Re: cups-filters-1.27 problem? [Re: Input Filter and LPD]

[Predrag Punosevac]

Stefan Sperling  wrote:

> On Sun, Jun 14, 2020 at 02:58:18PM +0200, Marcus MERIGHI wrote:
> > i think this belongs to ports@. though base lpd(8) is in the Subject:, 
> > this is probably abount cups-filters-1.27 failing on us.
> 
> > $ cat /etc/printcap.sh
> > #!/bin/sh -e
> > /usr/local/bin/a2ps -BRq --columns=1 -o - | \
> > /usr/local/bin/foomatic-rip --ppd \
> > /etc/foomatic/direct/brother-mfc9120cn.ppd \
> > -P brother-mfc9120cn
> 
> Try removing a2ps from your filter script. This no longer works as the
> cups-filters readme file suggests.

We must have been reading the different cups-filters readme file. This
is on 6.7 release




$OpenBSD: README,v 1.11 2020/02/22 14:51:47 ajacoutot Exp $

+---
| Running cups-filters on OpenBSD
+---

Bonjour broadcasts of shared, remote printers
=

The cups-browsed(8) daemon can be used to browse Bonjour (mDNS)
broadcasts of remote printers and create/remove local raw queues
pointing to these printers. This requires the avahi-daemon(8) to be
running first. To start it at boot time, the rc.conf.local(8)
pkg_scripts line can be edited as follow:

pkg_scripts=[...] messagebus avahi_daemon cups_browsed [...]

lpd(8) using direct printing with foomatic-rip(1)
=

The Foomatic framework supports direct printing which does not rely on
CUPS (nor any spooler) and can thus be used with lpd(8).

Complete documentation can be found at:

http://www.linuxfoundation.org/collaborate/workgroups/openprinting/database/
nospoolerdocumentation

Foomatic configuration
--
A specific PPD file has to be created for the printer. Refer to the
documentation that comes with the foomatic-db-engine package to learn
how to generate PPD files
(cf. /usr/local/share/doc/pkg-readmes/foomatic-db-engine).

# mkdir -p /etc/foomatic/direct
# foomatic-ppdfile -p 'Samsung-ML-2850D' -d 'Postscript' >
/etc/foomatic/direct/samsung-ml2850d.ppd

Alternatively, you can look under
/usr/local/share/foomatic/db/source/PPD
for existing PPD files matching your printer (in which case the PPD file
can be copied and/or gunzip'd under /etc/foomatic/direct).

Input filter script
---
Non-postscript files must be converted before being sent to
foomatic-rip(1). Several packages can be used for that, a2ps being the
most complete.
Note that a2ps(1) requires external helper tools for converting files.
They can be configured under /etc/a2ps{,-site}.cfg. By default,
converting images requires the ImageMagick package.

Here's a sample converter script:

---8<---
#!/bin/sh

/usr/local/bin/a2ps -BRq --columns=1 -o - | \
/usr/local/bin/foomatic-rip -P samsung-ml2850d
---8<---

Additional options can be passed to foomatic-rip(1).
e.g.
- to print in duplex mode: -o duplex
- to set the page size to letter: -o pagesize=letter

Default options can be set by editing the PPD file.

To use the above script as an input filter for lpd(8), see the next
sections (make sure the script is executable).

lpd(8): USB printer printcap(5) example
---
lp|samsung|Samsung-ML-2850D:\
:lp=/dev/ulpt0:\
:if=/path/to/script.sh:\
:sd=/var/spool/output:\
:lf=/var/log/lpd-errs:\
:sh:

lpd(8): network printer printcap(5) example
---
rp|samsung|Samsung-ML-2850D:\
:lp=9100@1.2.3.4:\
:if=/path/to/script.sh:\
:sd=/var/spool/output:\
:lf=/var/log/lpd-errs:\
:sh:
(where 1.2.3.4 is the printer IP address and 9100 the printer stream
port)

foomatic-rip(1) does *not* speak LPD (port 515).
If the printer does not support raw printing over port 9100, it must be
setup locally on a print server (see above for an example using USB)
then accessed over LPD by the clients (there is no need to setup any
print filter on the clients since it will run on the print server).



I did try your suggestions and I didn't get raw PostScript but I didn't
get anything either. The printer kept printing blank pages.


> Removing the a2ps line fixed PDF printing for me after upgrading to 6.7.
> 
> Yes, that means printing plain ASCII files is broken. But that can be
> worked around by manually converting ASCII to PDF first. And because
> a2ps never worked correctly with UTF-8 it was rather useless anyway.

I am old enough to remember the times all PostScript printers could
print raw ASCII files without fiddling with filters or even a spooling
system. For the most part I don't need UTF-8 so that was not an issue
for me.  However, the native lpr command doesn't have a switch for
duplex printing. That is expected as lpd is only a spool

Re: [S-mailx] .nailrc and Gmail

[Predrag Punosevac]

Predrag Punosevac  wrote:

> 
> Predrag Punosevac  wrote:
> 
> > Hi Steffen,
> > 
> > I apologize for a direct email. I am not sure new mailing list will
> > allow me to post. 
> > 
> > I kept my own v14.8.12 copy of nail after few things got broken on
> > OpenBSD.  I didn't notice that you adopted OpenBSD port and updated it
> > to 14.9 branch until today. Old port was behind the version I was using.
> > 
> 
> 
> Hi Steffen,
> 
> I apologize for cross posting. After upgrading my laptop to 
> 
> predrag@oko-mobile$ uname -a
> OpenBSD oko-mobile.int.bagdala2.net 6.7 GENERIC.MP#2 amd64
> 
> I felt it was the time for me to jump the ship and finally go with
> s-nail from the official ports tree. 
> 
> predrag@oko-mobile$ pkg_info s-nail
> Information for inst:s-nail-14.9.17
> 
> I got to the bottom of all "issues" I originally reported 
> 
> https://www.mail-archive.com/s-mailx@lists.sdaoden.eu/msg00948.html
> 
> in the thread. I used quotation marks around issues as in the hindsight
> there was really only one. All other issues were due to the fact that I
> didn't realize that you have completely rewritten s-nail and there is
> really not much in common with the original Heirloom mailx
> 
> http://heirloom.sourceforge.net/mailx.html
> 
> which I used for at least 15 years. That is not to say that the things
> don't work or they are worse. They just work different. It took me two
> full days of dicking with it to get a to get a hang of it.
> 
> First thing first you really trough me off the board with Ctrl+D instead
> of next line and a dot to sent the email. I have not read the code, and

I am a total moron! I read carefully mail(1). Ctrl+D is of course
correct. What I didn't realize is that mail always reads

predrag@oko$ more /etc/mail.rc
set append dot save asksub
ignore Received Message-Id Resent-Message-Id Status Mail-From
Return-Path Via

s-nail of course has no access to that file.

set dot 

in my .s-nailrc and s-nail behaves as expected. As an upshot I
"backported" 14.9.19 to 6.7 stable branch. As you can see it works
as expected.

Cheers,
Predrag



> even if I did I don't have sufficient programming background to
> understand design decession but I am using dot to sent emails since
> circa 1989 and that is a hard pill to swallow. That is why I kept
> reporting that sending email doesn't work. 
> 
> I noticed that 14.9.17 on 6.7 doesn't report that annoying message
> 
> There are new messages in the error message ring (denoted by ERROR),
> nail:   which can be managed with the `errors' command
> ERROR# ? 
> 
> I really like new configuration grammar. This is my not so minimal
> working example
> 
> predrag@oko-mobile$ more .mailrc 
> set ask
> set crt
> ignore message-id received date fcc status resent-date resent-message-id
> resent-from in-reply-to
> 
> set mailx-extra-rc=~/.nailrc
> 
> and this is dotnailrc file
> 
> account gmail {
>   set inbox=imaps://usern...@imap.gmail.com
>   set imap-use-starttls
>   set password="secret"
>   set folder=imaps://usern...@imap.gmail.com record="+[Gmail]/Sent Mail"
>   set from="Predrag Punosevac " \
>   mta=smtp://usern...@smtp.gmail.com:587 \
>   set smtp-use-starttls 
>   set smtp-auth="login" 
> # IMAP SHORTCUTS SECTION for standard Gmail folders
>   shortcut allmail +[Gmail]/All\ Mail
>   shortcut sent +[Gmail]/Sent\ Mail
>   shortcut spam +[Gmail]/Spam
>   shortcut trash +[Gmail]/Trash
>   }
> account cmu {
>   set inbox=imaps://username%40andrew.cmu@imap.gmail.com
>   set imap-use-starttls
>   set password="secret"
>   set from="Predrag Punosevac " \
>   mta=smtp://username%40andrew.cmu@smtp.gmail.com:587 \
>   set smtp-use-starttls 
>   set smtp-auth="login" 
> # IMAP SHORTCUTS SECTION for standard Gmail folders
>   shortcut allmail +[Gmail]/All\ Mail
>   shortcut sent +[Gmail]/Sent\ Mail
>   shortcut spam +[Gmail]/Spam
>   shortcut trash +[Gmail]/Trash
>   }
> account hotmail {
>   set inbox=imaps://username%40hotmail@imap-mail.outlook.com
>   set folder=imaps://username%40hotmail@imap-mail.outlook.com
>   set imap-use-starttls
>   set password="secret"
>   set from="Predrag Punosevac " \
>   mta=smtp://username%40hotmail@smtp-mail.outlook.com:587
>   set smtp-use-starttls 
>   set smtp-auth="login"
>   }
> 
> 
> # Binary options
> set askattach
> set askcc
> set 

Re: [S-mailx] .nailrc and Gmail

[Predrag Punosevac]

Predrag Punosevac  wrote:

> Hi Steffen,
> 
> I apologize for a direct email. I am not sure new mailing list will
> allow me to post. 
> 
> I kept my own v14.8.12 copy of nail after few things got broken on
> OpenBSD.  I didn't notice that you adopted OpenBSD port and updated it
> to 14.9 branch until today. Old port was behind the version I was using.
> 


Hi Steffen,

I apologize for cross posting. After upgrading my laptop to 

predrag@oko-mobile$ uname -a
OpenBSD oko-mobile.int.bagdala2.net 6.7 GENERIC.MP#2 amd64

I felt it was the time for me to jump the ship and finally go with
s-nail from the official ports tree. 

predrag@oko-mobile$ pkg_info s-nail
Information for inst:s-nail-14.9.17

I got to the bottom of all "issues" I originally reported 

https://www.mail-archive.com/s-mailx@lists.sdaoden.eu/msg00948.html

in the thread. I used quotation marks around issues as in the hindsight
there was really only one. All other issues were due to the fact that I
didn't realize that you have completely rewritten s-nail and there is
really not much in common with the original Heirloom mailx

http://heirloom.sourceforge.net/mailx.html

which I used for at least 15 years. That is not to say that the things
don't work or they are worse. They just work different. It took me two
full days of dicking with it to get a to get a hang of it.

First thing first you really trough me off the board with Ctrl+D instead
of next line and a dot to sent the email. I have not read the code, and
even if I did I don't have sufficient programming background to
understand design decession but I am using dot to sent emails since
circa 1989 and that is a hard pill to swallow. That is why I kept
reporting that sending email doesn't work. 

I noticed that 14.9.17 on 6.7 doesn't report that annoying message

There are new messages in the error message ring (denoted by ERROR),
nail:   which can be managed with the `errors' command
ERROR# ? 

I really like new configuration grammar. This is my not so minimal
working example

predrag@oko-mobile$ more .mailrc 
set ask
set crt
ignore message-id received date fcc status resent-date resent-message-id
resent-from in-reply-to

set mailx-extra-rc=~/.nailrc

and this is dotnailrc file

account gmail {
set inbox=imaps://usern...@imap.gmail.com
set imap-use-starttls
set password="secret"
set folder=imaps://usern...@imap.gmail.com record="+[Gmail]/Sent Mail"
set from="Predrag Punosevac " \
mta=smtp://usern...@smtp.gmail.com:587 \
set smtp-use-starttls 
set smtp-auth="login" 
# IMAP SHORTCUTS SECTION for standard Gmail folders
shortcut allmail +[Gmail]/All\ Mail
shortcut sent +[Gmail]/Sent\ Mail
shortcut spam +[Gmail]/Spam
shortcut trash +[Gmail]/Trash
}
account cmu {
set inbox=imaps://username%40andrew.cmu....@imap.gmail.com
set imap-use-starttls
set password="secret"
set from="Predrag Punosevac " \
mta=smtp://username%40andrew.cmu@smtp.gmail.com:587 \
set smtp-use-starttls 
set smtp-auth="login" 
# IMAP SHORTCUTS SECTION for standard Gmail folders
shortcut allmail +[Gmail]/All\ Mail
shortcut sent +[Gmail]/Sent\ Mail
shortcut spam +[Gmail]/Spam
shortcut trash +[Gmail]/Trash
}
account hotmail {
set inbox=imaps://username%40hotmail@imap-mail.outlook.com
set folder=imaps://username%40hotmail@imap-mail.outlook.com
set imap-use-starttls
set password="secret"
set from="Predrag Punosevac " \
mta=smtp://username%40hotmail@smtp-mail.outlook.com:587
set smtp-use-starttls 
set smtp-auth="login"
}


# Binary options
set askattach
set askcc
set autocollapse
set autosort=thread
set junkdb=~/.junkdb
set v15-compat
set tls-verify=strict

# String Options
set imap-keepalive=240
set imap-list-depth=5

# Reading HTML mail
set pipe-text/html="lynx -dump -force_html -stdin"

# Address Book
alias   somebody   someb...@gmail.com


For people who would be reading this email one comment. Without

set folder 

line you will not be able (not in a convenient way) to list your IMAP
folders. I feel new grammar is nicer than the old. I have been using
s-nail now for three days. I really, really like what you have done with
search options. The only time in my life I contemplated switching to nmh

https://www.nongnu.org/nmh/

was for its integration with standard shell tools (of course MH format
is really a big plus).

I see that you are planning to remove IMAP support and that would be a
tough thing for me to swallow.

I also noticed that you put a lot of time into documentation. Most of
the stuff is there but it feels a bit disorganized for my taste. I 

Re: Troubles with Bro

[Predrag Punosevac]

"dan (ddp)"  wrote:

> On Wed, Oct 9, 2019 at 2:06 PM Predrag Punosevac  
> wrote:
> >
> > Hi Ports,
> >
> > I am fully aware that this is very late in a release cycle so hopefully
> > this works as expected on 6.6 which I didn't test
> >
> > iris# uname -a
> > OpenBSD iris.int.autonsys.com 6.5 GENERIC.MP#5 amd64
> > iris# syspatch -l
> > 001_rip6cksum
> > 002_srtp
> > 003_mds
> > 004_bgpd
> > 005_libssl
> > 006_tcpsack
> > 007_smtpd
> > 008_swapgs
> > 009_resume
> > 010_frag6ecn
> > 011_expat
> > 012_sysupgrade
> > 013_unbound
> > 014_dhcpd
> >
> > iris# /usr/local/bin/broctl deploy
> > checking configurations ...
> > bro scripts failed.
> > error in /usr/local/share/bro/base/protocols/dce-rpc/./main.bro, line
> > 51: "redef" used but not previously defined (DPD::ignore_violations)
> >
> >
> > Notice that I only change the name of the interface in
> > /etc/bro/node.cfg per /usr/local/share/doc/pkg-readmes/bro
> >
> > iris# /usr/local/bin/broctl start
> > Warning: broctl node config has changed (run the broctl "deploy"
> > command)
> > starting bro ...
> > Error: bro terminated immediately after starting; check output with
> > "diag"
> >
> > iris# /usr/local/bin/broctl status
> > Warning: broctl node config has changed (run the broctl "deploy"
> > command)
> > Name Type   Host  StatusPidStarted
> > bro  standalone localhost stopped
> >
> >
> >
> > iris# /usr/local/bin/broctl diag
> > Warning: broctl node config has changed (run the broctl "deploy"
> > command)
> > [bro]
> >
> > No core file found and egdb is not installed.  It is recommended to
> > install egdb so that BroControl can output a backtrace if Bro crashes.
> >
> > Bro 2.6.4
> > OpenBSD 6.5
> >
> > Bro plugins: (none found)
> >
> >  No reporter.log
> >
> >  stderr.log
> > error in /usr/local/share/bro/base/protocols/dce-rpc/./main.bro, line
> > 51: "redef" used but not previously defined (DPD::ignore_violations)
> >
> >  stdout.log
> > max memory size (kbytes, -m) unlimited
> > data seg size   (kbytes, -d) 33554432
> > core file size  (blocks, -c) unlimited
> >
> >  .cmdline
> > -i em0 -U .status -p broctl -p broctl-live -p standalone -p local -p bro
> > local.bro broctl broctl/standalone broctl/auto
> >
> >  .env_vars
> > PATH=/usr/local/bin:/usr/local/share/broctl/scripts:/sbin:/usr/sbin:/bin:/usr/bin:/usr/X11R6/bin:/usr/local/sbin:/usr/local/bin
> > BROPATH=/var/spool/bro/installed-scripts-do-not-touch/site::/var/spool/bro/installed-scripts-do-not-touch/auto:/usr/local/share/bro:/usr/local/share/bro/policy:/usr/local/share/bro/site
> > CLUSTER_NODE=
> >
> >  .status
> > TERMINATED [atexit]
> >
> >  No prof.log
> >
> >  No packet_filter.log
> >
> >  No loaded_scripts.log
> >
> >
> >
> > Also
> >
> > Cron  /usr/local/bin/broctl cron
> >
> > Error: cannot start bro; check output of "diag"
> > Error: env: bash: No such file or directory
> >
> > starting bro ...
> >
> > This used to work as expected in 6.4 I shut down my installation as I
> > didn't need at that time but I now needed badly. I will upgrade to 6.6
> > as soon as it is released.
> >
> 
> I've had similar errors after upgrading the package. I usually have to
> uninstall and re-install it.
> Copying the configuration out of etc and replacing it after the
> re-install has worked fine for me.
> I haven't had time to look into the issue at all, so I don't have any
> clues as to why it actually happens.
> 

Sure enough uninstalling and re-installing fixed for me as well. Thank
you so much for quick heads-up.

Predrag


> >
> > Best,
> > Predrag
> >
> >

Troubles with Bro

[Predrag Punosevac]

Hi Ports,

I am fully aware that this is very late in a release cycle so hopefully
this works as expected on 6.6 which I didn't test

iris# uname -a 
OpenBSD iris.int.autonsys.com 6.5 GENERIC.MP#5 amd64
iris# syspatch -l  
001_rip6cksum
002_srtp
003_mds
004_bgpd
005_libssl
006_tcpsack
007_smtpd
008_swapgs
009_resume
010_frag6ecn
011_expat
012_sysupgrade
013_unbound
014_dhcpd

iris# /usr/local/bin/broctl deploy 
checking configurations ...
bro scripts failed.
error in /usr/local/share/bro/base/protocols/dce-rpc/./main.bro, line
51: "redef" used but not previously defined (DPD::ignore_violations)


Notice that I only change the name of the interface in 
/etc/bro/node.cfg per /usr/local/share/doc/pkg-readmes/bro

iris# /usr/local/bin/broctl start  
Warning: broctl node config has changed (run the broctl "deploy"
command)
starting bro ...
Error: bro terminated immediately after starting; check output with
"diag"

iris# /usr/local/bin/broctl status
Warning: broctl node config has changed (run the broctl "deploy"
command)
Name Type   Host  StatusPidStarted
bro  standalone localhost stopped



iris# /usr/local/bin/broctl diag  
Warning: broctl node config has changed (run the broctl "deploy"
command)
[bro]

No core file found and egdb is not installed.  It is recommended to
install egdb so that BroControl can output a backtrace if Bro crashes.

Bro 2.6.4
OpenBSD 6.5

Bro plugins: (none found)

 No reporter.log

 stderr.log
error in /usr/local/share/bro/base/protocols/dce-rpc/./main.bro, line
51: "redef" used but not previously defined (DPD::ignore_violations)

 stdout.log
max memory size (kbytes, -m) unlimited
data seg size   (kbytes, -d) 33554432
core file size  (blocks, -c) unlimited

 .cmdline
-i em0 -U .status -p broctl -p broctl-live -p standalone -p local -p bro
local.bro broctl broctl/standalone broctl/auto

 .env_vars
PATH=/usr/local/bin:/usr/local/share/broctl/scripts:/sbin:/usr/sbin:/bin:/usr/bin:/usr/X11R6/bin:/usr/local/sbin:/usr/local/bin
BROPATH=/var/spool/bro/installed-scripts-do-not-touch/site::/var/spool/bro/installed-scripts-do-not-touch/auto:/usr/local/share/bro:/usr/local/share/bro/policy:/usr/local/share/bro/site
CLUSTER_NODE=

 .status
TERMINATED [atexit]

 No prof.log

 No packet_filter.log

 No loaded_scripts.log



Also 

Cron  /usr/local/bin/broctl cron

Error: cannot start bro; check output of "diag"
Error: env: bash: No such file or directory

starting bro ...

