11 Dec 2022 17:02, you wrote to me:

The only problem running SQPACK is that it always renumbers the
message base. Not acceptable with nntp clients. :(

Is that because sqpack doesn't have an option to not renumber,

Yes.

Is sqpack no longer maintained?

something about the Squish format itself?

Nope, Fastecho for example can purge and pack without renumbering.
Both JAM and Squish.

Ok. I suppose you don't know of any other (standalone) tool that purges Squish bases...

(Just curious - I currently use JAM and FMail so I don't need it.)

Carlos

--- GoldED+/W32-MSVC 1.1.5-b20180707
* Origin: cyberiada point (2:341/234.1)

* Originally in FIDOSOFT.HUSKY
* Crossposted in FMAIL_HELP

Hi Carlos.

Is sqpack no longer maintained?

I think it is but nobody seems to be interested in fixing the renumbering issue. :(

something about the Squish format itself?

Nope, Fastecho for example can purge and pack without renumbering.
Both JAM and Squish.

Ok. I suppose you don't know of any other (standalone) tool that purges Squish bases...

I believe not. :(

(Just curious - I currently use JAM and FMail so I don't need it.)

'ftools maint -D' seems to renumber JAM base here. How do you purge messages in JAM without renumbering?

'Tommi

--- FMail-lnx64 2.1.5.0-B20230106
* Origin: -------------------> (2:221/6.600)

Hello Tommi,

Friday January 27 2023, Tommi Koivula wrote to Carlos Navarro:

Is sqpack no longer maintained?

I think it is but nobody seems to be interested in fixing the
renumbering issue. :(

It is not a bug. It is a missing feature. That is why its priority is low.

Michael

... node (at) f1042 (dot) ru
--- GoldED+/LNX 1.1.5-b20180707
* Origin: Moscow, Russia (2:5020/1042)

Hello, Michael Dukelsky.
On 27/01/2023 16.37 you wrote:

Hello Tommi,
Friday January 27 2023, Tommi Koivula wrote to Carlos Navarro:

Is sqpack no longer maintained?

I think it is but nobody seems to be interested in fixing the
renumbering issue. :(

It is not a bug. It is a missing feature. That is why its priority is low.

Nobody said it is a bug. :)


--
Tommi
--- Hotdoged/2.13.5/Android
* Origin: . (2:221/360.8110)

Tommi wrote (2023-01-27):

(Just curious - I currently use JAM and FMail so I don't need it.)

'ftools maint -D' seems to renumber JAM base here. How do you purge messages in JAM without renumbering?

AFAIK JAM has no "unique message identifier" (UMSGID) like Squish. If you purge messages from the beginning, BaseMsgNum can be set:

This field determines the lowest message number in the index file.
The value for this field is one (1) when a message area is first
created. By using this field, a message area can be packed (deleted
messages are removed) without renumbering it. If BaseMsgNum contains
500, the first index record points to message number 500.

BaseMsgNum has to be taken into account when an application
calculates the next available message number (for creating new
messages) as well as the highest and lowest message number in a
message area.

If you delete messages in the middle you get holes without renumbering.

Solution: use Squish (as in the software).

---
* Origin: War is Peace. Freedom is Slavery. Ignorance is Strength. (2:280/464.47)

Hello, Oli!

Friday January 27 2023 16:55, from Oli -> Tommi Koivula

AFAIK JAM has no "unique message identifier" (UMSGID) like Squish.

You are right, JAM does not call it UMSGID but you can always have a unique message number by adding the basemsgnum value to otherwise 1-based message number. Well, unless you screw up the JAM database by running a broken purge tool which will reset basemsgnum back to 1.

The Husky SMAPI library supports the UMSGID semantic for JAM message base as well. They could have done it more efficiently without reading out the whole index file for renumbering. My min UMSGID would be the basemsgnum. My max UMSGID would be the size of the index file divided by 8 plus the basemsgnum.

If you delete messages in the middle you get holes without
renumbering.

Correct and this is expected behavior.

A similar problem exists in the Usenet world and NNTP protocol (RFC3977). NNTP has to maintain unique message numbers, so the client can save lastread numbers locally, for example .newrc is the most popular file format for that.

NNTP server will report the total number of active messages in the group (similar to "activemsgs" in JAM and "num_msgs" in Squish) and the low and high water mark. What would happen when the message in the middle is deleted? The number of active messages will be decremented but lowest and highest number will stay the same. Deleting from the beginning can be done my increasing the lowest number.

Solution: use Squish (as in the software).

Or write/fix the Jam purge tool which will be purging messages from the beginning and adjusting basemsgnum accordingly but keeping the holes in the middle which is by design there. Squish has introduced doubly linked lists, which is a different design solution for removing messages in the middle. Squish has a drawback though, the fixed array of 9 replies messages in the thread.

