Re: Username length limitation policy

Hash: SHA1

Hello Kurt, gnome-infrastructure,

Kurt von Finck wrote:
> The policy is in place because it guarantees, *GUARANTEES* that all
>  services on GNOME production servers will be available to all
> users.

I keep hearing different information about this policy, but I do not
yet have the specifics.  Is there a wiki page or something describing
this policy in detail?  I assumed that 'mneptok' would be an
acceptable username, since it is less than 8 characters.  Why is it
not acceptable?  What are the details of the username policy?

> You cannot make the same guarantee if we switch. You're willing to
> be a guinea pig, but you'll be a guinea pig for a change in process
> that *is not necessary*. There is absolutely no good reason to
> change this policy. In fact, there's no reason at all (never mind a
> good reason) other than, "CJ wants to."

It will be necessary some day down the road, though.  Perhaps we are
not far enough down that road to even consider running out of address
space for usernames.  8^26 is not infinity.  If my brother, Chris
decides to join the GNOME team, he will run up against this issue,
since he can no longer use ccollier.  Where do we draw the line?  When
do we start considering issues?  It seems from your (implied) claim
that egoism and selfishness nullify requests that you are suggesting
issues not be petitioned if they directly impact the petitioner.

> It is your responsibility to show why a change in policy that
> serves the community well is necessary. Not desirable. Not cool.
> Necessary.

Then allow me to continue the petition on the part of other C.
Colliers in the world.  This policy effects them, now that I have
taken the obvious default username.

Allow me to recommend support of first last gnome org as a new policy,
grandfathering in the current usernames.  According to one of my
fellows from the #smuggle channel on


17:25 < eeejay> the trend today is [name] [surname] [domain] [tld]
17:25 < eeejay> i don't see why that shouldn't be a standard
17:25 < eeejay> it shouldn't get out of fashion

> It is NOT the responsibility for anyone involved in GNOME
> development or administration to explain to you why the policy is
> in place to YOUR satisfaction.

I'm not asking you as a GNOME developer or administrator.  I am asking
you as a member of the gnome-infrastructure mailing list.  My
understanding is that this list is a forum for discussion of GNOME
foundation policy.  I am asking on behalf of others currently on the
list with the same question, and on behalf of the future readers of
the archives who have the same question.

It was recommended to me by other members of the sysadmin team that I
discuss and debate the topic on this mailing list.  Thank you for
taking part.

> Again ...
> The policy is in place because it guarantees, *GUARANTEES* that all
>  services on GNOME production servers will be available to all
> users.
> That's the answer you get. If it doesn't satisfy you, well there's
> really nothing we can do about that, is there?

Thank you.  Are there any other details to the policy?

>> I am a patient man.  I view my contributions to GNOME as a
>> long-term investment.  A slight policy modification now will save
>> me many hours of tedious work over the span of the rest of my
>> life.  I view the further deprecation of gtkglarea during the
>> span of this policy amendment discussion as an acceptable cost,
>> when viewed in this perspective.
> Then be patient enough to make simple shell aliases like:
> alias menubar='ssh kvf menubar gnome org' alias container='ssh
> kvf container gnome org' alias window='ssh kvf window gnome org'
> alias button='ssh kvf button gnome org'

Behdad was kind enough to tell me about some ~/.ssh/config file syntax
that will help me cope with ccollier gnome org

Thank you for your recommendations as well.  I will consider them for
non-ssh services.

> My local username is not my GNOME username. It took me all of 5
> minutes to surmount this "problem."

Some of us are not as knowledgeable in this regard as others.  Thank
you again for assisting me with this issue.

>> In your own words,  "you can't have your desired username (which
>> I could not, either)."  It seems to me that it is in fact a
>> problem for someone other than me.  Ross also expressed that he
>> has come up against this issue as well.
> You're assuming that I wanted a username in excess of 8 characters,
>  which is not true.

I may have been, but more generally, I was thinking that your username
ran up against gnome username policies as mine had, and as Ross's had.

> When I joined I wanted the username "mneptok." I have used mneptok
> as an IRC nick for almost 15 years. The three major mneptok.* TLDs
> are registered to me. Probably more people on the Internet know me
> as mneptok than know me by my real name.

Why is this against the policy?  What rule does it conflict with?  Is
this rule documented somewhere?

> Well, it was against policy. Even though other users (like Ross and
> his "ninja" username) had been grandfathered in.
> Choice 1: Cause a huge stink, make sysadmins spend their time
> explaining policies to me rather than do actual productive work,
> and sulk like a child because I couldn't be "mneptok" on GNOME's
> servers.