This used to work as expected in 6.4 I shut down my installation as I
didn't need at that time but I now needed badly. I will upgrade to 6.6
as soon as it is released. 


Best,
Predrag

acme-tiny: Agreement URL does not match the current one

[Predrag Punosevac]

I am trying to migrate few web sites to https using free certificates
from Let's Encrypt. Since I am not crazy about installing official
certbot client I went for acme-tiny. It seems to be broken

# uname -a
OpenBSD horae.autonlab.org 6.0 GENERIC.MP#2 amd64

ValueError: Error registering: 400 {
  "type": "urn:acme:error:malformed",
  "detail": "Provided agreement URL
[https://letsencrypt.org/documents/LE-SA-v1.0.1-July-27-2015.pdf] does
not match current agreement URL
[https://letsencrypt.org/documents/LE-SA-v1.1.1-August-1-2016.pdf]";,
  "status": 400
}

This seems well known bug

http://lists.alioth.debian.org/pipermail/letsencrypt-devel/Week-of-Mon-20160808/000784.html

Best,
Predrag

Re: Inkscape core dump

[Predrag Punosevac]

Rafael Sadowski  wrote:

> On Sun May 29, 2016 at 10:21:07PM -0400, Predrag Punosevac wrote:
> > Running amd54 5.9 stable (upgraded from 5.8)
> > 
> > predrag@oko$ uname -a
> > OpenBSD oko.bagdala2.net 5.9 GENERIC.MP#4 amd64
> > 
> > predrag@oko$ pkg_info inkscape
> > Information for inst:inkscape-0.91p7
> > 
> > Comment:
> > SVG vector drawing application
> > 
> > Description:
> > Inkscape is a vector graphics editor, with capabilities similar
> > to Illustrator, CorelDraw, or Xara X, using the W3C standard Scalable
> > Vector Graphics (SVG) file format.
> > 
> > Maintainer: The OpenBSD ports mailing-list 
> > 
> > WWW: http://www.inkscape.org/
> > 
> > 
> > Trying to export SVG image as eps or ps will cause InkScape to crash
> 
> Thanks for the report. Could you try Erling's advice, if it's not work
> for you I'll debug inkscape.
> 

The same result. I should have said in my first e-mail that "export PNG"
works flawlessly. It looks like Inkscape is using ImageMagick as a
backend for image format conversions. Personally I always use
GraphicsMagick so for all I know it could be ImageMagick bug.


> > 
> > predrag@oko$ egdb
> > GNU gdb (GDB) 7.10.1
> > Copyright (C) 2015 Free Software Foundation, Inc.
> > License GPLv3+: GNU GPL version 3 or later
> > <http://gnu.org/licenses/gpl.html>
> > This is free software: you are free to change and redistribute it.
> > There is NO WARRANTY, to the extent permitted by law.  Type "show
> > copying"
> > and "show warranty" for details.
> > This GDB was configured as "x86_64-unknown-openbsd5.9".
> > Type "show configuration" for configuration details.
> > For bug reporting instructions, please see:
> > <http://www.gnu.org/software/gdb/bugs/>.
> > Find the GDB manual and other documentation resources online at:
> > <http://www.gnu.org/software/gdb/documentation/>.
> > For help, type "help".
> > Type "apropos word" to search for commands related to "word".
> > (gdb) core-file inkscape.core
> > [New process 9413]
> > [New process 20111]
> > [New process 8182]
> > [New process 21072]
> > Core was generated by `inkscape'.
> > Program terminated with signal SIGABRT, Aborted.
> > #0  0x0b579bd0887a in ?? ()
> > [Current thread is 1 (process 9413)]
> > 
> 
> coredumps without debug symbols are not helpful but your crash
> description is very helpful.
> 

I should have realized that the package was built without debug option.

Predrag

> Best regards,
> 
> Rafael

Inkscape core dump

[Predrag Punosevac]

Running amd54 5.9 stable (upgraded from 5.8)

predrag@oko$ uname -a
OpenBSD oko.bagdala2.net 5.9 GENERIC.MP#4 amd64

predrag@oko$ pkg_info inkscape
Information for inst:inkscape-0.91p7

Comment:
SVG vector drawing application

Description:
Inkscape is a vector graphics editor, with capabilities similar
to Illustrator, CorelDraw, or Xara X, using the W3C standard Scalable
Vector Graphics (SVG) file format.

Maintainer: The OpenBSD ports mailing-list 

WWW: http://www.inkscape.org/


Trying to export SVG image as eps or ps will cause InkScape to crash

predrag@oko$ egdb
GNU gdb (GDB) 7.10.1
Copyright (C) 2015 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later

This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show
copying"
and "show warranty" for details.
This GDB was configured as "x86_64-unknown-openbsd5.9".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
.
Find the GDB manual and other documentation resources online at:
.
For help, type "help".
Type "apropos word" to search for commands related to "word".
(gdb) core-file inkscape.core
[New process 9413]
[New process 20111]
[New process 8182]
[New process 21072]
Core was generated by `inkscape'.
Program terminated with signal SIGABRT, Aborted.
#0  0x0b579bd0887a in ?? ()
[Current thread is 1 (process 9413)]


Thanks!
Predrag

Re: rrdtool-chroot pango-basic-fc.so

[Predrag Punosevac]

Stuart Henderson  wrote:

> Ah, it seems that pango got simplified again. What files did you end up
> needing to copy into the chroot? Would you be able to try it with this diff?
> 

I think you are right. I ended up doing the following 2 lines from the
script

cp -p /usr/local/bin/rrdtool /var/www/usr/local/bin
for i in $(ldd /usr/local/bin/rrdtool | grep 'lib/' | awk '{
print $7 }'
) ; do \
cp -p $i /var/www/usr/lib/
done

to get a semi-functional syweb instance at home. I am too busy until
next weekend to look this again. 


Predrag

> Index: files/rrdtool-chroot.sh
> ===
> RCS file: /cvs/ports/net/rrdtool/files/rrdtool-chroot.sh,v
> retrieving revision 1.2
> diff -u -p -r1.2 rrdtool-chroot.sh
> --- files/rrdtool-chroot.sh   6 Apr 2014 17:01:34 -   1.2
> +++ files/rrdtool-chroot.sh   24 Apr 2016 01:03:36 -
> @@ -1,7 +1,7 @@
>  #!/bin/ksh
>  #
>  # make rrdtool(1) available in webserver chroot(8): copy in library
> -# dependencies and generate barebones pango/fontconfig/pango configuration
> +# dependencies and generate barebones fontconfig configuration
>  
>  # WARNING: /var/www/usr/libexec/ld.so and /var/www/lib/* may be shared
>  # by other executables in the chroot(8)
> @@ -12,12 +12,11 @@ if [ "$(id -u)" -ne 0 ]; then
>  fi
>  
>  do_enable() {
> - mkdir -p 
> /var/www{/etc/{pango,fonts/TTF},/usr/lib,/usr/libexec,${LOCALBASE}/bin,${X11BASE}/lib,/cache/fontconfig}
> + mkdir -p 
> /var/www{/etc/fonts/TTF,/usr/lib,/usr/libexec,${LOCALBASE}/bin,${X11BASE}/lib,/cache/fontconfig}
>   cp -p ${LOCALBASE}/bin/rrdtool /var/www${LOCALBASE}/bin
>   for i in $(ldd ${LOCALBASE}/bin/rrdtool | grep 'lib/' | awk '{ print $7 
> }') ; do \
>   cp -p $i /var/www/usr/lib/
>   done
> - cp -p ${LOCALBASE}/lib/pango/*/modules/pango-basic-fc.so 
> /var/www/usr/lib/
>   cp -p /usr/libexec/ld.so /var/www/usr/libexec/
>   cp -p /usr/X11R6/lib/X11/fonts/TTF/DejaVuSans.ttf /var/www/etc/fonts/TTF
>   cat << EOF > /var/www/etc/fonts/fonts.conf
> @@ -28,21 +27,16 @@ do_enable() {
>   /cache/fontconfig
>  
>  EOF
> - cat << EOF > /var/www/etc/pango/pango.modules
> -/usr/lib/pango-basic-fc.so BasicScriptEngineFc PangoEngineShape 
> PangoRenderFc common:
> -EOF
>  }
>  
>  do_disable() {
>   rm -rf /var/www/usr/local/bin/rrdtool \
>   /var/www/usr/libexec/ld.so \
>   /var/www/etc/fonts \
> - /var/www/etc/pango \
>   /var/www/cache/fontconfig
>   for i in $(ldd ${LOCALBASE}/bin/rrdtool | grep 'lib/' | awk '{ print $7 
> }') ; do \
>   rm -f /var/www/usr/lib/$(basename $i | sed -e 's,\.so.*,,')*
>   done
> - rm -f /var/www/usr/lib/pango-basic-fc.so
>   rmdir 
> /var/www{/etc/fonts,${LOCALBASE}/bin,${LOCALBASE},/usr/lib,/usr/libexec,/usr} 
> 2>/dev/null
>  }
>  

Re: rrdtool-chroot pango-basic-fc.so

[Predrag Punosevac]

Predrag Punosevac wrote:

> I must have done something stupid during the upgrade from 5.8 to 5.9
> (amd64)
> 
> # /usr/local/share/examples/rrdtool/rrdtool-chroot enable 
> cp: /usr/local/lib/pango/*/modules/pango-basic-fc.so: No such file or
> directory
> 
> Directory /usr/local/lib/pango is indeed missing. 
> 
> # pkg_info pango
> Information for inst:pango-1.38.1
> 
> However I don't see that the package contains /usr/lib/pongo/1.8/modules
> anymore (I downloaded and extracted package manually).
> 
> To add insult to injury it seems that rrdtool-chroot is not behaving
> correctly on my machine. After commenting out the line in rrdtool-chroot
> enable which is suppose to copy pango-basic-fc.so I noticed that rrdtool
> binaries were not coppied to /var/www/bin. 
> 
> I did that manually and now I can see nice graphs in the syweb of my
> home network. The only thing I noticed about symon/symux is that it has
> problems with sensor(cpu0.temp0) data
> 
> # /usr/local/share/examples/symon/c_smrrds.sh all  
> /var/www/symon/rrds/localhost/io_cd0.rrd exists - ignoring
> /var/www/symon/rrds/localhost/io_sd2.rrd exists - ignoring
> /var/www/symon/rrds/localhost/io_sd1.rrd exists - ignoring
> /var/www/symon/rrds/localhost/io_sd0.rrd exists - ignoring
> /var/www/symon/rrds/localhost/if_em0.rrd exists - ignoring
> /var/www/symon/rrds/localhost/proc_httpd.rrd exists - ignoring
> ERROR: parsing min val - Cannot convert '-U' to float
> could not create /var/www/symon/rrds/localhost/sensor_cpu0.temp0.rrd
> /var/www/symon/rrds/localhost/mbuf.rrd exists - ignoring
> /var/www/symon/rrds/localhost/pf.rrd exists - ignoring
> /var/www/symon/rrds/localhost/if_lo0.rrd exists - ignoring
> /var/www/symon/rrds/localhost/mem.rrd exists - ignoring
> /var/www/symon/rrds/localhost/cpu0.rrd exists - ignoring
> 
> 
> Also on syweb it seems that the following layouts are broken pf and
> total_firewall (I don't get graphs but it could be very well related to
> above problem with rrdtool running in chroot).
> 
> Any ideas?
> 
> Predrag

rrdtool-chroot script is broken for me on the freshly installed 5.9.
with the same error

rrdtool-chroot pango-basic-fc.so

[Predrag Punosevac]

I must have done something stupid during the upgrade from 5.8 to 5.9
(amd64)

# /usr/local/share/examples/rrdtool/rrdtool-chroot enable 
cp: /usr/local/lib/pango/*/modules/pango-basic-fc.so: No such file or
directory

Directory /usr/local/lib/pango is indeed missing. 

# pkg_info pango
Information for inst:pango-1.38.1

However I don't see that the package contains /usr/lib/pongo/1.8/modules
anymore (I downloaded and extracted package manually).

To add insult to injury it seems that rrdtool-chroot is not behaving
correctly on my machine. After commenting out the line in rrdtool-chroot
enable which is suppose to copy pango-basic-fc.so I noticed that rrdtool
binaries were not coppied to /var/www/bin. 

I did that manually and now I can see nice graphs in the syweb of my
home network. The only thing I noticed about symon/symux is that it has
problems with sensor(cpu0.temp0) data

# /usr/local/share/examples/symon/c_smrrds.sh all  
/var/www/symon/rrds/localhost/io_cd0.rrd exists - ignoring
/var/www/symon/rrds/localhost/io_sd2.rrd exists - ignoring
/var/www/symon/rrds/localhost/io_sd1.rrd exists - ignoring
/var/www/symon/rrds/localhost/io_sd0.rrd exists - ignoring
/var/www/symon/rrds/localhost/if_em0.rrd exists - ignoring
/var/www/symon/rrds/localhost/proc_httpd.rrd exists - ignoring
ERROR: parsing min val - Cannot convert '-U' to float
could not create /var/www/symon/rrds/localhost/sensor_cpu0.temp0.rrd
/var/www/symon/rrds/localhost/mbuf.rrd exists - ignoring
/var/www/symon/rrds/localhost/pf.rrd exists - ignoring
/var/www/symon/rrds/localhost/if_lo0.rrd exists - ignoring
/var/www/symon/rrds/localhost/mem.rrd exists - ignoring
/var/www/symon/rrds/localhost/cpu0.rrd exists - ignoring


Also on syweb it seems that the following layouts are broken pf and
total_firewall (I don't get graphs but it could be very well related to
above problem with rrdtool running in chroot).

Any ideas?

Predrag

syslog-ng 3.7

[Predrag Punosevac]

Dear Mr. Mestdagh,

What is the reason syslog-ng is not updated to 3.7.1? Is branch 3.7
still considered unstable for use on OpenBSD or there are some other
problems in updating the existing 3.6.4 port.

Thank you.
Predrag Punosevac

vifm

[Predrag Punosevac]

I noticed that vifm has been removed from the OpenBSD port three prior
to 5.8 release. Are there technical reasons for such decision (I
remember pains to initial import vifm into the port three due to the
code quality)? It looks like vifm got many new features which might have
triggered the decision. 

I just tried ranger instead of vifm but it seems overkill for an
occasional user like myself (I found vifm sometimes useful with ssh/tmux
sessions).

Best,
Predrag

Re: easy-rsa LibreSSL issues

[Predrag Punosevac]

Theo de Raadt wrote:

> Maybe you can talk to the authors nicely and see if they can find a
> better way...

I did. I think they took it pretty well. Please see below. However 
quick browse through

https://github.com/OpenVPN/easy-rsa/issues

revels that easy-rsa has never been tested upstream with LibreSSL.


Predrag




 Original Message 
Date: Wed, 28 Oct 2015 04:59:11 -0700
From: Eric Crist 
To: OpenVPN/easy-rsa 
Subject: Re: [easy-rsa] LibreSSL exposes misuse of $ENV  (#76)


I read through the entire thread and it would be nice for us to support both 
LibreSSL and PolarSSL in the future.  This is likely a 3.2 fix.

---
Reply to this email directly or view it on GitHub:
https://github.com/OpenVPN/easy-rsa/issues/76#issuecomment-151820651

easy-rsa LibreSSL issues

[Predrag Punosevac]

I just spent 30 minutes playing with easy-rsa which is shipped broken on
5.8 until I realized what was going on. I see that sthen has already
reverted easy-rsa to OpenSSL run dependency per comment

switch easy-rsa to using openssl to unbreak; libressl doesn't allow
$ENV:: in config files and easy-asa uses this heavily.

Moving forward should I even bother with easy-rsa and just use vanilla
libressl to generate certificates? What is the recommendation for this
port in the light of libressl "incompatibilities".

Predrag

Re: matplotlib Gtk3

[Predrag Punosevac]

Antoine Jacoutot  wrote:

> > There are actually a few other odd things with this port dependencies.
> > I'll have a look.
> > Thanks.
> 
> This should do it.
> For python2 we depend on gtk+2.
> For python3 we depend on gtk+3.
> The toolkit dependencies are somewhat optional (like py-qt4); and we don't 
> want to depend on all possible options for obvious reasons.
> So if you want python2 with gtk+3 then you need to manually install the 
> dependencies (in your case, py-gobject3 and gtk+3).
> 
Thanks Antoine! Manually installing py-gobject3 did the trick for me. I
already had gtk+3 on my system. I just got an e-mail from David Coppa
and it looks like he could not reproduce my problem. I would sware that
I have pristine non-broken python27 environment where I used pip install
only for youtube-dl but I might have done something stupid. It would not
have being the first time. 

> OK?
> 
> 
> 
> Index: Makefile
> ===
> RCS file: /cvs/ports/graphics/py-matplotlib/Makefile,v
> retrieving revision 1.54
> diff -u -p -u -p -r1.54 Makefile
> --- Makefile  4 May 2015 22:03:20 -   1.54
> +++ Makefile  1 Jul 2015 07:18:16 -
> @@ -3,7 +3,7 @@
>  COMMENT= Python charting and plotting API
>  
>  MODPY_EGG_VERSION =  1.4.2
> -REVISION =   2
> +REVISION =   3
>  DISTNAME =   matplotlib-${MODPY_EGG_VERSION}
>  PKGNAME =py-${DISTNAME}
>  CATEGORIES = graphics devel math
> @@ -15,15 +15,12 @@ MAINTAINER =  Daniel Dickman   # BSD-like
>  PERMIT_PACKAGE_CDROM=Yes
>  
> -WANTLIB += X11 Xau Xcomposite Xcursor Xdamage Xdmcp Xext Xfixes
> -WANTLIB += Xi Xinerama Xrandr Xrender ffi fontconfig freetype 
> -WANTLIB += m pixman-1 png pthread pthread-stubs ${MODPY_WANTLIB}
> -WANTLIB += stdc++ tcl85 tk85 xcb xcb-render xcb-shm z
> +WANTLIB += ${MODPY_WANTLIB} ${MODTK_WANTLIB}
> +WANTLIB += freetype m png stdc++ z
>  
>  MASTER_SITES=${MASTER_SITE_SOURCEFORGE:=matplotlib/}
>  
>  MODULES= lang/python \
> - devel/gettext \
>   x11/tk
>  
>  FLAVORS= python3
> @@ -31,7 +28,7 @@ FLAVOR?=
>  
>  MODPY_SETUPTOOLS = Yes
>  
> -RUN_DEPENDS= devel/py-tz${MODPY_FLAVOR} \
> +COMMON_DEPENDS=  devel/py-tz${MODPY_FLAVOR} \
>   devel/py-dateutil${MODPY_FLAVOR} \
>   devel/py-parsing${MODPY_FLAVOR} \
>   devel/py-six${MODPY_FLAVOR} \
> @@ -39,16 +36,19 @@ RUN_DEPENDS=  devel/py-tz${MODPY_FLAVOR} 
>   graphics/py-Pillow${MODPY_FLAVOR} \
>   ${MODPY_TKINTER_DEPENDS}
>  
> -.if ! ${FLAVOR:Mpython3}
> -WANTLIB += atk-1.0 cairo gdk-x11-2.0 gdk_pixbuf-2.0 gio-2.0
> -WANTLIB += glib-2.0 gmodule-2.0 gobject-2.0 gtk-x11-2.0 harfbuzz
> -WANTLIB += pango-1.0 pangocairo-1.0 pangoft2-1.0 pcre
> -LIB_DEPENDS += x11/gtk+2
> -RUN_DEPENDS += x11/py-gtk2
> +BUILD_DEPENDS=   ${COMMON_DEPENDS}
> +RUN_DEPENDS= ${COMMON_DEPENDS}
> +
> +.if ${FLAVOR:Mpython3}
> +RUN_DEPENDS +=   devel/py-gobject3${MODPY_FLAVOR}
> +# pygobject GTK3 introspection (not provided by indirect dependencies)
> +RUN_DEPENDS +=   x11/gtk+3
> +.else
> +RUN_DEPENDS +=   x11/py-gtk2
>  .endif
>  
> -BUILD_DEPENDS=   ${RUN_DEPENDS}
> -LIB_DEPENDS+=graphics/png
> +LIB_DEPENDS= ${MODTK_LIB_DEPENDS} \
> + graphics/png
>  TEST_DEPENDS =   ${RUN_DEPENDS} \
>   devel/py-nose${MODPY_FLAVOR}
>  
> 
> -- 
> Antoine

matplotlib Gtk3

[Predrag Punosevac]

I am having issue with matplotlib on 5.7 release. Trying to 

import matplotlib.pyplot as plt

gives 

 11 except ImportError:
---> 12 raise ImportError("Gtk3 backend requires pygobject to be 
installed.")
 13 
 14 try:

ImportError: Gtk3 backend requires pygobject to be installed.


