You are not logged in.
For SpaceNut re Oldfart1939's new topic set ...
You probably already noticed the approval of Oldfart1939 for the suggestion to move posts not by him to the Companion topic.
Yes. I endorse that concept. Otherwise too many other sidebar comments cloud things. That's a general comment and not only for the topic at hand.
With your assistance (or kbd512 in your absence) we can provide Oldfart1939 a clear runway to build up the content of an article that could (hopefully) then be submitted to a print publication (first priority) or to an online outlet (alternative).
(th)
Last edited by tahanson43206 (2021-04-27 11:17:08)
Online
Progress ... sort of ... I corrected the special function that checks posts and dates, so now it correctly stops the run if there is an error. Unfortunately, it is wrong, so once again, I'll have to study the log to understand what happened.
The ** good ** news is that all the potential hurdles/stops were jumped/passed up to that point.
Once we get past the function to check posts and dates, everything else is well proven from the Not Verified ID's.
I'll set this aside for tonight and look at it again tomorrow.
Edit#1 ... For SpaceNut ... I copied my posts from the primary, pasted them in the Companion, and deleted them from the primary.
Louis may not be able to do that (I'm not sure) so you might be able to help.
We should end up with just Oldfart1939 posting in the primary. We and ** all ** the other members can comment, suggest, critique, and do whatever else we are inspired to do to help him create an article that someone somewhere will want to print, and someone else will (hopefully) want to read.
(th)
Last edited by tahanson43206 (2021-04-27 18:38:37)
Online
For SpaceNut (and anyone else who may be interested) ...
This post is reserved for what I hope will be a breakthrough test run of WBA upgraded to Level 1...
It ran unassisted through to the Posts/Dates verification procedure, and hit a block there. I'm hoping the failure is due to something simple, like looking at the wrong variable at run time. I had previously tested the module in Step mode, and it seemed to work. I'll add a confirmation message, so the module confirms the ID it reviewed passes muster.
As a reminder, 4460 is "out there" waiting to be upgraded from banned to member, if it passes your final review.
Edit at 11:12 local time ... It turns out the failure yesterday was due to another script error. The script has to tell the program were data is stored, and it pointed to the wrong place. However, the program **did** act correctly, by confirming that the posts were not zero.
I've added more runtime reporting to allow the Posts/Dates function to report success.
We ** might ** get past the Post/Dates hurdle today!
Edit at 11:59 local time .... Holy Moley! The script appears to have finished!
4461 should be in the Bans table ready for your review and upgrade to member!.
For statistics ... The WBA on-screen display shows runtime of 3:39 ... that is nearly double the Level 0 performance, but the new Level 1 program performs a significant number of validations at points along the way.
We need to go over the log together, and if you are comfortable with the protections built into the program, we can consider giving it a bit of running room.
I'm taking a break now and will return later to post the log with notes I might be inspired to add.
(th)
Last edited by tahanson43206 (2021-04-28 13:28:11)
Online
The quote below is from the version of the ID Update script that just succeeded with ID 4461 in "run free" mode ...
This text will be the same for every run, so will not be repeated in future reports.
SearchTerm:Script used to upgrade a banned ID 2021/04/28
SearchTerm:Ban script used to upgrade an ID
SearchTerm:Banned ID upgraded to member status, script for
Update for Web Automation Report for 04-28-2021 at 11:52:23
WebBrowserAutomation
Version 2021.4.28.2Command Code stored at Array line: 1 W number is: 10000
Escape sequence stored at Array line: 2 <>
Command Code stored at Array line: 3 X
Escape sequence stored at Array line: 4 <>
Command Code stored at Array line: 5 W number is: 5000
Escape sequence stored at Array line: 6 <>
Command Code stored at Array line: 7 W number is: 1000
Escape sequence stored at Array line: 8 <>
Command Code stored at Array line: 9 CPX
Command Code stored at Array line: 10 T number is: 22
Escape sequence stored at Array line: 11 <>
Command Code stored at Array line: 12 W number is: 5000
Escape sequence stored at Array line: 13 <>
Command Code stored at Array line: 14 W number is: 1000
Escape sequence stored at Array line: 15 <>
Command Code stored at Array line: 16 W number is: 1000
Command Code stored at Array line: 17 SC at Row # 1
Command Code stored at Array line: 18 RC at Row # 1
Escape sequence stored at Array line: 19 <>
Escape sequence stored at Array line: 20 <>
Command Code stored at Array line: 21 W number is: 1000
Command Code stored at Array line: 22 CPL