IMHO, sometimes it is worthwhile to discuss and debate policies with
one's peers in order to become more familiar with the rules and
regulations of the organizations one is a part of.  I will not invite
you to take part in these discussions in the future, if this is what
you wish.

> Choice 2: Accept the policy and start being productive and helping
> in the efforts of my favorite Linux desktop.
> Guess which I chose?

My guess is that you feel you have chosen number 2.

>> Please re-state this.  I do not know what you mean by
>> "consequence."
> Nothing has broken or been delayed or otherwise been a major issue
> regarding username length. Well, until now.

I feel like the time spent discussing this policy is an acceptable
delay on my part.  I will not ask for your opinion in the future, if
you would prefer it this way.

>> I do not know what a "tempest in a teapot" is.
> tempest in a  teacup/teapot A great disturbance or uproar over a
> matter of little or no importance.

I feel that understanding the policies of the GNOME Foundation is of
utmost importance.  Thank you for helping me to understand better.

>> What would such a wall look like?  It seems to me that (at least)
>>  three contributors to the GNOME project have had problems with
>> this policy.  This sounds like a wall to me.
> It has not adversely affected the development of GNOME. More time
> has been spent explaining to you why this policy exists and
> justifying it than was ever spent trying to deal with this
> non-issue before.
> Does that sound like a wall to you? Are you willing to break down
> that wall and let us all move on to more productive work?

Discussion and debate on this topic is done, as everything in the
GNOME world is done, on a volunteer basis.  I expect that those
without the time to discuss it will not discuss it.  I appreciate you
taking the time to go over this with me.

>> My single selfish expectation from contributing to the Free
>> Software community is that my works be remembered, and my name be
>> remembered with them.  Therefore, I contribute Free Software
>> under the username cjcollier, not cjcollie, cjcoll, or v-cjcoll.
>> This issue is not stopping me from being a productive contributor
>> to the larger Free Software community.  I publish daily under the
>> Free Document License and General Public License as
>> cjcollier colliertech org   All I am asking from the GNOME
>> community is that I also be allowed to publish as
>> cjcollier gnome org 
> Ah, so it's about selfish pride. Not a good reason to abandon a
> policy that works perfectly well; well, except for you.

And this is the crux of my question.  Is it important to the GNOME
community, or is my request merely selfish pride?  I do not think I am
in a place to answer this question.  Your response makes me feel more
than a little ashamed.

> I think it's safe to assume that the majority of GNOME developers
> would rather not risk breaking services for everyone solely in
> order to assuage your ego.
> A e-mail alias is reserved for Foundation members.

What is a Foundation member?

> And your domain is, not

> And your real name in your e-mail client is CJ Adams-Collier.

Indeed.  After marrying my wonderful wife, Hannah, we changed our
name.  However, I registered years before I met
Hannah, and have stuck to the username despite changing my legal name.

> What's next? You want cj-adams-collier as a username?

not planning on it.

> Let.
> It.
> Go.

I don't think I quite understand.

>> I believe we should address the issue before we move on.  There
>> have been no arguments made for the current policy of username
>> length limitation other than conventional wisdom and a claim that
>> 8-character usernames are more scalable.
> Eight character usernames are more scalable.

How so?  8^26 is not a terribly large number, in the grand scheme of
things.  Add to this the limitation that was placed on your 'mneptok'
username (a <8 char username), and it seems that the 8-char username
policy will begin to come up against resistance sooner rather than
later.  I advise review of the policy before it becomes an issue
rather than forcing ourselves to fight fires as they spring up.

> Eight character usernames are know to work with the services we
> have.

Unless you have services I am not familiar with, so are much longer

> Eight character usernames present no potential problems or
> interruptions of GNOME development.

For a few more years at least, I agree.

> Eight character usernames were a policy decided upon by smart
> people that carefully weighed the issue.

So was the 640K barrier and the 32-bit IPv4 address space.  Sometimes
these things are outgrown.

> On the opposing side, > eight character usernames are something CJ
> wants because of self-described selfish reasons.

I think there are likely some other reasons, many of which I have stated.

> I think that's all that needs to be said. I think you need to let
> go, and move on.

Thank you for your opinion.  I respect it, but I am not yet convinced
that the issue has been resolved.

> ./k
> kurt von finck



- --
<cjcollier colliertech org>
+1 206 226 5809
Version: GnuPG v1.4.2.1 (GNU/Linux)


Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]