Typo in Ubuntu Code of Conduct 1.0

Do we need to change the code of conduct's version if we fix this typo?

>Do we need to change the code of conduct's version if we fix this typo?

Of course not, unless radical changes are being made !

Yes, the version has to be changed, since all digital signatures will be invalid. Everyone will have to sign it again.

On Mon, Dec 05, 2005 at 07:10:10PM -0000, Guilherme Salgado wrote:
> Do we need to change the code of conduct's version if we fix this typo?

I believe so, too. I've added it to the agenda for the next CC meeting:
https://wiki.ubuntu.com/CommunityCouncilAgenda

We've discussed this at the CC meeting. What happens is the following:

  a) Kamion will proofread the CoC and send us a diff to <email address hidden>

  b) We will need to increase the version number (I hope Kamion will tell us the version in his email too)

I asked if people needed to re-sign and they said there is no need as long as people can see a list of versions and a reason for change (does our data model support this?)

In the long term, the fact that people need to re-sign to become Ubunteros is a policy decision made on a case-by-case basis -- if we change the content we may require this.

Is this enough for us to move ahead here (pending Kamion's email?)

> I asked if people needed to re-sign and they said there is no need as long as people can see a list of versions and a reason for
> change (does our data model support this?)

Currently CoC model supports multiple versions, and you can modify the currentversion by editing the code (in ICodeOfConductConf),
it will demote all "ubuntite" flag, as supposed to be.

What we do not support yet is a 'comment' field in CodeOfConduct table, it might be helpful in cases like this.

Wonder if the following might find a place in the CoC, if and when plans materialize to change the content -

Be 'selfless' in whatever you do for the community - the returns, the rewards, truly are manyfold . Ubuntu and Free Software are valid proof of this !
If it is the name, the fame and the money that are more important to you, (along with their shackles - restrictive licenses and patents) . Well, what we have to offer here is something much more valuable, priceless - Freedom !
Come join us on the Launchpad.

More than guiding 'conduct' the above is meant to 'INSTILL' in a person the precise reason why 'he' or 'she' is contributing towards Free Software.
Further, the line 'Come join us on the Launchpad' (intended to sound more like an 'order') is directed towards passers-by with wrong notions about our community, to at least take a look at what we are upto, and perhaps spread the word of 'Freedom'.
The above may be modified in part or whole, or excluded totally if found to be unnecessary or irrelevant.

On Tue, Dec 06, 2005 at 04:52:48PM -0000, Alfred J.Tims wrote:
> Wonder if the following might find a place in the CoC, if and when plans
> materialize to change the content -

Alfred, the best place to take this suggestion is to the Community
Council. There's a wikipage that talks about the council meetings:

    https://wiki.ubuntu.com/CommunityCouncilAgenda

This is outside the scope of Launchpad (and this bug).

On Tue, Dec 06, 2005 at 04:44:09PM -0000, Celso Providelo wrote:
> Currently CoC model supports multiple versions, and you can modify the
> currentversion by editing the code (in ICodeOfConductConf), it will
> demote all "ubuntite" flag, as supposed to be.

The flag should not be reset, as I wrote above. I wonder if we can make
this optional.

Is there a way of knowing what specific version a person signed?

If there is we may be able to add a "trivial_change" flag to the CoC
version that doesn't force a flag reset.

The 'ubuntite' flag is build on-the-fly, by comparing the signed version of CoC with the 'current' version pointed by ICoCConf utility. If you want to keep the current CoC signature, simply do not change the current version in ICoCConf...

<idea>We may create a new field in ICoCConf called 'accepted' to be used in the validity checks for ubuntite flag</idea>

The 'ubuntite' flag is built on-the-fly, by comparing the signed version of CoC with the 'current' version pointed by ICoCConf utility. If you want to keep the current CoC signature, simply do not change the current version in ICoCConf...

<idea>We may create a new field in ICoCConf called 'accepted' to be used in the validity checks for ubuntite flag</idea>

I suck, the current implementation JUST support it, ubuntite checks if the user has a signature newer than ICoCConf.datereleased.
So just point ICoCConf.current to the new version and keep the daterelease as it is to maintain the current ubuntites certificates.
Sorry the mess.

Misc. thoughts on previous comments, and a general thought....

> must version # change?

Yes. Otherwise, electronic signatures made on the two "1.0" versions would not match (that, is the hashes would not match). It could be possible to substitute the "other 1.0" and cause someone's previous signature to fail. (Stress "could", haven't looked into this).

Best practice with signed electronic documents? Change the version number when changing the document. (Hmm, this is a best practice of document management in any medium.)

> is there a way of knowing what specific
> version a person signed?

Wow, I hope so! Otherwise, the electronic signature has limited utility! (In fact, one way to determine this would be to generate hashes of all versions on record and compare those to the hash in the person's signature. This just needs to be made automagic. :->)

> must everyone re-sign?

IMHO, no, as the change is not substantive, that is, is not a change in content. Some might argue that yes, everyone should re-sign, since the document is "different", that is, "not the same one".

Well, that depends on whether you are a human being or a computer: To a computer, the documents are different, since a few bits have changed. To a human, the documents are both different, since a different word is used, AND NOT different, since the meaning and intent are pretty darned clear.

Add in the fact that the Code of Conduct is on the Ubuntu web site and it is pretty clear to most human beings that the documents are the same, one just had a typo.

(There is a temptation in the electronic world to make "perfect evidence" when in the "real world" there is no such thing. We use the best evidence we can, for the circumstances. For example, in many jurisdictions, if the CoC were considered a contract, then private law would apply and the evaluation of evidence would be based on "balance of probabilities" and the "reasonable person test". We would not need perfect evidence, we would need to reasonably establish probable intent.)

Since these electronic processes are being used to support human beings, and not the other way around, IMHO it is unnecessary to re-sign because of a typo that does affect meaning.

Any such change will have to be evaluated on a case-by-case basis, by the CC, with input from the rest of us, and a policy decision will have to be made. A policy in this area would be a good idea....

It could look like this:

1. All CoC versions must have a revision number, and a canonical version of each revision.

2. The canonical version must be published on the web, and the location of the canonical version must be included in the CoC.

3. All changes to any version of the CoC will result in a new revision number for that version of the CoC.

4. The revision number of any version of the CoC is an integral part of that version of the document.

5. The Community Council will determine the fitness and suitability of any particular revision of the CoC for any particular purpose.

6. Changes to correct spelling, grammar, and presentation in any version of the CoC that do not affect the meaning and intent of the CoC will be considered minor changes and will not invali...

This was fixed with the upload of the 1.0.1 version of the CoC. Now we only need to fix bug 39547, so people can sign the 1.0.1 version.