Talk:Blind spots in Doom reject tables


Awesome. Just... awesome. Illdo 09:54, 28 August 2007 (UTC)

Awwwh, thanks.  (Certainly the last thing I could have pictured myself doing when I discovered the wiki!)    Ryan W 17:58, 28 August 2007 (UTC)

Possible infelicity[edit]

Certain "negligible" entries may have to be changed to "no effect" if they involve parts of sectors so small that no actor's center can actually be in them.  I can't tell from the code whether that's required of the target in order for a monster to attempt a shot, especially near a blockmap boundary.    Ryan W 22:36, 30 January 2009 (UTC)

Known anomalies count[edit]

The Bugs section of the maps listed here mention the count of "known anomalies": is that supposed to have any relation to the number of entries in their corresponding table here? Just wondering as the E1M1 page counts 7 and the table 11, E1M2 7 vs 13, E1M3 4 vs 6, and so on (no doubt, but I didn't check further yet). I was going to update the map pages to link directly to their corresponding table, and might as well correct the counts if appropriate. --Xymph (talk) 12:58, 14 May 2016 (CDT)

Maybe (and hopefully) User:Ryan W knows for sure as he seems to be the article creator, but I'm guessing off hand that this data here was derived from an automated analysis of the reject tables, and the sparse data on map articles was just created from contributors' memories of bad reject spots? I honestly don't know for sure. This article does mention that many reject spots are of no consequence because enemies and players cannot view each other from those locations anyway, so the map articles would probably exclude those as bugs due to our perceptually defined concept of what constitutes a bug. --Quasar (talk) 14:01, 14 May 2016 (CDT)
The angle that some entries here are not included in the counts did occur to me, but by what criterium indeed? Upon checking a handful of entries, it appears that only entries marked 'None' are excluded – 'Neglible' and upwards are counted. It was indeed Ryan W who made the map page edits for them in tandem with the tables here, at least for the few maps I checked. But even on that assumption, E1M4 brought to light an inconsistency: count 25 vs 21 entries (1 marked None) in the table. Typo? --Xymph (talk) 14:44, 14 May 2016 (CDT)
I'd say it deserves a real revisiting in light of the inconsistencies. --Quasar (talk) 19:06, 14 May 2016 (CDT)
In my edits, the only automation was invoking RMB to create a revised lump.  Spreadsheet formulas then "diff"ed it against the original (Sectors column), and the categorization and analysis was all pencil-and-paper (Effect column).  There is a great deal of algebra and I don't believe this has ever been done before, so any double-checking of the results is a great benefit IMO.
Other contributors have also added entries, but (anticipating Xymph's next question) there was no discussion as to whether we all used the same reasoning, so differences of opinion may well have arisen.  If a map article mentions a REJECT anomaly with no corresponding table, I'd echo Quasar's assumption that it was written from memory and should be verified.  There are apparently a few ringers which have become common knowledge over the years, however.
Excluding 'None' entries doesn't ring a specific bell, though it sounds reasonable (apart from the implied common sense on my part).  The 25 count for E1M4 may be a simple transcription error, reading from the wrong page of notes; my raw output definitely flags only 42 one-way cases.    Ryan W (usually gone) 21:39, 14 May 2016 (CDT)
Well, I once stated I wasn't going to "install an editor or other tools", and then deployed DeePsea (for secret sectors and other map details) and SLADE3 (for map views) as yet, but I am not going to touch RMB or double-check these tables. Really. :) It's beyond my available time, energy and interest. I'll just keep cross-checking the map page counts as I work through the rest of the IWAD. Up through E2M5, that E1M4 discrepancy was the only one, which I'll fix now. --Xymph (talk) 03:11, 15 May 2016 (CDT)
It's the kind of thing we might want to open an RFC on, as it's unlikely anybody has time for it right now x_x --Quasar (talk) 03:16, 15 May 2016 (CDT)