Command Code stored at Array line: 23 X
Escape sequence stored at Array line: 24 <>
Command Code stored at Array line: 25 W number is: 5000
Escape sequence stored at Array line: 26 <>
Command Code stored at Array line: 27 W number is: 1000
Escape sequence stored at Array line: 28 <>
Command Code stored at Array line: 29 CPX
Command Code stored at Array line: 30 T number is: 15
Command Code stored at Array line: 31 PC from Row: 1
Escape sequence stored at Array line: 32 <>
Command Code stored at Array line: 33 W number is: 5000
Escape sequence stored at Array line: 34 <>
Command Code stored at Array line: 35 W number is: 1000
Escape sequence stored at Array line: 36 <>
Command Code stored at Array line: 37 RC
Command Code stored at Array line: 38 CPL
Escape sequence stored at Array line: 39 <>
Command Code stored at Array line: 40 T number is: 20
Escape sequence stored at Array line: 41 <>
Command Code stored at Array line: 42 W number is: 5000
Escape sequence stored at Array line: 43 <>
Escape sequence stored at Array line: 44 <>
Command Code stored at Array line: 45 RC
Command Code stored at Array line: 46 CPL
Escape sequence stored at Array line: 47 <>
Command Code stored at Array line: 48 T number is: 22
Escape sequence stored at Array line: 49 <>
Command Code stored at Array line: 50 RC
Command Code stored at Array line: 51 CPS number is: 1
Escape sequence stored at Array line: 52 <>
Command Code stored at Array line: 53 T number is: 1
Escape sequence stored at Array line: 54 <>
Command Code stored at Array line: 55 W number is: 1000
Escape sequence stored at Array line: 56 <>
Command Code stored at Array line: 57 SC at Row # 4
Command Code stored at Array line: 58 PF Perform Function # 4
Escape sequence stored at Array line: 59 <>
Command Code stored at Array line: 60 T number is: 22
Command Code stored at Array line: 61 X
Command Code stored at Array line: 62 N number is: 5
Escape sequence stored at Array line: 63 <>
Command Code stored at Array line: 64 W number is: 1000
Escape sequence stored at Array line: 65 <>
Command Code stored at Array line: 66 SC at Row # 2
Command Code stored at Array line: 67 RC at Row # 2
Command Code stored at Array line: 68 T number is: 2
Command Code stored at Array line: 69 X
Command Code stored at Array line: 70 N number is: 5
Command Code stored at Array line: 71 X
Escape sequence stored at Array line: 72 <>
Command Code stored at Array line: 73 W number is: 1000
Escape sequence stored at Array line: 74 <>
Command Code stored at Array line: 75 SC at Row # 3
Command Code stored at Array line: 76 RC at Row # 3
Command Code stored at Array line: 77 T number is: 10
Command Code stored at Array line: 78 X
Command Code stored at Array line: 79 D
Command Code stored at Array line: 80 X
Escape sequence stored at Array line: 81 <>
Command Code stored at Array line: 82 W number is: 5000
Escape sequence stored at Array line: 83 <>
Command Code stored at Array line: 84 W number is: 1000
Escape sequence stored at Array line: 85 <>
Command Code stored at Array line: 86 RC
Command Code stored at Array line: 87 CPL
Command Code stored at Array line: 88 T number is: 22
Escape sequence stored at Array line: 89 <>
Command Code stored at Array line: 90 W number is: 1000
Escape sequence stored at Array line: 91 <>
Command Code stored at Array line: 92 CPS number is: 2
Escape sequence stored at Array line: 93 <>
Command Code stored at Array line: 94 T number is: 21
Escape sequence stored at Array line: 95 <>
Command Code stored at Array line: 96 W number is: 5000
Escape sequence stored at Array line: 97 <>
Escape sequence stored at Array line: 98 <>
Command Code stored at Array line: 99 CPL
Command Code stored at Array line: 100 T number is: 22
Escape sequence stored at Array line: 101 <>
Command Code stored at Array line: 102 W number is: 5000
Escape sequence stored at Array line: 103 <>
Command Code stored at Array line: 104 W number is: 1000
Escape sequence stored at Array line: 105 <>
Command Code stored at Array line: 106 CPS number is: 2
Command Code stored at Array line: 107 T number is: 1
Command Code stored at Array line: 108 B
Command Code stored at Array line: 109 T number is: 1
Escape sequence stored at Array line: 110 <>
Command Code stored at Array line: 111 W number is: 1000
Escape sequence stored at Array line: 112 <>
Command Code stored at Array line: 113 RC
Command Code stored at Array line: 114 PC from Row: 3
Command Code stored at Array line: 115 T number is: 1
Command Code stored at Array line: 116 X
Command Code stored at Array line: 117 D
Escape sequence stored at Array line: 118 <>
Command Code stored at Array line: 119 W number is: 5000
Escape sequence stored at Array line: 120 <>
Escape sequence stored at Array line: 121 <>
Command Code stored at Array line: 122 CPX
End of file: C:\DataDell\WebBrowserAutomation\20210428ScriptFromBansTableText.txt
The actual script contains comments not show in the array load report. The full script is available upon request.
