Darren Nickerson [Tue, 18 Jun 2002 19:28:48 +0000 (19:28 +0000)]
[Bug 291] Solaris DSO support
Solaris DSO support is broken when using Sun's native compiler.
Until it's fixed, DSO is only available when using gcc/g++
Darren Nickerson [Sat, 15 Jun 2002 18:00:06 +0000 (18:00 +0000)]
[Bug 305] gcc-3.x support
The combined effort of Robert & Lee has led to a viable first cut at gcc-3.x
support. This may require further work, but does not appear to break anything!
Darren Nickerson [Fri, 14 Jun 2002 19:31:22 +0000 (19:31 +0000)]
[Bug 290] Changes to compile on Solaris 8 with Forte 6.2
Thanks to Sean Kamath for tidying up support for Slowlaris' native compiler,
and to Lee for the committable adaptation
Darren Nickerson [Fri, 14 Jun 2002 19:23:58 +0000 (19:23 +0000)]
[Bug 202] faxrm & faxalter format string vulnerabilities
Collaborative effort to squash these, led by Lee. Thanks everyone!
This is a preliminary commit, which should be sanity-checked by anyone who
has a second, since we'd like to urge people to deploy 4.1.3 as a security
update
Darren Nickerson [Fri, 14 Jun 2002 19:03:43 +0000 (19:03 +0000)]
[Bug 303] faxinfo reports V.34-fax received files as 14.4
Faxinfo now supports correct faxinfo information for Super G3-received faxes,
thanks to some fine solo work from Lee. Awesome.
Darren Nickerson [Tue, 11 Jun 2002 21:25:16 +0000 (21:25 +0000)]
[Bug 302] support MultiTech Real-Time Fax Compression Conversion
Thanks to the close cooperation between Lee and Steve@MultiTech, this
is an almost overnight reality. Thanks fellas!
This patch implements RTFCC via a Class2RTFCC config option, and changes
the default prototype config for the MT5634ZBA/ZPX-PCI-V92 modems to
not receive in MMR..., since HylaFAX is apparently unable to receive an
MMR fax properly [Bug 223]
[Bug 275] MultiTech V92 V.34-Fax Class 2.1
Sets the default for these modems to not enable HDLC frame tracing, since
some firmware revisions incorrectly wrap the second line of long HDLC frame
traces
[Bug 301] adds Class2PTSQueryCmd
iFax Solutions has determined that some hardware does not get along with
the AT+FPS comment which Dmitry forced upon the Class 2.0 driver some time
ago. This patch creates a config parameter which allows us to revert to the
old behaviour for such devices. Many thanks to Lee for some fine diagnostic
work on this one.
[Bug 300] faxgetty segfaults if the sender uses a specially formatted TSI
Applied Vyacheslav's patch removing the format string vulnerability,
preventing faxgetty from segfaulting. Nice work team!
[Bug 295] Conversion of TIFF to PDF always assumes letter page size
John Florian and Lee have conspired to make faxrcvd more intelligent about
page sizes. Nice work gentlemen!
Darren Nickerson [Fri, 24 May 2002 16:32:37 +0000 (16:32 +0000)]
[Bug 292] Class 1 sending CRP handling
Vyacheslav pointed out that we don't handle the receipt of CRP in Class 1
sending. This patch fixes that. Thanks Lee!
Darren Nickerson [Fri, 24 May 2002 16:30:04 +0000 (16:30 +0000)]
[Bug 293] syslog FIFO can break timing
This patch eases the FIFO bottleneck for now. Multi-threading considered to
be the proper long-term solution? Thanks Vyacheslav and Lee!
Darren Nickerson [Fri, 24 May 2002 16:21:18 +0000 (16:21 +0000)]
[Bug 282] changes Class1SwitchingDelay to Class1SwitchingCmd
According to T.31 Appendix-II.1 we should be using +FRS instead of a software
pause (Class1SwitchingDelay) between when we receive an HDLC frame and before
we send one in response. This patch replaces Class1SwitchingDelay with
Class1SwitchingCmd
Darren Nickerson [Fri, 24 May 2002 15:44:14 +0000 (15:44 +0000)]
[Bug 277] here are the patches that Mandrake and Debian use
Merged in relevant vendor fixes for 64-bit compatibility.
Thanks Lee and Patrice for sifting through the detritus!
Darren Nickerson [Thu, 23 May 2002 21:09:11 +0000 (21:09 +0000)]
[Bug 275] MultiTech V92 V.34-Fax Class 2.1
Thanks to some tremendous engineering by Lee & Patrice, and the invaluable
collaboration of the fine folks @ MultiTech who continue to impress us with
their willingness to guide our efforts, we now have the first functional
implementation of class 2.1 (Super G3 or V.34). Also deserving a special
mention is Guillaume Legoupil, whose patch to add v.34 faxing to
class 2/2.0 was the starting point for this whole effort.
Congratulations to all of you for a fine piece of teamwork!
Darren Nickerson [Thu, 23 May 2002 20:38:44 +0000 (20:38 +0000)]
[Bug 258] faxsetup should fixupFaxUser if the homedir is incorrect
People upgrading to HylaFAX-4.1 or newer will now have the opportunity to
correct the user's passwd entry to have the correct homedir (new spooldir)
[Bug 281] adds prototype support for Eicon
Special thanks to Ruben Martinez & James Jordan @ EICON for the EICON
Diva T1/Pri loaner on which these prototypes were developed.
[Bug 201] MaxDials is broken with sendfax
It turns out that only the documentation was broken. Feature, not a bug.
Applies Lee's patch to clarify the documentation.
