I don't mind changing the docs because this is an area that is very
confusing and hard to explain and in need of an overhaul anyway.
Some of this information is in the Deployer's Guide and some of it is
scattered throughout the Developer's Guide, and there is much stale
information (for example lzt=history) that I have bugs against
anyway. Since I'm going to have to change it, I would rather change
it to something logical.
That does not address Henry's point about the code, however.
jrs
On Jan 18, 2007, at 11:19 PM, Henry Minsky wrote:
That sounds like a good idea. Unfortunately there are a lot of
places in the
code and docs (and wrappers and solo deployer) that are going to be
affected
by that change. Where's our new summer intern ? :-)
On 1/18/07, P T Withington <[EMAIL PROTECTED]> wrote:
I'd rather change lz(request)t(ype) to lzw(rapper). Better yet,
we've discussed a number of times trying to unify the lz? options,
perhaps by saying something like:
?lzoptions=runtime:dhtml,wrapper:html,debug:false,proxy:true
[I don't know for a fact that that is or is not a legitimate URI
query parameter syntax, or if : and , already have reserved meanings
in query parameters. Presumably we could come up with something
along those lines.]
Similarly, the litany of server requests that are lumped into the ?
lzt argument should be split out as something like:
?lzoptions=request:clear-cache
etc.
The biggest benefit is that we would remove a lot of pollution from
the query arg space (especially the non-lz-prefixed args like
profile, proxy, etc.).
If `runtime:dhtml,wrapper:html` is still confusing, I suggest we
change dhtml to ajax.
On 2007-01-18, at 20:36 EST, David Temkin wrote:
> With the addition of lzr=dhtml, lzt=html has become pretty
confusing.
>
> Any thoughts on deprecating lzt=html in favor of lzt=wrapper, or
> something along those lines?
>
>
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