(th)
Last edited by tahanson43206 (2021-04-28 11:33:23)
Online
For SpaceNut ...
Here is the runtime report from the test of WBA on 2021/04/28 ...
Text from Clipboard for the record: zzusaymeznxlxcl
Text from Clipboard for the record: http://newmars.com/forums/userlist.php? … rch=Submit
Text from Clipboard for the record: http://newmars.com/forums/profile.php?id=419463
Text from Clipboard for the record: zzusaymeznxlxcl
Text from Clipboard for the record: TestID04461
Text from Clipboard for the record: TestID04461@test.com
Text from Clipboard for the record: http://newmars.com/forums/profile.php?s … &id=419463
Total Command Lines found: 122
Total input Lines in script: 369Number of ID's processed: 1
Starting Number: 4461
Last Number of Run: 4462
Summary for Web Automation Report for 04-28-2021 at 12:06:57
Average time of Loop: 00:03:39Total time of Run: 00:11:48
Successful Run. No Warning Errors.
Looking forward, I'd like to know if you are willing to let the program run for some small number of ID's, such as 10?
Edit#1: I notice the log includes a printout of the email address of the banned ID.
Is that a concern? It might be ... please let me know ...
That is a script command (show clipboard) so it can be removed if you prefer.
Edit#2: Total time of the run includes the time needed to configure the program for the run.
(th)
Last edited by tahanson43206 (2021-04-28 13:27:21)
Online
edit ya thats not needed and I removed it...
So long as we do not do more than 100 a day go for it...and I will try to set the stage with getting them banned if that helps the program....
If I find an old user that was banned but posts are of value I will indicate by name change and setting the password to keep others out...Once we are past them we can restore that users name and set the ban once more....
Offline
For SpaceNut re 1056 ...
Thanks for the advice ... That print of the old email was for diagnostic purposes and I will remove it.
I just ran a sequence for 4462:
Text from Clipboard for the record: zzusaymezngldrk
Text from Clipboard for the record: http://newmars.com/forums/userlist.php? … rch=Submit
Text from Clipboard for the record: http://newmars.com/forums/profile.php?id=419097
Text from Clipboard for the record: zzusaymezngldrk
Text from Clipboard for the record: TestID04462
Text from Clipboard for the record: TestID04462@test.com
Text from Clipboard for the record: http://newmars.com/forums/profile.php?s … &id=419097Total Command Lines found: 122
Total input Lines in script: 369Number of ID's processed: 1
Starting Number: 4462
Thanks for thinking about what we might do when we run into a banned person with posts.
There are a number we definitely want to keep, for their contribution over an extended period.
I may be able to upgrade the program to skip over them safely.
It is easy for me to set the run size to any size you want, so I'll start with 100.
I have one more shakedown run to do ... I'm looking for messages that aren't showing up. They don't appear to be a ** problem **, but I'll feel better when they are showing up.
Edit at 20:37 local time ... that extra shakedown turned out to be a good idea.
There were two errors. I've fixed them and will now try a full speed test run again.
(th)
Last edited by tahanson43206 (2021-04-28 18:38:22)
Online
For SpaceNut ... the test run for 4463 now correctly examines the profile of the banned ID ...
I took out the print of the old email, and added a proof of match on dates and zeroi posts.
Here is what the bottom of the log looks like for 4463:
Text from Clipboard for the record: zzusaymeynpazbh
Text from Clipboard for the record: http://newmars.com/forums/userlist.php? … rch=Submit
Text from Clipboard for the record: http://newmars.com/forums/profile.php?id=419117
Text from Clipboard for the record: zzusaymeynpazbh
Post-Date Function Start: Reading Clipboard #4
PerformFunction: Valid ID: match in: egistered: 2017-02-26 (95.65.79.205)
Last post: Never
Last visit: 2017-02-26 17:55:24
Posts: 0 - Show all topics - Show all posts - Show all subscriptions
Admin noteSubmit When you update your profile, you will be redirected back to this page.
Jump to
Go
Powered by FluxBB
Text from Clipboard for the record: TestID04463
Text from Clipboard for the record: TestID04463@test.com
Text from Clipboard for the record: http://newmars.com/forums/profile.php?s … &id=419117Total Command Lines found: 121
Total input Lines in script: 370Number of ID's processed: 1
Starting Number: 4463
Last Number of Run: 4464
Summary for Web Automation Report for 04-28-2021 at 22:53:01
Average time of Loop: 00:03:38Total time of Run: 00:03:45
Successful Run. No Warning Errors.
