SF.net SVN: harbour-project:[14736] trunk/harbour

[snai...@users.sourceforge.net]

Revision: 14736
http://harbour-project.svn.sourceforge.net/harbour-project/?rev=14736&view=rev
Author: snaiperis
Date: 2010-06-10 16:12:39 +0000 (Thu, 10 Jun 2010)

Log Message:
-----------
2010-06-10 19:12 UTC+0300 Mindaugas Kavaliauskas (dbtopas/at/dbtopas.lt)
* harbour/examples/httpsrv/uhttpd.prg
* harbour/examples/uhttpd2/umain.prg
* optimized hb_socketRecv() calls
* optimized (removed) hb_socketSelect() calls, current API
has timeout parameter and make things simple

* harbour/examples/commouse/commouse.prg
! fixed error reporting (missing hb_comGetError()) parameter

* harbour/include/hbsocket.ch
* adjusted comment text

* harbour/contrib/hbtpathy/telepath.prg
! fixed missing nTimeout parameter in tp_send()

* harbour/source/rtl/hbcomhb.c
! fixed typo in hb_comSend()
! fixed hb_storni() paramter order bugs

Modified Paths:
--------------
trunk/harbour/ChangeLog
trunk/harbour/contrib/hbtpathy/telepath.prg
trunk/harbour/examples/commouse/commouse.prg
trunk/harbour/examples/httpsrv/uhttpd.prg
trunk/harbour/examples/uhttpd2/umain.prg
trunk/harbour/include/hbsocket.ch
trunk/harbour/src/rtl/hbcomhb.c

This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

[Viktor Szakáts]

Thank you very much Mindaugas.

> * harbour/include/hbsocket.ch
> * adjusted comment text

Actually I had it first how you reverted it now, but
intentionally changed it to functions where these constants
actually play role. Seems more self-documenting that way.

Viktor

[Mindaugas Kavaliauskas]

On 2010.06.10 19:17, Viktor Szak�ts wrote:
>> * harbour/include/hbsocket.ch
>> * adjusted comment text
>
> Actually I had it first how you reverted it now, but
> intentionally changed it to functions where these constants
> actually play role. Seems more self-documenting that way.

But there are more functions that return address (by reference:
hb_socketAccept(), hb_socketRecvFrom()), or accept address as parameter
(hb_socketBind(), hb_socketConnect(), hb_socketSentTo()). All of these
are related to address array index, so, all should be mentioned. I
thought it is to much to mention all of them, and
hb_socketGet{Sock|Peer}Name() does not seem to be the most common
functions from socket library.

Regards,
Mindaugas

[Viktor Szakáts]

>> Actually I had it first how you reverted it now, but
>> intentionally changed it to functions where these constants
>> actually play role. Seems more self-documenting that way.
>
> But there are more functions that return address (by reference: hb_socketAccept(), hb_socketRecvFrom()), or accept address as parameter (hb_socketBind(), hb_socketConnect(), hb_socketSentTo()). All of these are related to address array index, so, all should be mentioned. I thought it is to much to mention all of them, and hb_socketGet{Sock|Peer}Name() does not seem to be the most common functions from socket library.

I didn't know about the rest. If it's general index
for all "addr" type of array arguments, the two names
indeed don't make much sense. I will remove them.

Thanks for clarifying!

Viktor