KenDB3 wrote to All <=-
Don't get me wrong, I'm pretty impressed and excited about V4, I just
have a lot of work to do in order to translate some of my current
custom pages over to new formatting, etc. And I would rather have some
of that work done before doing upgrades.
just have a lot of work to do in order to translate some of my
current custom pages over to new formatting, etc. And I would rather
I'm in the same boat and running ecweb v3 just fine on a new (ish)
build of 3.17. Work fine here, with a little bit of tweaking.
The move from v3 to v4, particularly when it comes to custom pages, should be fairly painless. If you ever do feel like making the switch (which I strongly encourage) I'm around to help get things working / looking the way you want them to.
I was getting set up to do some upgrades, but I realized that over on VERT messages stopped being displayed on the old Web UI after Sat Nov 12 2016. I noticed that where you choose a particular sub, it shows dates for when it last got a message, like Synchronet Discussion shows message should be there from today.
(ie, when looking here: http://vert.synchro.net/msgs/subs.ssjs?msg_grp=DOVE-Net)
So, I guess my question is, once I upgrade to the new builds, does that mean the only way to support messages via the Web interface is to move to the Web-V4 UI being worked on be echicken?
I was getting set up to do some upgrades, but I realized that over
on VERT messages stopped being displayed on the old Web UI after Sat
Nov 12 2016. I noticed that where you choose a particular sub, it
shows dates for when it last got a message, like Synchronet
Discussion shows message should be there from today.
(ie, when looking here:
http://vert.synchro.net/msgs/subs.ssjs?msg_grp=DOVE-Net)
Hmm, you're right. In the list of DOVE-Net sub-boards, the Latest column says that the most recent message in Synchronet Discussion was posted on January 12, 2017. If you click through to view messages in that sub, the newest message is from November 12, 2016.
This could be a bug in the old web UI. Based on the timing, I would hazard a guess that it may not be interacting well with vote/poll messages or something along those lines.
So, I guess my question is, once I upgrade to the new builds, does
that mean the only way to support messages via the Web interface is
to move to the Web-V4 UI being worked on be echicken?
Not necessarily. I don't think there's been any statement that the stock web UI will no longer be maintained. However, updating to 3.17a will likely mean breaking that part of your web interface until somebody fixes that bug.
I'll take a quick look at it later, but am likely to give up if it's not obvious / quick / easy to fix.
I was getting set up to do some upgrades, but I realized that
over on VERT messages stopped being displayed on the old Web UI
after Sat Nov 12 2016. I noticed that where you choose a
particular sub, it shows dates for when it last got a message,
like Synchronet Discussion shows message should be there from
today. (ie, when looking here:
http://vert.synchro.net/msgs/subs.ssjs?msg_grp=DOVE-Net)
Hmm, you're right. In the list of DOVE-Net sub-boards, the Latest
column says that the most recent message in Synchronet Discussion was posted on January 12, 2017. If you click through to view messages in
that sub, the newest message is from November 12, 2016.
This could be a bug in the old web UI. Based on the timing, I would hazard a guess that it may not be interacting well with vote/poll
messages or something along those lines.
Hi there folks,
I was getting set up to do some upgrades, but I realized that over on VERT messages stopped being displayed on the old Web UI after Sat Nov 12 2016. I noticed that where you choose a particular sub, it shows dates for when it last got a message, like Synchronet Discussion shows message should be there from today.
(ie, when looking here: http://vert.synchro.net/msgs/subs.ssjs?msg_grp=DOVE-Net)
So, I guess my question is, once I upgrade to the new builds, does that mean the only way to support messages via the Web interface is to move to the Web-V4 UI being worked on be echicken?
Don't get me wrong, I'm pretty impressed and excited about V4, I just have a lot of work to do in order to translate some of my current custom pages over to new formatting, etc. And I would rather have some of that work done before doing upgrades.
Hello echicken,
On Thu Jan 12 2017 13:10:08, echicken wrote to KenDB3:
I was getting set up to do some upgrades, but I realized that
over on VERT messages stopped being displayed on the old Web UI
after Sat Nov 12 2016. I noticed that where you choose a
particular sub, it shows dates for when it last got a message,
like Synchronet Discussion shows message should be there from
today. (ie, when looking here:
http://vert.synchro.net/msgs/subs.ssjs?msg_grp=DOVE-Net)
Hmm, you're right. In the list of DOVE-Net sub-boards, the Latest column says that the most recent message in Synchronet Discussion was posted on January 12, 2017. If you click through to view messages in that sub, the newest message is from November 12, 2016.
This could be a bug in the old web UI. Based on the timing, I would hazard a guess that it may not be interacting well with vote/poll messages or something along those lines.
Do you think this could also have something to do with the fact that me and one
other seem to be having some issues regarding the last read pointer in random sub-boards resetting every time you login?