Based on this run, I think it is possible to try a short run of 10 ID's.
The test will run from 4464-4475.
Edit after midnight local time ... The test run with 10 ID's seemed to go well ... 4475 is sitting at the top of the Bans table.
I'll double check the log, and if everything looks good, I'll launch a run for overnight ...
4 minutes per ID * 100 >> 400 minutes / 60 >> 40 / 6 >> 20 / 3 >> 6.33 hours ....
It should be done before sunrise, and then FluxBB will time it out.
(th)
Last edited by tahanson43206 (2021-04-28 22:17:49)
Online
For SpaceNut ... a short run started last night only completed part of the series ...
I asked WBA to work from 4476-4500 and it stopped at 4494.
I'll study the log later this morning.
***
I note the interesting discussions underway in Louis' new topic and the car industry topic.
I was surprised to see Calliban offering some hints of endorsement of Louis' powered-road concept.
One definite advantage of such a road (I would imagine) would be further implementation of the computer regulated traffic patterns hinted at in current automated automobile research. Science fiction writers have been having fun with the idea for years.
Edit#1 (after looking at log) .... the cause of failure is not clear ... the WBA is on the wrong page. That could have been caused by a delay in server response. The checks caught the error and stopped the run.
However, I'll improve log reporting so that if this type of error occurs again, it may be possible to figure out what happened.
for now, I'm going to let the milestone of 4494 stand.
(th)
Last edited by tahanson43206 (2021-04-29 08:14:49)
Online
For SpaceNut re project ... The Level 1 version of WBA is showing promise...
I'd like to separate the robot from my ID ... Would you be willing to consider a special "moderator" ID for the robot?
What I have in mind is your power to unban ID's. I appear to have the power to ban ID's but I can't unban them.
We can regulate access to the robot ID by banning it after starting a run.
To begin a work session, you would unban the ID. I would log in using that ID and start the run.
At a separate computer, I would log in as tahanson43206 and ban the robot ID.
If my theory holds up, the robot could continue working, but no one (hacker or nation state) could log in using that ID.
When the run ends, the FluxBB server disconnects the session after 15 minutes of idle time.
The name we choose is up to you, but something like WBARobot might work.
Edit#1: Another possibility is the (oh so original) NewMarsRobot ....
The big advantage (for me) is that a long running job can be kept separate from my ID, because I am in and out frequently throughout the day.
A run for 100 ID's will take about 6 hours. I have adjustments to make before attempting a run of that size.
(th)
Last edited by tahanson43206 (2021-04-29 10:55:58)
Online
For SpaceNut re ID project ...
I'm starting to think about adaptive automation.
My classification system is probably outclassed by professional designs, but since I'm blissfully unaware of them, I'll continue on ...
Level 0 is brute force, like a lawnmower. The initial set of Not Verified ID's were processed at Level 0, and close human supervision was mandatory.
Level 1 is stop-on-sensor-input, and we have reached that level. A dishwasher or clothes washer would be examples.
Level 2 would be adaptive ... it would make decisions based upon sensor input, and attempt recovery in case an error occurs.
In order to improve Level 1, I am beginning to look at navigation on a web page, which is the domain of interest for this project.
It is possible to know with certainty where we are on a web page, using developer tools.
However, in my first experiments with these tools (using Chrome because it is so Developer friendly) I see hints of possibility but lots and lots of potential pitfalls that could spoil the program's day (or run in this case).
It is possible to ask for precise location (for example) by calling up the html source code for the element of interest [Inspect Element]
More importantly, because WBA can't visually "see" anything, it is dependent upon keyboard commands.
Chrome offers keyboard commands to access Inspect Element, but I have not yet found a keyboard command to get out of Inspect Element mode.
An encouraging sign is that Chrome highlights the element of interest (for example, Submit button) so Control/C can be used to capture that text.
The program could then confirm that it is "looking at" the Submit button, by running a string search on the copied text.
However, at ** this ** point, I have no idea how to get out of Inspect Element mode.
It may not even be necessary to close Inspect Element, if work can continue on the main page.
Edit#1: In case you're wondering, I'm pursuing this with some intensity because I was [surprised/taken-aback] to find the WBA program "looking" at my profile page, after the failure that happened last night. This happened because (a) something happened to disrupt the flow of events and (b) I am using ** my ** ID to run this automation. Fortunately the safeguards built into Level 1 prevented damage when the program found itself on the wrong page.
The tabbing that is needed to get from the address line to a location of interest is a brute force activity.
The program has no idea (at present) where it is on the working page, after a series of tabs.
It would be ** really ** helpful to know where it is, before issue an "enter" command (for example)
The Chrome [Inspect Element} command definitely has the potential to confirm the program is at the Submit button after tabbing, but (at this point) I have only a vague idea how to use that capability.
