[OLPC-Philippines] Sugar & XFCE

[Carlos Nazareno]

On Sat, Dec 6, 2008 at 4:35 AM, Edward Cherlin <echerlin at gmail.com> wrote:

Hi Edward!

>> 1) The Journal / lack of a real file manager:
> Simple remedy/workaround:
> In Terminal,
> yum install mc

Yes, I know about running midnight commander from the terminal and
have been using it for months now, but it is a "workaround". It is not
a replacement for a more modern file manager like that windows-like
UIs provide. It is neither a remedy nor a workaround because it
requires the user to drop to the terminal instead of it being just
launched from the Sugar UI.

Besides, doing "outside Sugar" file manipulation wreaks havoc on the
file indices the journal builds on local NAND and USB/sticks/memory
cards so this is very much problematic.

>> a) accumulation of too many no longer needed journal entries over time
>> which makes usage difficult
> Known issue. Alternatives in development.

Yeah, I gather from my Journal feedback a couple of months back. Just
mentioned these issues again as a refresher as to why the journal
needs more improvement in practice.

>> c) the difficulty in transferring multiple files, requiring the
>> terminal or MC to do extensive file transferring
> See file system workaround above.

Again it's a hacky-workaround but not a solution since it fails at
usability and built-in system integration.

>> 2) Apps need to be sugarized.
> Known issue. Alternatives in development.

News to me. Fantastic! :)

> This is not the purpose of the program. It is an education project,
> not a laptop or Linux marketing project. But there are workarounds for
> you. Just install one of the standard Linux distributions as an
> alternate boot on your XO. Or run Sugar on some other Linux system.

I'm afraid those are not an options for me as I signed on to test and
help develop edugames and to do a study on how to best author/develop
apps in Flash to run on the XO, and so I need to work within
"officially supported" OLPC frameworks on the XO as much as possible.

Our project proposal got approved by 1CC and we received 3 XO units towards it:
http://wiki.laptop.org/go/Projects/Flash_As_EduGaming_Platform

I'm still currently doing the rounds coordinating with fellow Flash
Actionscript developers here in the Philippines and doing testing and
investigation, after which, we'll invite other Flash developers
worldwide to contribute to the project.

> No Ministry of Education can put unsecured laptops in the hands of all
> of the children of a country. It is politically impossible. It is
> impossible to teach regular Linux to first-graders.
Regular Linux = terminal.

Gnome/KDE/XFCE = not regular linux! = noob friendlier than regular linux

With regards to traditional desktops, you'd be surprised at what young
kids are capable of.

Have proper extensive comparative usability studies been done on how
the XO's target demographic copes with Sugar and other traditional
GUIs?

> Software for elementary schools must be specially written regardless. Case closed.

Precisely! And that's why we need to make writing software content for
the XO much easier than the current process.

> We need to get Ministries of Education to contract with textbook
> writers and software developers for materials that meet their needs.
> We have had this problem since the beginnings of Computer Literacy in
> the 1970s, that everybody thinks they can pay for computers, and
> nothing else is needed--no teacher training, no specifically
> educational software, no integration with curriculum, no new
> textbooks--nothing. But they go right on paying print publishers for
> textbooks that completely fail to meet the children's needs.

Spot-on! THIS is the bigger and more relevant concern I see facing
OLPC and any other e/i/m-learning initiative.

I've begun talking with some of my old professors at Ateneo de Manila
University (http://www.ateneo.edu) and maybe coordinate with the other
top Universities about possible creation of content/open courseware
that can freely be distributed (copyright concerns: we can't just scan
existing textbooks and slap them onto the XO as PDFs) in the context
of the Philippine education setting.

>> The main difference was that back then in 2003 on my K6-2 Win98 rig, I
>> was running the Flash MX IDE, a text editor, the Opera 6 browser with
>> about 20 browser tabs open, Winamp and that CPU & Memory hog: Norton
>> Internet Security (antivirus + firewall) simultaneously.
>>
>> Barring virtual memory/swap space I don't see any reason why can't we
>> get similar performance out of the XO.
>
> You have hit on the essence of the problem. We are not going to run
> swap from flash memory. %-[ We may need to have a word with that
> fellow Moore about a $10 hard drive that draws less than a watt. ^_^

SD-cards are pretty affordable these days to the point that it becomes
trivial compared to the price of an XO (well, at least over here,
where we're near to Taiwan and China :P).

