[Building Sakai] Deprecation summary page

Clay Fenlason clay.fenlason at et.gatech.edu
Wed Mar 24 07:29:44 PDT 2010


Some comments in-line:

On Wed, Mar 24, 2010 at 10:14 AM, Aaron Zeckoski <azeckoski at unicon.net> wrote:
>> 4) I confess to a bit of uncertainty about the state of the two
>> Profile tools. I was under the impression in December [2] that the two
>> profiles couldn't work together, but it looks like Steve managed to
>> complete the work that allowed them to do so [3]. And yet something
>> about Aaron's description of removing remaining dependencies to
>> Profile 1 makes me wonder if in fact they can still be swapped with
>> sakai properties. Can someone confirm?
>>
>> In any event, since Steve says they can work together, I've altered
>> the proposal to simply say that Profile classic should be stealthed
>> and marked as deprecated.
>
> The issue is not with profile 1 itself but with other tools like
> roster that have a direct dependency on the profile 1 tool. Aside from
> tool to tool dependencies being a poor practice it also makes it
> impossible for someone to remove the profile 1 tool if they are not
> using it (without removing all the tools that depend on it). At an
> absolute minimum we need to remove these dependencies.
>

Right, but just so I'm clear (though I suspect this isn't really a
question for Aaron, I don't want the question to be lost), is it the
case that:

a) Profile classic is still in 2.7.x?
b) if you unstealth it, it should continue to function as before?

> There isn't really a way to mark a tool as deprecated.

In the policy draft I've put forward, tool deprecation entails:

1) Stealthing the tool
2) Listing it as deprecated in the release notes, and announcing the
same on-list
3) Removing it to contrib in the next feature release, assuming there
aren't extenuating circumstances that would argue for keeping it
around longer (eg. if a migration path still needs to be worked out)

~Clay

>
> On Wed, Mar 24, 2010 at 1:46 PM, Clay Fenlason
> <clay.fenlason at et.gatech.edu> wrote:
>> Some outstanding questions I'd like to get more comment on:
>>
>> 1) I've led off with a proposed deprecation policy draft. [1] Thoughts
>> about this?
>>
>> 2) I've tried to represent the JCR and Static Cover issues as well as
>> I understand them [1], but would welcome clarity from my betters. At
>> the time of this writing, we're still ping-ponging the issue of
>> whether to actually flag static covers as deprecated, but seem to have
>> come to agreement on the rest (retain them, but remove their use from
>> core code).
>>
>> 3) The Kernel Utils proposal seems to me unhelpfully vague. It does
>> not, for example, explain if anything will be done for 2.7. I'm
>> inclined to think that this would be appropriate for kernel 1.2, but
>> not kernel 1.1 (ie. Sakai 2.7). Is there a list of shovel-ready
>> 3rd-party utils?
>>
>> 4) I confess to a bit of uncertainty about the state of the two
>> Profile tools. I was under the impression in December [2] that the two
>> profiles couldn't work together, but it looks like Steve managed to
>> complete the work that allowed them to do so [3]. And yet something
>> about Aaron's description of removing remaining dependencies to
>> Profile 1 makes me wonder if in fact they can still be swapped with
>> sakai properties. Can someone confirm?
>>
>> In any event, since Steve says they can work together, I've altered
>> the proposal to simply say that Profile classic should be stealthed
>> and marked as deprecated.
>>
>> ~Clay
>>
>> [1] http://confluence.sakaiproject.org//x/EhsQB
>> [2] http://n2.nabble.com/Fwd-Building-Sakai-2-7-0-Profile2-v-Profile-tt4155769.html#a4155769
>> [3] http://confluence.sakaiproject.org/display/MGT/2.7+Deprecations?focusedCommentId=68164488#comment-68164488
>>
>> On Tue, Mar 23, 2010 at 4:31 PM, Clay Fenlason
>> <clay.fenlason at et.gatech.edu> wrote:
>>> I'd promised I was going to work this up this past weekend, but I've
>>> been sick the last few days, and it's been delayed. Still not quite
>>> finished, but I'm feeling poorly again and need to rest a bit, so I
>>> won't delay circulating the link [1]. The latest on JCR and Static
>>> covers still isn't there.
>>>
>>> Comments and other edits of course welcome.
>>>
>>> HTH
>>>
>>> ~Clay
>>>
>>> [1] http://confluence.sakaiproject.org//x/EhsQB
>>>
>> _______________________________________________
>> sakai-dev mailing list
>> sakai-dev at collab.sakaiproject.org
>> http://collab.sakaiproject.org/mailman/listinfo/sakai-dev
>>
>> TO UNSUBSCRIBE: send email to sakai-dev-unsubscribe at collab.sakaiproject.org with a subject of "unsubscribe"
>>
>
>
>
> --
> Aaron Zeckoski - Software Engineer - http://tinyurl.com/azprofile
>


More information about the sakai-dev mailing list