There is no such thing as IAtom

15 views
Skip to first unread message

Pepijn de Vos

unread,
Dec 5, 2010, 5:29:56 PM12/5/10
to clo...@googlegroups.com
tl;dr: Please add an interface to clojure.lang.Atom. kthxbye

I had the brilliant idea of using CouchDB for something equally brilliant, and if possible implement a Clojure view server. Turns out Clutch fits the bill perfectly, except that I would like to use Aleph, and Couch is using something called MVCC.

I haven't looked into the libs to deep, but I know the REST API, and what it does for updating is that you need to supply the current revision and if it doesn't match (i.e. there has been another update), the update fails. So then you need to get the new _rev and try again. Much the same way compare-and-set! works.

I thought it would be perfect to implement IAtom and IDeref to get the latest value and to swap! documents in a spin loop with a side-effect-free function, like atoms do.

But it turns out there is no interface for Atom and it is marked final. The same is true for Ref and Agent, but raek suggested this might be because they are more complex and integrated with the STM.

Is there a good reason why I'm not allowed to implement my own atom? If not, can it be added? Thanks.

Groeten,
Pepijn de Vos
--
Sent from my iPod Shuffle
http://pepijndevos.nl

Benjamin Teuber

unread,
Dec 6, 2010, 5:24:40 AM12/6/10
to Clojure
I guess it was Rich's intention to have swap! be used for real atoms
only so your code remains understandable - that's why it's called
swap! for atoms, alter for refs and alter-var-root for vars.

So why not define your own protocol for updating documents? If you
really want (usually bad idea, I guess) you could still extend atoms
or even refs to support this protocol, too.

Regards,
Benjamin

pepijn (aka fliebel)

unread,
Dec 6, 2010, 11:52:02 AM12/6/10
to Clojure
You can not extend them, as they are marked final.

Another point to consider is clojure-in-clojure. If that is ever going
to happen, one needs to be able to implement Atom as well. It is also
generally better to code to an interface rather than to an
implementation.

Alter and send also work differently from swap!, so I think it makes
sense to have separate functions for them. But CouchDB also exhibits
shared, synchronous, independent state, so I don't see the point in
prohibiting the use of the same interface.

Stuart Sierra

unread,
Dec 6, 2010, 12:58:50 PM12/6/10
to Clojure
On Dec 6, 11:52 am, "pepijn (aka fliebel)" <pepijnde...@gmail.com>
wrote:
> You can not extend them, as they are marked final.

You can extend Clojure protocols to final classes.

> Another point to consider is clojure-in-clojure. If that is ever going
> to happen, one needs to be able to implement Atom as well.

clojure-in-clojure will probably be defined in terms of protocols.
Whether or not "IAtom" is one of those I can't say.

-S

Alyssa Kwan

unread,
Dec 6, 2010, 1:08:42 PM12/6/10
to Clojure
+1

There is no STM integration with atoms. That's not a concern.

Just write your own Clojure core with your change. I did for durable
identities. <shamelessPlug>git://github.com/kwanalyssa/clojure.git</
shamelessPlug>

Seriously though, just use protocols.

Thanks,
Alyssa
> > Sent from my iPod Shufflehttp://pepijndevos.nl- Hide quoted text -
>
> - Show quoted text -

pepijn (aka fliebel)

unread,
Dec 7, 2010, 11:30:04 AM12/7/10
to Clojure
Thanks all.

How would I do this protocol extension thing? My guess is that I need
to define the IAtom protocol myself, and then do something like this?

(defprotocol IAtom
(compare-and-set! [])
...)

(extend-type clojure.lang.Atom
IAtom
(compare-and-set! [])
...
IDeref
(deref []))

But will that pass a Atom type hint? I will play with it and see for
myself. Thanks.
> > > Sent from my iPod Shufflehttp://pepijndevos.nl-Hide quoted text -
Reply all
Reply to author
Forward
0 new messages