Best Regards, Nil
--- GoldED+/LNX 1.1.5
* Origin: -=NIL BBS=- (1:16/101)

'ftools maint -D' seems to renumber JAM base here.

Same here. Wrong assumption.

How do you purge messages in JAM without renumbering?

I don't. Sorry... O:-)

Carlos

--- GoldED+/W32-MSVC 1.1.5-b20180707
* Origin: cyberiada point (2:341/234.1)

Hi Nil,

On 2023-01-27 11:08:46, you wrote to Oli:

Squish has a drawback though, the fixed array of 9 replies messages in
the thread.

And it removes the least significant bit of stored dates. So message dates only have even seconds, and thus sometimes messing with dupe checking on other systems, when forwarded messages have been stored in a squish message base before being forwarded to the links of a system. Btw: Hudson has the same problem.

Bye, Wilfred.

--- FMail-lnx64 2.1.5.2-B20230114
* Origin: FMail development HQ (2:280/464)

Wilfred wrote (2023-01-27):

Hi Nil,

On 2023-01-27 11:08:46, you wrote to Oli:

Squish has a drawback though, the fixed array of 9 replies messages
in the thread.

And it removes the least significant bit of stored dates. So message dates only have even seconds, and thus sometimes messing with dupe checking on other systems, when forwarded messages have been stored in a squish message base before being forwarded to the links of a system. Btw: Hudson has the same problem.

You keep repeating that myth. But it's in fact a problem with Husky's Franken-JAM-Squish-SMAPI and a JAM message base rescan / export. Squish bases do store the original header date string (__ftsc_date) which is used on rescans. That is part of the Squish format specification.

The problem is that SMAPI uses the 2-second DOS date format, but there is no __ftsc_date for JAM bases that could be used. IIRC the correct datetime is stored in the JAM base by hpt / SMAPI, it just loses the least significant bit on export.

Squish doesn't have this problem. I tested it with Squish and hpt, no dates were modified. Maybe there is some software that does it wrong, but in all the discussion nobody could name one.

This is also not a problem on pass-through tossing.

---
* Origin: War is Peace. Freedom is Slavery. Ignorance is Strength. (2:280/464.47)

27 Jan 2023 10:48, you wrote to me:

How do you purge messages in JAM without renumbering?

CrashMail II's 'crashmaint PACK' seems to work fine.

Carlos

--- GoldED+/W32-MSVC 1.1.5-b20180707
* Origin: cyberiada (2:341/234.1)

Hi Carlos.

How do you purge messages in JAM without renumbering?

CrashMail II's 'crashmaint PACK' seems to work fine.

Thanks, I'll check it some day.

Does it work ok in Linux/64bit? Jamnntpd does not, and they are related...

'Tommi

---
* Origin: rbb.fidonet.fi (2:221/360)

26 Apr 2023 18:38, you wrote to me:

How do you purge messages in JAM without renumbering?

CrashMail II's 'crashmaint PACK' seems to work fine.

Thanks, I'll check it some day.

Does it work ok in Linux/64bit? Jamnntpd does not, and they are
related...

No idea, sorry. I checked with a Win32 version.

Carlos

--- GoldED+/W32-MSVC 1.1.5-b20180707
* Origin: cyberiada (2:341/234.1)

Tommi wrote (2023-04-26):

Hi Carlos.

How do you purge messages in JAM without renumbering?

CrashMail II's 'crashmaint PACK' seems to work fine.

Thanks, I'll check it some day.

Does it work ok in Linux/64bit? Jamnntpd does not, and they are related...

I haven't tested it, but I believe so. Both are available for Debian in 64-bit. Do we have to file a bug report?

* Origin: This site requires JavaScript (2:280/464.47)

Hi Oli.

Does it work ok in Linux/64bit? Jamnntpd does not, and they are related...

I haven't tested it, but I believe so.

Not good enough for me. Not a matter of religion. :)

Both are available for Debian in 64-bit. Do we have to file a bug
report?

In this debian 64bit I tried "sudo apt install crashmail", then created a very basic crashmail.prefs with one area

area rbb.test.9 2:221/6 jam /bbs/piste/msgbase/rbb.test.9
keepnum 99
keepdays 99

crashmaint maint verbose
Processing rbb.test.9...
0 messages deleted by number, 4278124318 messages left
0 messages deleted by date, 4278124318 messages left

And the jam base is now broken. Hpt cannot work with it, Golded cannot work with it. :(

So yes, we have to file a bug report. Just test by yourself and go ahead.

'Tommi

--- FMail-lnx64 2.2.0.0
* Origin: ------------------------> (2:221/6.600)

* Originally in fidosoft.husky
* Crossposted in jamnntpd

Hi Oli.

Both are available for Debian in 64-bit. Do we have to file a bug
report?