Can anyone confirm this?

Predrag

Re: php-fpm-5.6 startup error

[Predrag Punosevac]

On 27 March 2015 at 09:48, frantisek holop  wrote:
on from the older version syntax:
>
> ; Unix user/group of processes
> ; Note: The user is mandatory. If the group is not set, the default user's 
> group
> ;   will be used.
> user = www
> group = www
>
> to the new php-fpm-5.6:
>
> ; Unix user/group of processes
> ; Note: The user is mandatory. If the group is not set, the default user's 
> group
> ;   will be used.
> listen.owner = www
> listen.group = www
> listen.mode = 0660
>
>
> could be a bit smoother...
> i personally dont find the new version more readable...
>
> $ sudo /etc/rc.d/php_fpm start
> php_fpm(failed)
> $ sudo /usr/local/sbin/php-fpm-5.6
> [26-Mar-2015 23:37:41] ALERT: [pool www] user has not been defined
> [26-Mar-2015 23:37:41] ERROR: failed to post process the configuration
> [26-Mar-2015 23:37:41] ERROR: FPM initialization failed
>
> adding back
>
> user = www
>
> for the pool seems to fix this.
>
> -f
> --
> i am not a dictator.  it's just i have a grumpy face.

Just to leave electronic trail for people who will be using 5.7 release.
This is broken in 5.7 release please add

user = www

Cheers,
Predrag

Re: Brother DCP-7030 printer - blank pages

[Predrag Punosevac]

hans wrote:
> I am trying to get the DCP-7030 to print on current/amd64.
> It is connected via USB, connects as (full dmesg below):
> 
> ulpt0 at uhub4 port 2 configuration 1 interface 0 "Brother Industries
> product 0x01ea" \
> rev 2.00/1.00 addr 2 ulpt0: using bi-directional mode ugen1 at uhub4
> port 2 \
> configuration 1 "Brother Industries product 0x01ea" rev 2.00/1.00 addr 2
> 
> I run a simple "lpd -ls" form rc.conf.local,
> and have some other printers where I print without problems.
> 
> $ cat /etc/printcap
> lp:lp=:rm=147.32.232.58:rp=lp:sd=/var/spool/output/lp:sf:sh
> mu:lp=:rm=147.231.88.24:rp=lp:sd=/var/spool/output/mu:sf:sh
> ws:\
>   :lp=/dev/ulpt0:\
>   :af=/etc/foomatic/brotherDCP7030.ppd:\
>  
>   :if=/usr/local/bin/foomatic-rip:\
>   :sd=/var/spool/output/ws:\
>   :sh:sf

foomatic-rip doesn't support LPD anymore. Have you read 

/usr/local/share/doc/pkg-readmes/cups-filters*


You can also try to follow this howto which I wrote (there should be a
copy on ports archive) 

http://daemonforums.org/showthread.php?t=8710

but I would strongly suggest you read documentation Antoine wrote.

Cheers,
Predrag

Re: Best practice for CUPS-less foomatic use?

[Predrag Punosevac]

I know that many people were very frustrated when upstream broke
foomatic-rip for LPD users. Thanks to Antoine Jacoutot many of us will
be CUPS free for years to come. I am leaving internet trace for people
who would be looking for the info but Antoine documentation

/usr/local/share/doc/pkg-readmes/cups-filters-1.0.54p2

is golden standard.

In the nut shell what I did:

0. Install foomatic-rip, drivers (in my case hpijs), a2ps, ImageMagick

1. Edit /etc/printcap as
lp|HP|HP-PhotoSmart_C5200:\
:lp=/dev/ulpt0:\
:if=/etc/foomatic-rip/script.sh:\
:sh:sd=/var/spool/output:\
:lf=/var/log/lpd-errs:

2. Create /etc/foomatic-rip.script.sh and make it executable

#!/bin/sh

/usr/local/bin/a2ps -BRq --columns=1 -o - | \
/usr/local/bin/foomatic-rip -P HP-PhotoSmart_C5200 --ppd \
/etc/foomatic-rip/direct/hp-photosmart_c5200_series-hpijs.ppd

3. Add hp-photosmart_c5200_series-hpijs.ppd to /etc/foomatic-rip/direct

4. Adjust permission on ulpt0 and ugen devices. 

5. Start lpd daemon and be happy camper.

Cheers,
Predrag

Re: virt-manager GtkVnc Error

[Predrag Punosevac]

Antoine Jacoutot  wrote:

> On Thu, Oct 16, 2014 at 11:28:20PM -0400, Predrag Punosevac wrote:
> > I am trying to use virt-manager on a snapshot both OS and ports from
> > 10th of August (should be very close to November 1st release).
> > 
> > predrag@oko$ uname -a
> > OpenBSD oko.bagdala2.net 5.6 GENERIC.MP#333 amd64
> > 
> > virt-manager compiles and runs fine. I can see the status of all KVM
> > hosts on the remote location (btw I am using openconnect to connect to
> > that location and it works like a charm). However trying to open vnc for
> > a particular machine gives error 
> > 
> > 
> > Error launching details: cannot import name GtkVnc
> 
> $ pkglocate GtkVnc 
> gtk-vnc-0.5.3p2:x11/gtk-vnc:/usr/local/lib/girepository-1.0/GtkVnc-2.0.typelib
> gtk-vnc-0.5.3p2:x11/gtk-vnc:/usr/local/share/gir-1.0/GtkVnc-2.0.gir
> 
> Did you try installing that?

It was missing. I would think that thouse packages should have been
pulled as a running dependency for virt-manager. No?

Predrag
> 
> -- 
> Antoine

virt-manager GtkVnc Error

[Predrag Punosevac]

I am trying to use virt-manager on a snapshot both OS and ports from
10th of August (should be very close to November 1st release).

predrag@oko$ uname -a
OpenBSD oko.bagdala2.net 5.6 GENERIC.MP#333 amd64

virt-manager compiles and runs fine. I can see the status of all KVM
hosts on the remote location (btw I am using openconnect to connect to
that location and it works like a charm). However trying to open vnc for
a particular machine gives error 


Error launching details: cannot import name GtkVnc

Traceback (most recent call last):
  File "/usr/local/share/virt-manager/virtManager/engine.py", line 803,
in _show_vm_helper
details = self._get_details_dialog(uri, uuid)
  File "/usr/local/share/virt-manager/virtManager/engine.py", line 760,
in _get_details_dialog
obj = vmmDetails(con.get_vm(uuid))
  File "/usr/local/share/virt-manager/virtManager/details.py", line 488,
in __init__
from virtManager.console import vmmConsolePages
  File "/usr/local/share/virt-manager/virtManager/console.py", line 27,
in 
from gi.repository import GtkVnc


Since this is a new port has anybody managed Vnc to work.

Most Kind Regards,
Predrag Punosevac

P.S. I also notice that memory, disk, and network can't be displayed.
CPU usage both guest and host works fine.

rsync 3.1.1 bug warning

[Predrag Punosevac]

I am under impression that OpenBSD will be shipping rsync version 3.1.1.
While backing up one of OpenBSD 5.5 servers rsync 3.1.0 to FreeBSD 10.0
rsync 3.1.1 with rsnapshot I bumped into a bug which causes rsync to
terminate communication. One way to "fix" it is NOT to use option -z.

It seems that bug doesn't affect communication with earlier versions of
rsync as I had no problem backing up bunch of Red Hat desktops to the
same FreeBSD storage server.

I hope it will be of some help to people.

Predrag

Re: RRD.so does not exist

[Predrag Punosevac]

Stuart Henderson  wrote:

> Ah, I'm sorry, I didn't spot that you were sticking to "close to 5.6".
> 
> > > I am just trying to build a single port using a port
> > > snapshot and the sytem from 10 of August which are suppose to be epsilon
> > > close to 5.6 release.
> 
> If you mean ports.tar.gz, I definitely recommend fetching the files from
> anoncvs to make sure they are as expected. Jason's suggested command to
> fetch a new tree is correct, but since you already have an unpacked tree
> you can 'cd /usr/ports; cvs up -Pd -r OPENBSD_5_6". There might not be
> any changes and this might not fix things, but at least the tree will
> then be in a known state that is expected to work.
> 
> The other thing is that you'll need any packages on the system to either
> be from the right snapshot (as long as I didn't lose track of which version
> is which, they will be dated July 31) or built from an OPENBSD_5_6 ports
> tree, if you have a mixture of packages from earlier snapshots then there
> may be problems.
> 
> > > Is rrdtool package broken for 5.6 release?
> 
> It isn't - I have it installed on a system.
> 
> $ ls -l /usr/local/bin/rrdtool
> -r-xr-xr-x  1 root  bin  49016 Jul 31 11:21 /usr/local/bin/rrdtool
> 
> 

Thanks Stuart! As usual you are above and beyond when it comes to
helping people in OpenBSD community. 

I think I know what happened. I got a ports tree one week too late to be
as close to release as I wanted. I picked that Ruby hiccup in collectd
from early August. 

Anyhow last time I did this for 5.5 release my end of February snapshot
is still running on half a dozen machines (207 days uptime and counting
:)).

Most Kind Regards,
Predrag

Re: RRD.so does not exist

[Predrag Punosevac]

Stuart Henderson  wrote:

> On 2014/10/09 08:44, Predrag Punosevac wrote:
> > I am trying to build collectd on the snapshot from a 10 of August which
> > should be very close to November 1st release.
> > 
> > # uname -a
> > OpenBSD oko.bagdala2.net 5.6 GENERIC.MP#333 amd64
> > 
> > 
> > 
> > As some of you know collectd package depends on rrdtool  which in turn
> > depends on package rrdupdate. 
> > 
> > # ls -l rrdtool
> > total 28
> > drwxr-xr-x  2 root  wheel   512 Aug 11 13:41 CVS
> > -rw-r--r--  1 root  wheel  3296 Aug 10 13:33 Makefile
> 
> That file ^^ is an old version, there have been several commits to it
> since August.
> 
> > -rw-r--r--  1 root  wheel   115 Oct  9 02:07 distinfo
> > drwxr-xr-x  3 root  wheel   512 Apr  6  2014 files
> > drwxr-xr-x  3 root  wheel   512 Mar  5  2014 patches
> > drwxr-xr-x  3 root  wheel   512 Jun 14 13:14 pkg
> > 
> > 
> > 
> > During the build I get this error 
> > 
> > 
> > Error:
> > /usr/ports/pobj/rrdtool-1.4.8/fake-amd64/usr/local/ \
> > lib/ruby/site_ruby/2.0/x86_64-openbsd/RRD.so does not exist
> > 
> > That is indeed the case. That looks like what this fella was getting
> > 
> > https://www.mail-archive.com/misc@openbsd.org/msg130224.html
> > 
> > Is there a work around this problem?
> > 
> > Most Kind Regards,
> > Predrag
> > 
> 
> "cvs up -PdA" and try building again.
> 

Are you sure that this is a good idea? I am not trying to follow 
current? If I check out current port tree I will have to rebuild
everything. I am just trying to build a single port using a port
snapshot and the sytem from 10 of August which are suppose to be epsilon
close to 5.6 release. Is rrdtool package broken for 5.6 release?

Most Kind Regards,
Predrag

RRD.so does not exist

[Predrag Punosevac]

I am trying to build collectd on the snapshot from a 10 of August which
should be very close to November 1st release.

# uname -a
OpenBSD oko.bagdala2.net 5.6 GENERIC.MP#333 amd64



As some of you know collectd package depends on rrdtool  which in turn
depends on package rrdupdate. 

# ls -l rrdtool
total 28
drwxr-xr-x  2 root  wheel   512 Aug 11 13:41 CVS
-rw-r--r--  1 root  wheel  3296 Aug 10 13:33 Makefile
-rw-r--r--  1 root  wheel   115 Oct  9 02:07 distinfo
drwxr-xr-x  3 root  wheel   512 Apr  6  2014 files
drwxr-xr-x  3 root  wheel   512 Mar  5  2014 patches
drwxr-xr-x  3 root  wheel   512 Jun 14 13:14 pkg



During the build I get this error 


Error:
/usr/ports/pobj/rrdtool-1.4.8/fake-amd64/usr/local/ \
lib/ruby/site_ruby/2.0/x86_64-openbsd/RRD.so does not exist

That is indeed the case. That looks like what this fella was getting

https://www.mail-archive.com/misc@openbsd.org/msg130224.html

Is there a work around this problem?

Most Kind Regards,
Predrag

Re: Mantis Bug Tracker on ONPP

[Predrag Punosevac]

Predrag Punosevac wrote:

>I would just like to leave electronic evidence that Mantis Bug Tracker
>works on the top of ONPP stack (OpenBSD, Nginx, PostgreSQL, PHP) without
>any problems. It works in the chroot of course. The only "tricky" part
>is manually creating config_inc.php from config_inc.php.sample file as
>well making sure that permissions are correct since Mantis is no longer
>in ports.
>

Actually even trickier part was getting e-mail notification to work. If
you need some help with Mantis BT please contact me via private e-mail. 
I have no idea if there is any interest in resurrecting a port in
particularly in the light of relatively recent discussion about web
applications. In any case I am very, very busy at the moment and even 
if I was not there are more interesting ports to work on than Mantis.

Cheers,
Predrag



>By the way if it was up to me I would use fossil and proxy fossil server
>with Nginx. Works like a charm. However my co-workers wanted Mantis
>due to its integration with CVS and Subversion. 

Re: pythontex question

[Predrag Punosevac]

I did a little bit more digging. man sem_open reveals that this might be
related to OpenBSD semaphore implementation and the fact that this code
has to be compiled with pthread lib linked. It is turning that this is
much more fundamental issue than I thought. I have found some thread on
Sage mailing list about this things. 

pythontex question

[Predrag Punosevac]

I just tried to compile some pythontex code I wrote a while ago.

predrag@oko$ uname -a
OpenBSD oko.bagdala2.net 5.4 GENERIC.MP#44 i386

The first surprise was that the pythontex was not installed. I was under
impression that it was the part of TeXLive 2012? No big deal. I went and
did manual installation. 

However I am stuck now with the python problems

predrag@oko$ pythontex mama.tex
This is PythonTeX v0.12
Traceback (most recent call last):
  File "/home/predrag/.pythontex/pythontex.py", line 54, in 
pythontex.main()
  File "/home/predrag/.pythontex/pythontex2.py", line 2323, in main
do_multiprocessing(data, temp_data, old_data, engine_dict)
  File "/home/predrag/.pythontex/pythontex2.py", line 972, in
do_multiprocessing
pool = multiprocessing.Pool(max_processes)
  File "/usr/local/lib/python2.7/multiprocessing/__init__.py", line 232,
in Pool
return Pool(processes, initializer, initargs, maxtasksperchild)
  File "/usr/local/lib/python2.7/multiprocessing/pool.py", line 138, in
__init__
self._setup_queues()
  File "/usr/local/lib/python2.7/multiprocessing/pool.py", line 232, in
_setup_queues
from .queues import SimpleQueue
  File "/usr/local/lib/python2.7/multiprocessing/queues.py", line 48, in

from multiprocessing.synchronize import Lock, BoundedSemaphore,
Semaphore, Condition
  File "/usr/local/lib/python2.7/multiprocessing/synchronize.py", line
59, in 
" function, see issue 3770.")
ImportError: This platform lacks a functioning sem_open implementation,
therefore, the required synchronization primitives needed will not
function, see issue 3770.



A quick Google reveals that this might be Python problem specific to
pythontex package and the fact that most likely written on amd64
platform. Is this true? I know that there are some people who know
infinitely more about Python than me on this mailing list. 


Most Kind Regards,
Predrag

P.S. I just checked 

https://github.com/jasperla/openbsd-wip

and I am very excited that TeXLive seems to be ready (I will try to do
testing after the New Year) but I am even more excited that many core 
scientific Python ports will be finally updated. Any progress in
creating the flavor of scipy, numpy, and friends for Python 3.3.1?

Re: x2goclient

[Predrag Punosevac]

Stuart Henderson  wrote:

> On 2013/12/22 03:15, Predrag Punosevac wrote:
> > I noticed that Rafael Sadowski was working on x2goclient
> > 
> > https://github.com/jasperla/openbsd-wip/blob/master/x11/x2goclient/Makefile
> > 
> > Are there any serious technical obstacles to commit the port? Due to
> > the fact that NX is no completely proprietary and close source and in
> > the light of my rather poor experience with FreeNX server I started
> > completely switching to X2Go.
> > 
>
> the biggest issue is listed in TODO..
>

Thanks Stuart! Unfortunately there are some much more serious things
which were communicated to me privately. I am glad that thing is not in
OpenBSD ports tree.


> | PKGNAME =x2goclient-${VERSION}
> | DISTNAME =x2goclient-${VERSION}
>
> just set DISTNAME
>
> | # GPL2
>
> use "GPLv2+" or "GPLv2 only" depending on whether it's v2-or-newer or v2-only
>
> | LIB_DEPENDS =x11/nx/nxproxy \
> | security/libssh \
> | x11/qt4 \
> | databases/openldap \
> | print/cups
>
> cups libs are now in the -libs subpackage
>
> | BUILD__DEPENDS =x11/nx/nxproxy \
>
> broken line, should only be one _
>
> | do-install:
> | ${INSTALL} ${WRKSRC}/x2goclient ${PREFIX}/bin
>
> should be INSTALL_PROGRAM
>
> (and whitespace in the port Makefile is not to usual standards)

x2goclient

[Predrag Punosevac]

I noticed that Rafael Sadowski was working on x2goclient

https://github.com/jasperla/openbsd-wip/blob/master/x11/x2goclient/Makefile

Are there any serious technical obstacles to commit the port? Due to
the fact that NX is no completely proprietary and close source and in
the light of my rather poor experience with FreeNX server I started
completely switching to X2Go.

Mantis Bug Tracker on ONPP

[Predrag Punosevac]

I would just like to leave electronic evidence that Mantis Bug Tracker
works on the top of ONPP stack (OpenBSD, Nginx, PostgreSQL, PHP) without
any problems. It works in the chroot of course. The only "tricky" part
is manually creating config_inc.php from config_inc.php.sample file as
well making sure that permissions are correct since Mantis is no longer
in ports.

By the way if it was up to me I would use fossil and proxy fossil server
with Nginx. Works like a charm. However my co-workers wanted Mantis
due to its integration with CVS and Subversion. 

dpb freezes on ruby-1.8.7 and 1-9.3

[Predrag Punosevac]

Dear All,

I was trying to build last night all packages for amd64 as I promised
two weeks ago to Mark Espie on that Supermicro server with 64 cores and
256 GB of RAM. I was using the snapshot from yesterday and the latest
port. dmesg has been sent to dmesg@ I just run 
./dpb -a

As I came this morning to my office I noticed that about 
3GB of packages has been build and that the built is hanging with the 
message 
http://www.devio.us/~ppunosevac/logs.tgz

This server is interesting as it has only one 600 GB Solid State HDD

Intel DC S3500 Series MCL (6Gb/s)

supposedly for better read and write times.  

I did couple of tests and executing mv command on the 500 MB directory
took 5 minutes. It took almost a minute to rm -rf that same directory.

What am I doing wrong? 


Best,
Predrag

P.S. I apologize to Mark for this long delay. The good news is that I
have so much other stuff to do that I could probably hang on to this
server for another 4-5 days before putting it into production so please
feel free to teach me something and use this machine for something
OpenBSD related. 