Any chance work can be done on the core OS using additional swap space
on an SD card if say, 2-4GB SD cards were to be used?

In fact, I'd like to see more support done towards running OSes
(including default XO OS) off SD Cards and to "ignore" NAND
altogether, because NAND wearing down & irreplacability is a real
eventual concern, and also the price of SD Cards are getting cheaper
and cheaper, thus solving the disk space problem of the XO.

> There are dozens of subnotebooks now with all of the features you are
> asking for, but with commercial markups on top. About half of the
> available offerings are for Linux, an unheard-of advance in the
> market. Check out liliputing.com. Enjoy.

It's a real credit and success to OLPC for creating the netbook/UMPC
ecosystem, and it's been said that if these other systems succeed at
OLPC's goals, then OLPC will have succeeded.

Nevertheless, it's the XO's success and usability towards the
education goals that we're more concerned OLPC devel-wise, aren't we?
;) I also hope you're not suggesting that OLPC abandon the education
niches/markets that all these other UMPC competitors are starting to
fill :)

> Actually, they can be shown the advantages of Sugar for the children,
> according to Walter Bender's experience. Nicholas Negroponte has read
> too much into questions about Windows from Ministry of Education
> employees.

Yes, but IMHO, Sugar is geared towards very young children, whereas
9-year olds and up are more than capable of mucking about with
Gnome/KDE/XFCE.

Heck, I've seen 6-8 year old street kids pool together money and
taking turns playing the Command and Conquer RTS game for an hour's
worth at an internet cafe. And ZOMGWTH: I've even gotten into a game
of Half-Life 2 Synergy mod (http://synergymod.net/) with a 5-year old
playing alongside us online and speaking on the mic (it was a
relatively disturbing experience, essentially a toddler playing
Counter-strike).

>> Now that alternative systems like the Deep Blue UMPC
>> http://www.ilikeblue.net/products/umpc.htm (1GHz Via processor, 1GB
>> ram, 40GB HD)
>> can be had for about $200 ($240 with WinXP) brand new in places like
>> here in the Philippines,
>
> It's a bit difficult to find details, but I see one model for $222,
> with resolution of 640x480. No thanks.

Actually, it's 800x480 (which is more than enough for educational
software and goals) and yup, that's the one:
http://technodrone.wordpress.com/2008/09/29/blue-h1-umpcnetbookmini-notebook/
and it's more like $200 these days.

And yep, XO's 1200x900 res just one of the reasons I still say it's a
superior piece of hardware despite the perceived lack of horsepower.

> Available is not good enough. To run standard computers, electricity
> has to be reliable and above all _cheap_ relative to the rest of the
> economy. In developing countries, it isn't, even in the cities.

Yes, but even at those speeds/power consumption, these other
power-hungry sub-laptops still consume less power than desktops and
makes them affordable enough for places like public schools here in
the Philippines (where the quality of education is a far cry from that
of private schools, and that cost of school materials, schoolbooks can
still be hardly afforded by a lot of the students from lower-income
families).

>> Delivering laptops to less-privileged students in more developed urban
>> areas (where efficient power usage is less of a concern) may not sound
>> as romantic as deploying them to infrastructure-poor remote rural
>> environments, but these places are no less relevant and laptops can
>> serve as an even more effective force multiplier because proper
>> support and infrastructure is more readily available.
>
> That's fine. Those schools can run Ubuntu or whatever on any hardware
> that suits them. That doesn't change the XO mission. Nor does it
> change the fact that there is no alternative as rugged as the XO, or
> that matches up to some of the XO's other virtues.

Agreed, and I still say that the XO more than fills enough this niche ;)

Cheers!

-Naz

-- 
Carlos Nazareno
http://www.object404.com

interactive media specialist
zen graffiti studios
naz at zengraffiti.com