The Purpose of the Tracker

The tracker is intended to maintain the status of issues. It is not a good place for discussion, because it displays all comments in most-recent order, and you must scroll through them to find the information you want. Try to keep comments to short summaries, and to refer discussion to mailing lists. Longer commentary is most useful if it takes the form of a patch to the manuals.

Patches should not be pasted into the comments, but uploaded as separate files. The comment should indicate which patch(es) is (are) considered the current candidate for application to XEmacs. The obvious exception is for patches submitted by email, but even then most MUAs allow you to attach a patch as a file.

If the OP doesn't provide a test or recipe for reproduction of a bug, then exactly one "me too" comment is appropriate, and the status field should be bumped to "verified".

Accessing the Tracker

You may access the tracker in one of three ways:

1. through the web interface,
2. through the e-mail gateway, or
3. using the command line tool.

The last is accessible only to administrators. Other users must use the web and e-mail interfaces. All three are explained below.

Issue Description and Classification

Issues are described by the required Title property and an optional Change Note. The Change Note can refer to Users, Messages, and other Issues by designators (like "user1", "msg32", or "issue222"). These are automatically converted to links by Roundup. (It is often important that a superseding Issue reference the Issue(s) it supersedes. This information is also available in the superseding Issue's history, but that's a little obscure.)

Issues are classified according to the required properties Type, Severity, Module, and Platform. Type defaults to "defect", and severity to "inconvenience". Platform currently has a default of "N/A", but this may change. The set of Modules includes several modules relating to the core C and Lisp code implementing the Emacs Lisp and editor primitives, modules relating to XEmacs services including the website, mailing lists, code repositories, and this tracker itself.

Issue life cycles and the development process

New issues may be submitted via the web or e-mail. The progress of work on the issue is summarized by the Status property, which takes on the following values.

Priority values

The priority is generally the "property" of the responsible Developer, or the XEmacs Review Board.

Users, Roles, and properties

In Roundup, each user, whether submitter, developer, or tracker administrator, must register a User with the tracker. The User is strongly associated with a list of email addresses used to identify incoming mail from that user. (The tracker may be upgraded to authenticate and authorize users via PGP.) The user's identity is encapsulated in an object named something like "User1" (that is in fact the original Admin user). So you can actually change both your primary address (which Roundup uses to contact you, as well as identify email submissions) and the alternate addresses (used to identify email submissions). However these addresses must be unique in the system since they identify users, and there are several conditions in which you may encounter conflicts. In that case get in touch with the tracker adminstrator, Stephen Turnbull stephen@xemacs.org.

Each user may have several Roles, to which are attached various privileges when that user is logged in to the tracker. The project roles, which have various privileges, are Anonymous, User, Developer, and Reviewer. These privileges are not necessarily strictly increasing, so make sure you have all the roles you need and are authorized to have. There is also a Role called Admin that authorizes tracker administrators. Each User object automatically receives the User role.

The Developer role is assigned to any user who would like to volunteer to work on some area of interest to them. The difference between a Developer and a User is that Developers' names appear in the Assigned-To menu. Any user can assign an issue to themself. Being a Developer means that you're volunteering to have others assign appropriate issues to you. If you're not working on XEmacs almost daily, and regularly contributing patches, we would prefer that you not take the Developer role -- you can already assign to yourself, and nobody else knows your expertise, so they won't assign to you anyway. It just clutters the menu.

If you haven't been active in XEmacs, but would like to contribute by working on some specific area, you are welcome to add the Developer role to your user. But remember that it's not useful unless others know what you're good at. If you're not well-known yet, post to the XEmacs Contributors list <xemacs-beta@xemacs.org> and introduce yourself, and describe the kinds of issues (bugs or features) you want people to assign to you.

Reviewers are allowed to do things like close issues that they don't own. Only the adminstrator can create Reviewers.

The Roles should not be confused with the issue properties of Actor (a User who changes an issue), Assigned To (a Developer responsible for resolving an issue), and Creator (the User of the original submitter of the issue). They do confer special privileges in some circumstances, but only for the associated issue. On the other hand, the privileges conferred by a Role are global.