Re: [UPDATE] py-scipy 0.12.0

[Predrag Punosevac]

On 04/12/13 11:42, Markus Bergkvist wrote:
>> Updates py-scipy to 0.12.0. Builds and installs fine on AMD64 and I'm
>> using it with some in-hourse projects with no obvious issues so far.
>>
>> Maintainer timeout.
>>
>> make regress fails, but so does the in-tree port too.
>> Comments?
>
> If it works for you, go ahead.
> Please take maintainership, or at least remove me.
>
Any update on this? I would gladly test diff however numpy has to be
fixed first (updated agenst 2.7.4 preferably to 1.7.1. release)

Thanks,
Predrag

Re: [UPDATE] py-numpy

[Predrag Punosevac]

On Fri, Apr 12, 2013 at 8:08 AM, Markus Bergkvist
 wrote:
> updates py-numpy to 1.7.0. Builds and installs fine on AMD64 and I'm using
> it with some in-hourse projects with no obvious issues.
>
> However, make regress gives these errors
>
>> FAILED (KNOWNFAIL=5, SKIP=10, failures=2)
>> Running unit tests for numpy
>> NumPy version 1.7.0
>> Python version 2.7.3 (default, Apr  4 2013, 16:00:07) [GCC 4.2.1 20070719
>> ]
>> nose version 1.1.2
>
>
> But doing the same with in-tree version, 1.6.0, I get
>
>> FAILED (KNOWNFAIL=3, SKIP=6, failures=12)
>> NumPy version 1.6.0
>> Python version 2.7.3
>
>
> Must fix or ignorable?

Could anybody give me an update on the status of this update? In the
mean time as we know Python is upgraded to 2.7.4 and numpy-1.7.1 has
been released.

Thank you,
Predrag

Python for scientist

[Predrag Punosevac]

After a bit of deliberation I am posting this question on ports as I
belive that is more appropriate than misc.

I am not trained in scientific computing but these days I find myself
more and more doing some of that stuff as a part of my day job. As
MATLAB doesn't run on OpenBSD I tend to use Python whenever possible
i.e.

numpy
scipy
matplotlib
ipython

Untill recently I assumed that I should run Python2.7 as some of the
above libraries might not work well on 3.3. In part this confusion is
due to the fact that Enthought Python which we use on Linux is based on
2.7 branch.

However, after a bit of investigation I belive that I am wrong and that
the above libraries work perfectly well on 3.3. There are many reasons
why I prefer 3.3. over 2.7 branch. 

Today for the first time I checked Makefiles for above libraries looking
for 3.2 flavor. I do not see that such thing exists. When running Python
(3.2 on this computer) shell and trying to import numpy for example 
give the error as the numpy is stored in

/usr/local/lib/python2.7/site-packages/numpy

Is anybody using numpy on 3.2? Does numpy and above ports need to be
build specifically to run on 3.3 (awaiting for that 3.3.1 patch)? 
Could anybody educate me little bit about these issues?

Most Kind Regards,
Predrag

Re: UPDATE: lang/python/2.7

[Predrag Punosevac]

Stuart Henderson  wrote:

> On 2013/04/11 16:18, Predrag Punosevac wrote:
> > Federico G. Schwindt wrote:
> > > The diff below updates python 2.7 to 2.7.4.
> > 
> > I just installed the lates i386 snapshot to play with this. The build
> > fails see below. Any suggestions? I am looking forward to 3.3.1 upgrade.
> > Feel free to send a private e-mail.
>
> Try cleaning your python directory and re-applying, there was some problem
> when you applied the diff as you still have a file which would have been
> removed by the diff.

I tried cleaning manually all original files. Now it looks a little bit
better since is not scramming about existing original files but it still
looking for a patch.


# cd 2.7
# make clean
===>  Cleaning for python-2.7.4
# make makesum
===>  Checking files for Python-2.7.4
`/usr/ports/distfiles/Python-2.7.4.tgz' is up to date.
# make checksum
===>  Checking files for Python-2.7.4
`/usr/ports/distfiles/Python-2.7.4.tgz' is up to date.
>> (SHA256) Python-2.7.4.tgz: OK
# make extract
===>  Checking files for Python-2.7.4
`/usr/ports/distfiles/Python-2.7.4.tgz' is up to date.
>> (SHA256) Python-2.7.4.tgz: OK
===> python-2.7.4 depends on: metaauto-* -> metaauto-1.0
===> python-2.7.4 depends on: autoconf-2.69 -> autoconf-2.69p0
===> python-2.7.4 depends on: groff->=1.21 -> groff-1.22.2
===> python-2.7.4 depends on: bzip2-* -> bzip2-1.0.6p0
===> python-2.7.4 depends on: libffi-* -> libffi-3.0.9p2
===> python-2.7.4 depends on: gdbm-* -> gdbm-1.10
===> python-2.7.4 depends on: tk->=8.5,<8.6 -> tk-8.5.13
===> python-2.7.4 depends on: tcl->=8.5,<8.6 -> tcl-8.5.13
===> python-2.7.4 depends on: db->=4,<5|db->=4v0,<5v0 -> db-4.6.21v0
===>  Verifying specs:  ssl panelw m crypto c util z readline pthread  ncursesw 
stdc++ expat ossaudio bz2 sqlite3 ffi   gdbm>=3 python2.7 c pthread  X11 tk85 
tcl85 python2.7 c pthread lib/db4/db>=4 python2.7 c pthread
===>  found ssl.19.0 panelw.5.0 m.8.0 crypto.22.0 c.67.0 util.11.4 z.4.1 
readline.3.0 pthread.17.0 ncursesw.12.1 stdc++.55.0 expat.10.0 ossaudio.3.1 
bz2.10.4 sqlite3.22.0 ffi.0.0 gdbm.4.0 python2.7.0.0 X11.15.1 tk85.0.10 
tcl85.1.3 lib/db4/db.5.0
===>  Extracting for Python-2.7.4
rm -r /home/predrag/ports/Python-2.7.4/Python-2.7.4/Modules/expat
# make patch
===>  Patching for Python-2.7.4
File to patch: 

Re: UPDATE: lang/python/2.7

[Predrag Punosevac]

Federico G. Schwindt wrote:
> The diff below updates python 2.7 to 2.7.4.

I just installed the lates i386 snapshot to play with this. The build
fails see below. Any sugestions? I am looking forward to 3.3.1 upgrade.
Feel free to send a private e-mail.

Best,
Predrag


===>  Extracting for Python-2.7.4
rm -r /home/predrag/ports/Python-2.7.4/Python-2.7.4/Modules/expat
===>  Patching for Python-2.7.4
===>   Ignoring patchfile patch-Lib_test_regrtest_py.orig
===>   Ignoring empty patchfile patch-Lib_test_test_dl_py
===>   Ignoring patchfile patch-Lib_test_test_dl_py.orig
===>   Ignoring patchfile patch-Lib_test_test_tempfile_py.orig
File to patch:
No file found--skip this patch? [n] y
2 out of 2 hunks ignored--saving rejects to Oops.rej
***>   patch-configure_in did not apply cleanly
2 out of 17 hunks failed--saving rejects to setup.py.rej
***>   patch-setup_py did not apply cleanly
*** Error 1 in . (/usr/ports/infrastructure/mk/bsd.port.mk:2579 
'/home/predrag/ports/Python-2.7.4/.patch_done': @if cd /usr/ports/lang/pytho...)
*** Error 1 in . (/usr/ports/infrastructure/mk/bsd.port.mk:1863 
'/usr/ports/packages/i386/all/python-2.7.4.tgz')
*** Error 1 in . (/usr/ports/infrastructure/mk/bsd.port.mk:2409 
'_internal-package')
*** Error 1 in . (/usr/ports/infrastructure/mk/bsd.port.mk:2389 'package')
*** Error 1 in . (/usr/ports/infrastructure/mk/bsd.port.mk:1880 
'/var/db/pkg/python-2.7.4/+CONTENTS')
*** Error 1 in /usr/ports/lang/python/2.7 
(/usr/ports/infrastructure/mk/bsd.port.mk:2389 'install')

Re: py-pip dilemma

[Predrag Punosevac]

Daniel Dickman  wrote:

> What ports are you looking at specifically? I'm working to port pandas so we 
> might have some overlap. See here:
> http://dickman.org/openbsd/patches/pandas/
>

numpy
scipy
matplotlib
ipython 


For starters. I know people have been working on ipython but qt4 flavor
is holding them back. I use those on daily basis. I use several other
modules but these four are the highest priority.

Cheers,
Predrag


> The main one I have left to do is matplotlib. If you'd want to test I'm happy 
> send patches over.
>
> On Dec 13, 2012, at 10:00 PM, Predrag Punosevac  wrote:
>
> > I am soliciting opinions about the "safety" of py-pip package
> > management system. I am using Python primarily for scientific
> > computing/prototyping. Many of standard scientific python modules in our
> > ports tree are a bit outdated (trying to update some of those ports is
> > on my todo list as I am sure it is on todo list of many people but busy
> > schedule is another story).  
> > 
> > Inspired by my recent experience with hs-cabal I decided to play with
> > py-pip on one of 5.1 installations which I will update soon.
> > Surprisingly or maybe not I was actually able to update most of my
> > python modules with exception of scipy which fails with linker problem 
> > 
> > ld: cannot find -lpython2.7
> > 
> > How safe is really using py-pip as a Python package manager? Is it
> > possible to reconfigure py-pip to behave like cabal which installs
> > modules in my home directory rather than in /usr/local/bin?
> > 
> > Best,
> > Predrag
> > 

py-pip dilemma

[Predrag Punosevac]

I am soliciting opinions about the "safety" of py-pip package
management system. I am using Python primarily for scientific
computing/prototyping. Many of standard scientific python modules in our
ports tree are a bit outdated (trying to update some of those ports is
on my todo list as I am sure it is on todo list of many people but busy
schedule is another story).  

Inspired by my recent experience with hs-cabal I decided to play with
py-pip on one of 5.1 installations which I will update soon.
Surprisingly or maybe not I was actually able to update most of my
python modules with exception of scipy which fails with linker problem 

ld: cannot find -lpython2.7

How safe is really using py-pip as a Python package manager? Is it
possible to reconfigure py-pip to behave like cabal which installs
modules in my home directory rather than in /usr/local/bin?

Best,
Predrag

haskell-pandoc

[Predrag Punosevac]

Alexander Polakov  wrote:

> * Predrag Punosevac  [121205 00:40]:
> > On Mon, Dec 03, 2012 at 10:32:56PM +0100, Matthias Kilian wrote:
> > > I'd like to add --enable-tests by default in ghc.port.mk, unless a
> > > hs-ports has set NO_REGRESS=Yes.
> > [...]
> > 
> > This is an OT. Any chance that one of you Haskell guys gets pandoc
> > ported to OpenBSD.
> > 
> > http://johnmacfarlane.net/pandoc/
> > 
> > P.S. I tried installing with cabal but it just returns a slue of
> > complains. 
> > 
>
> pandoc installs and works fine here on -current/i386.

I can confirm this! I just compiled cabal-install from ports on the i386
current.

$ uname -a
OpenBSD foo.my.domain 5.2 GENERIC.MP#107 i386

Inspite 

$ cabal update
Downloading the latest package list from hackage.haskell.org
Note: there is a new version of cabal-install available.
To upgrade, run: cabal install cabal-install


I didn't want to install new cabal-install I just went with the one from
ports. This is the result

Registering pandoc-1.9.4.5...
Preprocessing executable 'make-pandoc-man-pages' for pandoc-1.9.4.5...
[1 of 1] Compiling Main ( man/make-pandoc-man-pages.hs,
dist/build/make-pandoc-man-pages/make-pandoc-man-pages-tmp/Main.o )
Linking dist/build/make-pandoc-man-pages/make-pandoc-man-pages ...
/usr/local/lib/ghc/libHSrts.a(RtsFlags.o)(.text+0x2a2): In function
`copyArg':
: warning: strcpy() is almost always misused, please use strlcpy()
/usr/local/lib/ghc/libHSrts.a(RtsUtils.o)(.text+0x2f2): In function
`showStgWord64':
: warning: sprintf() is often misused, please use snprintf()
Installing library in /home/predrag/.cabal/lib/pandoc-1.9.4.5/ghc-7.4.2
Installing executable(s) in /home/predrag/.cabal/bin
Registering pandoc-1.9.4.5...
$ 

So it comes out successfully. 


Best,
Predrag



>
> -- 
> Alexander Polakov | plhk.ru

Re: Enable cabal test-suites for hs-ports by default?

[Predrag Punosevac]

On Mon, Dec 03, 2012 at 10:32:56PM +0100, Matthias Kilian wrote:
> I'd like to add --enable-tests by default in ghc.port.mk, unless a
> hs-ports has set NO_REGRESS=Yes.
[...]

This is an OT. Any chance that one of you Haskell guys gets pandoc
ported to OpenBSD.

http://johnmacfarlane.net/pandoc/

I use txt2tags in general but there are few situations where I would
like to be able to use pandoc (yes I know it is available on-line). 

Thank,
Predrag

P.S. I tried installing with cabal but it just returns a slue of
complains. 

Re: [UPDATE] print/acroread 9.5.1

[Predrag Punosevac]

> Fri, Apr 27, 2012 at 4:31 PM, David Coppa  wrote:

> > Hi,
> >
> > This is my diff to update print/acroread to the latest version.
> >
> > Before this can go in, our compat_linux needs to be patched for
> > PROT_GROWS{DOWN,UP} flags handling (see my other mails on ports@).
> > Dunno if Paul does want/has time to do it, otherwise I can try
> > by myself...
> >
> > Besides the issue above, acroread-9.5.1 works very well: I've
> > openend one of those giant redbooks from IBM and all the major
> > functionalities (search, zoom in, zoom out, links, etc. etc.)
> > appear to work just fine.

> Here's a new (final, I hope) revision:

> - Remove more useless cruft from the acroread script

> - Install .desktop files and icons in the right places

> - Install the acroread.1 manual page (groff is needed, lots of errors
> from "mandoc -Tlint")

>CIAO,
>David


I completely forgot about this David's post from May. What is the status
of this work?

Best,
Predrag

Re: linux related ports

[Predrag Punosevac]

Core dumps as promised. Also my machine got twice completely frozen
and right now Opera seems the biggest suspect.

$ gdb -c opera.core
GNU gdb 6.3
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you
are
welcome to change it and/or distribute copies of it under certain
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for
details.
This GDB was configured as "i386-unknown-openbsd5.2".
"/home/predrag/opera.core" is not a core dump: File format not
recognized
(gdb) q
$ gdb -c sh.core
GNU gdb 6.3
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you
are
welcome to change it and/or distribute copies of it under certain
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for
details.
This GDB was configured as "i386-unknown-openbsd5.2".
Core was generated by `sh'.
Program terminated with signal 11, Segmentation fault.
#0  0x1c000180 in ?? ()
(gdb)

Missing Makefiles

[Predrag Punosevac]

I have the latest ports three which I got via CVS from the anoncvs in
Paris France (Jasper's machine I think). I wanted to play with csync.
The Makefile is missing.  On the another hand /usr/ports/net/ocsync
looks good but I just want to test csync (I see some situations where
might be more useful than Unison which I currently use as requires no
sever side installation).

Also noticed 

/usr/ports/graphics/xpdf
/usr/ports/textproc/xpdf

The first one 

/usr/ports/graphics/xpdf

is missing Makefile. The second one 

/usr/ports/textproc/xpdf

builds and runs as expected.

Cheers,
Predrag

linux related ports

[Predrag Punosevac]

I was playing little bit with Opera 12.11 and on updated linux emulation
libraries. It works great for the most part. I am having troubles on
OpenBSD foo.my.domain 5.2 GENERIC.MP#107 i386

with cleaning private data. I can reliably reproduce the crash. The
message is 

linux: syscall epoll_create is obsolete or not implemented

The same thing happens if I want to turn off page caching and turn of 
"remember content on visited pages" from History (advanced preferences).
This things used to work on earlier version with mp kernel. HTML5 is
also a bit problematic (not that is very useful anyway due to non
existing alsa support infrastructure). I got a few times sh.core but I
forgot to run gdb. I will post next time when it crashes. I have not
played with Opera mail client which used to be very problematic (not
using it for e-mail but useful for RSS feeds).

Is linux emulation now good enough for things like Acroread 9.0 or
above. Are there any plans to update Acroread? I have a very few uses
for Acroread but there are some pdf forms and some pdf files which
include JavaScript that could use Acroread.

Thanks to all the developers involved in linux emulation and porting
related software.

Cheers,
Predrag

Re: UPDATE: TeX Live 2012

[Predrag Punosevac]

Dear Edd and Stuart,

Sorry I could not join the party earlier. I started testing TeXLive 2012
i386 on current. The diff 

http://junkpile.org/tl.diff.2

Stuart posted works great. TeXLive builds and runs fine. I tested many
complex documents. Some highlights of my early testing:

1. Semantic checkers chktex and lacheck work as expected
2. PSTrics (heavy user here) works as expected. My usual workflow
tex->dvi->ps->pdf works without a glitch.
3. bibtex, natbib works great have not checked biblatex will check. 
4. RevTeX4.1  class which I used to install manually works as expected.
5. ModernCV, modern letter, exams class works as expected.
6. I tested flyers, posters, badges, powerdot all sorts of very
complicated documents (floats, hyperref, index anything you can think
of).
7. Have not checked latexmk (I usually use home brewed Makefiles) on my
todo list.
8. I have not tested MetaPost on my todo list (I do not use it). I am
also planning to play with Metafont.
9. Asymptote not tested.
10. pdfjam tested.
11. tex4ht and htlatex tested.
12. LuaTeX not tested, ConTeXT not tested. I will play in next couple of
weeks.
13. XeTeX not tested!
14. detex tested. 
 
The most important for the last. I could not find tlgmr. I am working
under the assumption that you removed tlgmr from the build
intensionally? However I see lots of tl related binaries:

intltool-extractintltool-update
intltool-merge  intltoolize
installfont-tl  intltool-prepare

That is the first thing I am going to hit next couple of days. I
probably left out tons of other stuff too but this is all I could do in
one day.


IMHO TeXLive2012 works good enough to be imported into the current tree.
That would hopefully attract some more testers.

Thank you so much for your hard work.

Best,
Predrag


P.S. Do you want me to test on amd64 and sparc64? I usually stay with
i386 because of Opera. Any reports on loongson and sgi? I do not have
the hardware.

ipython-0.13

[Predrag Punosevac]

Has anybody tried to update existing ipython port to the newest stable
version 0.13?

Thank you,
Predrag

catppt coredump

[Predrag Punosevac]

Is anybody using catppt to extract information from ppt slides? This is
what I get when trying to run catppt on a slides somebody sent me


$ gdb -c catppt.core
GNU gdb 6.3
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you
are welcome to change it and/or distribute copies of it under certain
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for
details.
This GDB was configured as "i386-unknown-openbsd5.1".
Core was generated by `catppt'.
Program terminated with signal 11, Segmentation fault.
#0  0x1c00363a in ?? ()

My machine 

$ uname -a
OpenBSD oko.bagdala2.net 5.1 GENERIC.MP#188 i386

Can anybody recommend another command line tool or script which can be
used instead. I do not like to have OO on my computer so for now I am
forced to upload things on Microsoft skydrive.

Thanks,
Predrag

nvi commit question

[Predrag Punosevac]

Dear All,

I just noticed that Naddy has committed nvi to the ports tree. As a user
of nvi from the base could anybody shed light on the purpose of this
commit. I remember earlier discussions among developers to bring nvi
from the base in sync with upstream. Does this commit mean that this is
not going to happen any time soon? Are there any benefits from using nvi
from the ports tree as oppose to the one from the base for people who do
not need UTF-8 support? In another words is this just replacement for
nvi-m17n from the ports or is it  something that "regular" users like me
should care about.

Best,
Predrag 

Re: ConTeXt problem

[Predrag Punosevac]

Edd Barrett  wrote:

> On Mon, Apr 30, 2012 at 11:02:23PM -0400, Predrag Punosevac wrote:
> > $ context test.tex
> > 
> > mtxrun  | unknown script 'context.lua' or 'mtx-context.lua'
>
> Your example works fine here.
>
> Do you have texlive_texmf-context installed?

I tought I had. However, when I checked manually I had package built but
not installed. I added it and everything works fine here. Sorry for the
noise.

I also noticed your earlier discussion with Mark and few other
developers about ConTeXt. I support your idea about splitting and making
stand alone port of ConTeXt. Unfortunately due to the nature of my work
I have no genuine interest to switch from LaTeX. If Frantishek wants to
finish his work I can help testing it. Otherwise I will offer ConTeXt
guys on their mailing list ssh access to one of my OpenBSD machines if
they want to play with it.

