[Live-demo] Xubuntu 13.04 build test

Angelos Tzotsos gcpp.kalxas at gmail.com
Sun Apr 28 04:55:39 PDT 2013

working iso:
logs for the above build:

logs with our scripts enabled, causing failure:

On 04/28/2013 02:24 PM, Angelos Tzotsos wrote:
> Good news, I managed to get a custom iso working from 13.04 by 
> stripping down our build script to bare minimum.
> So we are back in business :)
> Cheers,
> Angelos
> On 04/28/2013 01:10 PM, Angelos Tzotsos wrote:
>> Hi,
>> I tried to make another build with 13.04, this time turning off all 
>> our scripts.
>> The result was the same, a broken kernel system.
>> It seems that our build process does not work for the new ubuntu 
>> versions, or needs updating.
>> This actually means that we would need to deal with it sooner (13.04) 
>> or later (14.04).
>> I found this new help page:
>> https://help.ubuntu.com/community/LiveCDCustomizationFromScratch
>> We originally used this page:
>> https://help.ubuntu.com/community/LiveCDCustomization
>> Angelos
>> On 04/26/2013 12:38 AM, Hamish wrote:
>>> Angelos wrote:
>>>> Personally I have not used LXDE, but I have heard that it is
>>>> very limited (some people call it a non-deskop but of course
>>>> DE flame wars are well known...).
>>> At least the 12.04 offering is fully featured and mature enough,
>>> the only thing that seemed half-baked was the login manager, but
>>> you only got that if you manually switched to it, they were using
>>> lightdm by default I think.
>>>> On the other hand, staying to 12.04.2 would be a safe
>>>> choice, it is pretty stable at the moment....
>>>> So our options:
>>>> 1. Xubuntu 13.04 (alpha1)
>>>> 2. Lubuntu 13.04
>>>> 3. Xubuntu 12.04.2
>>> I'm pretty easy with whatever choice people like to make from
>>> those.
>>> One big advantage of going to 13.04 is it means we have less
>>> package overrides and external ppas to deal with, which means
>>> a lot less work and packaging problems to deal with.
>>> Another is that it takes care of Cameron's drop the dead-wood
>>> idea without us having to start making ugly popularity contest
>>> decisions. A sure sign of a healthy project is one that provides
>>> updated packages in a timely fashion. Other metrics like Oholo
>>> activity have similar failings to using SLOC output to assess
>>> employee productivity. I'll again trot out PROJ.4 as the classic
>>> example of highly mature code which looks bad if your metric is
>>> based on change frequency, but is probably used by many millions
>>> of people every day if they know it or not; not many bug fixes
>>> as there aren't many bugs left.
>>> But it's probably still too early to say; time will tell if
>>> 13.04 is a good one or not. Since the early retirement date
>>> we'd have to upgrade again before too long.. so we would be
>>> making a lot of work for ourselves. :-/
>>> 2c,
>>> Hamish

Angelos Tzotsos
Remote Sensing Laboratory
National Technical University of Athens

More information about the Osgeolive mailing list