Darren Nickerson [Thu, 11 Apr 2002 18:17:06 +0000 (18:17 +0000)]
[Bug 265] Remote has no T.4 receiver capability error message
Error is now more descriptive of the particular T.4 incompatibility
(No Receiver). Thanks Patrice!
Darren Nickerson [Thu, 11 Apr 2002 17:55:30 +0000 (17:55 +0000)]
[Bug 256] faxcron can delete seqf
Faxcron will now leave seqf alone, as it should always have.
Thanks to Christian Recktenwald and Peter Samfest, who originally reported
the problem (and the fix).
[Bug 247] faxaddmodem shouldn't require the full ATI0 result in the prototype
Makes faxaddmodem less greedy . . . use regex to ignore sugar at the end of
ATI0 result where firmware revision lives. Nice work Lee!
[Bug 236] can't write to $PWD in configure, etc.
Applied Lee's patch to change the (often) reserved word PWD to PWDCMD,
which allows us to mess with it with impunity. Thanks!
[Bug 231] timedout jobs should have their notice field specifying the cause of failure
Applied your fix Patrice, fields now reads "Kill time expired". Thanks!
[Bug 184] Problems building HylaFax 4.1 on Digital Unix 4.0F
Backing out the changes which john recanted on the mailing list. Apparently
his compiler was broken at the time. Thanks for feeding this into bugzilla
Lee!
[Bug 67] ModemDialCmd in config.modem
Applies Lee's patch to remove @ from ModemDialCmd and reduce default ModemRate
to 19200 from 38400. Thanks to Michael Fengler for originally reporting
the problem.
Darren Nickerson [Fri, 15 Feb 2002 04:52:20 +0000 (04:52 +0000)]
[Bug 233] problems with Ghostscript Fontmap.GS file
Added runtime Fontpath/Fontmap detection, and check for ghostscript's
newer Fontmap.GS convention, with Fontmap as a fallback. Thanks Lee & Giulio!
Darren Nickerson [Fri, 15 Feb 2002 04:41:38 +0000 (04:41 +0000)]
[Bug 220] unblocked jobs are sent to the end of the send queue
Blocked jobs were being sent to the back of the bus. This should restore
them to a priority which more accuratly reflects their original injection
into the send queue. Contributed by Patrice Fournier.
Darren Nickerson [Fri, 15 Feb 2002 04:31:02 +0000 (04:31 +0000)]
[Bug 215] HylaFAX DNIS routing enhancement
Primarily a documentation update, other than passing the cid.name value
from faxgetty to faxrcvd. Thanks Lee!
Darren Nickerson [Fri, 15 Feb 2002 04:20:48 +0000 (04:20 +0000)]
[Bug 78] RFE: new status flag F=Failed
Change status of completed jobs to:
D - for transmitted jobs
F - for failed jobs.
Contributed by Patrice Fournier.
Darren Nickerson [Fri, 15 Feb 2002 04:04:49 +0000 (04:04 +0000)]
[Bug 199] adds prototype for Topic modems
More stamp collecting - this modem does not support +FTS=n, so
Class1PPMWaitCmd, Class1TCFWaitCmd and Class1EOPWaitCmd get setup explicitly.
Darren Nickerson [Fri, 15 Feb 2002 03:55:03 +0000 (03:55 +0000)]
[Bug 197] Class1 : +FTS=n must be used
Use +FTS=n instead of software pause() to improve Class1 timing, and revert
EOP-related pause to ~95 ms, its historic value which was recently (probably)
erroneously changed (see Bug 152). Thanks to everyone who contributed their
feedback & votes!
Tim Rice [Thu, 14 Feb 2002 05:18:53 +0000 (05:18 +0000)]
[config.guess] newest from ftp://ftp.gnu.org/gnu/config/config.guess
[configure] add OpenUNIX support
[configure, faxd/ModemServer.c++, man/hylafax-config.4f]
SCO priority scheduling fixes
[faxd/faxQCleanApp.c++] Bug 160: Bernd Proissl's rewrite of a
non portable patch by John Patrick Poet
[html/building.html] minor update for SCO
[Bug 211] adds ModemSoftResetCmdDelay
ModemSoftResetCmd in default init string may have been confusing modems.
This patch, developed on -devel and with Harald's counsel, provides a
forced pause immediately following ModemSoftResetCmd which is configurable
via the ModemSoftResetCmdDelay config param. Nice work all!
-dpn
[Bug 210] enhances/fixes Class 1 parseQuery
Applied a.pogoda's parser enhancement, a workaround for modems sending
(<item>,<item>,...) enclosed in brackets. Case in point, elsa microlink 56k
internet II
-dpn
[Bug 198] adds default support for new 3Com USR modems
Applied Lee's patch to address James Werkowski's observation that we were
not detecting "3Com USRobotics Courier V.Everything" correctly.
-dpn
[Bug 194] changes faxaddmodem default to use Class 1
This fairly major paradigm shift seems to accurately reflect the 'rough
concensus' of the -users and -devel list. Class 1, thanks in part to Dmitry's
fine work between 4.0 and 4.1, is a much more forgiving implementation.
Committed your patch Lee - thanks!
-dpn
[Bug 193] Class1SwitchingDelay default value adjustment
Reapplied this patch (accidentally reversedsomehow), which changes the default
from 150 to 75ms. This is needed for most ISDN modems. Thanks Lee!
-dpn
[Bug 184] Problems building HylaFax 4.1 on Digital Unix 4.0F
Applied Lee's reformat of John's patch, which should go a long way towards
cleaning up Digital Unix builds. Great work John - thanks!
-dpn
[Bug 157] - improved glare handling needed
Applied Lee's patch to help cope with situations where an incoming call RINGs
before the modem preparing an outbound fax is able to pick up the line with ATDT
-dpn