Activity properties

Several activity properties are automatically updated. These are Creator, Creation Date, Submitted, Actor, and Activity Date. The creator is the user who first submitted the issue, and the creation date the timestamp of submission. Because some issues have been imported in bulk from other databases (specifically the mailing list archives), it is useful to maintain a "submitted" field, which is initialized to the date of first submission (e.g., the RFC 2822 Date header for a mailing list or Usenet post). Actor and activity date are the user who most recently updated any of the properties of the issue, and the timestamp of the activity, respectively.

Procedure

1. Register your user (or use "Forgot Your Login?" to get a temporary password assigned if you already have one from a preloaded issue).
2. Use the "Show Open" query in the "Issues" block of the sidebar to list open queries. The unclassified issues will show up as "(no priority set)" at the top of the list.
3. Pick an item, click on the Title to edit it.
4. Review the issue, and
   1. set the module (required, no default)
   2. set the platform (required, default is "N/A" which is the best setting for Lisp bugs -- platform should be set appropriately for crashes, build problems, and GUI bugs); note that multiple platforms may be selected
   3. set the severity, which is an assessment from the user's point of view of how important the issue is (required, default is "inconvenience"); note that if you set the severity to "security", and save the issue, you will not be able to edit it again (unless you are a Reviewer or Tracker admin)
   4. (optionally) set keywords, creating them if necessary; try to avoid creating keywords as the interface is still clumsy; note that multiple keywords may be selected
   5. (optionally) set priority, which is a guide to the urgency with which the developers should work on the issue (default is "normal")
   6. (optionally) review and edit the title, for example to remove the "[Bug: " prefix from M-x report-xemacs-bug.

Setting priority:

1. Don't demote a bug you haven't experienced to "cosmetic" unless it's really really obviously cosmetic (eg a doc typo).
2. Don't promote a bug to "critical". At the moment there's no show to stop :-( so there are no showstoppers.
3. Don't change priority of an issue that has an Assigned To developer. That indicates the developer's assessment of when he'll get to it.
4. At most 25 urgent bugs are allowed. If your bug makes it into the top 25 in your opinion, and there are already 25 urgent bugs, then pick one to demote to "normal" priority. Don't worry about getting the "least important" one, but be polite: if you can't find one that the developers will clearly consider less important than yours, you probably should not be promoting yours!

Setting status:

1. Be conservative about updating status. You should only push status to "committed", "documented", or "closed" if you understand the issue and any related patches or documentation well enough to have written them yourself. (That doesn't mean you need to be able to write it in English!) If you find yourself doing this more than once or twice, consider nominating yourself for XEmacs Reviewer. (No joke, we always need more reviewers!)
2. An issue should not be closed until you've confirmed that the issue is documented in the appropriate places. The mailing list archives do not count as documentation!
3. Missing, incorrect, or hard-to-find documentation is a bug in itself, even if the related behavior is acceptable. Do not close an issue as "not a bug" simply because somebody explained why it works that way on the mailing list.
4. If you find a patch labelled "committed" in the tracker, it's OK to assign the issue to the committer as committing a patch is a definite assumption of responsibility. However, by default you should bump status only as far as "assigned".

Setting keywords:

Keywords are important!

1. Some keywords are workflow-related, the "has X" and "needs X" keywords in particular. The "has X" keyword means that such an item has been contributed, typically by the issue's creator, and needs to be confirmed before committing the change. The "needs X" keyword means that such an item must be developed and committed to resolve the issue.

   The "has X" and "needs X" keywords are not opposites. When a contributed X is verified and committed, the "has X" keyword should be removed. Similarly, when an X is developed and committed, the "needs X" keyword should be removed.

2. Other keywords refer to XEmacs behavior, and are helpful to users in filtering out issues unrelated to their problem.

   At present, the following keywords for general areas of XEmacs behavior are available: display, i18n, input/output, motion, performance, search, and security. Feel free to add more if appropriate, but as mentioned I'd like to keep the number small for now. Try to keep new keywords at a pretty high level of generality.

String and Numeric properties

These fields just take a simple text value, like It's broken.

Boolean properties

These fields take a value which indicates "yes"/"no", "true"/"false", "1"/"0" or "on"/"off".

Constrained (link and multilink) properties

Fields like "Assigned To" and "Keywords" hold references to items in other classes ("user" and "keyword" in those two cases.)

Sometimes, the selection is done through a menu, like in the "Assigned To" field.

