Re: [gentoo-dev] ML changes

[Kumba <kumba@gentoo.org>]

Kevin Lacquement wrote:
> 
> I'm not sure about stuff in -core becoming publicly accessible.  After 
> all, isn't it in the private list for a reason?  Perhaps summaries of 
> -core discussions being forwarded to -dev would be a better option. 
> However, I'm new to -dev, so if this is what already happens I don't know.


It's been a topic debated off and on on whether or not to keep -core locked away 
forever, but face it, even the CIA declassifies its dirty laundry every so 
often.  Now I'm not saying we should hold onto -core material for 30+ years, but 
I see no point in forever locking up the information on -core.  At minimum, it 
provides a historical look into how developers used to think.  Equally, this is 
why we need a sufficient time gap to let a majority of topics die off on -core 
before they become fodder for public consumption.  And why a marker being 
available to permanently lock certain threads/messages as needed.




> Here's where we want the non-devs to get access.  After all, not all 
> development and debugging is done by devs.  All the current devs were, 
> at one point, users.  Where did they get their start?  My bet is they 
> entered via the -dev mailing list, learned the ropes here, and 
> eventually earned their dev status.  If the -dev list is closed, where 
> do the new dev-wannabes learn the ropes and get their voices heard?

You missed the small mention of "open" in my first sentence.  I probably should 
have clarified what my definition of what "open" is, but it pretty much means no 
moderation on the -dev list so that users and developers could post.




> Would it perhaps be better to send announcements to -dev-announce, and 
> have that list forward to -dev?  That way we avoid issues if a subject 
> starts with [ANNONUCEMENT], for example


-dev-announce is a list proposed by another developer, and it's got its own bug 
number someplace (don't have it on hand ATM, however).  And technically, you 
wouldn't be forwarding the -dev-announce messages to -dev, because -dev-announce 
is essentially acting as a filter to -dev.  -dev would, in theory, contain ALL 
technical discussion related to the project.  -dev-announce would contain all 
announcements of certain, specific, technical things occurring within the 
project (and already talked about on -dev).  As a result, someone posting to 
-dev and wishing that post to also be forwarded to -dev-announce would attach 
[ANNOUNCEMENT]: to their subject line.  Not all devs are gonna wanna get into 
discussions, even technical ones.  Thus they can still monitor -dev-announce to 
keep abreast of things.

This method is no different really from the art of prefixing [PATCH]: to the 
subject line of an email on a kernel development list (or development list for 
any other software project) to indicate that the contents of the email includes 
a patch.  Even for LKML and linux-mips, there are tools in git that can target 
emails marked at patches, and automatically perform various feats of magic on 
them (such as stuffing the patches into a git queue of sorts).

This is why I don't think we could expect many problems from an announcement 
message.  Presumably, an announcement message would not be put out unless it'd 
already been discussed.  History, however, shows us that this is not always the 
case.  Thus, if some kind of a discussion were to arise from some kind of 
announcement, it likely wouldn't get forwarded to -dev-announce anyways (since 
replying to a mail would read as "Re: [ANNOUNCEMENT]", and it wouldn't get 
picked up by the automated mailer).  Furthermore, the -dev-announce list can 
probbaly be locked to only accept inbound mail from a specific host or address, 
itself tied to a script or bot of some kind.  If someone accidentally sent a 
message to -dev-announce, they would get a bounce back of some kind.




> If these messages will be machine-like, why not have them 
> machine-generated?  When you become a dev, someone (you?  the person 
> that -dev-ifie's you?) fills out a form, and the information from the 
> form is forwarded to the list.

We could automate it possibly, pulling data from the LDAP system used to auth 
devs to a number of gentoo systems.  Or someone in devrel could just take a few 
seconds to fill out a few fields in an email template and hit send.  I said 
impersonal because my mind is thinking technical == dry, white-paper-like 
material.  Either method works. but it's just a suggestion.  The more personal, 
emotion-filled (and I don't mean negative emotion-filled either) ones could go 
elsewhere, like to -project or such.


Cheers,


--Kumba

-- 
Gentoo/MIPS Team Lead

"Such is oft the course of deeds that move the wheels of the world: small hands 
do them because they must, while the eyes of the great are elsewhere."  --Elrond
-- 
gentoo-dev@gentoo.org mailing list