39895 - new account setup needs to parse GCOS/GECOS field

Status: RESOLVED FIXED

(SeaMonkey :: MailNews: Account Configuration, defect, P3)

Description

[Jeremy M. Dolan]

On unix systems, account entries often use a simple comma-delimited format
called 'GECOS' or 'GCOS' in the comment field, to store more data then just real
name. (phone, extention, etc).  Currently, when you set up a new account, it
takes the whole comment field as the default realname. What we really want is
everything before the first comma. May also be an desirable for Mac OSX port.

Comment 1

[Alec Flett]

add sspitzer to CC, this is the auto-fillin of the user's full name.

One situation I'm concerned about is non-GCOS systems that have entered people's 
full names as "Lastname, First" - is there a way to work around that or somehow 
know that this is or isn't a "GCOS" system?
Where can I read about GCOS? Is this stored in the passwd file?

Comment 2

[Jeremy M. Dolan]

"Lastname, First" entries in that field are very wrong. many other applications
attempt to check for comma delimited text there. For the very few systems that
have 'last, first', for some reason, having just their last name is better then
having all of the other information.
(ie, 'Smith' is better then 'Smith, John,107,555-5555,416-9826')

If an application were to want "last, first", it would need to switch them
itself.

Note that on a Slackware Linux 7.0, new accounts are created by default with
the GECOS field set to "real name,,,". This is also what 'chfn' will set the
field to, if the user doesn't specify entries for the room number/phone prompts.

someone should investigate on solaris, as well...i would assume GECOS is pretty
standard there, too.

Comment 3

[scottputterman]

moving to future milestone.

Comment 4

[Andrew Taylor]

Some systems allow an ampersand in the GCOS field, which represents the
username, capitalized.

For example:
andrew::500:500:& Taylor:/home/andrew:/bin/bash
Here the GCOS should be expanded to Andrew Taylor.

This is documented in passwd(4) on Solaris.

Comment 5

[Andrew Taylor]

Attachment: Patch to truncate gcos field at first comma and replace ampersand with capitalized username

Comment 6

[Alec Flett]

patch looks good, except you can use nsCAutoString to avoid all the conversions 
back and forth between unicode and ASCII

Comment 7

[Andrew Taylor]

Attachment: Same as first patch, but without the needless unicode conversion

Comment 8

[Alec Flett]

*** Bug 55123 has been marked as a duplicate of this bug. ***

Comment 9

[Dan Rosen]

oh, cool. i couldn't remember "gecos" as the magic name for unix user info. good
to see there's a fix in hand, too... adding keywords, cc. i'll review the patch
tomorrow if possible.

Comment 10

[scottputterman]

marking rtm-. Not common enough to get into rtm and the user always has the
ability to edit what we autofill if they are in this situation.

Comment 11

[Alec Flett]

we should check this into the tip when someone gets a chance.

Comment 12

[Jeremy M. Dolan]

just a responce to 'not common enough'...

just checked all the passwd files i have access to, a good 75% of user accounts
on linux and solaris machines have gcos fields in them. thats fairly common, by
my definition.

Comment 13

[Alec Flett]

well, I just checked all the unix systems I have access to, and none of them
used this system, so according to my sample space, this system doesn't exist :)

I'll check the patch into the tip so it makes it for mozilla 1.0, but netscape
is not taking this on their branch.

Comment 14

[Alec Flett]

add seth for review, I'll be super-reviewer, then this will go into the trunk

Comment 15

[Alec Flett]

massive reassign of account manager bugs -> sspitzer
please feel free to put me back on the CC if you have any questions/comments

Comment 16

[Alec Flett]

adding myself back to this bug so I can checkin when I get a review from seth.

Comment 17

[(not reading, please use [e-mail] instead)]

the patch looks good, but can you correct / add comments to explain what these
additional code does?

once we do that, r=sspitzer

Comment 18

[(not reading, please use [e-mail] instead)]

I forgot to add:

"sorry for the delay on this.  I've been focusing on RTM bugs."

Comment 19

[Alec Flett]

ok, back to me, marking fixed. thanks seth.

Comment 20

[Alec Flett]

fix now in, marking fixed.

Comment 21

[Dan Rosen]

no way this could get into rtm? seems like low-risk/very-high-profile to me...
am i deluding myself?

Comment 22

[(not reading, please use [e-mail] instead)]

deluded big time.  especially since the work around is to go and edit the text
field when it comes up.

Comment 23

[[e-mail] (gone - use [e-mail] instead)]

*** Bug 61141 has been marked as a duplicate of this bug. ***

Comment 24

[esther]

changing qa