It only seems to happen in random sub-boards, but I will go to/read the last message in the sub-board, and the next time I telnet in (it might be the next day, I don't remember), it will show me messages #1 again. I believe the only executable I'm running these days is SMBUTIL in a nightly event, or sbbsecho to
import/export FTN related stuff.
Could it possibly be an issue while using "Gate between net types" and having sbbsecho export Dovenet messages to my other system? Just trying to think of all possible scenarios here, and I'm not sure what the other guy that experienced this does in relation to what I'm doing, but..
Thanks man. I thought the dates looked coincidental.
I don't think so.
It only seems to happen in random sub-boards, but I will go to/read
the last message in the sub-board, and the next time I telnet in
(it might be the next day, I don't remember), it will show me
messages #1 again. I believe the only executable I'm running these
days is SMBUTIL in a nightly event, or sbbsecho to import/export
FTN related stuff.
Neither of those programs deal with user message/scan pointers.
Could it possibly be an issue while using "Gate between net types"
and having sbbsecho export Dovenet messages to my other system?
Just trying to think of all possible scenarios here, and I'm not
sure what the other guy that experienced this does in relation to
what I'm doing, but..
I don't think so.
Are you using a web UI that does stuff with msg/scan pointers (reads
and updates them)?
Are you using a web UI that does stuff with msg/scan pointers (reads and updates them)?
I'm using the latest ecwebv4, but I'm fairly certain I only posted one message to a test area awhile back. So any of the other sub-boards I'm seeing this in (DOVE-GEN being one of them, along with random ones in other networks as well), shouldn't have been affected.
Not necessarily. I don't think there's been any statement that the stock web UI will no longer be maintained. However, updating to 3.17a will likely mean breaking that part of your web interface until somebody fixes that bug.
Somehow I had the impression that Rob & Deuce wanted to replace the stock/runemaster web interface with your ecweb interface (and thus the Runemaster interface would not be maintained after some point). I think it
One of the main problems with replacing the stock web UI with my
latest one is feature parity. I don't know every special feature that
was added to the Runemaster interface over the years, so I can't say
that mine does everything that the old one did. Some of those unknown features were requested by people for reasons, so it wouldn't be fair
to make them disappear.
If that looks incorrect to you, please let me know. I do browse messages on my website fairly often now, and I'm not having problems like these, so I dunno.
I've been using V4 extensively and have only had issue with the POLLS posting incorrectly, never any pointer issues.
I'm using the latest ecwebv4, but I'm fairly certain I only
posted one message to a test area awhile back. So any of the
other sub-boards I'm seeing this in (DOVE-GEN being one of them,
along with random ones in other networks as well), shouldn't have
been affected.
For the record, the scan pointer can be updated when you're viewing messages, not just when you post. However, it should only happen when
you load a thread in which the newest message in that thread has a
number higher than your current scan pointer for that sub.
I was actually reluctant to add this feature, because it makes it a
bit too easy to miss new messages. I will however eventually add a
'new message scan' which should come in handy.
Are you using a web UI that does stuff with msg/scan pointers (reads and updates them)?
I'm using the latest ecwebv4, but I'm fairly certain I only posted one message to a test area awhile back. So any of the other sub-boards I'm seeing this in (DOVE-GEN being one of them, along with random ones in other networks as well),
shouldn't have been affected.
Re: Question about Voting and the old Web UI
By: Digital Man to Accession on Thu Jan 12 2017 12:48:12
Are you using a web UI that does stuff with msg/scan pointers (reads and updates them)?
FWIW I know that he's using my latest, and it does. However, when I saw mention of this the other day I went to double-check the one part of it that modifies the scan pointer. Perhaps I'm doing something wrong, but it looks fine to me.
Essentially, it's:
if (newest_message_header.number > msg_area.sub[sub].scan_ptr) {
msg_area.sub[sub].scan_ptr = newest_message_header.number;
}
(Additionally, it does not modify the 'last_read' property.)
If that looks incorrect to you, please let me know. I do browse messages on my website fairly often now, and I'm not having problems like these, so I dunno.
Not necessarily. I don't think there's been any statement that the stock web UI will no longer be maintained. However, updating to 3.17a will likely mean breaking that part of your web interface until somebody fixes that bug.
Somehow I had the impression that Rob & Deuce wanted to replace the stock/runemaster web interface with your ecweb interface (and thus the Runemaster interface would not be maintained after some point). I think it may have been in some messages on Dove-net a while back, but I don't remember exactly now.