Thanks,
Predrag


>
> % context test.tex
> mtx-context | run 1: luatex 
> --fmt="/home/edd/.texlive2011/texmf-var/luatex-cache/context/f919750633122078d706dd10f1cfd39b/formats/cont-en"
>  
> --lua="/home/edd/.texlive2011/texmf-var/luatex-cache/context/f919750633122078d706dd10f1cfd39b/formats/cont-en.lui"
>  --backend="pdf" "./test.tex"
> This is LuaTeX, Version beta-0.70.1-2012041909 (rev 4277) 
>  \write18 enabled.
> (test.tex
>
> ConTeXt  ver: 2011.05.18 18:04 MKIV  fmt: 2012.5.3  int: english/english
>
> system  > cont-new.mkiv loaded
> (/usr/local/share/texmf-dist/tex/context/base/cont-new.mkiv
> system  > beware: some patches loaded from cont-new.mkiv
> )
> system  > test.top loaded
> (test.top)
> fonts   > latin modern fonts are not preloaded
> languages   > language en is active
> {/usr/local/share/texmf-dist/fonts/map/pdftex/context/mkiv-base.map}
> fonts   > preloading latin modern fonts (second stage)
> (/usr/local/share/texmf-dist/tex/context/base/type-siz.mkiv) 
> (/usr/local/share/texmf-dist/tex/context/base/type-otf.mkiv){/usr/local/share/texmf-dist/fonts/map/dvips/lm/lm-math.map}{/usr/local/share/texmf-dist/fonts/map/dvips/lm/lm-rm.map}
> fonts   > virtual math > unable to resolve name mapsfromchar
> fonts   > fallback modern rm 12pt is loaded
> system  > begin file test.tex at line 5
> structure   > sectioning > title @ level 2 : 0.0 -> \CONTEXT 
> structure   > sectioning > section @ level 3 : 0.0.1 -> Text
> backend > xmp > using file 
> '/usr/local/share/texmf-dist/tex/context/base/lpdf-pdx.xml'
> pages   > flushing realpage 1, userpage 1, subpage 1
> system  > end file test.tex at line 14
>  
> )
> mkiv lua stats  > used config file  - 
> selfautodir:/share/texmf/web2c/texmfcnf.lua
> mkiv lua stats  > used cache path   - readable: 
> '/usr/local/share/texmf-var/luatex-cache/context/f919750633122078d706dd10f1cfd39b'
>  (order 1)
> mkiv lua stats  > used cache path   - readable: 
> '/home/edd/.texlive2011/texmf-var/luatex-cache/context/f919750633122078d706dd10f1cfd39b'
>  (order 2)
> mkiv lua stats  > used cache path   - writable: 
> '/home/edd/.texlive2011/texmf-var/luatex-cache/context/f919750633122078d706dd10f1cfd39b'
> mkiv lua stats  > input load time   - 0.047 seconds
> mkiv lua stats  > stored bytecode data  - 285 modules, 60 tables, 345 
> chunks
> mkiv lua stats  > cleaned up reserved nodes - 33 nodes, 9 lists of 423
> mkiv lua stats  > node memory usage - 27 glue_spec
> mkiv lua stats  > node list callback tasks  - 6 unique task lists, 5 
> instances (re)created, 2412 calls
> mkiv lua stats  > used backend  - pdf (backend for directly 
> generating pdf output)
> mkiv lua stats  > loaded patterns   - en::2
> mkiv lua stats  > callbacks - 2559 direct, 2978 indirect, 
> 5537 total
> mkiv lua stats  > lxml preparation time - 0.000 seconds, 0 nodes, 11 
> lpath calls, 0 cached calls
> mkiv lua stats  > result saved in file  - test.pdf
> mkiv lua stats  > loaded fonts  - 34 files: stmary10.afm 
> lmmono12-regular.otf lmmono8-regular.otf lmmono9-regular.otf 
> lmroman12-bold.otf lmroman12-regular.otf lmroman7-bold.otf 
> lmroman7-regular.otf lmroman9-bold.otf lmroman9-regular.otf lmsans10-bold.otf 
> lmsans12-regular.otf lmsans8-regular.otf lmsans9-regular.otf eufb10.tfm 
> eufb7.tfm eufm10.tfm eufm7.tfm msam10.tfm msam7.tfm msbm10.tfm msbm7.tfm 
> lmex10.tfm lmmi12.tfm lmmi7.tfm lmmi9.tfm lmmib10.tfm lmmib7.tfm lmsy10.tfm 
> lmsy7.tfm lmsy9.tfm rm-lmr12.tfm rm-lmr7.tfm rm-lmr

Re: ConTeXt problem

[Predrag Punosevac]

Just a follow up to my original message. ConTeXt problem with TeXLive
2011 on 5.1 seems to be OpenBSD specific. Check out the thread I started

http://tex.stackexchange.com/questions/53892/texlive-2011-context-problem

This might be interesting output for people who know more about TeXLive
Running mtxrun as a root users makes no difference. 



$ mtxrun --generate

resolvers   | resolving | variable 'SELFAUTOLOC' set to '/usr/local/bin'
resolvers   | resolving | variable 'SELFAUTODIR' set to '/usr/local'
resolvers   | resolving | variable 'SELFAUTOPARENT' set to '/usr'
resolvers   | resolving | variable 'TEXMFCNF' set to ''
resolvers   | resolving | variable 'TEXMF' set to ''
resolvers   | resolving | variable 'TEXOS' set to 'local'
resolvers   | resolving |
resolvers   | resolving | unknown configuration file 
'/usr/local/texmfcnf.lua'
resolvers   | resolving | unknown configuration file '/usr/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/share/texmf-local/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/share/texmf/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/texmf-local/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/texmf/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/share/texmf-local/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/share/texmf/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/texmf-local/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/texmf/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/share/texmf-local/web2c/texmfcnf.lua'
resolvers   | resolving | found configuration file 
'/usr/local/share/texmf/web2c/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/texmf-local/web2c/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/local/texmf/web2c/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/share/texmf-local/web2c/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/share/texmf/web2c/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/texmf-local/web2c/texmfcnf.lua'
resolvers   | resolving | unknown configuration file 
'/usr/texmf/web2c/texmfcnf.lua'
resolvers   | resolving |
resolvers   | resolving | loading configuration file 
'selfautodir:/share/texmf/web2c/texmfcnf.lua'
resolvers   | resolving |
resolvers   | resolving | locating list of 'home:.texlive2011/texmf-config' 
(runtime)
resolvers   | methods | resolver: method=locators, how=uri, scheme=tree, 
argument=tree:///home:.texlive2011/texmf-config
resolvers   | trees | locator '/home/predrag/.texlive2011/texmf-config' not 
found
resolvers   | resolving | locating list of 'home:texmf' (runtime)
resolvers   | methods | resolver: method=locators, how=uri, scheme=tree, 
argument=tree:///home:texmf
resolvers   | trees | locator '/home/predrag/texmf' not found
resolvers   | resolving | locating list of '/usr/local/share/texmf-config' 
(cached)
resolvers   | methods | resolver: method=locators, how=uri, scheme=file, 
argument=/usr/local/share/texmf-config
resolvers   | files | file locator '/usr/local/share/texmf-config' not found
resolvers   | resolving | locating list of '/usr/local/share/texmf-project' 
(cached)
resolvers   | methods | resolver: method=locators, how=uri, scheme=file, 
argument=/usr/local/share/texmf-project
resolvers   | files | file locator '/usr/local/share/texmf-project' not 
found
resolvers   | resolving | locating list of '/usr/local/share/texmf-fonts' 
(cached)
resolvers   | methods | resolver: method=locators, how=uri, scheme=file, 
argument=/usr/local/share/texmf-fonts
resolvers   | files | file locator '/usr/local/share/texmf-fonts' not found
resolvers   | resolving | locating list of '/usr/local/share/texmf-local' 
(cached)
resolvers   | methods | resolver: method=locators, how=uri, scheme=file, 
argument=/usr/local/share/texmf-local
resolvers   | files | file locator '/usr/local/share/texmf-local' found as 
'/usr/local/share/texmf-local'
resolvers   | resolving | hash '/usr/local/share/texmf-local' appended
resolvers   | resolving | locating list of '/usr/local/share/texmf-dist' 
(cached)
resolvers   | methods | resolver: method=locators, how=uri, scheme=file, 
argument=/usr/local/share/texmf-dist
resolvers   | files | file locator '/usr/local/share/texmf-dist' found as 
'/usr/local/share/texmf-dist'
resolvers   | resolving | hash '/usr/local/share/texmf-dist' appended
resolvers   | resolving | locating list of '/usr/local/share/texmf-sys' 
(cached)
resolvers   | methods | resolver: method=locators, how=uri, scheme=fil

SymPy

[Predrag Punosevac]

Has anybody tried porting SymPy (Symbolic Python) to OpenBSD?

Thank you,
Predrag

ConTeXt problem

[Predrag Punosevac]

I was wondering if anybody has noticed any problems with ConTeXt on 5.1 
epsilon close to release. The machine from which I am writing you is
running i386 OpenBSD 5.1 snapshot from 13th of February which should be
very close to the release version.

$ uname -a
OpenBSD oko.bagdala2.net 5.1 GENERIC.MP#188 i386

I have problem processing minimal working example of ConTeXt. The
TeXLive 2011 is the version which will be shipped with 5.1 

$ more test.tex
setuphead[title][style={\ss\bfd},
before={\begingroup},
after={John Doe, the author\smallskip\currentdate\bigskip\endgroup}]
 
\starttext
 
\title{\CONTEXT}
 
\section{Text}
\CONTEXT\ is a document preparation system for the \TEX\ typesetting
program. It offers programmable desktop publishing features and
extensive
 
\stoptext



Output

$ texexec test.tex

mtxrun  | kpse fallback with progname 'context' initialized in
0.12 seconds


$ context test.tex

mtxrun  | unknown script 'context.lua' or 'mtx-context.lua'


Best,
Predrag

FreeMat 4.1

[Predrag Punosevac]

Is anyone working on the patch which updates OpenBSD FreeMat 4.0 port
to current FreeMat release 4.1? The 4.1 was released 28th of November this
year.

Best,
Predrag

Re: old unused ports: x11/ogle and related

[Predrag Punosevac]

On 07/04/11 16:10, Antoine Jacoutot wrote:
> On Mon, 4 Jul 2011, Amit Kulkarni wrote:
>
>> Ogle and related ports
>>
>> x11/ogle
>
> AFAIK ogle was the only software that would provide me with decent DVD
> video playing.
>
>> x11/ogle-gui
>> x11/goggles
>

Maybe it is just me but I have never been able to get out of MPlayer
half decent
support for chapters. Ogle is still my default DVD player.

Predrag

Re: RSS/Atom feed reader

[Predrag Punosevac]

Predrag Punosevac  wrote:

> Landry Breuil  wrote:
>
> > > > newsbeuter depends on a lib which can't be ported to openbsd yet.
> > > >
> > > > There's also www/raggle..
> > > 
> > > And this is a port for Canto (http://codezen.org/canto/).
> >
> > I think it should use MODPY_ADJ_FILES if possible, and then i'm ok with it.
> >
>
> Thank you so much guys for all your kind answers. These are my comments:
>
> @Marcus 
> I still can not get that script (atom2rss) to work but since you 
> confirmed that it works I will try it again. I really like snownews.
>
> @Landry
> I played with www/raggle. The interface reminds me of calcurse. It does
> everything it needs to do (as described in man pages) except one thing. 
> I can NOT delete feeds. Namely keybinding Delete doesn't work. 
> Keybinding Home, End, Page Up, and Page Down also do not work but they
> have alternative keys (0,$,j,k). Delete does not have. I have just a
> regular PS2 keyboard (IBM Type M). Nothing fancy. US layout.
>
> Speaking of newsbeuter I had a feeling that you tried porting it. It is
> sad that things are written in such non-portable way.
>
> @David
> Thank you for the port of Canto. I am compiling as we speak.
>
> Cheers,
> Predrag Punosevac
>
>
>
>
> > Landry

You will be laughing at this one. My 3.5 year old daughter Ekaterina has
fixed my "Keybouard" problem with raggle. Namely, raggle doesn't
recognize Delete key on Type M keyboard. However, it DOES recognize
Del key which is equivalent to decimal point . (when NumLock is 
switched off). Next time before hitting sand I will ask her.

Cheers,
Predrag

Re: RSS/Atom feed reader

[Predrag Punosevac]

Landry Breuil  wrote:

> > > newsbeuter depends on a lib which can't be ported to openbsd yet.
> > >
> > > There's also www/raggle..
> > 
> > And this is a port for Canto (http://codezen.org/canto/).
>
> I think it should use MODPY_ADJ_FILES if possible, and then i'm ok with it.
>

Thank you so much guys for all your kind answers. These are my comments:

@Marcus 
I still can not get that script (atom2rss) to work but since you 
confirmed that it works I will try it again. I really like snownews.

@Landry
I played with www/raggle. The interface reminds me of calcurse. It does
everything it needs to do (as described in man pages) except one thing. 
I can NOT delete feeds. Namely keybinding Delete doesn't work. 
Keybinding Home, End, Page Up, and Page Down also do not work but they
have alternative keys (0,$,j,k). Delete does not have. I have just a
regular PS2 keyboard (IBM Type M). Nothing fancy. US layout.

Speaking of newsbeuter I had a feeling that you tried porting it. It is
sad that things are written in such non-portable way.

@David
Thank you for the port of Canto. I am compiling as we speak.

Cheers,
Predrag Punosevac




> Landry

docx2txt, Spreadsheet::XLSX, and xlsx2csv

[Predrag Punosevac]

Dear All,

I would like to say happy Thanksgiving for all U.S. subscribers to
this mailing list.
I have one quick question for Perl gurus and also one quick for Python gurus.


Ever since MS "introduced" new Office 2007 formats I keep receiving more and
more docx and xmlx. Unlike previous  things which came out of the same kitchen
these are in essence just simple XML files zipped together with style
sheets files.
I was looking for a very simple solutions to convert these documents
to txt and csv
files along the lines of catdoc. For a single files when people do not
care much about
format of the original documents simple sed scripts are enough. However, there
are two very simple Perl solutions which seems to go step further

docx2txt  (http://sourceforge.net/projects/docx2txt/)
and
Spreadsheet::XLSX
(http://search.cpan.org/~dmow/Spreadsheet-XLSX-0.13-withoutworldwriteables/)

In particular first one seems to be more polished and except for
unnecessary use of
Bash shell for a script that comes with it seems like a fairly simple
code to me.
Would anybody from Perl gurus be interested to make a quick ports (I know I can
use them as stand alone Perl scripts but it would be nice just to
install them as
packages). It seems silly to me that I take a shot at the ports when
there are so
many serious Perl users around.

The second question is for Python gurus. I am also aware of xlsx2csv
(Python script)
https://github.com/dilshod/xlsx2csv
which seems to be doing exactly what I want but using Python instead
of Perl (a bit
strange choice for me). The drawback is that Python is not the part of
the base but
in 99% of my desktop installations I do end up with Python anyway so
what a hack.
Anybody interested in the quick port? xlsx2csv seems more refine than similar
Perl module.

Most Kind Regards,
Predrag Punosevac

Re: FreeMat 4.0

[Predrag Punosevac]

I just checked FreeMat on 4.0 current. The font problem in the editor is
gone. Also pcode generator of pseudo code is disabled by upstream.
Good!!! It was useless anywhy. So far 4.0 is real improvement over 3.6.

Sorry for the noise. 

Predrag

Re: FreeMat 4.0

[Predrag Punosevac]

I am at work right now having no access to current but on 4.8 release
(i386) FreeMat 3.6 crashes when I trying to produce pcode. I will try
tomorrow to reproduce crash on current. Pcode is not that important
anyway but people should be aware.

Also built in editor is useless (not that I am really using it or need
it). It gives annoying warning that the selected font is not fixed
pitch. After the warning the editor works as expected but hangs the
application when I try to close it. I tried changing font manually in
editor preference to fixed pitch font but the problem persists.
I will try to reproduce this on current.

Best,
Predrag

P.S. Is anybody on the upstream mailing lists? I am guessing that above
are not OpenBSD specific problems and have to do with quality control.

Missing runtime dependencies

[Predrag Punosevac]

I would like to report missing runtime dependencies for two applications
on the snapshot of 22 of August which should be very close to the 4.8
release.


1. /audio/abcde is missing /audio/id3ed which is needed to encode ripped
audio tracks into mp3 files (actually id3ed is doing only tagging).

also

2. /productivity/googlecl is missing /devel/py-gdata

Both applications work flawlessly after adding dependencies manually.

Cheers,
Predrag

Re: FreeMat 4.0

[Predrag Punosevac]

Alexandr Shadchin  wrote:

> On Mon, Oct 11, 2010 at 11:19 AM, Predrag Punosevac
>  wrote:
> > Here are some preliminary results of testing FreeMat-4.0 on OpenBSD
> > snapshot of 6th of October (i386 kernel bsd.sp).
> >
> > Running the run_tests multi-platform regression test suite which is
> > shipped with the FreeMat hangs after completing test bbtest_lower.
> > Application itself doesn't hang and I can start editor but the test is
> > never fully completed. It would be good if somebody can run this test on
> > the faster machine with multi cores as FreeMat requires LLVM compiler
> > and gfortran.
> >
> > I also run the test on the snapshot of 22 of August which should be
> > epsilon close to 4.8 release. I tested FreeMat version 3.6 which will be
> > shipped with the release on i386, bsd.mp. FreeMat dumps the core during
> > the test. This is the output from gdb
> >
>
> I used to test this script http://koba.devio.us/distfiles/run-test
>
> amd64(Core i5)
> script - OK
> GUI - first run OK, second - hangs after completing test bbtest_lower
>
> i386(PIII)
> script and GUI - hangs after completing test bbtest_lower
>

Ok, 

I have an idea where might be the problem. I checked the list of tests
and the test bbtest_ls is the one which is executed after the
bbtest_lower. This is the list of all tests

http://freemat.sourcearchive.com/documentation/3.6plus-pdfsg-6/dir_30c3fe9066a850dcd28222cc06f369c0.html


Now look at the code for that test

bbtest_ls.m

% Regression test function (black blox) for FreeMat v3.6
% This function is autogenerated by helpgen.
function bbtest_success = bbtest_ls
  bbtest_success = 1;
NumErrors = 0;
try
  ls m*.m

catch
  NumErrors = NumErrors + 1;
end
if (NumErrors ~= 0) bbtest_success = 0; return; end
NumErrors = 0;
try
  ls('m*.m')

catch
  NumErrors = NumErrors + 1;
end
if (NumErrors ~= 0) bbtest_success = 0; return; end
NumErrors = 0;
try
  ls 'm*.m'

catch
  NumErrors = NumErrors + 1;
end
if (NumErrors ~= 0) bbtest_success = 0; return; end


So that is the regression test for Freemat 3.6. Is it possible that the
test hangs because Freemat 3.6 crashes on OpenBSD? This is the culprit 
of the 3.6 crash

#0  0x1c14f1f8 in QList::~QList ()

I think that has to do with Qt libraries. Can anybody with the
knowledge of Qt libraries confirm that?

Cheers,
Predrag



> -- 
> Alexandr Shadchin

Re: FreeMat 4.0

[Predrag Punosevac]

Here are some preliminary results of testing FreeMat-4.0 on OpenBSD
snapshot of 6th of October (i386 kernel bsd.sp).

Running the run_tests multi-platform regression test suite which is
shipped with the FreeMat hangs after completing test bbtest_lower.
Application itself doesn't hang and I can start editor but the test is
never fully complited. It would be good if somebody can run this test on
the faster machine with multi cores as FreeMat requires LLVM compiler
and gfortran. 

