Skip to topic | Skip to bottom
Home
Lily
Lily.LilyCoreChangelogr1.8 - 31 Dec 2005 - 11:52 - GaranceDrosehntopic end

Start of topic | Skip to actions
This will not be an exhaustive detailed list of changes. This will just list the highlights of user-visible (or programmer-visible, or admin-visible) changes made to the lily service at RPI.


  • Dec 2005:
  • -- For users:
    • new /set location ability from Coke. Includes support on the web-interface of lily to put up a Google-map of all locations set by any user. Locations (if /set) are also displayed by /finger.
    • added a few more drinks in /drink. More notably, the full list of available drinks is now printed as multiple lines wrapped at screen_width -- instead of as one extremely-long line (over 1600 characters!).
    • added new hearts game from Coke. Type /help hearts for more details. This is probably not 100% correct yet, but presumably any remaining bugs will be worked out before the official 2.8.0 release.

  • -- For users and client-developers:
    • many fixes for leaf-cmd processing (for the few lily clients which use it). There's still more work which would need to be done here. The biggest challenging for this is probably that @program is handled by the MOO program itself -- and not by any verbs in our lilyCore code!

  • -- For server-side programmers:
    • @eval is much more graceful about how it handles errors now. Nicer traceback output, works better with leaf-cmd processing, etc. Also, @eval errors are now displayed only to the user who typed in the command, and are no longer copied to the main server logfile.
    • @eval now supports a way to enter a very long statement on multiple lines (but it is still limited to a single statement). You need to end your @eval command with " @[", at which point @eval will read lines from the user for more data that it will append to the command. This feature comes in handy for some improvements to @dump.
    • @dump should now come very close to doing a correct, complete, and useful dump of any given object into a series of commands which would recreate that object from scratch. It also takes advantage of the new ' @eval @[ ' ability so that the commands it generates are much less likely to run into line-length limitations, either in MOO itself, or in various lily clients.
    • @dump now also supports a few filtering options (although you can only specify one of them on any single @dump command). Options include: noverbs justverbs verbdefs noprops justprops propdefs and propvals. Note that propvals will print out @eval statements to set all properties which the object did not inherit the value from it's parent. Without that option, some values are shown via the @property command that defines the property, and other values are shown via @eval commands.
    • @dump is now generally smarter about how it parses the parameters given it, such as recognizing @dump $admin as a request to dump object #3, instead of using the dumb return from tostr("$admin") and thinking that you wanted to dump object #0.
    • @eval will now recognize a return value which is a list of just "long-ish" strings, and print out those strings one-per-line. This makes it much easier to read things like help entries (although ?gethelp does an even better job for help entries, this is good for return values LIKE help entries...).
    • @eval also now supports a time option, so you can do things like the following example. Note that it's reporting the values for seconds_left() and ticks_left(), both of which get reset on a call to suspend();. So this doesn't tell you how much you've used, it only tells you how close you came to running out.
              @eval time $code_utils:parse_objref("$admin_disc.owner")
              time=> seconds_left = 6 (of 6);  ticks_left = 64866 (of 64997)
              => #3


  • Nov 2005:
  • -- For users:
    • major rewrite of the /group command. Among many other changes, the notification messages which it returns for 'add' and 'delete' are much much better. And the listing done for multiple groups is now wrapped at screen-width by default. Also supports a new 'by name' option, so you can ask for ' /group by name ' to have the member-names for each group displayed in alphabetical order.
    • some improvements to processing and parsing in the /unignore command.
    • add a few new options to the /ignore command, which can be used to reduce how much is printed when the user wants to list out current ignores. A plain /ignore (without any parameters) will list all ignores in effect for your account (both of you ignoring others, and others ignoring you). The new options of 'by me' , 'by others' , and 'by count' to reduce how much is printed. E.g.:
        /ignore
        (you are currently ignoring Garance {in lily-dev}, and being ignored by Garance {in PA-general})
        /ignore by me
        (you are currently ignoring Garance {in lily-dev})
        /ignore by others
        (you are currently being ignored by Garance {in PA-general})
        /ignore by count
        (you are currently ignoring 1 user, and being ignored by 1 other user)

  • -- For server admins:
    • improvements to the $account command, mainly to make it easier to display or locate accounts which have been *lock*-ed out. Several cosmetic improvements.

  • -- For server-side programmers:
    • improve @show obj :
      • verbs and new-property-names are now listed as a comma-separated list (instead of one verb per line), which is wrapped at screen-width. "new-property-names" being the properties that the object itself defines, as opposed to those property-names that it gets from the parent.
      • property-values are checked before display. In the case of some lists, the value may be written out as multiple lines (one line per element in the list), or as a comma-separated list which is wrapped at screen-width. Before this change, all properties were printed on a single line, and for some lists that resulted in a line which was WAY too long to handle. Example: All your memos are stored in a single property value and RPI allows users to have up to 32767 bytes of memos (while other sites might allow even more).
    • add the @history command, for displaying the [changelog] section of a verb.
    • add the @explain command, for displaying the [rationale] section of a verb.
    • Right now @history and @explain only work on an #obj:verb reference, but I do intend to expand on that.


Note that many changes have been made since the most recent official release of the lilycoreDB which are not (yet?) listed here. The most recent official release is lilycoreDB version 2.6.3, done back in January 2001.

-- GaranceDrosehn -- Initial versions.
to top


Lily.LilyCoreChangelog moved from Lily.ChangeLog on 06 Dec 2005 - 09:04 by GaranceDrosehn - put it back
You are here: Lily > DeveloperGuide > NeededForRelease280 > LilyCoreChangelog

to top

Copyright © 1999-2009 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding this site? Contact Christopher Masto <chris@masto.com>.