atom feed21 messages in nz.co.katipo.lists.kohaRe: [Koha] LibLime Enterprise Koha Q&A
FromSent OnAttachments
Joshua FerraroSep 15, 2009 2:15 pm 
Kyle HallSep 15, 2009 2:35 pm 
Chris CormackSep 15, 2009 2:52 pm 
Chris NighswongerSep 15, 2009 3:08 pm 
Joann RansomSep 15, 2009 3:14 pm 
Piotr WejmanSep 15, 2009 3:26 pm 
Brendan GallagherSep 15, 2009 9:07 pm 
paul POULAINSep 16, 2009 1:15 am 
Liz ReaSep 16, 2009 11:01 am 
Daniel GrobaniSep 16, 2009 2:19 pm 
Reed WadeSep 16, 2009 3:00 pm 
Chris NighswongerSep 16, 2009 3:25 pm 
Owen LeonardSep 16, 2009 6:51 pm 
paul POULAINSep 17, 2009 12:44 am 
paul POULAINSep 17, 2009 1:11 am 
Thomas DuklethSep 18, 2009 9:36 am 
Thomas DuklethSep 18, 2009 9:43 am 
Thomas DuklethSep 18, 2009 10:44 am 
MJ RaySep 22, 2009 2:56 am 
paul POULAINSep 22, 2009 6:22 am 
Joe AtzbergerSep 22, 2009 8:05 am 
Subject:Re: [Koha] LibLime Enterprise Koha Q&A
From:MJ Ray (mj@phonecoop.coop)
Date:Sep 22, 2009 2:56:05 am
List:nz.co.katipo.lists.koha

"Thomas Dukleth" <koha@agogme.com> wrote:

On Thu, September 17, 2009 07:44, paul POULAIN wrote:

Daniel Grobani a écrit :

"We will not be making our git repo public as it contains customer-sensitive data."

Just a short answer: That's a so poor, so illogic, so technically irrelevant, and Josh is a so smart developer, that I simply can't believe LL made it that way !

I can believe that people do many poorly conceived things when they are in a hurry. I also know that Git is a very robust tool which can be used to manage much more than source code data.

Yes, this is a main reason for the difficulty of republishing private forks. Not only do you have to make sure that the git HEAD doesn't contain confidential data that was added in error, but also that the visible history doesn't show it.

With things like Electronic Data Interchange and securing RFID terminals on private networks, and (not wishing to insult past helpers) with contractors who aren't familiar with the all-is-open FOSS/co-op way of working, it's disappointingly common for things like account details or encryption certificates to end up in the code. They shouldn't be there, but on a non-FOSS ILS, it probably wouldn't matter as long as they get cut out before the full release.

The punishments for publishing personal data are scary enough that to be sure, I've now resorted to making patch files and applying those to the gitorious tree that chris has kindly made available to us. It's a damn slow process and we only had 3 workers on Koha at any time - I shudder to think how big a reconcilement Josh faces - but we are committed to making our Koha a proper public superset of the mainline and we will get there. I thought the US privacy laws were less severe than ours, but I guess companies are more litigious and would sue if LibLime published any account details.

Even so, our fork was only a temporary convenience - which has actually turned out to be inconvenient enough that we'll not do it like that again - and we are committed to merging. It would be really good to see a similar commitment to merge with the mainline from LibLime and not just publish periodically.

Then there are the foundation and asset management questions for LL...

Regards,