Hello,
Now seems to be a good time (even though the B-Test1 schedule seems to be pushing back) to think about the next major items to be addressed for B-Test2 (and later) milestones. Here is rough summary:
* Upto one VM instance per machine. Right now, you can create more than one VMs just by clicking the shooting star icon. It has some implication in the tactics of "background activity", which Bert asked but I think we didn't get definitive answer.
* Address the tickets on the dev.laptop.org. Some short-term items are listed on the trac system. Takashi says he is looking at the gettext. (I'll list some items from Jerome Peace.)
* Look and feel adjustment. OLPC's look is changing. We need to find a good timing to do it so that we don't end up with doing the same thing not too many times. This includes, perhaps, move some menu items to OLPC's chrome, change the way that halos work, etc.
* Authoring tool for mentoring. Upon EventPlaybackSpace (or something), a tool that a designer and developers can use should be built.
* Contents. Working with designers to make different tutorial contents built upon the above item.
* Voice over machinary. Hooking a better sound codec than GSMCodec with Squeak. Ogg Vorbis comes to my mind, but we need to check again with OLPC folks.
* Video. Diego says he updated the V4L2 binding. We will test it and make sure it works with the camera.
* Script shaping. A reasonable way is to write a text renderer that uses Pango primitives.
There can be more... Some might think that there are items we can do for B-Test1, but I think we are now safe to start concentrating on the B-Test2.
Any suggestions and feedback is welcome.
-- Yoshiki
Hi Yoshiki -
This list looks like a great start. I agree with all of these topics as they correspond to issues that we need to address for one reason or another. The two extra issues I'd put on this list are:
* Generic "sensor support" (along the lines of world stethoscope) if that makes any sense at all (e.g., if OLPC will have the means to plug in cheap sensible sensors)
* OLPC-centric interface, meaning that we need to spend some time integrating the OLPC-specific means for opening projects, sharing them (both locally and remotely) providing means for data exchange between activities etc.
Cheers, - Andreas
Yoshiki Ohshima wrote:
Hello,
Now seems to be a good time (even though the B-Test1 schedule seems to be pushing back) to think about the next major items to be addressed for B-Test2 (and later) milestones. Here is rough summary:
Upto one VM instance per machine. Right now, you can create more than one VMs just by clicking the shooting star icon. It has some implication in the tactics of "background activity", which Bert asked but I think we didn't get definitive answer.
Address the tickets on the dev.laptop.org. Some short-term items are listed on the trac system. Takashi says he is looking at the gettext. (I'll list some items from Jerome Peace.)
Look and feel adjustment. OLPC's look is changing. We need to find a good timing to do it so that we don't end up with doing the same thing not too many times. This includes, perhaps, move some menu items to OLPC's chrome, change the way that halos work, etc.
Authoring tool for mentoring. Upon EventPlaybackSpace (or something), a tool that a designer and developers can use should be built.
Contents. Working with designers to make different tutorial contents built upon the above item.
Voice over machinary. Hooking a better sound codec than GSMCodec with Squeak. Ogg Vorbis comes to my mind, but we need to check again with OLPC folks.
Video. Diego says he updated the V4L2 binding. We will test it and make sure it works with the camera.
Script shaping. A reasonable way is to write a text renderer that uses Pango primitives.
There can be more... Some might think that there are items we can do for B-Test1, but I think we are now safe to start concentrating on the B-Test2.
Any suggestions and feedback is welcome.
-- Yoshiki _______________________________________________ Etoys mailing list Etoys@laptop.org http://mailman.laptop.org/mailman/listinfo/etoys
Andreas,
Thank you for responding!
This list looks like a great start. I agree with all of these topics as they correspond to issues that we need to address for one reason or another. The two extra issues I'd put on this list are:
- Generic "sensor support" (along the lines of world stethoscope) if
that makes any sense at all (e.g., if OLPC will have the means to plug in cheap sensible sensors)
Yes. Markus Gaelli expressed the interest in sound support, and we should leverage World Stethoscope, the sensor, and some tone modification.
- OLPC-centric interface, meaning that we need to spend some time
integrating the OLPC-specific means for opening projects, sharing them (both locally and remotely) providing means for data exchange between activities etc.
The sharing part of this probably includes the digital signature issue.
-- Yoshiki
Yoshiki Ohshima wrote on Mon, 30 Oct 2006 10:08:45 -0800
Now seems to be a good time (even though the B-Test1 schedule seems to be pushing back) to think about the next major items to be addressed for B-Test2 (and later) milestones.
It seems to me that the two largest issues to be addressed before the final release (middle of next year?) are IPv6 support and dealing with Arabic and Thai scripts. Neither is very easy.
-- Jecel
Am 31.10.2006 um 14:39 schrieb Jecel Assumpcao Jr:
Yoshiki Ohshima wrote on Mon, 30 Oct 2006 10:08:45 -0800
Now seems to be a good time (even though the B-Test1 schedule seems to be pushing back) to think about the next major items to be addressed for B-Test2 (and later) milestones.
It seems to me that the two largest issues to be addressed before the final release (middle of next year?) are IPv6 support and dealing with Arabic and Thai scripts. Neither is very easy.
Also, some performance improvements would not hurt. For example, it would be good to support the 5-6-5 bitmap format used by the X server directly. That would mean we need to distinguish between 15 bpp and 16 bpp - until now, 16 bpp meant 5-5-5.
- Bert -
Am 31.10.2006 um 15:06 schrieb Bert Freudenberg:
Am 31.10.2006 um 14:39 schrieb Jecel Assumpcao Jr:
Yoshiki Ohshima wrote on Mon, 30 Oct 2006 10:08:45 -0800
Now seems to be a good time (even though the B-Test1 schedule seems to be pushing back) to think about the next major items to be addressed for B-Test2 (and later) milestones.
It seems to me that the two largest issues to be addressed before the final release (middle of next year?) are IPv6 support and dealing with Arabic and Thai scripts. Neither is very easy.
Also, some performance improvements would not hurt. For example, it would be good to support the 5-6-5 bitmap format used by the X server directly. That would mean we need to distinguish between 15 bpp and 16 bpp - until now, 16 bpp meant 5-5-5.
One more - we need an idea what should happen when "sharing" the Etoys activity.
- Bert -
Hi
- Address the tickets on the dev.laptop.org. Some short-term items are listed on the trac system. Takashi says he is looking at the gettext. (I'll list some items from Jerome Peace.)
I think it is good opportunity because translator's effort will be significant reduced. Current etoy's translation sysytem is a bit broken. I am getting started this issue. I think it includes these things.
- "Untranslated items" should be automatically generated by source code: Because there are a number of unused untranslated items in the dictionary in Squeak. In gettext way, untranslated items are extracted by _() macro. Also, those should be extracted by #translated method in source code of Squeak, originally. But it is broken because sometimes receiver of #translated is a variable instead of a "literal". It would be easy to fix even though there are many of such codes.
- Better gettext importer / exporter: I uploaded Diego's gettext exporter again in the image. But I think it is good if it supports multiple line, or handling header information.
- Translation for a project: Current translation dictionary is for the image. But how about a project? If we attach a dictionary by a project. One tutorial project file can be read in different languages.
- certainly, web interface for translation is useful.
Cheers,
- Takashi
etoys-dev@lists.squeakfoundation.org