Is the Simputer like a PC?
No. The Simputer is *NOT* a personal computer. It could however be a
Is the Simputer like a Palm?
Again no! The Simputer is much more powerful than a Palm. For
example, in terms of screen size (320x240), memory capabilities (32MB
RAM) and the OS (GNU/Linux).
How do I enter text? Can I attach a keyboard?
There are two options on the base simputer for entering text: one is
a soft keyboard, that can be brought up onthe touch screen and you poke
at it to enter one character at a time. The second option is to use a
novel character entry software called tap-a-tap which is similar in
spirit to graffitti, but quite distinct (no patent infringement:-). But
if you insist on entering tons of text using the Simputer, you may be
able to attach a USB keybaord. We don't recommend the Simputer as a mass
What features set the Simputer apart from other handhelds?
The smart card reader in the Simputer. The Information Markup
Language (IML) that is (amongst other things) smart card aware. The use
of extensive audio in the form of text-to-speech and audio snippets.
Can I buy a Simputer from the trust?
The trust does not manufacture the simputer. It will license it to
When will the Simputer be available?
If all goes well, by March 2002 you should be able to buy one of
What will the Simputer cost?
We expect the Simputer to cost about Rs 9000 when the volumes are
upwards 100,000 units.
What does MAIT have to do with the Simputer?
Nothing! Except that one of our trustees happens to be the president
of MAIT. Also, one of our trustees is a life member of the KSCA but the
cricket association has nothing to do with the Simputer either :-)
What processor does the Simputer run on?
It runs on an Intel strong-arm chip. The chip is known for its low
Can I create a Beowulf cluster using many Simputers?
You must be a /.er; in which case you know the answer!
What does the simputer run on?
Three AAA battries or off the mains. It can also use rechargeable
batteries, but the charger is not built in.
What is IML?
IML stands for Information Markup Language, though it could well be
an Illiterate Markup Language. The language has been created to suit the
unique needs of the Simputer. The markup itself is however, XML based.
Another Markup Language? Why not HTML?
(a) Because HTML is not currently as versatile as IML.
(b) Because we don't control the standards and hence cannot make the
necessary changes tailored to our requirements
(c) And IML is an XML application, and so does follow the Internet
Doh! What about it?
Note: Several of the following questions were posed by
firstname.lastname@example.org who maintains www.opencollector.org.
Are there any problems in applying a GPL-like license to a design
that ultimately derives from an Intel reference design? Or will this
design be replaced?
We don't see any problems. The Intel reference design is just that, a
reference design. Intel holds no claim to the products developed using
their reference design. Our design is very different to the Intel
reference design. Where the two designs might look alike, as in how the
Flash is connected to the SA1110 for example, there is only one way to
do it and this is clearly indicated in the data sheets, apart from the
The Simputer license requires users to ensure that 'no third party
can receive or read the specifications from you without having first
read and agreed to the terms of this SGPL'. Other groups working on free
hardware design licenses have suggested such things before (basically,
using trade secret law) but the counter argument has been that it will
not be possible to seek redress against the person who receives the
specifications in this situation, only against the person (possibly
unknown) who passed them on. Do you see the use of such clauses (in
addition to copyright law) as essential to protecting free hardware
The problem we faced in coming up with a suitable protection model
for the Simputer is the fact that being essential a hardware
specification, the elements sought to be protected were not strictly
copyrightable. Consequently, protection mechanisms such as the copyleft
principle used in the GNU GPL do not fully apply. It was therefore
important that the Simputer GPL utilised a stronger mix of copyright and
contract than was used by free software licenses. Ultimately, this is
the trade-off. The Simputer GPL, in order to protect the unique
intellectual property of the Simputer had to accommodate the
shortcomings of trade secret law.
Having said that, we do not feel that the Simputer GPL is any less
enforceable than other more traditional GPL's. For one, every person who
uses the specification is deemed to have read and agreed to the SGPL.
While we do make the transmitter of the information liable to disclose
the specification along with the license, it would not be a open for a
recipient to say that he/she is not bound by the terms of the SGPL
merely because the version he/she received from an unnamed third party
did not include the SGPL. The only exception to this liability is
someone who, using clean room operating procedures, comes up with
something similar to the Simputer - but that is a problem that all
patent holders face as well and we have no special solution to suggest.
Finally, since the SGPL contains a mix of protections under
copyright, patent trademark and trade secret law, we feel that when all
these components and brought into play simultaneously the SGPL will
achieve its goal of protecting the Simputer specifications
There seems to be a possible tension between the use and creation of
free software and designs, and the goal of development of an indigenous
IT industry. How do you see the relationship developing?
We see no conflict, but synergy. Linux has shown that software
capabilities have only increased with the availability and use of Unix
in the early days of computing in India and more recently with Linux. By
making available the SImputer hardware under SGPL, we believe that the
barrier to entry to hardware system development will be lowered.
There have been a few free designs of rather similar type before: the Japanese Morphy One (http://www.morphyone.org/), the Pengachu (http://www.media.mit.edu/
~rehmi/pengachu.html) and more recently the Brazilian 'VolksComputer' (though in this case it is unclear if the design itself will be free). Can such projects build on one another, or are the differences in national needs too great?
We believe that uniquely local needs require uniquely local
What steps can you take to keep the costs of such a device as low as
SGPL is one such mechanism. Since multiple players will hopefully be
involved in the manufacture of Simputers, volumes will increase and
costs will be brought down due to competition. Businesses would make
money on solutions, rather than base hardware.