This should carry the same throughout the web and the telnet BBS though, right? Also, only when logged in as a user. Either way, I don't think
ecwebv4 is the issue here at this point. Some sub-boards throw me back to the first message in a newscan on a daily basis, if not more than that. The last time I actually logged into ecwebv4 and posted a test message was a few weeks ago. I couldn't see anything ecwebv4 related that would cause
What if newest_message_header.number is invalid (e.g. NaN or 0xffffffff or
msgbase.last_msg)? Other than that possibility, the logic is fine.
I'm using the latest ecwebv4, but I'm fairly certain I only posted
one message to a test area awhile back. So any of the other
sub-boards I'm seeing this in (DOVE-GEN being one of them, along
with random ones in other networks as well), shouldn't have been
affected.
Posting doesn't (or shouldn't) affect new-scan pointers, just reading messages.
(Additionally, it does not modify the 'last_read' property.)
If that looks incorrect to you, please let me know. I do browse
messages on my website fairly often now, and I'm not having
problems like these, so I dunno.
Yeah, I dunno either. It could also be if they are using anything that modifies their msgs.cnf files and incorrectly/accidentally changes the pointer index files for sub-boards. That'd mess up everyone's scan pointers and configurations for that/those subs.
There's still some bugs in the legacy/runemaster web UI with regards to the new voting feature. I may need the occasional reminder or push (like this one) to take another look at fixing them.
Don't get me wrong, I'm pretty impressed and excited about V4, I just
have a lot of work to do in order to translate some of my current
custom pages over to new formatting, etc. And I would rather have some
of that work done before doing upgrades.
I understand. You could disable and avoid the voting feature, but eventually you're going to want that too.
I've been using V4 extensively and have only had issue with the
POLLS posting incorrectly, never any pointer issues.
I may have fixed something the other day re: posting poll messages, but I haven't tested it extensively. I'll take another look at it sometime soon.
Posting doesn't (or shouldn't) affect new-scan pointers, just reading messages.
Posting doesn't (or shouldn't) affect new-scan pointers, just
reading messages.
After what we've already discussed, I'm now completely confused. I
just newscanned all areas (using DDMsgReader) this morning, updated
the BBS entry in the BBS list, and logged off. I then telnetted in
about 45 minutes later to find that the HOME_COOKING, and STATS echos
on Fidonet were reset to message #1 when newscanning. I'm also fairly certain I received new messages in both of those areas during the time between telnet sessions.
That said, I took a look at my logs, and only three events ran during
that time as well, BINKIT (which runs binkit.js as a client poll, not
as a server), FIDOIN, and FIDOOUT.
Also, a VERT QWK callout was done, but no new messages were received
at the time, and I don't see any reason as to why a QWK callout for Dovenet would have any affect on Fidonet whatsoever.. so that can more than likely be ruled out.
Hello Digital,
On Thu Jan 12 2017 18:23:44, Digital Man wrote to Nicholas Boel:
Posting doesn't (or shouldn't) affect new-scan pointers, just reading messages.
After what we've already discussed, I'm now completely confused. I just newscanned all areas (using DDMsgReader) this morning, updated the BBS entry in
the BBS list, and logged off. I then telnetted in about 45 minutes later to find that the HOME_COOKING, and STATS echos on Fidonet were reset to message #1
when newscanning. I'm also fairly certain I received new messages in both of those areas during the time between telnet sessions.
That said, I took a look at my logs, and only three events ran during that time
as well, BINKIT (which runs binkit.js as a client poll, not as a server), FIDOIN, and FIDOOUT.
Also, a VERT QWK callout was done, but no new messages were received at the time, and I don't see any reason as to why a QWK callout for Dovenet would have
any affect on Fidonet whatsoever.. so that can more than likely be ruled out.
Hello Digital,
On Fri Jan 13 2017 08:10:56, Accession wrote to Digital Man:
Posting doesn't (or shouldn't) affect new-scan pointers, just
reading messages.
After what we've already discussed, I'm now completely confused. I
just newscanned all areas (using DDMsgReader) this morning, updated
the BBS entry in the BBS list, and logged off. I then telnetted in about 45 minutes later to find that the HOME_COOKING, and STATS echos on Fidonet were reset to message #1 when newscanning. I'm also fairly certain I received new messages in both of those areas during the time between telnet sessions.
That said, I took a look at my logs, and only three events ran during that time as well, BINKIT (which runs binkit.js as a client poll, not as a server), FIDOIN, and FIDOOUT.
Also, a VERT QWK callout was done, but no new messages were received
at the time, and I don't see any reason as to why a QWK callout for Dovenet would have any affect on Fidonet whatsoever.. so that can more than likely be ruled out.
And to add to the above, I just telnetted back in after a VERT QWK callout that
received two messages in DOVE-ADS and one message in DOVE-HAM (these are just my internal codes, but I'm sure you understand which ones they are), and now a whole mess of sub-boards were reset to message #1 across 4 networks, and I can't find a relation whatsoever except that all of the sub-boards that this to
happen to, seems to happen each time this occurs.
