Hello Michael!

I'm switching from regualr Nodelist/Nodediff to the DailyLists, all works fine but nlupdate refuses to update it daily because (obviusly) the file is younger than 7 days.
I've looked at the documentation but I haven't found a way to force the update.

8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--

---------- Wed 12 May 2021, nlupdate/lnx 1.9.0-cur 27-01-17
1 07:43:54 Start
C 07:43:54 Trying to update /home/husky/nodelist/Z2DAILY.131
C 07:43:54 /home/husky/nodelist/Z2DAILY.131 younger than 7 days, no update necessary
C 07:43:54 Trying to update /home/husky/nodelist/Z2PNT.127
C 07:43:54 /home/husky/nodelist/Z2PNT.127 younger than 7 days, no update necessary
1 07:43:54 End

8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--

Is there any way to force it?

Many thanks! :)

Ciao!
Fabio

--- GoldED+/LNX 1.1.5-b20180707
* Origin: ]\/[imac Rebirth Boss Point (2:335/364.1)

Hello Michael!

12 May 21 12:17, you wrote to me:

No, nltools were developed many years before the daily nodelist
appeared and did not change since then.

Well, it's time to change it, isn't it? :P

I use the following script where '/etc/fido_profile' contains exported variables with some directories used later in this and other scripts.

[...]

The script is called by htick when it receives a file from Z2DAILY
file echo:

ExecOnFile Z2DAILY Z2DAILY.??? /home/mike/bin/change-z2daily

Thank you Michael, very kind of you. I'll take a look at it as soon as I can. :)

Ciao!
Fabio
--- GoldED+/LNX 1.1.5-b20180707
* Origin: ]\/[imac Rebirth Boss Point (2:335/364.1)

Hello Fabio!

12 May 21, Fabio Bizzi wrote to Michael Dukelsky:

No, nltools were developed many years before the daily nodelist
appeared and did not change since then.

Well, it's time to change it, isn't it? :P

Why? What for? Do you know what a official weekly nodelist based tracker would do with routed netmails to new added nodes of the daylist?

Did you weigh up the pros against the cons?

Weekly diff is ~5k, daily list is ~190k x 7. The data efficiency, power consumtion and disk wear is decreased by 26600%.

nlupdate looks to be designed to fetch a diff from the fileecho. This is done to protect the diff files via tic file security. The 7 day block looks like a protection against revision mismatch:

3148 30. Apr 11:11 NODEDIFF.A20
3618 7. Mai 00:19 NODEDIFF.A27
2263 29. Jan 01:05 NODEDIFF.A29
2297 5. Feb 01:05 NODEDIFF.A36

With list 127 on day 130 the diff file A29 would be selected but it does contain diff 029 and not 129 -> error. The correct next diff would be A34.

Regards

Kai

--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Fabio!

14 May 21, Fabio Bizzi wrote to Kai Richter:

Why? What for? Do you know what a official weekly nodelist based
tracker would do with routed netmails to new added nodes of the
daylist?

It's obviuos, it would bounce them back to the originating node.

Yes. I had that twice this year.

Did you weigh up the pros against the cons?

yes, more or less, and I'd like the possibility to have a daily
nodelist updated tracker, it suits my RC needs.

Well, i'm not into *C stuff but i found in the policy: "A Regional Coordinator does not perform message-forwarding services for any nodes in the region."

Ah, i see, you are NC/Host too. "The Network Coordinator is responsible for maintaining the list of nodes for the network, and for forwarding netmail sent to members of the network from other FidoNet nodes."

Weekly diff is ~5k, daily list is ~190k x 7. The data efficiency,
power consumtion and disk wear is decreased by 26600%.

Are you seious? :D

Yes.

Nowdays, with ssd and gigabit ethernet links, you are speaking about
it?

You can check the facts on your system. Yes, bandwith is great and ssd are big, the trend of the world is massive data tranfser. The point is: What for? The relation of data created to data needed.

You said your RC needs. I don't know what the advantage of daily lists for an RC/NC is but i don't have to. They must be really important if you risk to send mail from a node that does exsist in the daily but not yet on the official weekly nodelist to a node. If that node does have a tracker it would do as described above, the mail will bounce.

nlupdate looks to be designed to fetch a diff from the fileecho.
This is done to protect the diff files via tic file security. The
7 day block looks like a protection against revision mismatch:

[...]

Daily lists don't have diffs. ;)

The question is what is your conclusion?

Mine is a misuse of a software that is not designed for the job.

I don't want to change your mind, probably the weekly nodelist suits
your needs, my need, instead, is to have a daily updated nodelist.

Why can't that wait for weekly revision? When a new nodenumber is requested i would add it to my local privat nodelist. This would enable mail exchange between the new and my system. Any agreement could be done. The configuration could be tested and finalized. The new node can select and read the echos he wanted and dive into threads smoothly. After 8 days in worst case his nodenumber is published in the nodelist and he could start public writing.

I think that's a win/win situation. The new node got time to avoid to jump in on the wrong foot (avoid "I really put my foot in it".). The NC got time to see what kind of support the new node could need and direct him to the right echoareas.

The daily option will tell to nlupdate to change the +7 days with a +1
day only for that nodelist and we can check that it can do it only
with a FullUpdate directive.

If the new node of the daily nodelist would write in public any netmail answers could bounce on the next tracker that does not use the daily list. When that happend to me my question was "whats wrong with the routing, my mail bounced?"

To avoid that all fidonet systems have to switch to daily nodelists. Then the whole network effort is increased by the 26600%.

