[Building Sakai] 2.9.0 DB Conversion, MySQL script error

[Steve Swinsburg]

So how is a new developer meant to know all of these unwritten rules if they aren't described anywhere? When starting out, we point everyone to the programmers cafe and the install guide, but then they are to also follow another set of rules that they don't know anything about? That doesn't fly for an open community.

I disagree that there are too many requirements. To get up and running with Sakai you need to: Install some supporting software which is probably on your system already, get the code, build it and run it. Everything else is customisation from the default.

I'm not fussed about the recommendation being a rule. If the guide is followed as is, your installations won't break. uPortal has similar wording in their install guide: https://wiki.jasig.org/display/UPM/03+MySQL+Configuration 

> The case of the columns is totally irrelevant to both Oracle and MySQL.  
>  

My point is that if the rule you proposed was enforced on table creation, not just updates, you wouldn't have that mess. 






On 25/01/2013, at 11:19 AM, Sam Ottenhoff <ottenhoff at longsight.com> wrote:

> 
> 
> 
> On Thu, Jan 24, 2013 at 7:06 PM, Steve Swinsburg <steve.swinsburg at gmail.com> wrote:
> That rule is not written anywhere for Sakai developers to follow, it is simply a best practice. Whereas the recommendation to use lower case table names on a MySQL setup is. And has been in the install guide since 2007.
> 
> 
> We have lots of unwritten rules as a community Steve.  Not breaking default MySQL Linux installs seems like an easy one.
> 
> If you want to pursue turning the recommendation into a rule, go for it.  I think we have enough install requirements already.
> 
>  
> And the way the 'rule' is written means it applies only to modifications to a table, not to the initial creation of the table. Why shouldn't the initial table creation be standardised on one case? Why only make people that modify a table follow the rule?
> 
> 
> Because we're an agglomeration of many tools from many parties.  It would be great if we had a set of standards to begin with, but we don't.
> 
>  
> 
> Perhaps then it could avoid SQL mayhem like this table, which has every combination of case type in its fields:
> 
> describe chat2_channel;
> 
> 
> 
> The case of the columns is totally irrelevant to both Oracle and MySQL.  
> 
>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/sakai-dev/attachments/20130125/255519a4/attachment.html