Re: FSE_ARG_INT64 argument in fsevent
Re: FSE_ARG_INT64 argument in fsevent
- Subject: Re: FSE_ARG_INT64 argument in fsevent
- From: Arun prabu <email@hidden>
- Date: Sun, 02 Dec 2012 17:56:39 +0530
Please do not do that. There is no need to do that.
The FSEvents framework provides file-level events
now (and has since 10.7).
That is very good and going to be very helpful. I have some 10.5 machines as well. Will there be any problem if /dev/fsevents is used in 10.5.
Another question. Whenever a file is modified, the events are sent only after the file is closed. Is there a way to get events everytime the file is flushed instead of when the file is closed. Because, there are some applications which frequently writes data to a file. The data is flushed at regular intevals but the file is closed when the application exits. In such case, content modified event will be received at the end eventhough data is already flushed to the disk at multiple intervals.
The behaviour is same in /dev/fsevents and also FSEvents API.
On Sat, Dec 1, 2012 at 12:58 AM, Dominic Giampaolo
<email@hidden> wrote:
> I am using /dev/fsevents to receive file system events. For each event, there is an argument of type
>
Please do not do that. There is no need to do that.
The FSEvents framework provides file-level events
now (and has since 10.7).
> FSE_ARG_INT64 which is 64 bits and the value keeps increasing for each event. It see different explation for this across internet. Some say this is not implemented and some say it represents timestamp. How should FSE_ARG_INT64 be interpreted.
>
Do not depend on the value in any way. It is not any
form of a timestamp at all. It is simply a counter
that increases.
--dominic
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Filesystem-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden