In the last couple of weeks I've had to erase the stateful partition on several occasions, upgrade to the Beta Channel and even upgrade to the Dev channel (which, by the way, happens to be very unstable). In addition, I've had to block automatic updates, which is the same as saying go to manual override. It's a mercy that I haven't had to reverse the polarities or change the Dilithium Crystals. No, I'm not re-enacting an episode 1960s' Star Trek, rather I've been trying to get my Chromebooks to work.
Thursday, 15 December 2011
Chrome Os and its Serious Product Failure
I know that in my last post about Chromebooks I wrote that I intended to delve into the depths of the Google Apps Management Console, however the next day they stopped working.
You see Chromebooks update themselves automatically, as many things do these days, but unfortunately, version 15 didn't work. Version 14 had worked fine, swimmingly even. You might possibly say dreamily. The students, even after two sessions had started to: Love. Their. Chromebooks. But version 15 let us down. A crucial part of the Chromebook experience had stopped working - the sign in screen. So I upgraded to the Beta channel - version 16 and that did work. Then to the Dev (Development) channel but that didn't work either.
It was something to do with the proxy setup in my school and authority. You see, Birmingham runs a system where a local 'Squid' server, based at the school, links to a central parent server somewhere in the depths of a shady building in the city centre. With this setup the local server caches the internet sites that students visit. This means that the first student that visits a website will bring the content both onto their computer and onto the Squid where much of the content is cached. Any students that follow up will then have a much faster experience because most of the content can be delivered to them down the metres of cable on a 100Mb connection, rather than down miles of fibre on a 10Mb connection. The central parent server has the job of filtering out unwanted websites, which it generally does a very good job of doing.
Something to do with that setup didn't work in Chrome Os 15 - it just didn't hold any of the proxy information or network information, so any new user wanting to use the Chromebook couldn't do so. The Chromebooks were dead in the water. As the guy at Google told me - it looks like a serious product failure.
However the marvellous thing about Google is that their product support is absolutely brilliant. If you ever want an experience of being spoken to politely and humbly by people who really know what they're talking about - go talk to Google product support. Within a couple of days some flash drives containing version 14 of Chrome Os were winging their way towards me from Dublin. The instructions, which included the use of the paperclip shown in the photograph, demanded that I block the update server: it would be a nightmare if I fixed the Chromebooks by downgrading them to version 14, only for them to re-update themselves to version 15. The technical guys in Birmingham were equally quick at sorting this - responding to the request to block the update server within a day and also providing me with some insightful tips on how proxy servers work (which in fact helped me right the paragraph above).
It amuses me somewhat that, after all that technical stuff about proxy servers, it should be a paperclip that I would need to sort the problem out. I've all but had the Chromebooks inside out over the last couple of weeks and it has finally been a paperclip (and version 14 of Chrome Os on a flash drive) that has fixed them.
Onwards and upwards then - I'll have a full set of working Chromebooks by tomorrow and I'll look forward to version of 18 of Chrome Os with considerable anticipation.