Where the input is not a simple menu selection, we use a comma-separated list of values to indicated which values of "user" or "keyword" are interesting. The values may be either numeric ids or the names of items. The special value "-1" may be used to match items where the property is not set. For example, the following searches on the issues:

assignedto=richard,george

match issues which are assigned to richard or george.

assignedto=-1

match issues that are not assigned to a user.

assignedto=2,3,40

match issues that are assigned to users 2, 3 or 40.

keyword=user interface

match issues with the keyword "user interface" in their keyword list

keyword=web interface,e-mail interface

match issues with the keyword "web interface" or "e-mail interface" in their keyword list

keyword=-1

match issues with no keywords set

Date properties

Date-and-time stamps are specified with the date in international standard format (yyyy-mm-dd) joined to the time (hh:mm:ss) by a period .. Dates in this form can be easily compared and are fairly readable when printed. An example of a valid stamp is 2000-06-24.13:03:59. We'll call this the "full date format". When Timestamp objects are printed as strings, they appear in the full date format.

For user input, some partial forms are also permitted: the whole time or just the seconds may be omitted; and the whole date may be omitted or just the year may be omitted. If the time is given, the time is interpreted in the user's local time zone. The Date constructor takes care of these conversions. In the following examples, suppose that yyyy is the current year, mm is the current month, and dd is the current day of the month.

• "2000-04-17" means <Date 2000-04-17.00:00:00>
• "01-25" means <Date yyyy-01-25.00:00:00>
• "2000-04-17.03:45" means <Date 2000-04-17.08:45:00>
• "08-13.22:13" means <Date yyyy-08-14.03:13:00>
• "11-07.09:32:43" means <Date yyyy-11-07.14:32:43>
• "14:25" means
• <Date yyyy-mm-dd.19:25:00>
• "8:47:11" means
• <Date yyyy-mm-dd.13:47:11>
• the special date "." means "right now"

When searching, a plain date entered as a search field will match that date exactly in the database. We may also accept ranges of dates. You can specify range of dates in one of two formats:

1. English syntax:

   [From <value>][To <value>]
   

   Keywords "From" and "To" are case insensitive. Keyword "From" is optional.

2. "Geek" syntax:

   [<value>];[<value>]
   

Either first or second <value> can be omitted in both syntaxes.

For example, if you enter string "from 9:00" to "Creation date" field, roundup will find all issues, that were created today since 9 AM.

The <value> may also be an interval, as described in the next section. Searching of "-2m; -1m" on activity field gives you issues which were active between period of time since 2 months up-till month ago.

Other possible examples (consider local time is 2003-03-08.22:07:48):

• "from 2-12 to 4-2" means <Range from 2003-02-12.00:00:00 to 2003-04-02.00:00:00>
• "FROM 18:00 TO +2m" means <Range from 2003-03-08.18:00:00 to 2003-05-08.20:07:48>
• "12:00;" means <Range from 2003-03-08.12:00:00 to None>
• "tO +3d" means <Range from None to 2003-03-11.20:07:48>
• "2002-11-10; 2002-12-12" means <Range from 2002-11-10.00:00:00 to 2002-12-12.00:00:00>
• "; 20:00 +1d" means <Range from None to 2003-03-09.20:00:00>
• "2003" means <Range from 2003-01-01.00:00:00 to 2003-12-31.23:59:59>
• "2003-04" means <Range from 2003-04-01.00:00:00 to 2003-04-30.23:59:59>

Interval properties

Date intervals are specified using the suffixes "y", "m", and "d". The suffix "w" (for "week") means 7 days. Time intervals are specified in hh:mm:ss format (the seconds may be omitted, but the hours and minutes may not).

• "3y" means three years
• "2y 1m" means two years and one month
• "1m 25d" means one month and 25 days
• "2w 3d" means two weeks and three days
• "1d 2:50" means one day, two hours, and 50 minutes
• "14:00" means 14 hours
• "0:04:33" means four minutes and 33 seconds

Simple support for collision detection

Item edit pages remember when the item was last edited. When a form is submitted, the user will be informed if someone else has edited the item at the same time they tried to.