I also run the test on the snapshot of 22 of August which should be
epsilon close to 4.8 relese. I tested FreeMat version 3.6 which will be
shipped with the release on i386, bsd.mp. FreeMat dumps the core during
the test. This is the output from gdb


GNU gdb 6.3
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i386-unknown-openbsd4.8"...
(no debugging symbols found)

Core was generated by `freemat'.
Program terminated with signal 11, Segmentation fault.
Reading symbols from /usr/lib/libpthread.so.12.1...done.
Loaded symbols for /usr/lib/libpthread.so.12.1
Reading symbols from /usr/lib/libz.so.4.1...done.
Loaded symbols for /usr/lib/libz.so.4.1
Reading symbols from /usr/local/lib/libarpack.so.1.0...done.
Loaded symbols for /usr/local/lib/libarpack.so.1.0
Reading symbols from /usr/local/lib/libfftw3.so.5.0...done.
Loaded symbols for /usr/local/lib/libfftw3.so.5.0
Reading symbols from /usr/local/lib/libfftw3f.so.5.0...done.
Loaded symbols for /usr/local/lib/libfftw3f.so.5.0
Reading symbols from /usr/local/lib/libumfpack.so.0.0...done.
Loaded symbols for /usr/local/lib/libumfpack.so.0.0
Reading symbols from /usr/local/lib/libamd.so.0.0...done.
Loaded symbols for /usr/local/lib/libamd.so.0.0
Reading symbols from /usr/local/lib/liblapack.so.5.0...done.
Loaded symbols for /usr/local/lib/liblapack.so.5.0
Reading symbols from /usr/local/lib/libblas.so.1.0...done.
Loaded symbols for /usr/local/lib/libblas.so.1.0
Reading symbols from /usr/local/lib/libavcall.so.1.0...done.
Loaded symbols for /usr/local/lib/libavcall.so.1.0
Reading symbols from /usr/local/lib/libpcre.so.2.3...done.
Loaded symbols for /usr/local/lib/libpcre.so.2.3
Reading symbols from /usr/local/lib/libportaudio.so.1.0...done.
Loaded symbols for /usr/local/lib/libportaudio.so.1.0
Reading symbols from /usr/lib/libcurses.so.11.0...done.
Loaded symbols for /usr/lib/libcurses.so.11.0
Reading symbols from /usr/local/lib/libQtOpenGL.so.7.0...done.
Loaded symbols for /usr/local/lib/libQtOpenGL.so.7.0
Reading symbols from /usr/local/lib/libQtNetwork.so.8.0...done.
Loaded symbols for /usr/local/lib/libQtNetwork.so.8.0
Reading symbols from /usr/local/lib/libQtXml.so.8.0...done.
Loaded symbols for /usr/local/lib/libQtXml.so.8.0
Reading symbols from /usr/local/lib/libQtSvg.so.7.0...done.
Loaded symbols for /usr/local/lib/libQtSvg.so.7.0
Reading symbols from /usr/local/lib/libQtGui.so.9.0...done.
Loaded symbols for /usr/local/lib/libQtGui.so.9.0
Reading symbols from /usr/local/lib/libQtCore.so.8.0...done.
Loaded symbols for /usr/local/lib/libQtCore.so.8.0
Reading symbols from /usr/X11R6/lib/libGLU.so.7.0...done.
Loaded symbols for /usr/X11R6/lib/libGLU.so.7.0
Reading symbols from /usr/local/lib/libg2c.so.8.1...done.
Loaded symbols for /usr/local/lib/libg2c.so.8.1
Symbols already loaded for /usr/lib/libpthread.so.12.1
Reading symbols from /usr/lib/libstdc++.so.50.0...done.
Loaded symbols for /usr/lib/libstdc++.so.50.0
Reading symbols from /usr/lib/libm.so.5.2...done.
Loaded symbols for /usr/lib/libm.so.5.2
Reading symbols from /usr/lib/libc.so.56.0...done.
Loaded symbols for /usr/lib/libc.so.56.0
Reading symbols from /usr/X11R6/lib/libGL.so.10.0...done.
Loaded symbols for /usr/X11R6/lib/libGL.so.10.0
Reading symbols from /usr/lib/libsndio.so.3.3...done.
Loaded symbols for /usr/lib/libsndio.so.3.3
Reading symbols from /usr/X11R6/lib/libfreetype.so.17.1...done.
Loaded symbols for /usr/X11R6/lib/libfreetype.so.17.1
Reading symbols from /usr/X11R6/lib/libXi.so.11.0...done.
Loaded symbols for /usr/X11R6/lib/libXi.so.11.0
Reading symbols from /usr/X11R6/lib/libXrender.so.5.0...done.
Loaded symbols for /usr/X11R6/lib/libXrender.so.5.0
Reading symbols from /usr/X11R6/lib/libXinerama.so.5.0...done.
Loaded symbols for /usr/X11R6/lib/libXinerama.so.5.0
Reading symbols from /usr/X11R6/lib/libfontconfig.so.7.0...done.
Loaded symbols for /usr/X11R6/lib/libfontconfig.so.7.0
Reading symbols from /usr/lib/libexpat.so.9.0...done.
Loaded symbols for /usr/lib/libexpat.so.9.0
Reading symbols from /usr/X11R6/lib/libXext.so.11.0...done.
Loaded symbols for /usr/X11R6/lib/libXext.so.11.0
Reading symbols from /usr/X11R6/lib/libX11.so.13.0...done.
Loaded symbols for /usr/X11R6/lib/libX11.so.13.0
Reading symbols from /usr/loc

Re: FreeMat 4.0

[Predrag Punosevac]

Alexandr Shadchin  wrote:

> Try this. Unpack in ports/math

Apparently somebody doesn't know how to apply diff properly(: The above 
port works like a charm on i386 snapshot of 6th of October.

I put the port in 

http://www.devio.us/~ppunosevac/ports/

and I put the binary packages in

http://www.devio.us/~ppunosevac/packages/

Actually with all dependencies installed compiling of FreeMat last less
than 30 minutes on the failry slow P4. I will be running quite a bit
m-code next couple of days on it so I will report problems. I will also 
try to compile FreeMat in next couple of days on sparc64.

Alexandr, thank you so much for your work!

Cheers,
Predrag

P.S. For people who are curious about FreeMat 4.0 on 4.8 release this is
the output

# pkg_add freemat-4.0.tgz
Can't install freemat-4.0 because of libraries
|library X11.14.0 not found
| /usr/X11R6/lib/libX11.so.13.0 (system): bad major
|library c.57.0 not found
| /usr/lib/libc.so.56.0 (system): bad major
|library ncurses.12.0 not found
| /usr/lib/libncurses.so.11.0 (system): bad major
Direct dependencies for freemat-4.0 resolve to: amd-2.2.0p2 pcre-8.02p1
fftw3-3.2.2 fftw3-float-3.2.2 portaudio-svn-1406p1 umfpack-5.1.0p3
qt4-4.6.3p2
Full dependency tree is amd-2.2.0p2 libxml-2.7.6 pcre-8.02p1
tremor-20080326 gstreamer-plugins-base-0.10.30p0 png-1.2.44 tiff-3.9.4
glib2-2.24.1p2 fftw3-common-3.2.2 fftw3-3.2.2 jpeg-8b libgamin-0.1.10p3
libmng-1.0.10 orc-0.4.6 pango-1.28.1 libiconv-1.13p1 cairo-1.8.10p0
fftw3-float-3.2.2 sqlite3-3.6.23p0 portaudio-svn-1406p1 libogg-1.2.0
glitz-0.5.6p2 libtheora-1.1.1p0 lcms-1.18a libvorbis-1.3.1
gettext-0.18.1 umfpack-5.1.0p3 gstreamer-0.10.30p0 qt4-4.6.3p2


Apparently it will take some backporting to get things working. It is
not worty since FreeMat is so complex that it needs much more testing on
4.8 current.

>
> -- 
> Alexandr Shadchin
>

Re: FreeMat 4.0

[Predrag Punosevac]

Alexandr Shadchin  wrote:

> On Mon, Sep 27, 2010 at 8:53 AM, Predrag Punosevac
>  wrote:
> > I tried to apply the patch to the snapshot (i386) of 22nd of August. I
> > know, I know that is not the way to do it. I will do tomorrow fresh
> > installation of the latest snapshot. For what is worthed the patch
> > failed (probably expectedly).
> >
> ...
> >
> > Best,
> > Predrag
> >
>
> Try to update the ports tree, as far as I can see the diff imposed not
> quite right.

I tried to compile FreeMat 4.0 on the snapshot of 6th of October. All
dependencies are compiled fine (it took 8 hours on a slow machine) but
FreeMat fails to compile. It looks like there is some problem with diff
for patches or I am doing something very wrong. Please see the output
below. 


> Or as an option, I compiled the package
> http://koba.devio.us/distfiles/freemat-4.0.tgz (-current) 12 MB
>
Is the package for i386 or amd64? I recall you testing freemat on amd64.
If it was for i386 it is NO go on 4.8 release as it requires newer
versions of some libraries. I have not tried to install on the snapshot
of 6th of October (4.8 current).

Best,
Predrag

P.S. Sorry that it took me so long to test it. I was just too busy with
my work. 


> -- 
> Alexandr Shadchin

===>  Verifying specs: lib/qt4/QtCore lib/qt4/QtCore GL GLU ICE SM X11 Xext c m 
ncurses pcre pthread stdc++ z amd fftw3 fftw3f portaudio umfpack QtGui 
QtNetwork QtOpenGL QtSvg QtXml GL GLU ICE SM X11 Xext c m ncurses pcre pthread 
stdc++ z amd fftw3 fftw3f portaudio umfpack QtGui QtNetwork QtOpenGL QtSvg QtXml
===>  found lib/qt4/QtCore.8.0 GL.10.0 GLU.7.0 ICE.9.0 SM.8.0 X11.14.0 
Xext.11.0 c.57.0 m.5.2 ncurses.12.0 pcre.2.3 pthread.12.1 stdc++.50.0 z.4.1 
amd.0.0 fftw3.5.0 fftw3f.5.0 portaudio.1.0 umfpack.0.0 QtGui.9.0 QtNetwork.8.0 
QtOpenGL.7.0 QtSvg.7.0 QtXml.8.0
===>  Extracting for freemat-4.0
===>  Patching for freemat-4.0
2 out of 2 hunks failed--saving rejects to CMakeLists.txt.rej
***>   patch-CMakeLists_txt did not apply cleanly
===>   Ignoring patchfile patch-CMakeLists_txt.orig
===>   Ignoring empty patchfile patch-configure_in
===>   Ignoring patchfile patch-configure_in.orig
===>   Ignoring empty patchfile patch-libs_libCore_Makefile_in
===>   Ignoring patchfile patch-libs_libCore_Makefile_in.orig
===>   Ignoring empty patchfile patch-libs_libFreeMat_Print_cpp
===>   Ignoring patchfile patch-libs_libFreeMat_Print_cpp.orig
===>   Ignoring empty patchfile patch-libs_libXP_Editor_cpp
===>   Ignoring patchfile patch-libs_libXP_Editor_cpp.orig
*** Error code 1

Stop in /usr/ports/math/freemat (line 2257 of 
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 1596 of 
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 2152 of 
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 2132 of 
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 1627 of 
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 2132 of 
/usr/ports/infrastructure/mk/bsd.port.mk).

Re: FreeMat 4.0

[Predrag Punosevac]

Alexandr Shadchin  wrote:

> On Wed, Sep 22, 2010 at 01:41:42AM -0400, Predrag Punosevac wrote:
> > FreeMat 4.0 is almost a year old now. Has anybody tried to bump up the
> > port from 3.6 to 4.0? I tried manually editing Makefile and compiling
> > 4.0 on the snapshot from 22nd of August (i386) but the patches do not
> > apply cleanly and the compilation fails. My main motivation for using
> > FreeMat 4.0 is better compatibility with MathLab m-code.
> > 
> > Best,
> > Predrag
> > 
> > P.S. GNU Ocatve appears also outdated so I have not even bothered to
> > play with it.
> >
>
> Hi!
>
> This update package Freemat to the latest release 4.0.
> Tested on amd64.
>
> --
> Alexandr Shdachin
>  
I tried to apply the patch to the snapshot (i386) of 22nd of August. I
know, I know that is not the way to do it. I will do tomorrow fresh
installation of the latest snapshot. For what is worthed the patch
failed (probably expectedly).

===>  Extracting for freemat-4.0
===>  Patching for freemat-4.0
2 out of 2 hunks failed--saving rejects to CMakeLists.txt.rej
2 out of 2 hunks failed--saving rejects to CMakeLists.txt.rej
2 out of 2 hunks failed--saving rejects to CMakeLists.txt.rej
***>   patch-CMakeLists_txt did not apply cleanly
===>   Ignoring patchfile patch-CMakeLists_txt.orig
===>   Ignoring empty patchfile patch-configure_in
===>   Ignoring patchfile patch-configure_in.orig
===>   Ignoring patchfile patch-configure_in.rej
===>   Ignoring patchfile patch-configure_in.rej.orig
===>   Ignoring empty patchfile patch-libs_libCore_Makefile_in
===>   Ignoring patchfile patch-libs_libCore_Makefile_in.orig
===>   Ignoring patchfile patch-libs_libCore_Makefile_in.rej
===>   Ignoring patchfile patch-libs_libCore_Makefile_in.rej.orig
===>   Ignoring empty patchfile patch-libs_libFreeMat_Print_cpp
===>   Ignoring patchfile patch-libs_libFreeMat_Print_cpp.orig
===>   Ignoring patchfile patch-libs_libFreeMat_Print_cpp.rej
===>   Ignoring patchfile patch-libs_libFreeMat_Print_cpp.rej.orig
===>   Ignoring empty patchfile patch-libs_libXP_Editor_cpp
===>   Ignoring patchfile patch-libs_libXP_Editor_cpp.orig
===>   Ignoring patchfile patch-libs_libXP_Editor_cpp.rej
===>   Ignoring patchfile patch-libs_libXP_Editor_cpp.rej.orig
*** Error code 1

Stop in /usr/ports/math/freemat (line 2242 of
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 1581 of
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 2137 of
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 2117 of
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 1612 of
/usr/ports/infrastructure/mk/bsd.port.mk).
*** Error code 1

Stop in /usr/ports/math/freemat (line 2117 of
/usr/ports/infrastructure/mk/bsd.port.mk).
# 

Best,
Predrag

Re: FreeMat 4.0

[Predrag Punosevac]

Alexandr Shadchin  wrote:

> On Wed, Sep 22, 2010 at 01:41:42AM -0400, Predrag Punosevac wrote:
> > FreeMat 4.0 is almost a year old now. Has anybody tried to bump up the
> > port from 3.6 to 4.0? I tried manually editing Makefile and compiling
> > 4.0 on the snapshot from 22nd of August (i386) but the patches do not
> > apply cleanly and the compilation fails. My main motivation for using
> > FreeMat 4.0 is better compatibility with MathLab m-code.
> > 
> > Best,
> > Predrag
> > 
> > P.S. GNU Ocatve appears also outdated so I have not even bothered to
> > play with it.
> >
>
> Hi!
>
> This update package Freemat to the latest release 4.0.
> Tested on amd64.
>
> --
> Alexandr Shdachin
>  
Thank you so much!!! You made my day. I will give it a spin
tonight on i386.

Predrag Punosevac

FreeMat 4.0

[Predrag Punosevac]

FreeMat 4.0 is almost a year old now. Has anybody tried to bump up the
port from 3.6 to 4.0? I tried manually editing Makefile and compiling
4.0 on the snapshot from 22nd of August (i386) but the patches do not
apply cleanly and the compilation fails. My main motivation for using
FreeMat 4.0 is better compatibility with MathLab m-code.

Best,
Predrag

P.S. GNU Ocatve appears also outdated so I have not even bothered to
play with it.

Re: TeXLive bugs?

[Predrag Punosevac]

Jachim Schipper wrote:
> Hint: the answer is in the fine manual, in paragraph 2.3 in fact.
> Joachim
Thanks!!! Problems are fixed.
Predrag

TeXLive bugs?

[Predrag Punosevac]

Dear Edd,

I know that you are very busy but I was wondering if you could tell me
anything from the top of your head about the following "problem".

I am trying to generate bunch of good looking business letters using the
following template 

http://www.matthewjmiller.net/files/cover_letters/cover_example.tex
http://www.matthewjmiller.net/files/cover_letters/uiuc_logo.eps (for
logo which should be put into figures/UILogoLG3L.eps

which I used for past three years for the correspondence. The final
output is supposed to look like 

http://www.matthewjmiller.net/files/cover_letters/cover_example.pdf



Unfortunately on the 

OpenBSD oko.bagdala2.net 4.8 GENERIC.MP#361 i386

of 22nd of August using texlive_texmf-full-2009p2, 
texlive_texmf-minimal-2009p1, and texlive_base-2009p4, I get that my 
signature is moved to the right (like it is used in continental Europe).
Also the date is flushed right on the mini page containing senders
information (which absolutely unacceptable as it is messing up document
big time). 

I would guess that there is a switch that I can use to flush my
signature to the left again. What about that stupid date.

Most Kind Regards,
Predrag Punosevac

P.S. I was expecting problems with Powerdot on my latest installation
because it is full of bugs but it is actually working better than
previous.) 

Re: surf and netsurf core dumps

[Predrag Punosevac]

I attached backtraces for both web browsers. I can easily reproduce
crash of Netsurf

http://ftp.openbsd.org/pub/OpenBSD/snapshots/packages/i386/

Surf doesn't crash but just goes into the infinite loop on the stupid
page which is suppose to worn that the browser is not supported. 
The page works fine in Midori, and XXXterm doesn't bypasses page
completely. So it is Surf issue.

Cheers,
Predrag

gdb netsurf netsurf.core

snip

(gdb) where
#0  0x032a9899 in g_type_check_value ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#1  0x03290314 in g_object_get_property ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#2  0x1c04fec1 in nsgtk_widget_get_visible (widget=0x80607b10)
at gtk/gtk_compat.c:36
#3  0x1c0505f7 in nsgtk_download_update (force_update=0)
at gtk/gtk_download.c:442
#4  0x069d8d16 in g_source_get_current_time ()
   from /usr/local/lib/libglib-2.0.so.1803.0
#5  0x069d8517 in g_main_context_dispatch ()
   from /usr/local/lib/libglib-2.0.so.1803.0
#6  0x069dbd7e in g_main_context_check ()
   from /usr/local/lib/libglib-2.0.so.1803.0
#7  0x069dc3a5 in g_main_context_iteration ()
   from /usr/local/lib/libglib-2.0.so.1803.0
#8  0x00f8fd43 in gtk_main_iteration_do ()
   from /usr/local/lib/libgtk-x11-2.0.so.1403.0
#9  0x1c052d47 in gui_poll (active=true) at gtk/gtk_gui.c:467
#10 0x1c03a6ff in netsurf_main_loop () at desktop/netsurf.c:169
#11 0x1c05387d in main (argc=Cannot access memory at address 0x0
) at gtk/gtk_gui.c:412

gdb surf surf.core
 
snip

(gdb) where
#0  0x1c00338e in drawindicator ()
#1  0x1c003592 in update ()
#2  0x0483465a in g_cclosure_marshal_VOID__PARAM ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#3  0x048279c9 in g_closure_invoke ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#4  0x0483c2b4 in g_signal_parse_name ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#5  0x0483e428 in g_signal_emit_valist ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#6  0x0483e779 in g_signal_emit () from
/usr/local/lib/libgobject-2.0.so.1803.0
#7  0x0482c011 in g_object_type_init ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#8  0x0482873f in g_enum_types_init ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#9  0x0482d1af in g_object_notify ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#10 0x0cb5dd7c in
WebKit::FrameLoaderClient::postProgressEstimateChangedNotification ()
from /usr/local/lib/libwebkit-1.0.so.2.2
#11 0x0c7ce344 in WebCore::ProgressTracker::finalProgressComplete ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#12 0x0c7ce3ea in WebCore::ProgressTracker::progressCompleted ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
---Type  to continue, or q  to quit---
#13 0x0c7a9b63 in WebCore::FrameLoader::checkLoadCompleteForThisFrame ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#14 0x0c7a9fb8 in WebCore::FrameLoader::recursiveCheckLoadComplete ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#15 0x0c791148 in WebCore::DocumentLoader::removeSubresourceLoader ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#16 0x0c7d823f in WebCore::SubresourceLoader::didCancel ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#17 0x0c7d3975 in WebCore::ResourceLoader::cancel ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#18 0x0c7d4962 in WebCore::ResourceLoader::cancel ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#19 0x0c7914a2 in WebCore::cancelAll ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#20 0x0c794082 in WebCore::DocumentLoader::stopLoading ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#21 0x0c7a4372 in WebCore::FrameLoader::stopAllLoaders ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#22 0x0c7aa1e1 in WebCore::FrameLoader::stopForUserCancel ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#23 0x0cb7d8f4 in webkit_web_view_stop_loading ()
   from /usr/local/lib/libwebkit-1.0.so.2.2
#24 0x0cb80a09 in webkit_web_view_dispose ()
---Type  to continue, or q  to quit---
   from /usr/local/lib/libwebkit-1.0.so.2.2
#25 0x0482a070 in g_object_run_dispose ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#26 0x0a496cfe in gtk_object_destroy ()
   from /usr/local/lib/libgtk-x11-2.0.so.1403.0
#27 0x1c003c60 in destroyclient ()
#28 0x04834f3f in g_cclosure_marshal_VOID__VOID ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#29 0x048279c9 in g_closure_invoke ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#30 0x0483c2b4 in g_signal_parse_name ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#31 0x0483e428 in g_signal_emit_valist ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
#32 0x0483e779 in g_signal_emit () from
/usr/local/lib/libgobject-2.0.so.1803.0
#33 0x0a496fe1 in gtk_object_destroy ()
   from /usr/local/lib/libgtk-x11-2.0.so.1403.0
#34 0x0a5a0104 in gtk_widget_hide ()
   from /usr/local/lib/libgtk-x11-2.0.so.1403.0
#35 0x0a5add36 in gtk_window_set_focus ()
   from /usr/local/lib/libgtk-x11-2.0.so.1403.0
#36 0x0482a070 in g_object_run_dispose ()
   from /usr/local/lib/libgobject-2.0.so.1803.0
---Type  to continue, or q  to quit---
#37 0x0a496cfe in gtk_object_de

Re: textproc/mupdf

[Predrag Punosevac]

I did quite a bit abuse on the latest mupdf package. It works like a
charm. Some issues I noticed on couple documents in the past are fixed.
Quick question for Stuart and other people who use it. 