What boggles me is that it's all over 4-5 networks, and only one has QWK networking enabled (Dovenet), whereas all the others are FTN only.
If Buckaroo and myself are the only ones experiencing this, is there a specific
way to completely wipe the sub-boards clean aside from removing all of the .hash, .ini, .shd, and .sid files in /sbbs/data/subs?
Last, not sure if this matters much, but I have one sub-board (0n-code) that seems to create extra files in /sbbs/data/subs:
[axisd@hardwiredpi subs]# ls -al 0n-code*
-rw------- 1 root root 0 Nov 27 18:26 0n-code.sdt
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.sdt
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.shd
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.sid
-rw------- 1 root root 0 Nov 27 18:26 0n-code.shd
-rw------- 1 root root 0 Nov 27 18:26 0n-code.sid
If I delete the three odd ones (.sdt.*), they show back up at some point or another. I've checked the sub-board config in SCFG as well as areas.bbs, and there seems to be no problems in either of them.
I suggest removing DDMsgReader from the equation for a while, just as
a test.
If Buckaroo and myself are the only ones experiencing this, is
there a specific way to completely wipe the sub-boards clean aside
from removing all of the .hash, .ini, .shd, and .sid files in
/sbbs/data/subs?
Pointers are stored in sbbs/data/subs, so cleaning the subs won't
resolve anything.
Last, not sure if this matters much, but I have one sub-board
(0n-code) that seems to create extra files in /sbbs/data/subs:
[axisd@hardwiredpi subs]# ls -al 0n-code*
-rw------- 1 root root 0 Nov 27 18:26 0n-code.sdt
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.sdt
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.shd
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.sid
-rw------- 1 root root 0 Nov 27 18:26 0n-code.shd
-rw------- 1 root root 0 Nov 27 18:26 0n-code.sid
If I delete the three odd ones (.sdt.*), they show back up at some
point or another. I've checked the sub-board config in SCFG as well
as areas.bbs, and there seems to be no problems in either of them.
Should your event command-lines. They're 0-byte files, so most likely you're trying to run smbutil or another smb-related program with a bad path to the sub-board base-filename.
Hello Digital,
On Fri Jan 13 2017 12:10:32, Digital Man wrote to Accession:
If Buckaroo and myself are the only ones experiencing this, is
there a specific way to completely wipe the sub-boards clean aside
from removing all of the .hash, .ini, .shd, and .sid files in
/sbbs/data/subs?
Pointers are stored in sbbs/data/subs, so cleaning the subs won't resolve anything.
I was only wondering in case there are some messed up messages causing (that chksmb isn't seeing) this from when the voting was first introduced. I'm fairly
certain I did updates during the process while it was unfinished.
Last, not sure if this matters much, but I have one sub-board
(0n-code) that seems to create extra files in /sbbs/data/subs:
[axisd@hardwiredpi subs]# ls -al 0n-code*
-rw------- 1 root root 0 Nov 27 18:26 0n-code.sdt
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.sdt
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.shd
-rw------- 1 root root 0 Jan 5 07:09 0n-code.sdt.sid
-rw------- 1 root root 0 Nov 27 18:26 0n-code.shd
-rw------- 1 root root 0 Nov 27 18:26 0n-code.sid
If I delete the three odd ones (.sdt.*), they show back up at some
point or another. I've checked the sub-board config in SCFG as well
as areas.bbs, and there seems to be no problems in either of them.
Should your event command-lines. They're 0-byte files, so most likely you're trying to run smbutil or another smb-related program with a bad path to the sub-board base-filename.
My event command line for MSGMAINT is "%!smbutil%. mp1000 *.shd" without quotes, and starts in the ../data/subs directory. I haven't messed with it so it's most likely unchanged from the original install, and my logs show:
Jan 13 05:01:00 hardwiredpi synchronet[329]: evnt Running native timed event: MSGMAINT
Jan 13 05:01:01 hardwiredpi synchronet[329]: evnt Timed event: MSGMAINT returned 0
So it should be running correctly, no?
Aside from that, I don't post any messages to that area using smbutil either. Since those were dated Jan 5, I'll delete them and keep an eye out to see if they show up again. Is it possible that the SBLMAINT event (which runs "sbl%. -m") would do some kind of a scan of message sub-boards? I disabled that awhile
back due to it failing every time it ran (since I recompiled it for the Pi).
| Sysop: | Winzlo |
|---|---|
| Location: | Minnesota, USA |
| Users: | 11 |
| Nodes: | 16 (0 / 16) |
| Uptime: | 495934:29:45 |
| Calls: | 82 |
| Files: | 1,070 |
| D/L today: |
27 files (11,920K bytes) |
| Messages: | 286,923 |