RE: Interop issue: Proposal for fixing lock token provision

Julian makes several good points.

One is that we need to address the If header length problem in 2518bis
in any case.  Even if we did provide a separate header that allowed
invalid lock tokens, we still need to use the If header when we want
to check for a list of valid lock tokens, or when we want to check for
a list of etags.

Another good point is that we don't have to argue about whether a
client should be allowed to submit invalid lock tokens, because once
the If header length problem is addressed (which we have to do
anyway), then the "If: <tag> (<token>)(Not <token>)" provides this
functionality without defining a header with new semantics.

In an earlier message in this thread, the following points were raised
as being problems with the "If:...Not..."  approach:

   PROBLEM #1: Servers may not support the OR, and the NOT, correctly,
   because most clients don't currently use this. This solution HAS NOT
   been proven, it is only theoretical. It might not work.

It will work against all existing servers that are correctly
implemented and that don't have an If header length problem, unlike
the "new header" approach, which is guaranteed to fail against all
existing servers.  If the servers are going to be updated,
updating them to correctly implement existing semantics seems
like a sensible approach to take, especially when the existing
semantics are a "logical or" and a "logical not", built-in operators
in most programming languages.

   PROBLEM #2: What if multiple locks are required (e.g. moving a
   collection that has multiple locked resources? What if the URLs are
   long? The IF header becomes very long and may be truncated by some
   proxies. E.g.

   If: <http://www.ics.uci.edu/users/f/fielding/index.html> 
   (<opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf6>
   NOT <opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf6>)
   <http:// www.ics.uci.edu/users/f/fielding/anotherfile.html>
   (<opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf7>
   NOT <opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf7>)
   <http:// www.ics.uci.edu/users/f/fielding/thirdfile.html>
   (<opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf8>
   NOT <opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf8>)

This is the If header length problem.  Once we fix the If header
problem (which we have to do anyway), there no longer is a need
for a new header to get this functionality.

   PROBLEM #3: Uck, this is really complicated. While it's useful to know
   that a client can do this with some existing servers, provided they
   handle it correctly, surely we can do this in a simpler mechanism that
   is less prone to interoperability problems. A simple header that allows
   the client to supply lock tokens to use is semantically equivalent to
   the above example, but shorter, simpler, and easier to implement. Also
   it can be split across multiple lines.
   Use-Lock-Tokens:
   <opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf6>, 
   <opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf7>
   Use-Lock-Tokens:
   <opaquelocktoken:f81d4fae-7dec-11d0-a765-00a0c91e6bf8>

Computer programs are really good at spitting out the same string
twice, surrounded by some constant characters.  Given the verbosity of
XML, and the fact that the Internet is used to stream video, cutting
down the number of characters required to identify lock tokens does
not strike me as a high priority task.  And since any client that
wants to interoperate with an existing server has to be able to
generate the If form of the request anyway, this is no simplification
for an interoperable client that has to be able to generate both forms
of the request.

Cheers,
Geoff


-----Original Message-----
From: Julian Reschke [mailto:julian.reschke@gmx.de]

Some thoughts:

- it doesn't make sense to raise the header length problem as argument
  in favor of adding a new header. If the If header length is a
  problem, it needs to be fixed anyway -- no matter how the other
  issues are treated

- As far as I now understand the client issue (BTW: why aren't the
  developers of this or these clients not participating in this
  dicussion??), they want to avoid to re-issue a request that
  specified a lock token just because the lock went away in the
  meantime. I *still* don't see why this is a problem -- if the lock
  was theirs, it's not supposed to go away without reason (such as bad
  timeout value). If the lock *wasn't* theirs, this is a case of "lock
  stealing", which certainly doesn't require to be optimized, right?

- if a client really really wants the request to succeed either case
  (lock being present or not), can't he simply submit an If header
  production such as:

If: <http://www.foo.bar/resource1> (<locktoken:a-write-lock-token>)(Not
<locktoken:a-write-lock-token>))

?

Received on Tuesday, 8 October 2002 17:18:56 UTC