Avoidable delays

I’d forgotten to post this bunch of bits, this rant I wrote on the 16th.   Better to have this up among the bitclouds than on my laptop’s suspiciously slow, possibly damaged hard drive.

I was browsing news about results of the senatorial race, when a paragraph in yesterday’s article by Tricia Aquino caught my attention:

“Many of the BEI (Board of Election Inspectors) of the clustered precincts, after having a hard time [transmitting], or else did not try hard enough to transmit the PCOS election returns from their precincts… took their flashcards and carried these to the municipal and city canvassing centers,” said PPCRV chairperson Henrietta de Villa after consulting with Commission on Elections Commissioner Christian Lim about the slow pace.”

“Are they even supposed to be doing that?” I thought, taking the flash memory cards (presumably containing election returns) out of the PCOS machines? The Smartmatic website specifically says their devices are “designed to transmit the electoral data in a secure and unidirectional manner, with no need of human intervention, and in accordance to the requirements of the electoral body.”

These communication networks either work or they don’t; there is no “try”.  Unless the PCOS devices are transmitting results using the cell phone network, and they don’t have a signal – or are unable to “register” to cell towers.

“So, why aren’t they able to transmit?”  It doesn’t make sense that local protests would delay transmission of results via the data cellular network, as Commissioner Brilliantes explained yesterday (unless those protests involved burning cell antenna towers).  I can’t imagine why I’m still seeing under 75% of all returns from the unofficial counts being reflected in news summaries, as PLDT’s Mon Isberto has said that their network remained fully operational, so that “as a result, transmission of election returns that have been coursed through our facilities have been handled with dispatch.”

 

Since I can’t trust the marketing pages on Smartmatic’s site (and because the COMELEC website gives no technical information about how the system is intended to work), I have to wonder: Is it safe to trust that the content of those PCOS memory cards, having been removed from the PCOS machines, can’t be altered?  We need to know whether the election returns data on those cards is encrypted, and whether that encryption can resist being reversed, or “broken” during the time that those returns aren’t processed in the normal course of events.  The longer it takes for those results to be reflected in the national tally, the higher the likelihood that an attempt to “brute-force” the encryption (which I can only assume exists) on those cards will succeed.  If the content can be decrypted, it can be modified.  And with stakes this large, the cost of this technical hack is small beans, in relative terms (think several dozen computers, or the cost of hiring of a crooked engineer).

 

And why should transmission using any of the GSM networks even be a problem at all?  Didn’t COMELEC avail of cell tower coverage maps from telcos during their logistics planning for these elections?  Did they not consider having an alternative data transmission channel – say, landline DSL modems – as a backup, for voting precincts that have unreliable cell service?  It’s not like the precincts move around, so that it’s hard to tell in advance that this might be a problem.  They can’t be unaware that cell phone jamming equipment can be bought from overseas, so that provision has to be made against the possibility that some idiot would try to disrupt the electronic canvass. The only memo I found on the COMELEC site dealing with technical issues doesn’t describe what to do in the event of transmission failures.

If all this sounds like a backhanded critical swipe at the COMELEC, it can’t be helped; not after hearing the head commissioner dismiss the usefulness of a source code review by saying the PCOS machines ‘reads codal language’, not the source code.  There are no ‘codal instructions’, only ‘machine instructions’; that hamfisted effort to explain (i.e. dumb down) the discussion, the imprecision, leaves me cold. And more than a little angry, in fact, for which I offer no apology.

The COMELEC could have avoided these problems with suitable technical preparation. Some of which they could have taken with advice from a much larger audience: us, the voting public.  With 50 million registered, adult voters, there was no need to dumb down and minimize public discussion of how the electronic voting process was intended to work.

There are thousands of network engineers from whom COMELEC could have elicited feedback on the design of the overall system, and who would have identified this possible source of delay – and a means to work around it (i.e. get the supplier to provide dialup modems).  Thousands of programmers who work with embedded systems, who could have told the rest of us citizens whether or not the system could be trusted with our vote – and whether the integrity of the ERs would be preserved despite removing the flash cards from the PCOS devices.  Goodness, it’s 2013 – they could have used their website to publish a forum, for people to discuss the technical implementation.  The inability of COMELEC to be transparent, and to elicit citizen feedback is, right now, sorely testing the integrity of the vote.

Setting up

I’ve ripped apart my copy of Wylie and Barrett’s Advanced Engineering Mathematics in nice signature-length sections, and have turned it’s first four chapters into nice iPad albums. Ditto for Harry Lass’s Vector and Tensor Analysis.

Over the next four weeks I’ll find either that I’d just wasted eight hours (and one perfectly good copy of W&B) of my previous weekend, or that I can still recover my old maths chops and get on with writing my own MRI scan image synthesis library.