Hello, All.

I'm having a really, really weird problem with Squish and OS/2. Squish/2 will run fine for a while, then suddenly, I'll get this:

+ 27 Dec 00:20:38 SQSH Begin, Squish/386 v1.11 (mem=-1443808K - main=00020010) # 27 Dec 00:20:38 SQSH Scanning messages
# 27 Dec 00:21:03 SQSH Packing from D:\Msgs\Private\Netmail (42 msgs)
: 27 Dec 00:21:04 SQSH Packed=0
27 Dec 00:21:04 SQSH ZIPing mail for 1:140/1 (4941 bytes)
! 27 Dec 00:21:04 SQSH Archiver error: Can't run archiver (-1)
! 27 Dec 00:21:05 SQSH Archiver command was: "d:\arc\zip -9Dj d:\bt\out\ff8600c7.mo0 d:\bt\out.sq\1cf9bc00.pkt"
! 27 Dec 00:21:05 SQSH Packet d:\bt\out.sq\1cf9bc00.pkt moved back to d:\bt\out.sq\1cf9bc10.out
27 Dec 00:21:05 SQSH ZIPing mail for 63:63/1 (1354 bytes)
! 27 Dec 00:21:05 SQSH Archiver error: Can't run archiver (-1)
! 27 Dec 00:21:05 SQSH Archiver command was: "d:\arc\zip -9Dj d:\bt\out.03f\ffd300c7.mo0 d:\bt\out.sq\1cf9bc11.pkt"
! 27 Dec 00:21:05 SQSH Packet d:\bt\out.sq\1cf9bc11.pkt moved back to d:\bt\out.sq\1cf9bc12.out
+ 27 Dec 00:21:05 SQSH End. Toss=0 (0/s), sent=0 (0/s), mem=-1443924K

Here's the weird part:

-If I do this using GoldEd+/EMX at the command prompt (using a batch file), this error occurs.

-If I use FleetStreet, which is PM (GUI) based, this doesn't happen. I use a REXX script to do this under FleetStreet.

-If I use Squish/386, there are no errors at all.

I am wondering if there's a problem with OS/2 running out of handles (I've seen
it before on here, I think; I know it's rare, but something is making me think about that specific problem).

Any ideas or suggestions? Any help would be appreciated.

Thanks,
Sean

// outpostsysop@earthlink.net | http/ftp/telnet: outpostbbs.us
--- GoldED+/EMX 1.1.5-21020
* Origin: Stranded at the Outpost... (1:18/200)

Hi Sean,

Well I can't really comment too much other than what you probaby already know...

+ 27 Dec 00:20:38 SQSH Begin, Squish/386 v1.11

BTW, WHICH executable are you using SQ386 or Squish/2? BOTH list as /386 in the
log and that can makes things confusing.

+ 27 Dec 00:20:38 SQSH Begin, Squish/386 v1.11 (mem=-

This says that Squish is running out of memory... in a big way (a negative quantity???). Something else on your system is doing this, so its probably not a Squish issue. One thing with Squish/386, I found that using "Buffers Medium" resulted in errors, so I have only been using Squish/2 and "Buffers Large" since very early on in Squish 1.11 days. This was even a Scott Dudley recommendation, given the known issues with the /386 specific environment...

! 27 Dec 00:21:05 SQSH Archiver error: Can't run archiver (-1)

Its unable to get the memory it needs to run, so it confirms the above situation)...

-If I do this using GoldEd+/EMX at the command prompt
(using a batch file), this error occurs.

Are you using the EMX handles bug fix? Namely the line in config.sys -

SET EMXOPT=-h256 -!0x800

EMX is not perfect, it does have issues at times, I have moved away from using EMX wherever possible because of these memory usage issues I keep seeing.
The worst one was BinkD/2, the EMX version would run for between 22 - 72 hours before refusing to run BinkD again until a reboot. I now use the non-EMX version of BinkD/2 and it runs for months at a time...

I am wondering if there's a problem with OS/2 running
out of handles)

Its more like EMX consuming the handles and not releasing them to me.

I hope this helps................pk.
--- Maximus/2 3.01
* Origin: Another Good Point About OS/2 (3:772/1.10)

Hello, Peter.

On 27 Dec 04 at 21:18, Peter Knapper wrote to Sean Dennis:

BTW, WHICH executable are you using SQ386 or Squish/2? BOTH list as
/386 in the log and that can makes things confusing.