Lists of Items

The first thing you'll see when you log into Roundup will be a list of open (ie. not resolved) issues. This list has been generated by a bunch of controls under the covers but for now, you can see something like:

The screen is divided up into three sections. There's a title which tells you where you are, a sidebar which contains useful navigation tools and a body which usually displays either a list of items or a single item from the tracker.

You may either register or log in. Registration takes you to:

Once you're logged in, the sidebar changes to:

You can now get to your "My Details" page:

Display, edit or entry of an item

Create a new issue with "create new" under the issue subheading. This will take you to:

Editing an issue uses the same form, though now you'll see attached files and messages, and the issue history at the bottom of the page:

These screens currently require a geometry of about 1000x750 for comfort.

Searching Page

Most values in the search form are entered via popup menus. The All Text and Title fields are free text. The ID field is an item designator. Creation Date and Activity are dates, times, or intervales. See entering values in your tracker for an explanation of what you may type into the search form.

Most popup menus have two special values: "don't care" and "unselected". "don't care" means that any value is acceptable. "unselected" means that no value have been set for this field. This is normally possible only for issues submitted by email. Use of the "unselected" value is useful to find issues which need to be updated to provide the information necessary to usefully search for them. The status menu in the search form has a third value: "unresolved." This means any of the values except "done/needs work" and "closed" (including the special value "unselected"). This is a good way to find current issues to work on.

You can also select one field to group on and one to sort on initially. The groups will be ordered according to the order for values of the "group on" property. Within each group, the items will be ordered according to the order for values of the "sort by" property. You can change or refine the grouping and sorting with the controls at the bottom of each search results page. If there is a large number of results (by default, more than 50), the list will be broken up into multiple pages. The query itself, as well as the sorting and grouping (including any refinements) will persist as you page through the list.

/issue?status=unread,in-progress,resolved&
    keyword=security,ui&
    @group=priority,-status&
    @sort=-activity&
    @filters=status,keyword&
    @columns=title,status,fixer

Access Controls

User access is controlled through Permissions. These are are grouped into Roles, and users have a comma-separated list of Roles assigned to them.

Permissions divide access controls up into answering questions like:

• may the user edit issues ("Edit", "issue")
• is the user allowed to use the web interface ("Web Access")
• may the user edit other user's Roles through the web ("Web Roles")

Any number of new Permissions and Roles may be created as described in the customisation documentation. Examples of new access controls are:

• only managers may sign off issues as complete
• don't give users who register through e-mail web access
• let some users edit the details of all users

Subject-line information

The subject line of the incoming message is examined to find one of:

1. the item that the message is responding to,
2. the type of item the message should create, or
3. we default the item class and try some trickiness

If the subject line contains a prefix in [square brackets] then we're looking at case 1 or 2 above. Any "re:" or "fwd:" prefixes are stripped off the subject line before we start looking for real information.

If an item designator (class name and id number, for example issue123) is found there, a new "msg" item is added to the "messages" property for that item, and any new "file" items are added to the "files" property for the item.

If just an item class name is found there, we attempt to create a new item of that class with its "messages" property initialized to contain the new "msg" item and its "files" property initialized to contain any new "file" items.

The third case above - where no [information] is provided, the tracker's MAIL_DEFAULT_CLASS configuration variable defines what class of item the message relates to. We try to match the subject line to an existing item of the default class, and if there's a match, the message is related to that matched item. If not, then a new item of the default class is created.

Sender identification

If the sender of an e-mail is unknown to Roundup (looking up both user primary e-mail addresses and their alternate addresses) then a new user will be created. The new user will have their username set to the "user" part of "user@domain" in their e-mail address. Their password will be completely randomised, and they'll have to visit the web interface to have it changed. Some sites don't allow web access by users who register via e-mail like this.

E-Mail Message Content

Roundup only associates plain text (MIME type text/plain) as messages for items. Any other parts of a message are associated as downloadable files. If no plain text part is found, the message is rejected.

To do this, incoming messages are examined for multiple parts:

• In a multipart/mixed message or part, each subpart is extracted and examined. The text/plain subparts are assembled to form the textual body of the message, to be stored in the file associated with a "msg" class item. Any parts of other types are each stored in separate files and given "file" class items that are linked to the "msg" item.
• In a multipart/alternative message or part, we look for a text/plain subpart and ignore the other parts.