Is there a way to configure auxiliary application which mupdf will use
as it is the case with Xpdf and .xpdfrc. I use video and http hyper
links in some presentations and I noticed that mupdf is looking for
Firefox. I want to force it to use XXXterm. I forgot what was happening
when I tried to open video hyperlinks because until now I had to use
Xpdf for those documents as there were not rendered properly with mupdf
for other reasons. 

Cheers,
Predrag

Re: NEW/UPDATE:net/sic-1.1

[Predrag Punosevac]

Matthias Kilian  wrote:

> Hi,
>
> On Sun, Aug 22, 2010 at 12:04:07PM -0400, Predrag Punosevac wrote:
> > sic is minimalistic irc client. This port has been submitted at least
> > four or five times in the past couple of years by various people but
> > never committed. The only new thing here is that sic has a new version
> > 1.1. I am resubmitting the port to create Internet archive. If somebody
> > cares to commit even better.
>
> - Don't patch the makefile or config.mk (unless you completely replace
>   it by something that works ;-) Instead, you can force some settings via
>   MAKE_FLAGS. For example, I did this:
>
>   V = .1
>   DISTNAME =  sic-$V
>   ...
>   CFLAGS +=   -I. -DVERSION=\\\"$V\\\" -D_GNU_SOURCE
>   MAKE_FLAGS ="CC=${CC}" \
>   "CFLAGS=${CFLAGS}" \
>   "LDFLAGS=${LDFLAGS}"
>
>   This will also allow to override CC (in case one wants to experiment
>   with a different compiler).
>
> - The suckless install target sucks (as always). Better write your own
>   do-install: that uses ${INSTALL_PROGRAM} resp. ${INSTALL_MAN},
>   which has the additional benefit that ${INSTALL_PROGRAM} automatically
>   strips the program unless DEBUG is defined. However, you also
>   have to get the version number into sic.1 (using a post-patch
>   target, for example).
>
> - The man page claims that the default port is 6667, but actually sic
>   initializes the port variable to "ircd" and passes that as servname
>   to getaddrinfo(3) later. This doesn't work unless you have a
>   matching entry in services(5).
>
> - Furthermore, if getaddrinfo(3) fails, sic uses strerror(3) instead of
>   gai_strerror(3) to produce an error message. This leads to totally
>   confusing error messages, like in:
>
>   $ sic -h 127.0.0.1
>   error: cannot resolve hostname '127.0.0.1': Invalid argument
>
> Ciao,
>   Kili

Great suggestions!!! I will try to get back to it in next two-three
days if for no other reason but play little bit more with Makefile which
is about as simple as it gets.  

I wrote the patch for config.mk almost two years ago when I was
originally learning about porting software from 

http://www.openbsd.org/papers/opencon07-portstutorial/index.html

I learnt quite a bit more about Make utility in the mean time and your
e-mail make so much sense. 

Cheers,
Predrag

Re: surf and netsurf core dumps

[Predrag Punosevac]

"Anthony J. Bentley"  wrote:

> > Netsurf crashes when I try to download a package from the OpenBSD
> > web-site. As you know netsurf doesn's support JavaScript so for all I
> > know even this crash might be "normal behavior".
>
> Netsurf 2.5 (what we have) is pretty unstable, and they weren't happy
> with it... the next version (still in SVN) should be better, as they've
> been concentrating mostly on bugfixes.
>
> What link were you using to download from? I'd like to send it upstream.
>

I was using 

http://ftp.openbsd.org/pub/OpenBSD/snapshots/packages/i386/

to download texlive_texmf-minimal because the pkg_add would time out for
some reason and I would end up only with the partial package. I changed
the mirror and everything is fine. I am hopping to give Landry backtrace
of crashes in next hour or so. I had no issues adding any other package
with pkg_add.


> > The stupid web-site which I use
> > for work 
> > 
> > https://pipeline.aug.edu/cp/home/displaylogin
>
> I can't help here, but as a student I can empathize. A lot of schools
> (including both of mine) use Sungard software, and it is terrible...

We are in the process of phasing it out and replacing it with "superior
Microsoft products like Exchange". Anyhow, I picked XXXterm for my my
main browser on the new installations . I have no idea how Marco does it
but it XXXterm works flawlessly with the crappiest sites. It loaded
above site without any warnings. If somebody could teach me how to print
a web-page or two from XXXterm, I would not keep Midori around.

I will not be playing with surf and netsurf much longer. I like idea
behind netsurf and it is much, much more usable than Dillo but
JavaScript is show stopper for me. 

Cheers,
Predrag

NEW/UPDATE:net/sic-1.1

[Predrag Punosevac]

Hello ports@

sic is minimalistic irc client. This port has been submitted at least
four or five times in the past couple of years by various people but
never committed. The only new thing here is that sic has a new version
1.1. I am resubmitting the port to create Internet archive. If somebody
cares to commit even better.

Cheers,
Predrag


sic-1.1.tgz
Description: Binary data

surf and netsurf core dumps

[Predrag Punosevac]

Dear All,

I would like to report core dumps of obscure web browsers surf and
netsurf.

Netsurf crashes when I try to download a package from the OpenBSD
web-site. As you know netsurf doesn's support JavaScript so for all I
know even this crash might be "normal behavior".


Surf (which is based on WebKit) crashes when I try to open link pipeline
on 

http://www.aug.edu/

Namely there is an annoying JavaScript which is supposed to report that
my browser is not supported. Surf goes into infinite loop on that
script. For the record Midori works fine so this is really problem with
the Surf itself.

Lastly, I would like to ask a question. Is there a way to mask Midori
like IE or Firefox (just like in Opera). The stupid web-site which I use
for work 

https://pipeline.aug.edu/cp/home/displaylogin

reports technical issues when I try to log with Midori. I get the same
nonsense when I use Opera which identify itself as Opera. However when I
mask it as Firefox or IE site works just fine.

I also included dmesg for the machine (as you can see it is the latest
snapshot). Packages were installed using pkg_add to the latest snapshot
packages. The date of the installation disk and packages is identical
08/21/2010.

Cheers,
Predrag



OpenBSD 4.8-current (GENERIC) #138: Fri Aug 20 23:24:58 MDT 2010
dera...@i386.openbsd.org:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel(R) Pentium(R) 4 CPU 1.80GHz ("GenuineIntel" 686-class) 1.80 GHz
cpu0: 
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM
real mem  = 1072754688 (1023MB)
avail mem = 1045245952 (996MB)
mainbus0 at root
bios0 at mainbus0: AT/286+ BIOS, date 03/01/02, BIOS32 rev. 0 @ 0xffe90, SMBIOS 
rev. 2.3 @ 0xf0450 (61 entries)
bios0: vendor Dell Computer Corporation version "A03" date 03/01/2002
bios0: Dell Computer Corporation OptiPlex GX240
acpi0 at bios0: rev 0
acpi0: sleep states S0 S1 S3 S4 S5
acpi0: tables DSDT FACP SSDT APIC BOOT
acpi0: wakeup devices VBTN(S4) PCI0(S5) USB0(S3) USB1(S3) PCI1(S5) KBD_(S3)
acpitimer0 at acpi0: 3579545 Hz, 24 bits
acpimadt0 at acpi0 addr 0xfee0: PC-AT compat
cpu0 at mainbus0: apid 0 (boot processor)
cpu0: apic clock running at 99MHz
ioapic0 at mainbus0: apid 1 pa 0xfec0, version 20, 24 pins
ioapic0: misconfigured as apic 0, remapped to apid 1
acpiprt0 at acpi0: bus 0 (PCI0)
acpiprt1 at acpi0: bus 2 (PCI1)
acpicpu0 at acpi0
acpibtn0 at acpi0: VBTN
bios0: ROM list: 0xc/0xc800 0xcc800/0x3800
pci0 at mainbus0 bus 0: configuration mode 1 (bios)
pchb0 at pci0 dev 0 function 0 "Intel 82845 Host" rev 0x03
intelagp0 at pchb0
agp0 at intelagp0: aperture at 0xe800, size 0x800
ppb0 at pci0 dev 1 function 0 "Intel 82845 AGP" rev 0x03
pci1 at ppb0 bus 1
vga1 at pci1 dev 0 function 0 "NVIDIA GeForce2 MX" rev 0xb2
wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
ppb1 at pci0 dev 30 function 0 "Intel 82801BA Hub-to-PCI" rev 0x12
pci2 at ppb1 bus 2
cbb0 at pci2 dev 7 function 0 "TI PCI1225 CardBus" rev 0x01: apic 1 int 16 (irq 
255), CardBus support disabled
cbb1 at pci2 dev 7 function 1 "TI PCI1225 CardBus" rev 0x01: apic 1 int 17 (irq 
255), CardBus support disabled
emu0 at pci2 dev 9 function 0 "Creative Labs SoundBlaster Live" rev 0x07: apic 
1 int 18 (irq 11)
ac97: codec id 0x83847609 (SigmaTel STAC9721/23)
ac97: codec features 18 bit DAC, 18 bit ADC, SigmaTel 3D
audio0 at emu0
"Creative Labs PCI Gameport Joystick" rev 0x07 at pci2 dev 9 function 1 not 
configured
xl0 at pci2 dev 10 function 0 "3Com 3c905B 100Base-TX" rev 0x30: apic 1 int 19 
(irq 11), address 00:50:da:1e:f5:ec
exphy0 at xl0 phy 24: 3Com internal media interface
xl1 at pci2 dev 12 function 0 "3Com 3c905C 100Base-TX" rev 0x78: apic 1 int 18 
(irq 11), address 00:06:5b:b0:ba:2b
exphy1 at xl1 phy 24: 3Com internal media interface
cardslot0 at cbb0 slot 0 flags 0
pcmcia0 at cardslot0
cardslot1 at cbb1 slot 1 flags 0
pcmcia1 at cardslot1
ichpcib0 at pci0 dev 31 function 0 "Intel 82801BA LPC" rev 0x12
pciide0 at pci0 dev 31 function 1 "Intel 82801BA IDE" rev 0x12: DMA, channel 0 
wired to compatibility, channel 1 wired to compatibility
wd0 at pciide0 channel 0 drive 0: 
wd0: 16-sector PIO, LBA48, 76319MB, 156301488 sectors
atapiscsi0 at pciide0 channel 0 drive 1
scsibus0 at atapiscsi0: 2 targets
sd0 at scsibus0 targ 0 lun 0:  ATAPI 0/direct removable
sd0: drive offline
wd0(pciide0:0:0): using PIO mode 4, Ultra-DMA mode 5
sd0(pciide0:0:1): using PIO mode 3, Ultra-DMA mode 2
atapiscsi1 at pciide0 channel 1 drive 0
scsibus1 at atapiscsi1: 2 targets
cd0 at scsibus1 targ 0 lun 0:  ATAPI 5/cdrom 
removable
wd1 at pciide0 channel 1 drive 1: 
wd1: 16-sector PIO, LBA48, 38165MB, 78163247 sectors
cd0(pciide0:1:0): using PIO mode 4, Ultra-DMA mode 2
wd1(pciide0:1:1): using PIO mode 4, Ultra-DMA mode 5
uhci0 at pci0 dev 31 function 2 "Intel 82801BA USB" rev 0x12: apic 1 int 19 
(irq 11)
ichiic0 at pci0 dev 31 function 3 "Intel 82801BA SMBus" r

Re: meta-packages questions

[Predrag Punosevac]

On Thu, Oct 29, 2009 at 12:38 PM, frantisek holop  wrote:
> anyway, i am fine with a "monolithic" texlive package, i am
> sure it was difficult to put it together as it is (good
> job Edd).

I have to admit first that due to the nature of my work I have to use
LaTeX and plain TeX. I know that ConTeXt is undergoing
rapid development and that if you ara involved in that process you might
have a need to use something on the cutting edge.

I have couple questions for you.

1. Is there are relatively painless way to updated just certain fonts
and components of TeXLive as it is the case with MikTeX. I know that 
the answer to that question used to be NO but that was before.

Even  if the answer is to previous question is NO you probably could get
a way with installing texlive-base, a few other dependences, and fonts 
that you use. You would then just hush TeX tree and you should be able 
to go.  You can check the minimal ConTeXt dependences from this Debian 
web-site

http://packages.debian.org/etch/texlive-context


Now look also this Wiki article from TeXLive web-page

http://wiki.contextgarden.net/TeX_Live

"TeXLive in principle offers no option to update between release cycles.
But of course, you will want to update ConTeXt as soon as Hans releases
a new version. This is easy to achieve: the installation process has 
created a skeleton directory /usr/local/texlive/texmf-local. It is 
empty right after installation. cd to this directory, download the 
latest cont-tmf.zip from the pragma website. Then unzip this archive 
and run

texexec --make --all

again. This will update your installation and the formats to the latest 
available version. When you want to update again later, simply repeat 
the process, overwriting all older files when unzipping."


2. Are you sure that you have to install TeXLive to begin with? I have
heard that lots of people who use ConTeXt have alternative method of 
installing ConTeXt to Unix.

I would check these documents 

http://wiki.contextgarden.net/Installation

and in particular 

http://wiki.contextgarden.net/ConTeXt_Minimals

Best,
Predrag

Re: NetBSD ports

[Predrag Punosevac]

> On Tue, Oct 06, 2009 at 12:12:34PM +0400, Alexander Bubnov wrote:
> I am sorry. Really, I don't want to have support here. I just would like to
> know relationship between pkgsrc and original OpenBSD port system.

There is NO relationship! NetBSD didn't have pkgsrc when Theo forked 
OpenBSD. OpenBSD ports system was originally inspired by FreeBSD ports.
Maybe Mark can recall what he was exactly thinking when he got idea for
OpenBSD packaging system.

>From the top of my head I am not sure if Alistair Crooks was inspired
by FreeBSD ports as well but I know that he created NetBSD pkgsrc 
based on experience he gained creating a packaging system for one
of proprietary Unix-es in mid nineteens. That was a paid job by the way.

All those information are freely available on the net. Just Google 
before rubbing people the wrong way.

Best,
Predrag

Re: NetBSD ports

[Predrag Punosevac]

On Tue, Oct 06, 2009 at 12:12:34PM +0400, Alexander Bubnov wrote:
> Hello everybody!
>
> Probably I am going to ask a stupid question but it is very interesting for
> me. Because I would like to help BSD projects.
> Why OpenBSD does not use pkgsrc of NetBSD project as default ports? I guess
> work can be faster in case port system is shared between BSD projects
> including FreeBSD. NetBSD ports are ported to many Oses so I would prefer
> these port system.


You may read about philosophical differences between pkgsrc and our
ports system from the interview
Mark Espie (the father of OpenBSD ports) gave for the tenth
anniversary of pkgsrc

http://www.netbsd.org/gallery/10years.html


I will give you couple more practical reasons.

I was able to bootstrap pkgsrc on 4.4 and consider myself one of the
few who in the recent past
have played with pkgsrc on OpenBSD.

It took some searching of NetBSD mailing lists for me to  bootstrap
pkgsrc on OpenBSD
due to a nasty bug observed by NetBSD developers.

Once bootstrap, I could not reliably build any application which
required X to run. The problem was
of course that NetBSD was still using XFree86. Now there was an option
which will allow you to use
native X server but that would not make port building any more
reliable. Things might have improved
a bit as they switched to XOrg.


pkgsrc to my knowledge has not been tested by a single pkgsrc
developer since 3.7 (according to NetBSD mailing lists).
It is kind a fanny that they expect pkgsrc just to work without any
testing. Don't you think so?

There are two common arguments used by people who in the past (2004
and earlier) have used or at least tried
to use pkgsrc on OpenBSD.

One is the number of ports (pkgsrc officially has more ported
software) and the second one is ability to tweak the
application.

The first claim doesn't hold the water. Just browse trough pkgsrc and
you will see that most of their software is
much older than our current port three. Often the pkgsrc has not been
updated for ages. Often large peaces of software
has been broken in many peaces giving you the impression that they
have more stuff.
Speaking of tweaking it is absolutely unnecessary for common users.
IMHO the one who need real tweaking will have
no problem to use OpenBSD ports system and to tweak the specific
application to their liking.


Speaking non-Wintel architectures I will wage you $100 that I will
build more software using our ports three on SGI, SUN,
or even VAX than you will be able to do that with pkgsrc.


I do not want to sound all negative. pkgsrc is a fine packaging system
but I would like to see a person who could reliably build software
with it outside of NetBSD and DragonFly platforms.  In all honestly it
is very portable but that doesn't mean that it can be used on OpenBSD
platform or any other platform  for that matter without serious
testing and bug hunting. You can trace the pain of DragonFly
developers from their mailing listand how much effort went into making
pkgsrc fully functional on DragonFly after they made the decision to
adopt pkgsrc as the official package management.

Best,
Predrag

Re: mplayer continued...

[Predrag Punosevac]

> On 2009/10/01 17:27, Edd Barrett wrote:
> > > didn't the discussion come out in favor of disabling win32-codecs?
> > 
> > If that is what you guys want, then so be it. I did not think a concrete
> > decision was ever met. Shall I remove the comment? Anything else need to
> > be done if we want to do this?

> yeah, the configure options need to be changed.

> > ah, well I suppose it doesn't need to be done now, we can always
> > do that later.

> istr espie was going to check his video archives.  this update was
> started months ago.  imo, it should wait until this gets sorted out,
> because otherwise it will never change.o, it should wait until this gets 
> sorted out,
> because otherwise it will never change.

I know that Jake and many other people are in favor of removing 
win32-codecs and that we are waiting for Mark's input. I personaly
trust Jake's judgment but for the curious I can produce a content which
MPlayer will not be able to handle without win32-codecs.

It is streaming live TV program which I usually watch  

rtsp://helix.beotel.net/encoder/rts1.rm

The MPlayer version from the current is able to play the sound but
will not display video at least on AMD64.

Edd, can you reproduce my claim on your latest version of MPlayer?

Best,
Predrag

Re: console imap client

[Predrag Punosevac]

Nail (Heirloom mailx) wins by a mile :-)

On Tue, Aug 25, 2009 at 3:21 AM, LEVAI Daniel wrote:
> Hi!
>
>
> Anyone could recommend me a ported console imap client, or if there isn't any
> ported, a simply working?
>
> Thanks in advance,
>
> Daniel
>
>
> --
> LÉVAI Dániel
> PGP key ID = 0x4AC0A4B1
> Key fingerprint = D037 03B9 C12D D338 4412  2D83 1373 917A 4AC0 A4B1
>
>

Re: Firefox 3.0.13 on 4.5 Stable

[Predrag Punosevac]

Donald Allen  wrote:

> You wrote: "Every time I've gotten a diff ready they find another security
> hole
> and do another release (why do people use this again?)..."
>
> I, too, have begun to wonder why Firefox is so popular, particularly since
> the major version number went to 3. Then I remind myself that the vast
> majority of the world's computers run Windows. So, putting such thoughts
> about the negative correlation between popularity and quality aside, do you
> have a suggestion for another web browser to run on 4.5 (amd64)? I don't
> believe Opera is open-source, it is not available as a package, and my
> experience with it is that it's got it's own set of problems. Epiphany has
> never struck me as a good alternative (and I don't run Gnome, though I do
> use other Gnome-based apps, such as Gnucash), but perhaps I should give it
> another look?
>
> /Don Allen
@Don,

Midori looks very promising but I am terrified by the long list of 
security issues with the WebKit. OpenBSD has up to date version of both 
WebKit and Midori. Other than that pretty much everything Gecko based 
will have more or less the same problems as Firefox. Konqueror uses 
KHTML which is quite mature. It does compile better on rare 
architectures than Firefox. Opera is as you know Presto based and if
nothing it has a very small market share which makes it unlike target.
I am curious myself what kind of a advice you will get from serious 
people. Yes, I am completely aware what Dillo2, Elinks, and Links2 
can do I some times use Links2 for graphical browsing. If I use text
browser I use Lynx which I think has by far the cleanest and most 
secure code. 

Best,
Predrag

@ Martynas and Steven
Thanks for your kind answers. If the Xenocara has been already bumped
for 4.7 that means that unless William or somebody else post diff to
4.6 people who do fresh installation of October 1st release will be
stuck with Firefox full of security problems. Am I correct?

