component Milestone Type Created
#11645 Gcompris show artifacts in xo-1.75 x window system defect Feb 17, 2012

Hi, Uruguay is requesting include Gcompris in their new image with the xo-1.75, and testing it, I have found different graphic artifacts. Should be good to know if are driver or libraries issues, and if we can solve them. To test it do:

Download and install http://arm.koji.fedoraproject.org/packages/gcompris/9.5/1.fc14/armv5tel/gcompris-9.5-1.fc14.armv5tel.rpm

Execute gcompris. There are little line artifacts in the main screen, but if you go to "discovering activities" (second icon), maze, and select the second or third maze game, can see the penguin is only displayed when points to the bottom, like if the routine used to rotate the image does not work.

I have tested with or without fullscreen. There are option --cursor to use the standard cursor. These errors do not happen in xo-1.5.

#10311 XO-1 os851 rotate button shows momentary backlight off and image data glitch x window system Future Release defect Aug 23, 2010

Pressing rotate button while activity view is shown.

On XO-1 release 8.2.1 build os802, pressing the rotate button causes only a redraw of the Sugar activity view.

On XO-1 release candidate for 10.1.2 build os851, pressing the rotate button causes a momentary loss of backlight, vertical bars, followed by redraw of the Sugar activity view.

(The end result is okay, but the means by which it got there was unpleasant.)

#11051 Glitch in icons in Journal (xo 1.5) x window system defect Jul 8, 2011

This glitch does not happen every time, but can be reproduced.

Tested in 871 and 872.


Go to the Journal

Attach a pendrive, the Journal will display a toolbar at the bottom.

Go to the detail view of any object in the journal.

Return to the listview in the Journal.

#11922 pygame drawing artifact in xo-1.75 not assigned defect Jun 8, 2012

The top of the pygame area show a fast redraw.

You can see it in Physics activity, moving with the "hand" tool a object across the top of the screen, or testing the activity hello-accelerometer cloning git://git.sugarlabs.org/hello-accelerometer/hello-accelerometer.git

lucks (11 matches)

Ticket Summary Component Milestone Type Created
#1860 Kuku high scores aren't preserved after restarting kuku-activity 8.2.0 (was Update.2) defect Jun 29, 2007

OFW: Q2C18 Build: 451

Steps to reproduce:

  1. Play the game until you get the win screen.
  2. Close the activity with the X button.
  3. Restart the Kuku activity.
  4. Your high score will be 0 again.

#1672 Need feedback after question is answered or all lives are lost. kuku-activity 8.2.0 (was Update.2) defect Jun 9, 2007

The user should be given feedback either after they have selected a correct answer, or have run out of lives.

#1692 Kuku mouse support kuku-activity Future Release defect Jun 13, 2007

Most of the first kids to play Kuku Anakula went to the mouse. We need to add basic mouse support - perhaps just flashing up key instructions when they try to use the mouse.

#1787 kuku timer, lives, and score need to be smaller kuku-activity Opportunity defect Jun 21, 2007

make these smaller to leave more white space on the left-hand side

#1789 kuku tiles need a border when displaying an image kuku-activity Opportunity defect Jun 21, 2007

make a small white border around images

#1469 Kuku image facing in right direction kuku-activity Opportunity enhancement May 13, 2007

According to SJ, the frog in number munchers would change the direction it was facing if it was up against the edge of the grid. For example, moving to the right edge of the grid, it would stay right facing, but change to left facing if moved up or down. Let's do something like that.

#1470 Editing support for content editing kuku-activity Opportunity enhancement May 13, 2007

Kent Quirk suggested we add editing support for people to add content (questions and answers).

  • possible abiword plugin for editing
  • simple markup (wikitext?) so that can edit in whatever application

might be the best so teachers can share materials

  • Kuku needs to be robust on reading these files.

#1664 Need individual grid tiles instead of tic-tac-toe like grid kuku-activity Opportunity enhancement Jun 8, 2007

Need to form smaller rectangles and offset them to give the appearance of individual tiles.

#1762 kuku player icons kuku-activity Opportunity enhancement Jun 19, 2007

make the player icons appear in succession rather than with a number beside them

#1859 Add a "Ready" screen in Kuku kuku-activity Opportunity enhancement Jun 29, 2007

The XO's sticky keys, coupled with occasional lag between questions/screens, can make it seem as though the user hasn't selected the answer even after pressing the key, when he actually has...so he'll press it again, and end up pecking twice (once for the correct answer he selected, and once again on the new screen that appeared AFTER he pressed the key, causing him to peck the incorrect answer on the new screen without his even realizing it). This caused me to accidentally pick the wrong answer quite a few times (after victory screens in particular). Maybe making a "Ready!" message on each screen and cause no key presses to do anything until the Ready message has displayed?

