<DIV>Hi Kaws,</DIV>
<DIV> </DIV>
<DIV>I may need to add some notes;</DIV>
<DIV> </DIV>
<DIV>1. The default .rpm version of libtiff in RH9 does not seem to contain some functional entries required for libunicall / libmfcr2. Therefore when compiling libunicall "unresolved references....." will happen. To solve this, I needed to recompile libtiff from the tarball. When installing libtiff, the default directories will be under /usr/local, so you will need to link /usr/lib/libtiff staff to /usr/local/lib/libtiff staff.</DIV>
<DIV> </DIV>
<DIV>2. The installed libtiff from tarball still does not contain tif_dir.h and tiffiop.h in right place (i.e. /usr/include or /usr/local/include), so you need to copy these staff from "libtiff" directory under the untarred directory. In fact you do not need to use Steve's header files if you do like this.</DIV>
<DIV> </DIV>
<DIV>3. At least the latest version of spandsp, libunicall and libmfcr2 can be compiled against the correct libtiff installation. The libspandsp should be installed first among all others as libunicall and libmfcr2 requires libspandsp (though Steve's web starts from libunicall).</DIV>
<DIV> </DIV>
<DIV>4. I still cannot run TESTCALL application correctly for Brazil, though. I wonder if this happens country specific or not, so please input your result in your country's implementation.</DIV>
<DIV> </DIV>
<DIV>HOSIN<BR><BR><B><I> kaws elchamal <alfdz16@hotmail.com></I></B> wrote:</DIV>
<BLOCKQUOTE class=replbq style="PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #1010ff 2px solid"><BR>hello HO SIN<BR><BR>thank you<BR><BR>l'll try this solution and i'll inform you<BR><BR>See you<BR><BR>kaws<BR><BR>><BR>>Hi Kaws,<BR>><BR>>First I installed libtiff from tarball downloaded from (it might be ok):<BR>>http://www.remotesensing.org/libtiff/<BR>><BR>>Later, I noticed Steve had made it downloadable when I carefully read his <BR>>site:<BR>>http://www.opencall.org/installing-spandsp.html<BR>>The required header files - tiffiop.h and tif_dir.h were in the very first <BR>>version "/pub/spandsp/spandsp-0.0.1". I only looked in for the latest <BR>>version......<BR>><BR>>HOSIN<BR>><BR>> kaws elchamal <ALFDZ16@HOTMAIL.COM>wrote:<BR>><BR>>Dear HO SIN,<BR>><BR>>many thanks for "L" solution you give me...<BR>><BR>>I have also libtiff troubles with RH9, how did you resolve them ?<BR>><BR>>R2 protocol I use work
only with AB bits(CDbits ares fixed) and it run<BR>>perfectely with the testcall.c, i'll try with chan_unicall and i'll give <BR>>you<BR>>the result...<BR>><BR>>*** GOOD luck Steve for updating the patch of last asterisk ***<BR>><BR>>thank you for your help<BR>><BR>>Kaws<BR>><BR>> ><BR>> >Message: 10<BR>> >Date: Sat, 27 Nov 2004 02:26:12 +0800<BR>> >From: Steve Underwood<BR>> >Subject: Re: ^3 [Asterisk-Users] problem with chan_unicall.c for<BR>> > MFC/R2 with asterisk<BR>> >To: Asterisk Users Mailing List - Non-Commercial Discussion<BR>> ><BR>> >Message-ID: <41A77544.2030300@coppice.org><BR>> >Content-Type: text/plain; charset=ISO-8859-1; format=flowed<BR>> ><BR>> >HO SIN wrote:<BR>> ><BR>> > > Dear Kaws, and Dear Steve,<BR>> > ><BR>> > > I think I am facing the same kind of trouble. This kaws's case is<BR>> > > simple; one line
inserted by patch designates wrong header file name<BR>> > > of "unical.h", therefore the objects are not compiled. The reality<BR>> > > is "unicall.h", and<BR>> ><BR>> >OK, fixed that one.<BR>> ><BR>> > > after modifying it, then I faced trouble that the compile does not go<BR>> > > through as some of Asterisk structure definition (for example,<BR>> > > ast_channel) seemed to have changed after around October version and<BR>> > > unicall.c could not compile properly. In fact the latest version of<BR>> > > Asterisk does not compile with the current unicall. I tracked back<BR>> > > to the version around late September.<BR>> ><BR>> >I am in the process of updating to the latest asterisk, but right now<BR>> >chan_unicall will only work with versions of * a month or more old.<BR>> ><BR>> > ><BR>> > > My environment is RH9, and resolved libtiff trouble, and
the compile<BR>> > > finally went through, but it still does not operate properly (does not<BR>> > > understand line signalling - AB bits correctly).......hmmm<BR>> > ><BR>> ><BR>> >Which version of R2 are you using? Most use the ABCD bits in the same<BR>> >way. China and Thailand are different. The code should be allowing for<BR>> >those. It certainly works OK for China.<BR>> ><BR>> >Steve<BR>> ><BR>><BR>>_________________________________________________________________<BR>>MSN Actions Solidaires : volontaire El'étranger<BR>>http://www.msn.fr/actionssolidaires/Default.asp?Ath=f<BR>><BR>><BR>><BR>>---------------------------------<BR>>Do you Yahoo!?<BR>> Meet the all-new My Yahoo! ETry it today!<BR><BR>_________________________________________________________________<BR>MSN Actions Solidaires : volontaire El'étranger
<BR>http://www.msn.fr/actionssolidaires/Default.asp?Ath=f<BR><BR></BLOCKQUOTE><p>
<hr size=1>Do you Yahoo!?<br>
The <a href="http://my.yahoo.com">all-new My Yahoo!</a> – What will yours do?