Looks like I checked the results in vain.
Error messages do not interest anyone.

Nearly all workunits processed from batch 48. Batch 49 already dry.

and... what's next...

---

The display mode filtered by K (spt_list.php?k=18) had errors. These errors surfaced from old, testing, batches where the app wasn't stable yet. To remedy this, i deleted:
all of batch=42 (error of sieve invocation)
all where k=0 (output processing error)
all where k in (17,19) and batch=43 (incorrect detection of odd tuples)
the two "2573" tuples in batch 45.
All the "low" batches that were deleted were already checked by XAVER.
Wrong results from batch 45 were already recalculated.
Natalia Makarova

Batch 50: 516663064778180900 .. 520000062098180900 -1.
16000 tasks and 16499 tasks.
Increased the formal memory requirement as per #3070.

Tomáš Brada
You have not deleted duplicates.
For example, 16-tuples duplicates

500272700245273201: 0 28 108 112 120 148 190 208 210 228 270 298 306 310 390 418
500272700245273201: 0 28 108 112 120 148 190 208 210 228 270 298 306 310 390 418
500272700528288777: 0 26 42 66 90 104 110 150 266 306 312 326 350 374 390 416
500272700528288777: 0 26 42 66 90 104 110 150 266 306 312 326 350 374 390 416
500272793205171263: 0 18 56 60 78 116 120 138 176 194 198 236 254 258 296 314
500272793205171263: 0 18 56 60 78 116 120 138 176 194 198 236 254 258 296 314
500274166044532933: 0 6 16 28 48 76 96 126 130 160 180 208 228 240 250 256
500274166044532933: 0 6 16 28 48 76 96 126 130 160 180 208 228 240 250 256

Can you check for repeatability and delete duplicates?

Tomáš Brada
What determines the need to divide tasks into batchs?

Is it possible to do continuous tasks generation, as is done in the PADLS TOTAL subproject?
In the SPT subproject tasks generation is simpler than in PADLS.

You have not deleted duplicates.

I do not know how to do that correctly. However I found a crude solution. Try now.

I do not know how to do that correctly. However I found a crude solution. Try now.

Now I do not see duplicates.

What determines the need to divide tasks into batchs?

Batches serve as a logical grouping of workunits. I can change parameters for new tasks, but it is harder for already submitted tasks. Dividing them into batches allows me to quickly filter the rellevant results.
For me, it is easier to generate bunch of workunits manually than to have a program that does that continuously. With right modification to boinc server software, it would be easier to generate on-the-fly, but that is not yet reality.

Batches serve as a logical grouping of workunits. I can change parameters for new tasks, but it is harder for already submitted tasks. Dividing them into batches allows me to quickly filter the rellevant results.
For me, it is easier to generate bunch of workunits manually than to have a program that does that continuously. With right modification to boinc server software, it would be easier to generate on-the-fly, but that is not yet reality.

However, the division of tasks into batches leads to inevitable downtime.

You wrote

Also I noticed that the project is running dry. Time to generate more workunits.

https://boinc.progger.info/odlk/forum_thread.php?id=49&postid=4816#4816

Nearly all workunits processed from batch 48. Batch 49 already dry.

https://boinc.tbrada.eu/forum_thread.php?id=3055&postid=3797#3797

Maybe to create the next batch in advance, without waiting for the zeroing of the previous batch?

Very well. New batch should be sent in advance, before previous one ends. I am currently waiting for feedback on 3070, but even if it does not come, I will send more workunits as soon as they run out.
Good news: old, processed boinc result assets are being deleted again!

Batch 51: Count: 48000
First: 520000062098180900
Next : 524928702098180900

The first square was found in the project!

520210977238677833: 0 6 104 110 210 216 234 240 314 320 338 344 444 450 548 554

520210977238677833+
0  6  210  216 
104  110  314  320 
234  240  444  450 
338  344  548  554 
K= 554 S= 1108

Interesting: the square consists of 8 consecutive pairs of sexy primes.

Illustration

Found the second square!

521980328345305811: 0 42 56 60 98 102 116 158 270 312 326 330 368 372 386 428

521980328345305811+
0  42  60  102 
56  98  116  158 
270  312  330  372 
326  368  386  428 

S = 2087921313381224100

Batch 52: 524928702098180900 .. 530000067298180900-1
Count: 49390

Batch 53: few short workunits checking past interval boundaries.
In the past version I am not sure the boundaries were checked correctly, they likely were, but it is easy to check. Current version correctly checks boundaries.

Nothing seems to be moving here

51 spt-m 0 1411 0 0 12

Found the first 24-tuple in the project!

k=24

528050771271601307: 0 14 20 86 92 114 140 144 174 182 186 204 212 230 234 242 272 276 302 324 330 396 402 416

Found the first 24-tuple in the project!
k=24

528050771271601307: 0 14 20 86 92 114 140 144 174 182 186 204 212 230 234 242 272 276 302 324 330 396 402 416

This tuple was probably found by nenym. Lucky one!