This lag is not limited to pecking - moving, and pretty much all commands, are subject to occasional lag that can interfere with gameplay. It would be nice if this was fixed.

#1784 kuku font scaling kuku-activity Opportunity task Jun 21, 2007

Need to scale fonts in the developing environment to look like what is on the XO.

mako (1 match)

Ticket Summary Component Milestone Type Created
#6398 mikmik server implementation needed distro Future Release enhancement Feb 11, 2008

Distributed wikis work best with a reliable place to browse shared namespaces (using either a browser or a dedicated activity/client).

A typical example would be a school of people all working together to build help documents, notes about their town, and class notes from their classes. While the total set of things everyone in town is working on might be larger than any individual wants to store on their own XO, they would at the same time want to browse this full collection from the school's collective cache.

See http://wiki.laptop.org/go/MikMik#Use_cases for details.

mcfletch (1 match)

Ticket Summary Component Milestone Type Created
#1701 Make pygame.display.set_mode set the mode olpc-games 8.2.0 (was Update.2) enhancement Jun 14, 2007

It would be nice if the set_mode method actually set the mode rather than just returning the current mode as set by the wrapper. Can it do this?

<lincolnquirk> coderanger, could the pygame.display.set_mode method be hijacked in the wrapper to make it ask GTK for a resize first, then forward the resize onto Pygame?

<coderanger> lincolnquirk: I tried that, but the pygame init is so slow that there is a noticeable jitter

<coderanger> lincolnquirk: You could try to do some monkeying so that the eventbox pygame is painting in is realized but not visible

<coderanger> it has to be shown (realized) before you can pull the window ID off it

mlj (4 matches)

Ticket Summary Component Milestone Type Created
#3052 bunny ears are stiff mechanical 8.2.0 (was Update.2) defect Aug 28, 2007

some (20% of the ones I've tried) C-test bunny ears are very stiff, making the laptop very hard to open.

#836 Gen 2 wishlist hardware enhancement Jan 26, 2007

Please add your wishes for the Gen1.5 machine here. This laptop will be mass produced starting in early 2008 with a rubber housing. The primary modifications currently under consideration are mechanical, although a touchscreen is also planned, as well as a movable camera Please add here anything that you wish to be considered for this XO release.

  • Mary Lou

Please add suggestions as separate trac items, and then add the bug to this collector bug, just as cjb did below.

#2298 I can move the glass of the panel on B3 and B4. hardware Future Release defect Jul 19, 2007

This is found by Quanta HW team: Gary Chiang.
Use your hand to touch the panel and move it.
You will find out the glass in the frame of the panel can be move.
This is cause by the panel module, not mechanical issue.
We found the same issue on B3 and B4 machine.

#477 Improvement needed in Gamepad Buttons hardware enhancement Nov 18, 2006

Hi, Folks,

We're not satisfied with the current gamepad implementation on the B1 units. The problems here are threefold. First, the implementation of the left gamepad controller needs a bit of tweaking to ensure consistency across all units (we currently see a failure rate of ~2.5% on the gamepad switches). Second, the electrical design of the current gamepad buttons should be enhanced for N-Key rollover, which means that it will be possible to depress "many" buttons simultaneously without generating "ghost" / false button triggering. Finally, the EC's scanning logic for the gamepad buttons is currently broken - simple firmware enhancements are required to implement the gamepad scanning logic properly.

All of these features should be implemented in the B2 build.

morgs (1 match)

Ticket Summary Component Milestone Type Created
#6248 Presence service broken with NetworkManager-0.7 (NM D-Bus API change) presence-service 9.1.0-cancelled defect Jan 29, 2008

I'm testing with 0.7.0. No buddies on the mesh view and no telepathy-gabble.log is created.

The presence service log shows that the getDevices call is failing (the method has been renamed to GetDevices apparently). It looks like the network manager Dbus interface changed quite a bit, because getProperties is not working anymore either.

As a quick workaround:

diff --git a/src/psutils.py b/src/psutils.py
index 6b8ec95..ae705f4 100644
--- a/src/psutils.py
+++ b/src/psutils.py
@@ -154,9 +154,8 @@ class IP4AddressMonitor(gobject.GObject):
         sys_bus = dbus.SystemBus()
         self._watch = sys_bus.watch_name_owner(NM_SERVICE, self._nm_owner_cb)
-        if not sys_bus.name_has_owner(NM_SERVICE):
-            addr = self._get_address_fallback()
-            self._update_address(addr)
+        addr = self._get_address_fallback()
+        self._update_address(addr)
     def do_get_property(self, pspec):
         if pspec.name == "address":

Interesting comment in the code:

class IP4AddressMonitor(gobject.GObject):
    """This class, and direct buddy IPv4 address access, will go away quite soon"""

1 2
Note: See TracReports for help on using and creating reports.