$200 still sounds like it might be expensive for poor communities -
will the government be providing financial aid for purchases?
We hope government and large multilateral organizations will use the
Simputer as a platform for various IT initiatives, indirectly making it
affordable for poor communities to get access to Simputers.
We have also recognized that even $200 could be too high and such
products may need to be subsidized. However, we have added a SmartCard
as a prime method of enabling the "sharing" of such devices. Rural
communities could own several devices and hire these out for usage to
individuals based on the ownership of a SmartCard. Each user's Smart
Card would contain the minimum "personalization" information required to
log into a Community Server which would maintain personalized data about
the user. You can treat this as some sort of "roaming profile"
information maintained in a smart card.
This model of sharing would bring down the cost of the Simputer to
that of owning only a simple smart card, and paying for the usage of a
Shared Simputers could be made available in rural schools, community
halls or other such areas where common facilities are usually found.
There is often a technical conflict between low-cost and
repairability (for example, I imagine you will not be socketing
devices!). Will the Simputer be more of a throwaway or a repairable
device; and if the second, how will you achieve this?
No devices are socketed. We haven't really worked on the
manufacturability/maintainability aspects of the Simputer yet. However,
in developing countries nothing is "throwaway". We expect proliferation
of Simputers to encourage spread of associated spare
parts/repair/maintainance activities. You can see that we are real
Why did you decide not to include a storage device (other than the
smart card) in the list of potential add-ons, especially where network
links may be unreliable?
First of all, the Simputer is more likely to be used in a
client/server environment where it is primarily an access device and the
Storage requirements will be met by storage at the Server. Internal
storage would have consumed substantially more power and real-estate.
The Smart Card should not be seen as a storage medium of any
significant capacity, though capacities could increase as technology
advances. It is better viewed as a "personalization" and security
Internal storage, of limited capacity, is already available through
Flash memory. Though limited in size, it is still substantially higher
than the Palm-based PDAs that are so popular.
We have also decided to use the USB as the medium for access to
external peripherals. Products like the M-Systems Disk-on-Key Flash Disk
are now available on USB. They can provide reliable storage ranging from
16MB to 1GB in capacity. So we do not see the lack of an internal
storage as a major problem.
Will applications downloaded over remote links be charged for on an
Application Service Provider model, or will these also be free?
Both models are possible. The Simputer platform is not just a
platform to take IT to the masses, but it is suitable for profit-making
enterprises. Thus the Simputer is aimed at reaping Digital Dividends
from the digital divide.
Your documentation mentions the 'tapAtap' as a means of text entry.
What is this?
tapAtap is a character-input application, similar to but distinct
from graffitti. Characters are entered using a 3x3 grid by tapping grid
Is there a risk that use of the custom IML will restrict the
services available to users?
Is it possible to obtain an IML implementation to use on other
platforms? Instead of 15 and 16, I have added a few questions that
someone else asked of IML and their answers.
What are the most useful things that people outside India can do to
Give it a critical appraisal. Use the Simputer if found appropriate.
Propagate SGPL-like models for other hardware components.
What are the special features of the IML-based browser you have
developed for the Simputer?
The novelty is in the Information Markup language (IML), the browser
that we have is one implementation. we designed the IML after we
concluded that the user-interface of the simputer must have the
- uniformity across diverse applications
- ease of use
- support for multilingual text and speech output
- support for smart card usage
- transparent access to remote/local resources.
The further requirements on IML were:
- ease of application development
- use of internet standards
- platform independence
What is the application development environment that you are
offering on the Simputer?
Is there a simulator available for the Simputer so that I can start
Application development for the simputer can be done on any platform:
linux, windows, solaris, MacOS. This is because any simputer application
can be viewed as a blackbox that reads in IML and outputs IML. Thus you
do not need to own a Simputer to create applications for it. The basic
Simputer itself is poorly suited for application development, since it
does not have a keyboard or a large display.
Further, IML browsers can run on any linux machine or Windows machine
at this time, and can be written easily for other platforms. If you hav
such a browser, the application development can proceed without
consideration of the underlying hardware.
In addition, any X application that restricts the display to 240x320
will directly run on the Simputer. For example, we have xclock, rxvt,
xeyes and other standard applications already running.
Have you insulated the application developer from the underlying
hardware design? If so how?
Would applications already developed on other handhelds like the
Palm and Handspring be portable to the Simputer environment? If so, how
have you ensured that?
We have not looked at the porting of such applications for several
reasons. First, we do not have expertise in palm application
development. Second, the application scenario that we initially envision
for the simputer is quite different from that of the palm. Third,
internally, the simputer is a full-fledged linux machine and not a PDA.
Has the use of Linux kernel helped in terms of time to market, cost
savings, savings on code size etc? Please elaborate.
Most certainly. The code being developed for Intel's Assabet board,
as well as the code developed for the LART project has been invaluable
to us. The touch screen driver came from the Compaq iPAQ effort. We have
been downloading the kernel upgrades and device drivers, and making
changes to reflect our hardware interconnects. To write these from
scratch would make it infeasible to even conceive of such a project.
Savings on code size is not our immediate priority, since there is
generous amount of RAM and flash memory on the simputer.
A design objective for the Simputer was partition the design heavily
in favor of software. What were the advantages of this design paradigm?
Linked to the previous question. Linux and the free software movement
are a critical part of the Simputer vision. However, we also intend to
bring in a similar model for hardware.