I feel less stupid now for posting original question since that is 
very important information which will affect my approach to 4.6.

Re: Firefox 3.0.13 on 4.5 Stable

[Predrag Punosevac]

William Yodlowsky  wrote:

> On 23 August 2009 at 16:31, Predrag Punosevac  wrote:
>
> > Does anyone has diff for compiling Firefox 3.0.13 on 4.5 Stable.
> > Currently, I am running 3.0.12 which has multiple vulnerabilities.
> > I tried to edit version number in the Makefile  for 3.0.12 but 
> > apparently there was some other changes so the compilation fails after
> > about 30 minutes. 
>
> Firefox, and what it depends on, are being worked on.  I am trying to 
> work out the least invasive way of updating the libraries it depends
> on without forcing updates on more than a dozen other packages... there
> is more to it than just firefox.
>
> Every time I've gotten a diff ready they find another security hole
> and do another release (why do people use this again?)...

Personally, I have two reasons to use Firefox. The first is because
Opera doesn't run on bsd.mp kernel and the second is that Blackboard
CMS is so badly coded that Opera can not properly display the 
grade book with my students.

Now back to Steven's post. I actually spent last couple of hours 
playing with the current ports three on a machine running stable and
I have a very specific gestion for Steve. How did you solve this

Missing library for pthread-stubs
Missing library for xcb
Fatal error
*** Error code 1

It looks to me by looking over log files that a library has been
added to Xenocara between 4.5 and 4.6. Now I when I try to compile
lets say Firefox35 the compiler starts screaming. 

Steve could you be so kind and explain me what you did with it.

Cheers,
Predrag

P.S. I actually completely trashed this "stable" system which looks to
me like a good excuse to move to 4.6. 

Re: Firefox 3.0.13 on 4.5 Stable

[Predrag Punosevac]

"Steven Surdock"  wrote:

> > -Original Message-
> > From: Predrag Punosevac [mailto:punoseva...@gmail.com]
> > Sent: Sunday, August 23, 2009 10:12 PM
> > To: Steven Surdock; ports@openbsd.org
> > Subject: Re: Firefox 3.0.13 on 4.5 Stable
> > 
> > "Steven Surdock"  wrote:
> > 
> > > > -Original Message-
> > > > From: owner-po...@openbsd.org [mailto:owner-po...@openbsd.org] On
> > > Behalf Of
> > > > Martynas Venckus
> ...
> > > > > From owner-ports+m36...@openbsd.org Sun Aug 23 23:34:53 2009
> ...
> > >
> > > If one were feeling masochistic, wanted to run -stable and
> > > mozilla-firefox-3.0.13 they could try the -current ports.  But,
> since
> > > there might be other dependencies from other ports, would one run
> 'make
> > > show-required-by' to determine what other ports/packages would be
> > > affected by bumping up sqlite3, nspr & nss?  Is there a better way?
> > > Thanks!
> > >
> > > -Steve S.
> > That is absolutely not true! There has been great changes in the base
> > of the system and many libraries have been bumped. There is absolutely
> > 0% that you can get a way with it.
> ...
> > I have a port of Firefox 3.0.12 which is backported for 4.5 stable so
> > I tried to fiddle little bit with it without success. For simple
> > ports you often can get a way with it.
> > 
> > Most Kind Regards,
> > Predrag Punosevac
> > 
> > P.S. To all n00bs who read this. Please do not try this things and if
> > you try to not ask anybody for help since you have unsupported
> systems.
>
> I agree, don't ask for help.  I did say if one was feeling
> masochistic...
>
> ssurd...@builder02$ uname -a
> OpenBSD builder02.internal.engineered-net.com 4.5 GENERIC#3 i386
> ssurd...@builder02$ pkg_info | egrep 'mozill|nss|nspr|sqli'
> mozilla-firefox-3.0.13 redesign of Mozilla's browser component
> nspr-4.8Netscape Portable Runtime
> nss-3.12.3.1p0  libraries to support development of security-enabled
> apps
> sqlite-2.8.17p1 Embedded SQL implementation
> sqlite3-3.6.16  embedded SQL implementation
>
> Naturally, this is completely unsupported.
>
So actually you did get a way with it :-) I am having a hard time to
believe that you used vanilla versions from current ports tree on
4.5 stable? If that is not the case could you be so kind to share
above ports with another poor soul.

Most Kind Regards,
Predrag Punosevac

P.S. It is definitely worth at least a beer.

Re: Firefox 3.0.13 on 4.5 Stable

[Predrag Punosevac]

"Steven Surdock"  wrote:

> > -Original Message-
> > From: owner-po...@openbsd.org [mailto:owner-po...@openbsd.org] On
> Behalf Of
> > Martynas Venckus
> > Sent: Sunday, August 23, 2009 6:53 PM
> > To: ports@openbsd.org; punoseva...@gmail.com
> > Subject: Re: Firefox 3.0.13 on 4.5 Stable
> > 
> > > From owner-ports+m36...@openbsd.org Sun Aug 23 23:34:53 2009
> ...
> > 
> > To use Firefox 3.0.13 you will have to:
> > - Backport sqlite3->=3.6.13.
> > - Backport nspr-4.8.
> > - Backport nss-3.12.3.1.
> > - Backport mozilla-firefox-3.0.13.
> > 
> > Or you can avoid all the trouble and use current instead.
>
> If one were feeling masochistic, wanted to run -stable and
> mozilla-firefox-3.0.13 they could try the -current ports.  But, since
> there might be other dependencies from other ports, would one run 'make
> show-required-by' to determine what other ports/packages would be
> affected by bumping up sqlite3, nspr & nss?  Is there a better way?
> Thanks!
>
> -Steve S.
That is absolutely not true! There has been great changes in the base
of the system and many libraries have been bumped. There is absolutely 
0% that you can get a way with it.

However it is true (hopefully n00bs are not reading this) that you will
be able to build current port tree on 4.6 release until next major
library bump from the base of the system. For the 1st of October 
release that probably means early November at best. Look for log files
and kernel developer mailing lists.

I have a port of Firefox 3.0.12 which is backported for 4.5 stable so
I tried to fiddle little bit with it without success. For simple 
ports you often can get a way with it.

Most Kind Regards,
Predrag Punosevac

P.S. To all n00bs who read this. Please do not try this things and if
you try to not ask anybody for help since you have unsupported systems.

Re: Firefox 3.0.13 on 4.5 Stable

[Predrag Punosevac]

Martynas Venckus  wrote:

> > From owner-ports+m36...@openbsd.org Sun Aug 23 23:34:53 2009
> > Date: Sun, 23 Aug 2009 16:31:55 -0400
> > From: Predrag Punosevac 
> > To: ports@openbsd.org
> > Subject: Firefox 3.0.13 on 4.5 Stable
> >
> > Does anyone has diff for compiling Firefox 3.0.13 on 4.5 Stable.
> > Currently, I am running 3.0.12 which has multiple vulnerabilities.
> > I tried to edit version number in the Makefile  for 3.0.12 but
> > apparently there was some other changes so the compilation fails after
> > about 30 minutes.
>
> MFSA 2009-42 and MFSA 2009-43 were addressed with rebase of the new
> NSS package.
>
> To use Firefox 3.0.13 you will have to:
> - Backport sqlite3->=3.6.13.
> - Backport nspr-4.8.
> - Backport nss-3.12.3.1.
> - Backport mozilla-firefox-3.0.13.
>
> Or you can avoid all the trouble and use current instead.

Thank you so much for your kind and prompt answer. Since you are the 
OpenBSD gurry for Firefox could you be so kind and tell me your 
recommendation for 4.6. Do you advise people to switch to 3.52 or you 
are advise me to stick with 3.0.13 until 3.5 series becomes more 
stable and less buggy. 

Most Kind Regards,
Predrag

Firefox 3.0.13 on 4.5 Stable

[Predrag Punosevac]

Does anyone has diff for compiling Firefox 3.0.13 on 4.5 Stable.
Currently, I am running 3.0.12 which has multiple vulnerabilities.
I tried to edit version number in the Makefile  for 3.0.12 but 
apparently there was some other changes so the compilation fails after
about 30 minutes. 

Sorry for the noise,
Predrag

Re: New: misc/vifm

[Predrag Punosevac]

Landry Breuil  wrote:

> On Sat, Aug 08, 2009 at 04:22:08PM +0200, Tobias Ulmer wrote:
> > First try that doesn't crash on amd64 within seconds.
> > 
> > Removes the most evil things gcc warns about and moves malloc out of the
> > signal handler. There is without a doubt, much more to repair...
>
> For sure.. with your patches, it still explodes on sparc64 at the same point..
> volunteers for code reading & fixing & forwarding patches upstream needed :)
>
> Landry

I would like to bring to your attention two facts:

1. Somebody did try to port vifm to OpenBSD few years ago. There is 
the reason vifm is still not in ports.

2. Good luck with fixing things upstream. My fellow Tucsonian who 
coded vifm did that in order to learn programming language C :-)
He is retired these days so am not sure if he is interested in 
brushing up his programming skills.

Best,
Predrag

Re: opera10 experimental build with bundled qt4 (and some linux emul questions)

[Predrag Punosevac]

frantisek holop  wrote:


I am no expert on Linux compatibility layer but similar issue is
well documented with libasound.so.2 by people who were playing with
Skype on OpenBSD. The solution is to get the missing library from 
a Linux system preferably Fedora Core 4 as the rest of OpenBSD 
compatibility. That would allow you to run the Opera. I suspect that
missing library has something to do with ALSA but nobody is expecting
Flash 10 and sound to work anyway. 

There is port of a library needed for now obsolete Maple V to run which
was not the part of fedora-core. Mark ported it separately when 
he got Maple V to work. Current Maple 12 is a very long shot to work
on OpenBSD even without Java.



> alternatively, has anyone tried to run opera in freebsd emulation?
> (like netbsd does)
>

This is very interesting question. I would really like to hear of 
anybody who is actually using any FreeBSD binaries now.
My understanding that FreeBSD compatibility is for FreeBSD 4.xxx or 
at best FreeBSD 5.xxx so I am guessing that is the dead code. 



> i also see that netbsd has multiple emul platforms for opera (freebsd5,
> solaris-sparc, and linux i386).  would be perhaps easier following
> netbsd in having suse libs instead of fedora core?
>
I think if you look archive little bit you will see by the statements
of Nikoly Sturm and few other developers that Linux compatibility is
more or less dead code. NetBSD amd64 will also allow you to run i386
binaries unlike OpenBSD which is an honest fully 64 bit OS. You could
develop some kind a sandbox on amd64 to run i386 binaries including 
Linux 32 bit binaries. I am having very hard time to see how would
that be commensurate with the goals of OpenBSD project. I think, I came
across statements by developers who explicitly stated the same things.

I am also having very hard time to see advantage of suse libraries 
over fedora. Quite on the contrary from where I sit Linux compatibility 
for me means RedHat (Fedora) compatible. There are so many other million
times more interesting things to do develop on  OpenBSD than to chase 
Linux compatibility, Opera, and ALSA. 

Cheers,
Predrag

Re: xpdf core dump

[Predrag Punosevac]

Matthias Kilian  wrote:

> Did you check wether the pdf file you're using is the same that
> sthen@ mentioned and that doesn't cause crashes for him and for me?
> I just want to be sure that we're testing the same file.
>
> $ md5 bpu02659.pdf
> MD5 (bpu02659.pdf) = b68c0a47869ab4147c35460d88366c96
>

Yes, Kili it is that particular file. You have the right file. I
checked again including md5 sum. It crashes the latest xpdf on 4.5
stable i386 running bsd.mp kernel. 

$ pkg_info xpdf
Information for inst:xpdf-3.02.3

Best,
Predrag

WIP:print/min12xxw

[Predrag Punosevac]

Hi Ports,

I attached the WIP port of the printer driver min12xxw for the 
Konica/Minolta PagePro 1[234]xxW printers. To my knowledge this is 
the only open-source printer driver not currently ported to OpenBSD.
I do not own the hardware so I have no means of really testing the 
port. I did this as an exercise as I am learning more about ports 
infrastructure. I hope somebody actually find a use of it.

Best,
Predrag




min12xxw.tgz
Description: Binary data

Re: xpdf core dump

[Predrag Punosevac]

Matthias Kilian  wrote:

> Welcome to the wonderful world of N+1 slightly different versions
> and forks of xpdf (and poppler) ;-)
>
I think, I am just saying goodbye to xpdf after probably 20 years. I can
not even recall when I started using xpdf. I am now mupdf guy thanks
to Stuart and the gang!!!


> I don't have patches for xpdf. If you want to test the latest xpdf,
> just try the -current port (it should build cleanly even on 4.5).
>

I have tested the latest xpdf on 4.5 stable. I can reproduce the same
crash with the newest version. I am installing current at this very 
moment on a machine which will be used for testing purposes. I would
be surprised that if xpdf behaves differently on current. 


> But I've patches for texlife and kdegraphics that need testing:
>
> http://openbsd.dead-parrot.de/print_texlive_base.diff
> http://openbsd.dead-parrot.de/x11_kde_graphics3.diff
>
> Note that they don't contain Miod's patch to SplashXPath.cc, because
> texlive just doesn't contain the splash stuff, and kdegraphics
> contains a different fix (obviously taken from poppler).
>

Point taken. I will be testing TeXLive with your patches. I use TeXLive
no less than 3-4 hours every day. 


> > hmm, that url doesn't work (probably because of the session id), but
> > assuming this is the same document, it works on xpdf (amd64 -current) for
>
> I tried the -current port without Miod's diff, and even the OPENBSD_4_5
> version on amd64 and i386, and it didn't crash. This seems to be
> one of those nice bugs that are difficult to reproduce.
>

This is very interesting because I can reproduce the bag with the 
latest version compiled on 4.5 stable. Do you remember last month
your conversation with Matt when he was complaining about the crash
on the similar document? I just could not reproduce it on 4.4/amd64
stable running bsd.mp kernel.



> Ciao,
>   Kili 

Thanks Kili,

Predrag

Re: xpdf core dump

[Predrag Punosevac]

Brynet  wrote:

> Predrag wrote:
> > mupdf doesn't compile on stable
>
> Here are the patches you'll need on 4.5.
>
> -Brynet
Hi Brynet,

I can just say wow! I am just blown away. It took no more than 6 hours 
since I reported a problem with XPDF for OpenBSD community to suggest 
better alternative solution which doesn't use problematic library for
rendering and even backported it to stable. Your patches worked as charm.
I tested already mupdf on 4.5 stable and it rocks. I put mupdf to test 
on the same HP manual which crashed XPDF. No problems. It worked like 
a charm. Then I tested mupdf with some very complicated slides 
generated by Powerdot class of LaTeX presentations. Epdfview, and GGV 
have serious problems with that documents. Mupdf worked as charm. 
I can just say big thanks to Stuart who ported the mupdf to OpenBSD. 
Mupdf+pdfjam is all I personally need to deal with PDF documents.

Thanks,
Predrag

Re: xpdf core dump

[Predrag Punosevac]

William Yodlowsky  wrote:

>
> I have an xpdf package for 4.5-stable/i386 for xpdf-3.02.3 at:
>
> http://openbsd.rutgers.edu/xpdf-3.02.3.tgz
>
> SHA1 (xpdf-3.02.3.tgz) = 2de911e02efe00b71866245d091726aba30c087d
>
> This may become what goes into official -stable.  Does it work for your
> problem pdf?
>
> Thanks.

Hi William,

xpdf-3.02.3 depends on xpdf-utils-3.02.3 so I installed both packages
from your web-site. Unfortunately, new version of xpdf does not fix
the problem. xpdf is still dumping core on that HP manual full of
graphics By the way, if you are submitting xpdf to stable note that 
xfig-3.2.5p2 which is backported to stable due to the security reasons
requires xpdf to run so you would have to build  xfig-3.2.5p2 against
xpdf-3.02.3 and xpdf-utils-3.02.3. I had no other dependency issues
on my particular installation. 

Most Kind Regards,
Predrag 

Re: xpdf core dump

[Predrag Punosevac]

>in -current, pkg_add mupdf. the -current port of this will most likely
>build on 4.5 too if you want to try it.

Hi Stuart,

mupdf doesn't compile on stable 

cc -c -o /usr/obj/ports/mupdf-0.4/build-i386/ximage.o -Wall -std=c99 
-I/usr/X11R6/include/freetype2 -I/usr/X11R6/include -I/usr/local/include -O2 
-pipe  -DHAVE_JASPER -DNEED_STRLCPY -DNEED_STRSEP -DNEED_GETOPT -Iapps/unix 
-Ifitz -Imupdf -Iapps apps/unix/ximage.c

...skipped mupdf for lack of libmupdf.a...
*** Error code 1

I guess it has a problem to link to libmupdf.a library. It is probably
not too difficult to patch so that it can compile on stable. It looks 
very interesting. 

By the way I forgot to say in my previous message that I am running
bsd.mp kernel on 4.5/i386 stable. 

Somebody reported about month ago problem with xpdf on i386 and I
recall not being able to reproduce the problem on amd64.
I checked the same document with Acroread and renders without problems.
I am mentioning that because unlike Opera and few other Linux binaries,
Acroread works rock stable on bsd.mp kernel. I was very surprised. 

Cheers,
Predrag

xpdf core dump

[Predrag Punosevac]

Hi Ports,

I think that this is a well known issue but I would like to document it
anyway. I am getting xpdf core dumped when I try to see the following
document

 
http://h2.www2.hp.com/bizsupport/TechSupport/CoreRedirect.jsp?redirectReason=DocIndexPDF&prodSeriesId=30377&targetPage=http%3A%2F%2Fbizsupport1.austin.hp.com%2Fbc%2Fdocs%2Fsupport%2FSupportManual%2Fbpu02659%2Fbpu02659.pdf

The document is rendering without problems when I use epdfview.

I am running 4.5/i386 stable on this particular desktop.

Cheers,
Predrag

P.S. Kili, if you have those patches for xpdf could you please post them
on one place or send them to my private email. I will be testing this
week HPLIP so I would like to test xpdf as well. 

Re: opera 10 beta with unite

[Predrag Punosevac]

"Aaron W. Hsu"  wrote:

> On Fri, 26 Jun 2009 22:39:58 -0400, frantisek holop   
> wrote:
>
> > i wonder how many people use opera on openbsd...
>
> I wonder how many people would use it if there were a native build with a  
> proper license that would permit redistribution of the packages and such  
> on CD-ROM. Would these packages be distributed with the CDs?
>
>   Aaron W. Hsu
>
What are you talking about? Opera is a proprietary close source
software. It would take a lot more than change of license for  
Opera to be available on CD. It has to be Open Source project. 
My crystal ball is telling me Opera is not going open source any time 
soon.

I would be happy to sign your petition for the native build 
of Opera for OpenBSD. I do fell little bit claustrophobic having
only choice of Firefox and Konqueror. I hope by 1st of November Midori
gets better but WebKit is the most insecure of four search engines. 

I am also really curious how would native build work on bsd.mp kernel.
Did anybody try to use tiny-proxy+Opera on bsd.mp? Does tiny-proxy solve
the problem of locking? In my experience that seems to be the issue 
with bsd.mp and linux comp rather than with Opera per say. 

Is it possible to use native aspell with Opera? I didn't see the Linux
version of aspell available among ports but installing Linux version
of aspell shouldn't be too difficult.

Cheers,
Predrag

Re: javascript

[Predrag Punosevac]

igor denisov  wrote:

> Hello there,
>
> Tried:
> links -enable-javascript 1 www.some-site.com
> and it says no javascript.
> Read help and it sayed javascript is supported. Is it compiled with 
> javascript?
> --
> igor denisov.
>
> --
> E75. ??? ?? ? ?.
> http://www.nokia.ru/microsites/e75
Yes it is compiled with Javascript option.
Links have only partial implementation of Javascript. Some sites will 
report that you have no Javascript on Links due to the fact that they
can not identify browser. Google for Links 2 and Javascript to see
about options which must be set in links.conf to avoid that. 

That is really not OS issue. You shuld ask about that on Links mailing
lists. If you need full implementation of 
Javascript use Firefox, Opera, Konqueror, or Midori.

Cheers,
Predrag

Re: Fwd: new version of scrypt-1.1.4

[Predrag Punosevac]

Stuart Henderson  wrote:

> no time to look further now; build fails:
>

I did look at it. Maybe this can give you some clue. When you comment
out 

# SEPARATE_BUILD =  concurrent

it builds fine. I have to check the documentation what this 
separate build mean.

Cheers,
Predrag