In this debian 64bit I tried "sudo apt install crashmail", then created a very
basic crashmail.prefs with one area

area rbb.test.9 2:221/6 jam /bbs/piste/msgbase/rbb.test.9
keepnum 99
keepdays 99

crashmaint maint verbose
Processing rbb.test.9...
0 messages deleted by number, 4278124318 messages left
0 messages deleted by date, 4278124318 messages left

And the jam base is now broken. Hpt cannot work with it, Golded cannot work
with it. :(

So yes, we have to file a bug report. Just test by yourself and go ahead.

I repeated the same procedure in this Raspian OS 32bit RPI 3B+ system. Works very nice. So 64bit is to blame.

'Tommi

--- GoldED+/LNX 1.1.5-b20220409
* Origin: Linux 6.1.21-v7+ armv7l (2:221/360)

Hello Tommi,

On Sat Apr 29 2023 19:19:30, you wrote to Oli:

Both are available for Debian in 64-bit. Do we have to file a bug
report?

In this debian 64bit I tried "sudo apt install crashmail", then
created a very basic crashmail.prefs with one area

area rbb.test.9 2:221/6 jam /bbs/piste/msgbase/rbb.test.9
keepnum 99
keepdays 99

crashmaint maint verbose
Processing rbb.test.9...
0 messages deleted by number, 4278124318 messages left
0 messages deleted by date, 4278124318 messages left

And the jam base is now broken. Hpt cannot work with it, Golded cannot work with it. :(

So yes, we have to file a bug report. Just test by yourself and go
ahead.

If I'm not mistaken, the crashmail debian stuff was maintained by RJ Clay (as was most FTN related debian packages), who hasn't been around in at *least* 2 years, and a small search does show someone with his name unfortunately passed away on October 28, 2021. The bug report may not go anywhere for you if he was the only person with read/write access. You may want to try one of the forks (or create your own) and proceed from there. Just an FYI.

Regards,
Nick

... "Take my advice, I don't use it anyway."
--- GoldED+/LNX 1.1.5-b20230205
* Origin: thePharcyde_ distribution system (Wisconsin) (1:154/10)

Hi Nicholas.

30 Apr 23 09:35, you wrote to me:

And the jam base is now broken. Hpt cannot work with it, Golded cannot
work with it. :(

So yes, we have to file a bug report. Just test by yourself and go
ahead.

If I'm not mistaken, the crashmail debian stuff was maintained by RJ Clay (as
was most FTN related debian packages), who hasn't been around in at *least* 2
years, and a small search does show someone with his name unfortunately passed
away on October 28, 2021. The bug report may not go anywhere for you if he was
the only person with read/write access. You may want to try one of the forks
(or create your own) and proceed from there. Just an FYI.

Too bad there is broken software in debian repositories...

I think I stay in SMAPI things and JamNNTPd/32bit (and Crashmaint/32bit maybe if really needed).

'Tommi

--- GoldED+/LNX 1.1.5-b20220409
* Origin: Linux 6.1.21-v7+ armv7l (2:221/360)

Nicholas wrote (2023-04-30):

And the jam base is now broken. Hpt cannot work with it, Golded
cannot work with it. :(

So yes, we have to file a bug report. Just test by yourself and go
ahead.

If I'm not mistaken, the crashmail debian stuff was maintained by RJ Clay (as was most FTN related debian packages),

Yes, he was the maintainer.

who hasn't been around in at *least* 2 years,

There are also no github commits after 2019.

and a small search does show someone with his name
unfortunately passed away on October 28, 2021.

which might also be another person.

The bug report may not go anywhere for you if he was the only person
with read/write access.

The bug reports are public. Even if nobody will fix it, we have at least a documentation of the problem. The package is orphaned an maintainer is "Debian QA Group".

https://tracker.debian.org/pkg/crashmail https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=982204

You may want to try one of the forks (or create
your own) and proceed from there. Just an FYI.

I think the repository from RJ Clay is the most recent one (but last time I checked as last year, IIRC).

---
* Origin: This site requires JavaScript (2:280/464.47)

Tommi wrote (2023-04-29):

Hi Oli.

Does it work ok in Linux/64bit? Jamnntpd does not, and they are
related...

I haven't tested it, but I believe so.

Not good enough for me. Not a matter of religion. :)

;)

Both are available for Debian in 64-bit. Do we have to file a bug
report?

In this debian 64bit I tried "sudo apt install crashmail", then created a very basic crashmail.prefs with one area

area rbb.test.9 2:221/6 jam /bbs/piste/msgbase/rbb.test.9
keepnum 99
keepdays 99

Had the area been created by crashmail or another tosser?

crashmaint maint verbose
Processing rbb.test.9...
0 messages deleted by number, 4278124318 messages left
0 messages deleted by date, 4278124318 messages left

And the jam base is now broken. Hpt cannot work with it, Golded cannot work with it. :(

Did you try to compare the .j?? files from before and after running crashmaint?

Did you try to run crashmaint a second time over the broken JAM base?

---
* Origin: This site requires JavaScript (2:280/464.47)