(th)
Last edited by tahanson43206 (2021-04-29 11:19:12)
Online
For SpaceNut re ID project ... I added some output to the log file and tried a short run from 4495 to 4500.
The run ended after 4496. I'm looking forward to seeing what the enhanced log may show.
I'm guessing timing is going to turn out to be the issue. That is the only variable I can't control.
On the other hand, the program ** might ** have found an ID with posts. That ** has ** to happen at some point.
Or the ID under review may have logged in after registration date.
Edit at 17:20 local time ...
After looking at the log, I concluded the halt was due to a delay of the Internet beyond the planned Wait.
The program halted as it was supposed to do, when it did not find the right text in the address bar.
I reran with no changes and the series completed. I am including here a sample of the upgraded log:
Starting Sequence for ID: 4500
Text from Clipboard in Step: zzusafmeutzhc2k
Text from Clipboard in Step: http://newmars.com/forums/userlist.php? … rch=Submit
Text from Clipboard in Step: http://newmars.com/forums/profile.php?id=419739
Text from Clipboard in Step: zzusafmeutzhc2k
Post-Date Function Start: Reading Clipboard #4
PerformFunction: Valid ID: match in: egistered: 2017-03-12 (95.65.79.205)
Last post: Never
Last visit: 2017-03-12 17:25:54
Posts: 0 - Show all t
Text from Clipboard in Step: TestID04500
Text from Clipboard in Step: TestID04500@test.com
Text from Clipboard in Step: http://newmars.com/forums/profile.php?s … &id=419739
Completed Sequence for ID: 4500
Total Command Lines found: 121
Total input Lines in script: 370Number of ID's processed: 4
Starting Number: 4497
Last Number of Run: 4500
Summary for Web Automation Report for 04-29-2021 at 17:25:24
Average time of Loop: 00:02:44Total time of Run: 00:18:47
Successful Run. No Warning Errors.
Edit at 19:32 local time ... I took a look at the Z* group and it appears there are only 0 post ID's in the next 100.
I'll launch a full 100 run this evening, unless I find a message from you suggesting otherwise.
At this point, I'm feeling increasingly confident the program will fail gracefully, although I definitely would prefer to run it under a dedicated ID that can be banned when it is not in use.
(th)
Last edited by tahanson43206 (2021-04-29 17:35:11)
Online
Have completed the unbanning of id's 4460 - 4499 and all is well as these are now ready for use.
Offline
For SpaceNut re #1063
Thank you for giving those newly refurbished ID's their diplomas! They are ready for new members, as soon as we can find them!
Last night I launched a full 100 run, and it halted after only 5 were completed. The issue seems to be confined to a single point in the script, so I've added a wait that should have been there but somehow was overlooked. I'll try again this evening. My idea is that the Internet should be less busy in the US at night.
Edit#1: Thinking ahead a bit ...
In my scale of automation, Level 0 is a lawnmower, or in our case, WBA as it was when it processed the Not Verified ID's
Level 1 is a dish washer or a clothes washer, or perhaps a telephone answering machine.
A telephone answering machine is definitely above basic level 1, where WBA is now. A telephone answering machine is able to recover from some errors.
however, I ** do ** have to reset mine periodically, when the little microcontroller gets lost in the bits.
I have been exposed to Level 2 automation, and worked on a project at that level for several years. It was a business application, and the team varied from two to four, but I was there for much of it.
I ** think ** WBA can be enhanced to move slightly above basic Level 1. What I have in mind is adapting to a halt due to an asynchronous event that is late.
The crude way I'm thinking about approaching this is by setting up a command in the script that would have a format along the lines of:
TryUntilMatch .... I admit to being inspired by the tried and true Try-Catch-Fail structure that's been part of formal languages for quite some time.
I don't like the fail part, but that is already present in Level 1 basic.
As I've noted before, this project seems to be evolving toward something like interpreted basic from the dawn of the computer age.
Thinking out loud here ... The commands of the script are stored in an array, and each step has a number.
The failure I've been seeing recently is around step 45 ... Step 45 is where the failure is detected and the run halts.
The ** actual ** failure occurred earlier, during the steps that asked the remote server to supply some data.
If the estimated time provided for the remote server is insufficient, then the script (as presently written) will look for data that is not present.
A way to handle this is to set up a loop over a range of steps already set in the array.
In ** that ** case, the script might contain a command with a format like this:
ResumeHere
If a failure occurs, the failure processor could look to see if there is a ResumeHere earlier in the command array.
If there is, the Step Loop could be reset to that Pointer value ...
That ** could ** work, but it might result in an infinite loop.
I've created ** lots ** of those over the years.
