Template talk:Medway watermills diagram

Message moved from User talk:Pldx1
There is a problem with the diagram now. Your edit has caused it to finish at Farleigh Lock. I've resisted the temptation to revert you. Please try to fix in the next 24h or I will restore the diagram back to the original pending a fix. Mjroots (talk) 09:36, 22 March 2016 (UTC)

Continuation
Dear User:Mjroots. You left me a message on my talk page. I am replying here because your message underlines a problem that goes far beyond individuals. Your are totally right when being infuriating when a template you created and maintained across the years, starting 2009, becomes messed because someone modified something somewhere without taking care of the consequences. But... I am not the culprit ! Here are the statuses as of now (Tue 22 March 2016) of the previous revisions of Template:Medway_watermills_diagram. Here, means template overflow,   don't compile,  is OK.

When I applied my modifications, I have checked the results of &#123;&#123;Medway_watermills_diagram|part=1&#125;&#125; where part is one of main, upper, lower, middle. And I haven't see any problem (but maybe, I am wrong, and I would be ashamed of that). On the contrary, I have also seen that the whole map (all=1) was on overflow, and that its rendering stopped somewhere. But, remember, I am here because I crossed this file at Category:Pages where template include size is exceeded: I haven't be surprised that removing a simple if structure hasn't be enough to step back below the overflow boundary. Nevertheless and at your demand, I will investigate further to see why even the 19:23, 28 May 2014 version, by User:YLSS, has turned on overflow. In any case, stepping back to any prior version wouldn't fix the problem. Best regards. Pldx1 (talk) 13:18, 22 March 2016 (UTC)


 * Thanks for this. I'm sure we can get it worked out eventually. Mjroots (talk) 13:40, 22 March 2016 (UTC)

Dear User:Mjroots. I have investigated further. If we want that Template:Medway_watermills_diagram goes back beyond the Template include size limit (while not moving to routemap), we have to And thereafter either: We are near the limit ! Pldx1 (talk) 12:05, 23 March 2016 (UTC)
 * 1) Using BS-table instead of BS-map
 * 2) Replacing the BSn and BSn-2 by their old counterparts
 * 1) killing three BSn lines
 * 2) or replacing five BS3 by their BSrow  expansions.

When you are satisfied with the result, you can compile the */parts by substituting the blocs, obtaining a routemap from routemaps, or a a*/src from BS-map, that can be compiled afterwards into the final routemap. OK, this requires a grand total of 50 sub-templates:  35 blocs + 5*3 for the (*,*/src,*/parts) relative to each of the 5 flavours (all, main, upper, lower,middle). Better forecast an attrition war around the theme "who has ever seen so many sub-templates ?" Pldx1 (talk) 13:57, 23 March 2016 (UTC)
 * And now the solution: replace all the BSrow in the BSn by their 2014 expression and the Post-expand include size becomes: 2090898/2097152 bytes. Obviously, I am not advocating to do that. The best method to build templates like the Template:Medway watermills diagram is rather to write separately each element, in whatever language you please. Here, we have 35 of them, like User:Pldx1/Bs-map/Medway/bloc16. Then you assemble them like in User:Pldx1/Bs-map/Medway/middle/parts that mostly contains


 * If that will fix it, go ahead. What concerns me is that the diagram displays correctly. Mjroots (talk) 16:28, 23 March 2016 (UTC)