As for the improved name resolution:

Kenton Varda wrote:
> On Sun, May 17, 2009 at 6:57 AM, Chris Kuklewicz <turingt...@gmail.com
> <mailto:turingt...@gmail.com>> wrote:
>
>
>     What do people think?
>
>
> You're right, this should have been handled too.  Oh well, I'll stick
> it on my TODO list for a later release.
I am quite happy to have helped.  The two name resolution functions were
side by side in my code; making the decision to fix only one looked
odd.  I will immediately support resolving extendee names to Messages,
ignoring Fields and other things.

As for the "packed" fields, I just now got my Haskell version to the
next stage:
  (1) new new runtime and converter both compile with "packed" support
  (2) it can convert the new unittest.proto into Haskell code with
"packed" support
  (3) the generated Haskell code compiles against new runtime with
"packed" support
  (4) it has regenerated its own descriptor.proto and been recompiled
         (enums needed an extra line to get packed fields efficiently)

So the next stage is to test the behavior and see if it can
inter-operate with itself and with packed files from protobuf-2.1.0.

Making the extension fields also "packable" was tedious but did not
require redesigning anything.  Whew.  The "unknown" field support did
not need updating at all.

As for the newly exposed field number constants:  I cannot make them a
proper enum data type in Haskell because those are closed definitions
and so could not include any of the extension fields outside the
message's own proto file.   I could still make them type safe constants,
but these could not be used as targets of a case statement.  The data is
available through reflection, so I will wait to implement anything else
until an actual person comes to me with a use case that I can make
design decisions for.

As for delimiting messages by prepending the length: I already had these
commands, so all I did was change the documentation from "author's
extension" to "compatible with protobuf-2.1.0".  Not that I actually
tested it...

-- 
Chris


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Protocol Buffers" group.
To post to this group, send email to protobuf@googlegroups.com
To unsubscribe from this group, send email to 
protobuf+unsubscr...@googlegroups.com
For more options, visit this group at 
http://groups.google.com/group/protobuf?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to