Squish/2.

using Squish/2 and "Buffers Large" since very early on in Squish 1.11

I have always used that too ever since I switched to OS/2.

Are you using the EMX handles bug fix? Namely the line in config.sys

I did set that up and will try it out.

reboot. I now use the non-EMX version of BinkD/2 and it runs for
months at a time...

I may back down to GoldEd 3.0.1 (the last true OS/2 native version).

Its more like EMX consuming the handles and not releasing them to me.

That makes perfect sense though because if you noticed what I'd said, it didn't
happen when I used FleetStreet and the DOS/386 version of Squish.

Once again, thanks... hopefully this will solve the problems!

Later,
Sean

// outpostsysop@earthlink.net | http/ftp/telnet: outpostbbs.us
--- GoldED+/EMX 1.1.5-21020
* Origin: Stranded at the Outpost... (1:18/200)

Sean Dennis wrote in a message to Peter Knapper:

That makes perfect sense though because if you noticed what I'd
said, it didn't happen when I used FleetStreet and the DOS/386
version of Squish.

Well id you try going back to the DOS version of Squish?
--- timEd 1.10
* Origin: Buffalo Online * Lancaster NY * buffalo.sytes.net (1:3407/4)

*** Quoting Robert Wolfe from a message to Sean Dennis ***

Well id you try going back to the DOS version of Squish?

Yes, but it's SLOW to run through everything. I'm working on hunting down this
problem.

Later,
Sean

--- Telegard/2 v3.09.g2-sp4/mL
* Origin: Outpost BBS - outpostbbs.us (1:18/200)

Sean Dennis wrote in a message to Robert Wolfe:

*** Quoting Robert Wolfe from a message to Sean Dennis ***

Well id you try going back to the DOS version of Squish?

Yes, but it's SLOW to run through everything. I'm working on
hunting down this problem.

Ah, ok. FWIW, I trust you've visited the new Maximus home page? :)

--- timEd 1.10+
* Origin: Buffalo Online * Lancaster NY * buffalo.sytes.net (1:3407/4)

*** Quoting Robert Wolfe from a message to Sean Dennis ***

Yes, but it's SLOW to run through everything. I'm working on
hunting down this problem.

I fixed it thanks to Herbert Rosenau giving me something to add to my CONFIG.SYS. I have to say that I haven't visited your new site, but it's not for the lack of wanting to-it's for the need of remembering. :)

My board is running well, but I have to tweak OS/2 a bit-seems I over tweaked it and it's causing the board to run "jerky" when someone logs on... even over the LAN.

Later,
Sean

--- Telegard/2 v3.09.g2-sp4/mL
* Origin: Outpost BBS - outpostbbs.us (1:18/200)

Sean Dennis wrote in a message to Robert Wolfe:

Yes, but it's SLOW to run through everything. I'm working on
hunting down this problem.

Ah, understand.

I fixed it thanks to Herbert Rosenau giving me something to add to
my CONFIG.SYS. I have to say that I haven't visited your new site,
but it's not for the lack of wanting to-it's for the need of
remembering. :)

Heh, http://www.maximusbbs.net ;)

My board is running well, but I have to tweak OS/2 a bit-seems I
over tweaked it and it's causing the board to run "jerky" when
someone logs on... even over the LAN.

Yeah, I have that same problem with my Max and Max/NT setup here with ComIP 3.1
running under NT 4 Enterprise Server (with all available updates, of course).

--- timEd 1.10+
* Origin: Buffalo Online * Lancaster NY * buffalo.sytes.net (1:3407/4)

*** Quoting Robert Wolfe from a message to Sean Dennis ***

Yeah, I have that same problem with my Max and Max/NT setup here with ComIP 3.1 running under NT 4 Enterprise Server (with all available updates, of course).

I just reinstalled SIO. Somehow, something got broken somewhere but now it's fine. I also switched external protocol drivers and it's all working now, thankfully.

Later,
Sean

--- Telegard/2 v3.09.g2-sp4/mL
* Origin: Outpost BBS - outpostbbs.us (1:18/200)

Sean Dennis wrote in a message to Robert Wolfe:

I just reinstalled SIO. Somehow, something got broken somewhere
but now it's fine. I also switched external protocol drivers and
it's all working now, thankfully.

Very cool. :)

Peace,
Robert
--- timEd 1.10+
* Origin: Buffalo Online * Lancaster NY * buffalo.sytes.net (1:3407/4)