On Thu, Mar 24, 2005 at 05:43:55AM +0100, Siward de Groot wrote: > * functions FAMMonitor* take an argument 'void* userData', but it is > not described in this manpage.
Hi Siward, Sorry that it took such a long time to get back to you. Here are some comments to your reports. The userData is described in the manpage. The second sentence in the "FAMMonitorDirectory, FAMMonitorFile" section says that "The parameters to this function are ... a user data pointer." > * manpage says that struct FAMEvent's member fr has type FAMRequest, > but the compiler disagrees (and it's not a pointer to a FAMRequest > either). I am not sure why you think fr does not have type FAMRequest because that is exactly what the definition in fam.h says. > * i noticed that famd can report multiple change events for a single > change, atleast that's what i saw when i saved a modified file with > xemacs, while simply touch-ing the file resulted in only one event; > maybe it repors changes in filesize and modificationtime separately? > In that case, it could be usefull to document that in the manpage. >From how I understand things, the behavior you observe should be a bug. See #239540 for another description of your behavior. > * manpage says, in description of FAMOpen/FAMClose, that "Variable > char* appName should be set to name of your application.", however, > there is no such variable (at least it is not described in this > manpage). Okay, a fix is forthcoming. > * if i can believe the description in the manpage, > FAMSuspend/ResumeMonitor are completely superfluous, as they have > identical result to simply not requesting events. I think you are right. > * error-reporting is completely broken, according to this manpage, > and there is no mention of what kind of errors could occur, > which makes it kinda difficult to decide how to work around it. I agree, but there is little I can do about that. > * it seems only way to get events in a non-blocking way is to set an > alarm ; but what if alarm goes off while app is receiving an event ? > does this break things or is it atomical ? This is a Unix programming problem not specific to fam. You can find discussions to this issue in the book "Advanced programming in the Unix environment" by Stevens. > * see also refers to fstat(1), which doesn't exist because man fstat > is in section 2; not really a big problem, but it makes my 'mkmyman' > script fail to find it's whatis. Good catch. Fix forthcoming. > * it doesnt include any examples ; if it had, i wouldn't have had to > write my testprogram, so i'm attaching that in case you would like > to include it. Thanks for supplying the example. However, I think it is not really suitable for the man page -- instead I can put it in the libfam-dev package documentation directory if you clean the file up (apply proper indentation and such) and put a DFSG-free license at the top of the source > thanks for maintaining fam, it does work :-) Thanks for submitting the bug report! -- Chuan-kai Lin http://web.cecs.pdx.edu/~cklin/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
