[9fans] abaco and gmail

[9fans] abaco and gmail

[ron minnich]

gmail keeps claiming I have cookies off, but they seem to be on in
webfs. Anyone tried gmail and abaco yet?

thanks

ron

Re: [9fans] abaco and gmail

[Federico Benavento]

there is a bug in the cokies code webfs.

On 2/16/07, ron minnich <rminnich@gmail.com> wrote:
> gmail keeps claiming I have cookies off, but they seem to be on in
> webfs. Anyone tried gmail and abaco yet?
>
> thanks
>
> ron
>


-- 
Federico G. Benavento

Re: [9fans] abaco and gmail

[Steve Simon]

I too cannot get gmail to work with abaco,
though cookies work fine with other sites I
access - strange.

Does plan9's pop client work properly with Gmail these days?

When gmail was first released it would mark messages as read after
they had been accessed once - this once was plan9's pop client
just scanning the inbox, so messages would disappear as soon
as the client realised they where there.

-Steve

Re: [9fans] abaco and gmail

[erik quanstrom]

can you describe the bug?

- erik

Re: [9fans] abaco and gmail

[erik quanstrom]

i think what you want is a pop fetcher.  since the point
of mail/fs pop client is to leave the messages on the server,
i think it would make more sense to write mail/fetch or somesuch
to handle cases like this

i don't use gmail, but it seems like it would be worth fixing.
can somebody tell me what webfs is doing wrong?

also, one fundamental problem with the current webfs is that
the interface claims that it handles base url+relative url,
but it doesn't so abaco (and others) try to do it itself.
unfortunately urls are needlessly hairy creatures (c.f.
webfs/url.c:/^merge_relative_path) and abaco's simple algorithm
doesn't always work.

i have a fix for this problem, but it's not been accepted into
sources.

- erik

Re: [9fans] abaco and gmail

[Gorka Guardiola]

So javascript is working already?.
-- 
- curiosity sKilled the cat

Re: [9fans] abaco and gmail

[erik quanstrom]

it's part of ozinferno.

- erik

Re: [9fans] abaco and gmail

[Gorka Guardiola]

It was a legitimate question. No gmail on a web page without javascript.

On 2/16/07, erik quanstrom <quanstro@coraid.com> wrote:
> it's part of ozinferno.
>
> - erik
>


-- 
- curiosity sKilled the cat

Re: [9fans] abaco and gmail

[erik quanstrom]

i don't know how fgb is doing, but after reading the ecmascript spec, i was
considerably less hopeful that i'd have the time to implement it.  imho, the worst
feature is that it requires having a runtime scanner and parser because javascript
can generate code at runtime and call it.  also, javascript requires that one map
page elements to javascripts object model -- one can theoretically reference any
element of a page from javascript.  

this tight coupling of the browser and javascript engine means that importing an
engine doesn't sound like even less fun than writing one.

on the other hand, perhaps this is an inspiration for a followon to acme.
with the right model, it would be considerably more fun to implement.

- erik

Re: [9fans] abaco and gmail

[ron minnich]

On 2/16/07, Gorka Guardiola <paurea@gmail.com> wrote:
> It was a legitimate question. No gmail on a web page without javascript.

no, there is a 'simple mode' you can use without javascript. But
something about gmail and abaco is not working.

ron

Re: [9fans] abaco and gmail

[Bruce Ellis]

a tight coupling is transparent in (oz/any)inferno.

On 2/17/07, erik quanstrom <quanstro@coraid.com> wrote:
> i don't know how fgb is doing, but after reading the ecmascript spec, i was
> considerably less hopeful that i'd have the time to implement it.  imho, the worst
> feature is that it requires having a runtime scanner and parser because javascript
> can generate code at runtime and call it.  also, javascript requires that one map
> page elements to javascripts object model -- one can theoretically reference any
> element of a page from javascript.
>
> this tight coupling of the browser and javascript engine means that importing an
> engine doesn't sound like even less fun than writing one.
>
> on the other hand, perhaps this is an inspiration for a followon to acme.
> with the right model, it would be considerably more fun to implement.
>
> - erik
>