So there needs to be a counter associated with the ResumeHere command ... the format could be:
ResumeHere:10
That is actually doable, I think....
The program could decrement the parameter (in this case 10) each time it does a reset
If the parameter reaches zero then the program would deliver a full stop as usual.
In the near term, I'll try a run with a new wait added at the point of failure.
It should have been there in any case.
Edit at 14:39 after Friday Volunteer duty ...
I've been thinking (again) about the means by which a simple Try-Catch-Else-Finish structure might be added to the existing simple script.
Goto came to mind as a concept I remember from Dartmouth basic, which itself (would have surely ) drew concepts from the assembler languages of the time.
Goto was structured into the primitive assembler languages of the time as a physical address in the machine memory to which the address mechanism would go if a test was true or false, or in the case of a simple Goto, without any test at all.
It this case, I am thinking that the row of a script command would work perfectly well as a reference point for a Goto.
Asking Google:
A label in a programming language is a sequence of characters that identifies a location within source code. In most languages labels take the form of an identifier, often followed by a punctuation character (e.g., a colon).
Label (computer science) - Wikipedia
en.wikipedia.org › wiki › Label_(computer_science)
Here is a summary of script commands at a point in development of WBA:
http://newmars.com/forums/viewtopic.php … 63#p178963
Neither Goto nor Label are present in the list for Level 1
I am edging toward an attempt to go beyond basic Level 1 ...
At this point I'm not clear on how a Try-Catch-Else-Finish concept might work.
Perhaps a simpler If-Then-Else structure might be easier to implement.
IF is not present in the list either ....
Perhaps only Label:## is needed ... ## represents the number of retries to permit before calling a halt
The program could be written to automatically reset to the nearest earlier step containing Label:##
Edit a 16:00 local time ... work has started on the 2021.05.01.1 version of WBA ...
At this point, the About documentation is updated to show the new Directive (Label:##)
In addition, code was added to recognize a Label:## statement in the script, but at this point nothing is being done with it.
Todo items include:
1) Finish processing directive from script file to store a row in the command array
2) Update the runtime procedure to understand the new directive
What I'm thinking about here is to simply report the label and it's parameter in the log
3) Update all three Compare procedures to deal with a Label properly
4) Update the Script processor to refuse to run if there is no Label present
That last item is the first hint this program (WBA) might incorporate some more sophisticated script checking than it does not.
At a minimum the program needs to (or at least ** could **) check to be sure there is at least on Label:## command in the array.
If there is NO such label in the script, and if the program code is going to look for it, then the program code should immediately fail if none is found.
(th)
Last edited by tahanson43206 (2021-04-30 14:06:50)
Online
For SpaceNut .... this is going to seem pretty small potatoes after the recent posts about the Moon Lander competition, the fate of humanity in the open plains of Mars, and the history of Wiki and NewMars.
So here we are, back talking about what to do when we hit the first banned ID that has posts or which has logged in at least once.
I'd like to offer a suggestion, while keeping in mind that you have already suggested something ...
I'd like to suggest we simply change the name of the ID by inserting a hyphen (dash) in front.
The effect on the ID itself will be minimal ... the posts will not be disturbed.
Readers who look at the posts will see the original name.
Someone might ask why there is a dash in front of the name, and we could tell them, but I'll bet no one will ask.
The benefit is that the ID will sort to the beginning of the queue of banned ID's, and since i am working from the back, eventually all the banned iD's that remain will have dashes in front of the name.
The posts will be undisturbed.
***
Now I'll go back and reread your suggestion, which was just a few days ago.
It was post 1056
edit ya thats not needed and I removed it...
So long as we do not do more than 100 a day go for it...and I will try to set the stage with getting them banned if that helps the program....
If I find an old user that was banned but posts are of value I will indicate by name change and setting the password to keep others out...Once we are past them we can restore that users name and set the ban once more....
Adding a dash in front would "change the name" without changing it's appearance to humans.
I don't think it is necessary to change the password, if we leave them banned.
All their posts will remain available.
If we set a dash in front, there is no need to rename them after we are done.
***
Just FYI ... I'm going to launch another 100 attempt shortly.
The script was adjusted to wait at the point of failure (near Step 45).
It'll be interesting to see if the run gets any further.
(th)
Last edited by tahanson43206 (2021-04-30 20:15:31)
Online
The small group 4500 to 4505 are all set for use
Offline
For SpaceNut re #1066
I'll try a run from 4506-4600 shortly.
First though, I'd like to thank you for your (amazing to me) success in finding so many fascinating (again, to me for sure) links to past and present forum activities.
I don't see anything that approaches what I've been trying to do or to suggest, might be a better way of putting it.