Regards

Kai

--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Kai wrote (2021-05-15):

I don't want to change your mind, probably the weekly nodelist suits
your needs, my need, instead, is to have a daily updated nodelist.

Why can't that wait for weekly revision?

Because it's not 1990 anymore. There is not really an advantage in artificial delays. If you like the old way, that's fine, but don't try to push your (Fidonet) world view on everybody else.

And it's not only about new nodes, existing entries might also change at any time. I mean we don't update DNS or dynDNS on a weekly basis only and then wait another 7 days to correct the error one accidentally made with the MX record to receive email again.

Daily updates are already slow.

---
* Origin: . (2:280/464.47)

Hello Oli!

16 May 21, Oli wrote to Kai Richter:

There is not really an advantage in artificial delays. If you like the
old way, that's fine, but don't try to push your (Fidonet) world view
on everybody else.

And it's not only about new nodes, existing entries might also change
at any time.

Daily updates are already slow.

Who is pushing his view to fidonet now? Fidonet is no real time world. It's concept was and is store and forward. I like the improvements due to fido over IP but speed is not the holy grail of fidonet.

I mean we don't update DNS or dynDNS on a weekly basis only and then
wait another 7 days to correct the error one accidentally made with
the MX record to receive email again.

What are you talking about? Which FTS is this MX stuff? If found mx in FTS-5000 and 5001 but not MX.

Fidonet does not update DNS at all. This is why we use dynDNS services. The dynDNS services provide the real time path to the URIs published in the nodelist. Those dynDNS services are designed for real time and allow fidonet to operate with stable destinations.

A change from weekly to daily nodelists would have severe consequences. The official sequence for offline systems is one revision cycle on hold, on the next cycle down and on the third cycle removed. This would offer a timeframe of two weeks for a broken node system to order spare parts, repair the system and recover online status. A daily revision would reduce that time to two days.

This would force sysops to monitor their systems in real time and to hold a spare part stock. We are still a hobby network. Some nodes can take care of their system on weekends only. Yes, there are many things that can do it faster and better and on any place in the world.

Why can't that wait for weekly revision?

Because it's not 1990 anymore.

Lol! Hey, then why are you using fidonet. ;-)

Regards

Kai

--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

On 16 May 21 11:00:02, Kai Richter said the following to Oli:

A change from weekly to daily nodelists would have severe consequences. The official sequence for offline systems is one revision cycle on hold, on the next cycle down and on the third cycle removed. This would offer a timefram of two weeks for a broken node system to order spare parts, repair the syst and recover online status. A daily revision would reduce that time to two d

LOL!!! What are you smoking?

Nick
--- Renegade vY2Ka2
* Origin: Joey, do you like movies about gladiators? (1:229/426)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

I don't see any need

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

I don't see any need
to send one liners

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

I don't see any need
to send one liners
as fast as we can.

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

I don't see any need
to send one liners
as fast as we can.

If a new node can't wait a week

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

I don't see any need
to send one liners
as fast as we can.

If a new node can't wait a week
then a quickshot node

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Nick!

16 May 21, Nick Andre wrote to Kai Richter:

LOL!!! What are you smoking?

I don't.

I don't see any need
to send one liners
as fast as we can.

If a new node can't wait a week
then a quickshot node
is gone faster than he came.

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

Hello Fabio!

17 May 21, Fabio Bizzi wrote to Kai Richter:

Daily lists don't have diffs. ;)

Mine is a misuse of a software that is not designed for the job.

My conclusion is that is a misuse for you, but it's simple, you can
still run weekly nodelists and leave everyone else to do exactly what
they wants.

I do. And i do talk about that topic because of...

It's obviuos, it would bounce them back to the originating node.

Yes. I had that twice this year.

Wow, really annoying...

...your definitin is annoying behavior too.

I don't want to change your mind, probably the weekly nodelist
suits your needs, my need, instead, is to have a daily updated
nodelist.

Why can't that wait for weekly revision?

I wrote it in my previous message, I'ts useful when I onboard new
nodes to the net, so, every node in my Region can send routed netmails through my node to the new node.

The same will happen with the weekly nodelist. I do know "why" we have a nodelist, the questions is "daily". Your reason can be solved with a private nodelist on your system. It would publish the new node to the tracker and the mail isn't bounced. Annoying behavior solved without "official" nodelist.

I'll kind ask the new sysop to no contact your node/points until the weekly new nodelist will be released, so no any bounces to you. :D

Good advice! Look at that situation. The node is there but you ask him not to publish an echomail to other nodes. Then why do we need daily nodelists?

A new node can crash to any known nodes in the nodelist. I'm not sure how strict the trackers are, if they don't check the destination only then a new node could route his contact information to nodes he needs to talk with before he is in the weekly nodelist.

If the new node of the daily nodelist would write in public any
netmail answers could bounce on the next tracker that does not
use the daily list. When that happend to me my question was
"whats wrong with the routing, my mail bounced?"

Is It too hard to look at your nodelist and check if the node is in?

I did and he was not in. That's the reason why the tracker bounced the netmail.

And you're supposing that everyone would answer with a netmail to an echomail message.

No, not everyone. If one does then he could get an annoying bounce.
If nobody does then again: Why do we need daily lists?

To avoid that all fidonet systems have to switch to daily
nodelists. Then the whole network effort is increased by the
26600%.

For me, it's a conspiracy of the strong powers ...

For me it's a review what's going on. The use of daily lists is an unreliable seven days limited advantage for new nodes only. Nodes pull and process informations that they already have except one missing line each day.

Regards

Kai
--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)