A word about Jega's register
I am an open source software advocate. I was therefore very interested to learn that Professor Attahiru Jega's Direct Data Capture (DDC) machine - a fancy name for a basic tools collection - is an open source database programme, which runs on an Ubuntu Linux system configured on quite ordinary PC hardware.
The System Architect of INEC's OpenVR, Nyimbi Odero, is a Kenyan who has lived and worked in Nigeria for many years; all the other members of his volunteer project team appear to be Nigerians. I know all this because it was easy to find out. I simply asked a member of the registration team to click on the About menu item on the registration screen to enable me read its contents. Open source software philosophy permits me to do so.
The database contains the following fields: Surname, First Name, Other Names, Date of Birth, an indication of whether the birth date given is approximate, Gender, Address, Occupation, Phone Number, and national ID. Additionally, a fingerprint reader used to capture prints from all ten fingers, the results embedded into the database for each individual record, as is a small-sized snapshot taken from the camera attached to the PC.
I did not see that the DDC setup contained any active network components, that is to say what was recorded was not being transmitted to a central machine in a remote location, nor did I see that it was equipped with a GPS device to track the machine's changes in physical location, or to record the location of registration of each voter.
The first thing to determine, when conducting a registration is the eligibility of the intended registrant. I hadn't heard or read that the Electoral Commission had asked anyone to come to the registration centres with proof of age.
It's usually clear to see whether a person is an adult or under aged, but it is not as easy to tell that a woman is Kenyan and not Nigerian. I saw that Professor Jega had taken care of the adult test; he will ask and trust he will be supplied with the truth, hence the question about date of birth.
I was disappointed to see that Jega had made no attempt at all to ascertain the more difficult question about nationality, and place of birth. There is the National ID field in the database record, yes, but it doesn't count because a great many Nigerians have not been issued with one, and many who have do not walk around with knowledge of its number. This failure to properly determine eligibility opens the door for ineligible persons to register.
Despite that lapse, the process was simple and straightforward: A queue position-number given; fingerprints and photograph taken; questions answered; a registration slip containing a unique identifier and photo printed; slip laminated and transformed into a voter's card. And then, curiously, card details entered into a very large Manual Voters Register (MVR) along with the registrant's thumbprint.
Considering how simple the DDC hardware and software setup it seems a huge waste of money that INEC opted to purchase laptops whereas the same results can be achieved using much cheaper netbooks. The laptop in use was a Zinox, made in Nigeria version, definitely brand new.
Above all I was on the lookout for any clues that would indicate that a secure and trusted electoral system has been put in place; one in which it is possible, in the event of a dispute about election results, to verify them again and again.
In the context of our recent history, a trusted system should make the bizarre so-called re-run elections unnecessary. The About screen contained the reassuring, if sexist, slogan "One Man One Vote and One Vote Only" but there was not much else to suggest that such a system is now in place. The very useful Linux machines, which can aid such a process, appear thus far to be seriously under-utilised.
I have wondered given that the machines are not networked, what the fingerprint capture aims to achieve. Perhaps fingerprint verification will be a preliminary part of the voting process; that would be a good thing. It could also be that the captured data will be consolidated post-registration so as to detect multiple registrants, but considering that scores of thousands of DDCs are involved, that is a less than optimal approach likely to prove tricky.
The MVR also suggests a disturbing lack of confidence in the technology. And as with the last census where teams were sent out to gather data, I have also wondered whether safeguards were put in place to ensure that all of the captured data made it back to the Commission's offices, and that unauthorised copies were not made.
I do realise that Jega has worked very hard and is under some serious time constraints. Nonetheless, there is one thing I can urge him not to overlook. It's been reported that the voters register is to be posted on the Commission's website for verification. That will be helpful provided that the information is only accessible via a query to the consolidated database. To verify, one should simply enter his voter ID and the system shall return his name and place of registration.