My concern for some time has been that the forum software does NOT permit knowledge to be stored in a retrievable form. Every magnificent post by notable authors just flows under the bridge. That's not a problem for ordinary posts that have a short shelf life, but work by RobertDyck, GW Johnson, kbd512, Oldfart1939 and (I am sure) many others was (and is) good enough to be remembered but more importantly, available for new members to read so they can ask for ** new ** information.
I'd like GW Johnson's archive of stored knowledge about space navigation to become available in a form that can translate into usable skill in human brains, both now and in the future.
I'd like RobertDyck's vision of the Large Ship to slowly transform into a real vessel, plying the Mars/Earth route.
Noah's vision is not yet fully formed, but it too has potential to become a set of plans that real people would actually follow in the Real Universe.
The Wike examples you showed us seem to be for information purposes, although the ambition to store poetry was a surprising (to me for sure) extension of the concept.
(th)
Last edited by tahanson43206 (2021-04-30 20:43:49)
Online
check out the other wiki link I found as its brought the information back up to its near complete levels ….
I think our current dual topics is the way to go until Mars society is willing to invest in writer for the wiki as thats a full time job by itself bring in the information from the topics.
As you saw the wiki page had the links back to the forum discussions pages of course the command lines changed just a little but I am able to find the old topics as to the reading of the embedded links change that is required to show our current discussions in the forum.
Offline
For SpaceNut re #1060 and stop after one ID
First, thanks for your additional suggestions for reading ... The idea the forum and the wiki were linked is one I had not thought of ... I'm not sure about that, but it sure is American Egalitarianism in full display.
I appreciate your endorsement of the primary/companion topic concept, and hope our current set of active members will be inspired to make it work. It's like having a sculptor's studio where the master is working on a block of stone, and the entire neighborhood is welcome to stop by and put suggestions on the bulletin board, and occasionally to talk to the master directly, if he(or she) so chooses.
***
The program stopped after one ID ... it failed at a different point in the script, but again, it looks to me as though the Internet was slow getting the response from the server back before the timer expired.
In this case, I verified that the programmed wait was five seconds, and normally FluxBB is back long before then, but Internet traffic may be slow in my provider network.
The version upgrade I'm about to work on (tomorrow) will have the ability to deal with such delays by starting over at a logical point in the sequence. I'm still trying to puzzle out how a simple looping system would work.
Edit at 7:24 May 1st ... The run to 4600 succeeded! Average time per ID was 3 minutes 38 seconds.
Completed Sequence for ID: 4600
Total Command Lines found: 123
Total input Lines in script: 373Number of ID's processed: 90
Starting Number: 4511
Last Number of Run: 4600
Summary for Web Automation Report for 05-01-2021 at 07:25:41
Average time of Loop: 00:03:38Total time of Run: 07:17:25
Successful Run. No Warning Errors.
(th)
Last edited by tahanson43206 (2021-05-01 05:30:12)
Online
For SpaceNut .... after reading Calliban's recent post in the MIT suit topic, I decided to look for a topic that might contain FAQ
There ** is ** none!
We could create one and encourage members to publish links to their posts there.
Then, when a new member asks a question, as should happen because there is so much to learn, we (members) can suggest searching the FAQ topic.
However, we should ** not ** rest there ... If the new member finds the topic is not covered at all, then we can help the new member to study the subject and to create an answer, and post a link to it in the FAQ column.
In addition, and **this** is the key benefit that might arise from this initiative, if we decide to pursue it ... if the new member finds an FAQ that ** sort of ** answers the question, then we can support the new member in adding to the collection the refinement or clarification the new member believes is needed.
(th)
Online
For SpaceNut (and anyone else who might be interested in the ID Recovery project) ...
The run last night went well! I posted the summary, but I can report that the detail collected for each ID is sufficient to answer a question if you ever have one. The key information you can provide for a lookup is the date recorded by WBA as it processes the ID in question.
A log with that date will be stored on the hard drive of the computer that does the work, and since the logs are fairly compact, I expect they'll not fill up the hard drive before we've exhausted our supply of Spammer created ID's.
***
I've been thinking about the problem of how to implement the next (tiny) increment of automation .... We are now solidly at Level 1. The program will stop if a single bit shows up in the wrong place at any of the numerous check points.
However, I'm interested in attempting to provide some minimal re-try capability ...
Last night, I realized that there are sections of the script where recovery makes no sense.
For those, I would propose placing [Label:0] as a row in the command array. At run time, if a failure occurs and a search for a preceding Label: is initiated, the zero parameter would signal that an immediate halt is appropriate.
However, in contrast, the entire sequence of finding and updating an ID can be repeated as many times as the Internet makes necessary, right up until the [Submit] of the updates.
For that situation, a Label:999 could be set at the top of the array, in row 1 ....