If the message is a response to a previous message, and contains quoted sections, then these will be stripped out of the message if the EMAIL_KEEP_QUOTED_TEXT configuration variable is set to 'no'.

Address handling

All of the addresses in the To: and Cc: headers of the incoming message are looked up among the tracker users, and the corresponding users are placed in the "recipients" property on the new "msg" item. The address in the From: header similarly determines the "author" property of the new "msg" item. The default handling for addresses that don't have corresponding users is to create new users with no passwords and a username equal to the address.

The addresses mentioned in the To:, From: and Cc: headers of the message may be added to the nosy list depending on:

ADD_AUTHOR_TO_NOSY

Does the author of a message get placed on the nosy list automatically? If 'new' is used, then the author will only be added when a message creates a new issue. If 'yes', then the author will be added on followups too. If 'no', they're never added to the nosy.

ADD_RECIPIENTS_TO_NOSY

Do the recipients (To:, Cc:) of a message get placed on the nosy list? If 'new' is used, then the recipients will only be added when a message creates a new issue. If 'yes', then the recipients will be added on followups too. If 'no', they're never added to the nosy.

Mail gateway script command line

Usage:

roundup-mailgw [[-C class] -S field=value]* <instance home> [method]

The roundup mail gateway may be called in one of three ways:

• with an instance home as the only argument,
• with both an instance home and a mail spool file, or
• with both an instance home and a pop server account.

It also supports optional -C and -S arguments that allows you to set a fields for a class created by the roundup-mailgw. The default class if not specified is msg, but the other classes: issue, file, user can also be used. The -S or --set options uses the same property=value[;property=value] notation accepted by the command line roundup command or the commands that can be given on the Subject line of an e-mail message.

It can let you set the type of the message on a per e-mail address basis.

PIPE:

In the first case, the mail gateway reads a single message from the standard input and submits the message to the roundup.mailgw module.

UNIX mailbox:

In the second case, the gateway reads all messages from the mail spool file and submits each in turn to the roundup.mailgw module. The file is emptied once all messages have been successfully handled. The file is specified as:

mailbox /path/to/mailbox

POP:

In the third case, the gateway reads all messages from the POP server specified and submits each in turn to the roundup.mailgw module. The server is specified as:

pop username:password@server

The username and password may be omitted:

pop username@server
pop server

are both valid. The username and/or password will be prompted for if not supplied on the command-line.

APOP:

Same as POP, but using Authenticated POP:

apop username:password@server

Using with the shell

With version 0.6.0 or newer of roundup (which introduced support for multiple designators to display and the -d, -S and -s flags):

To find all messages regarding chatting issues that contain the word "spam", for example, you could execute the following command from the directory where the database dumps its files:

shell% for issue in `roundup-admin -ds find issue status=chatting`; do
> grep -l spam `roundup-admin -ds ' ' get messages $issue`
> done
msg23
msg49
msg50
msg61
shell%

Or, using the -dc option, this can be written as a single command:

shell% grep -l spam `roundup get messages \
    \`roundup -dc find issue status=chatting\``
msg23
msg49
msg50
msg61
shell%

You can also display issue contents:

shell% roundup-admin display `roundup-admin -dc get messages \
           issue3,issue1`
files: []
inreplyto: None
recipients: []
author: 1
date: 2003-02-16.21:23:03
messageid: None
summary: jkdskldjf
files: []
inreplyto: None
recipients: []
author: 1
date: 2003-02-15.01:59:11
messageid: None
summary: jlkfjadsf

or status:

shell% roundup-admin get name `/tools/roundup/bin/roundup-admin \
      -dc -i /var/roundup/sysadmin get status issue3,issue1`
unread
deferred

or status on a single line:

shell% echo `roundup-admin get name \`/tools/roundup/bin/roundup-admin \
         -dc -i /var/roundup/sysadmin get status issue3,issue1\``
unread deferred

which is the same as:

shell% roundup-admin -s get name `/tools/roundup/bin/roundup-admin \
         -dc -i /var/roundup/sysadmin get status issue3,issue1`
unread deferred

Also the tautological:

shell% roundup-admin get name \
   `roundup-admin -dc get status \`roundup-admin -dc find issue \
       status=chatting\``
chatting
chatting

Remember the roundup commands that accept multiple designators accept them ',' separated so using '-dc' is almost always required.

Back to Table of Contents