But INEC should be wary of placing such large-scale sensitive information in the hands of our domestic criminals notorious for online fraud, and unknown foreign enemies.
Source:234next
The System Architect of INEC's OpenVR, Nyimbi Odero, is a Kenyan who has lived and worked in Nigeria for many years; all the other members of his volunteer project team appear to be Nigerians. I know all this because it was easy to find out. I simply asked a member of the registration team to click on the About menu item on the registration screen to enable me read its contents. Open source software philosophy permits me to do so.
The database contains the following fields: Surname, First Name, Other Names, Date of Birth, an indication of whether the birth date given is approximate, Gender, Address, Occupation, Phone Number, and national ID. Additionally, a fingerprint reader used to capture prints from all ten fingers, the results embedded into the database for each individual record, as is a small-sized snapshot taken from the camera attached to the PC.
I did not see that the DDC setup contained any active network components, that is to say what was recorded was not being transmitted to a central machine in a remote location, nor did I see that it was equipped with a GPS device to track the machine's changes in physical location, or to record the location of registration of each voter.
The first thing to determine, when conducting a registration is the eligibility of the intended registrant. I hadn't heard or read that the Electoral Commission had asked anyone to come to the registration centres with proof of age.
It's usually clear to see whether a person is an adult or under aged, but it is not as easy to tell that a woman is Kenyan and not Nigerian. I saw that Professor Jega had taken care of the adult test; he will ask and trust he will be supplied with the truth, hence the question about date of birth.
I was disappointed to see that Jega had made no attempt at all to ascertain the more difficult question about nationality, and place of birth. There is the National ID field in the database record, yes, but it doesn't count because a great many Nigerians have not been issued with one, and many who have do not walk around with knowledge of its number. This failure to properly determine eligibility opens the door for ineligible persons to register.
Despite that lapse, the process was simple and straightforward: A queue position-number given; fingerprints and photograph taken; questions answered; a registration slip containing a unique identifier and photo printed; slip laminated and transformed into a voter's card. And then, curiously, card details entered into a very large Manual Voters Register (MVR) along with the registrant's thumbprint.
Considering how simple the DDC hardware and software setup it seems a huge waste of money that INEC opted to purchase laptops whereas the same results can be achieved using much cheaper netbooks. The laptop in use was a Zinox, made in Nigeria version, definitely brand new.
Above all I was on the lookout for any clues that would indicate that a secure and trusted electoral system has been put in place; one in which it is possible, in the event of a dispute about election results, to verify them again and again.
In the context of our recent history, a trusted system should make the bizarre so-called re-run elections unnecessary. The About screen contained the reassuring, if sexist, slogan "One Man One Vote and One Vote Only" but there was not much else to suggest that such a system is now in place. The very useful Linux machines, which can aid such a process, appear thus far to be seriously under-utilised.
I have wondered given that the machines are not networked, what the fingerprint capture aims to achieve. Perhaps fingerprint verification will be a preliminary part of the voting process; that would be a good thing. It could also be that the captured data will be consolidated post-registration so as to detect multiple registrants, but considering that scores of thousands of DDCs are involved, that is a less than optimal approach likely to prove tricky.
The MVR also suggests a disturbing lack of confidence in the technology. And as with the last census where teams were sent out to gather data, I have also wondered whether safeguards were put in place to ensure that all of the captured data made it back to the Commission's offices, and that unauthorised copies were not made.
I do realise that Jega has worked very hard and is under some serious time constraints. Nonetheless, there is one thing I can urge him not to overlook. It's been reported that the voters register is to be posted on the Commission's website for verification. That will be helpful provided that the information is only accessible via a query to the consolidated database. To verify, one should simply enter his voter ID and the system shall return his name and place of registration.
But INEC should be wary of placing such large-scale sensitive information in the hands of our domestic criminals notorious for online fraud, and unknown foreign enemies.
Source:234next
No comments:
Post a Comment