If a failure occurs in any of the steps up to [Submit] the program could simply reset the step loop to 1.
However, once a record has been updated to the new ID format, it's going to be difficult to figure out what to do in various situations.
The Label:0 option will be handy for that situation.
***
In the update for May 1, I'll make a tweak to improve the appearance of the log entry for the special Function to validate users.
However, more importantly, we need a way to automatically stop a run when we exhaust the ID's whose names are greater than "TestID"
In the Not Verified run, I was managing the operation manually, so I could see that situation coming up and deal with it.
In the present case, the program is converting from the spammer ID to the TestID format, and as it chews away at the the query results, it is inevitably going to hit the TestID records it's been creating in that run.
That would presumably create a loop that would continue until the limit for that run is reached, which could be as many as 99 records.
To head that possibility off at the pass, I'm planning to add a "Stop At" feature to the program ...
I'll put "TestID" into the setup window at run time, and configure the program to halt when it finds a User ID with that name.
At that point, hopefully the query can be adjusted to find records with Username values less than TestID.
Edit#1: A quick check of the Bans query shows it will be several days before the Stop At feature is needed.
***
Best wishes for a successful and productive weekend in New Hampshire!
***
The foot tall grass is battened down here. I hope the local farmers are doing as well with their crops as my backyard has been showing is possible.
(th)
Last edited by tahanson43206 (2021-05-01 08:49:19)
Online
For SpaceNut ... tonight's run stopped for a legitimate reason ...
ZREverett
User activity
Registered: 2017-08-29 (14.140.110.198)
Last post: 2017-08-29 09:49:13
Last visit: 2017-08-29 09:49:14
Posts: 1 - Show all topics - Show all posts - Show all subscriptions
Admin note
This ID has 1 post
The program is supposed to stop if the ID has more than 0 posts, and it did!
Now we get to decide what we want to do with the ID.
If you want to save it, adding a dash in front of the Username will allow it to stay visible to forum readers.
On the other hand, if the ID was banned because it created a spam post, perhaps there is something else you might wish to do.
I'll close things down and look for guidance tomorrow.
(th)
Online
Pakistan spammer....having no posts as it was removed....
Convert as there is nothing to save....
Great job by the way on the programming efforts....
Offline
For SpaceNut re #1073 ...
Thanks for your encouragement! It means a lot in this virtual "world" !!!
***
Thanks too, for taking a look at the one post ID ... We are likely to run into more of these special cases, so runs of 100 seem unlikely. This is different from the Not Verified, who by definition had no posts or logins. However, the system (program and hardware) now seem reliable enough to launch more than once a day. The target for a given day can remain 100, but the number of launches can be as many as are needed.
The other variable is your availability to look at situations when they arise.
I ** did ** notice you've been busy! The number of banned ID's has increased from 70 pages to 74, and that's ** despite ** converting some of them to members.
***
Follow up I converted our Pakistani friend's handiwork into 4653. However, I was unable to update the posts count.
Is that something that you can reset? Does it matter?
I'll start another attempt to reach 4700 shortly.
Edit at 11:50 local time ... I checked on WBA and found it logged out ...
Halt occurred at step 60 for znvljhtl
User activity
Registered: 2020-02-18 (85.202.194.48)
Last post: Never
Last visit: 2020-02-19 02:39:02
Posts: 0 - Show all topics - Show all posts - Show all subscriptions
Admin note
This is further confirmation the new validation subroutine is working as intended.
This ID has zero posts, but the dates do not match.
Since the last visit was the immediate next day after registration, but no posts were created in over a year, ** and ** since you had already banned this ID, I'm going to complete the upgrade manually and restart WBA.
(th)
Last edited by tahanson43206 (2021-05-02 09:56:32)
Online
For SpaceNut ... 4672 got bobbled (probably by me) when the program stopped for the Last Visit date mismatch.
I found two 4673, so went back and fixed one of them so it is now 4672.
However, you have already upgraded it's siblings to member status.
Please go back and see if 4672 is qualified for upgrade.
I noted your work starting above 4700 so will make sure not to go beyond that point (except for 4701, if it remains vacant)
Nice to see GW Johnson's post of the obituary for Dr. Webb!
What a career! And nice to see the wording of the obituary!
Edit a few minutes later ... We have a new situation ...
zmaxime.re
User activity
Registered: 2020-11-06 (113.22.118.0)
Last post: 2020-11-07 00:00:09
Last visit: 2020-11-07 00:00:18
Posts: 10 - Show all topics - Show all posts - Show all subscriptions
Admin note
This ID may be a candidate to be renamed with a dash in front, if you want to keep it.
(th)
Last edited by tahanson43206 (2021-05-02 10:25:03)
Online