Meer privaatheid met twee webblaaiers

Ek is vroegaand dié week by 'n vriend se huis toe my Google-selfoon my daaraan herinner dat ek so vier minute nodig het om by my volgende afspraak uit te kom. Hoe weet Google, Facebook en ander so baie van ons en wil ons iets daaraan doen? Oor die selfoon kan ek nie veel sê nie maar ek glo dis daar 'n hopelose saak, soos Richard Stallman onlangs in 'n lesing by Wits verduidelik het. Op 'n gewone rekenaar het 'n mens egter 'n bietjie meer hoop. Die beste is om TOR te gebruik vir Internet-toegang maar in my ervaring is dit dikwels maar stadig.

'n Baie eenvoudiger manier om jouself 'n bietjie minder te laat naspoor is om eenvoudig twee verskillende webblaaiers (byvoorbeeld twee van Google Chrome, Opera, Firefox, Safari of Internet Explorer) te gebruik. Op die één, sê maar Google Chrome, gebruik 'n mens alles waarvoor jy moet aanteken, byvoorbeeld jou GMail en Facebook. Moet dan geen ander webbladsye in Google Chrome oopmaak nie! Wanneer jy dit sou doen, kan GMail en Facebook baie inligting oor jou bekom en die webbladsy wat jy besoek kan ook. Gebruik eerder byvoorbeeld Opera om ander webbladsye in oop te maak, insluitend jou bank. Jy sal sien dat daar dan geen +'e van Google of Facebook-boodskappe wat jou vertel wat jou vriende van hou meer in Opera gaan verskyn nie. Hierdie werk omdat webbladsye wat jy in Google Chrome oop het slegs inligting van en oor ander webbladsye in Google Chrome kan kry of uitruil maar nie daarbuite nie. Alles wat jy dan in Opera doen hoort dus nie direk deur Facebook of Google aan jou gekoppel te kan word nie.

When neither isn't at Kulula.com

The options are A, B, neither which is not the same as none of the above... Seriously?

If this is a Kulula joke, it is now becoming very tired!

Invlug-Internet op Emirates se A380

My perfekte somervakansie is verlede week afgerond deur 'n vlug op die boonste dek van 'n Emirates A380 vanaf Hong Kong na Doebai. Ek was heel gretig om die Internet-fasiliteit tydens die vlug te gebruik maar die WiFi-netwerk het 'n mens na 'n betaalbladsy herlei. Sover geen probleem, maar die tarief van $7,50 vir 5 megagreep het my laat terugdeins. Ek het later gelees dat dié diens, soos die aanboord-selfoonnetwerk, verskaf word deur middel van 'n enkele satellietverbinding en dit is natuurlik verstaanbaar nie goedkoop nie hoewel ek dink dit sou gaaf wees indien Emirates byvoorbeeld 'n sekere hoeveelheid data gratis per passasier sou verskaf (veral vir dié wat bo sit)! Nietemin, met Vodacom sou dieselfde hoeveelheid data op 'n buitelandse netwerk 'n mens R640,00 [sic] kos en ek verstaan dat die Emirates-diens byvoorbeeld beeldmateriaal saampers om die dataverbruik te verminder. Dit is seker nie hopeloos onredelik om vir $7,50 per (lang) vlug 'n WhatsApp-verbinding op die selfoon te hê nie. Ek sou graag weet hoeveel van die beskikbare kapasiteit van die satelliet-verbinding van vliegtuig na grond in praktyk deur Emirates-passasiers gebruik word.

When Unison just does too much

Unison File Synchronizer has served me well over many years, helping to keep my work and personal life up-to-date on several computers in at least three different locations. Lately, however, I have been trying to keep a very large collections of large files in sync on two different external hard drives and Unison is just too slow for this because it is so thorough. If you are essentially just adding files to a collection, the much easier way to keep directories A and B synchronized is

cp -nrv A/* B/
cp -nrv B/* A/

which simply places all files in (subdirectories of) A/ in B/ and vice versa, skipping files that already exist. If the transfer is interrupted and restarted however, it will not resume the copying of a file that had been partially copied earlier but otherwise this is a very simple and easy way to get the two-way archiving job done!

Lightspeed on Google Latitude?

The sometimes erratic location reporting in Google Latitude is interesting, not least because I like to think about possible measures of consistency for the time-location pairs that it generates. This week produced a good example of a dodgy location report that should have been eliminated. I was at Sydney airport for several hours prior to the late departure of my flight to Johannesburg on Tuesday. My location was logged correctly inside the airport many times, including at the departure gate, the last time at 09:54. There was one more timestamp inside Australia, at 10:08 after which the next recorded location was roughly 15 hours later in Johannesburg. The last Australian location however, was near Brisbane, more than 1000 km from Sydney!

Now, except for NASA, no-one can get from Sydney to southern Queensland in 14 minutes so this timestamp is obviously inconsistent. Furthermore, the fact that it was followed by 15 hours of nothing would make it a very solid candidate for discarding. I am quite sure that the reason for it was a mobile WiFi hotspot (possibly on a mobile phone just arriving from Queensland) the location of which Google had mapped in Queensland. I have observed similar problems in South Africa with mobile WiFi devices but surely a bit of data processing could really improve things quite a bit!