[Mark]

In recent months a change was made in one of the Mapster revisions that totally washed out the screen when adjusting F11 brightness. I reported it and it was toned down but it still is not as effective as it used to be. The first 1 or 2 brightness levels are usable but anything after that is still just white-washing the screen and not so much a proper brightening. Added to that is when using anything other than brightness level 1, the tile selector and ability to copy/paste tiles disappears but only in Polymer. So far I've gone back as far as May of this year and these issues are present.

Attached thumbnail(s)

• 

[Mark]

Another glitch lately is if I use the undo function in 3D view the polymer lighting disappears. I have to either switch to 2D and back to 3D for it to come back or press ' + x to disable and re-enable the lighting.

As mentioned by someone else in the discord, I too on occasion will see the skybox replaced by a texture when in certain areas. Its isolated and rare but it did happen to me a couple times.

I'm also experiencing a problem that I haven't tested outside of this one project to see if somehow its map related. In 3D mode I press S to place a sprite and it doesn't show in 3D but does show in 2D. But if I place a model there or a copy of a previously placed wall sprite first, it shows then I change it to the sprite I want. I'll have to see if this happens in other projects using this same revision.

[Bruno]

Yep, I also encounter the following issue : in mapster32, when I create a new sector, or if I insert a point on wall, the parallax sky appear and my wall disappear, replaced by the parallax sky.
Community members told me that polymer isn't update anymore. It seems that polymer lack of efficiency. Fps rate very slowdown for most of users.

[Mark]

Even if its not officially updated, if it gets broken in a new revision it should get fixed. Its been working good for my projects since 2009. I'm staying with it. It has been running decently on my 13 year old computer with a 5 years old video card. I recently bought a newer system, just an i5 cpu with an Nvidia 1080 video card and all my stuff is running great. Framerates have double or quadrupled in some cases. Aside from some other limitations I'm very happy with polymer. Except with these bugs that are popping up in the last year or so and not getting fixed.

[Mark]

I've been getting that problen too but I thought it might be something I did wrong in the map. But the problem did start at the same time I started using the latest Mapster. Could be coincidence.

[Bruno]

Yes now polymer seems to run well too on my new system. I remember when I released my 2013 remake : it was a framerate killer! Indeed, polymer charms me too. I discovered it with DN Eternity and dukeplus. It was mind-blowing for me.

[Mark]

I remember all those explosions at the start of your project bringing my computer to it's knees and begging for relief.
For my own copy I deleted over half of them and changed some of the large spotlights to point lights. It helped a lot and didn't destroy the overall look of the map.

[Bruno]

😂
It was too much indeed!
Working without polymer, then, was quite a challenge to keep the mood I was aiming for the street.
Brunohh3.map still in progress. Slowly due to my CON learning for my Serious Sam project. I'm working on the last room of that third Brunohh map : after escaping cinema basement without any weapon, you find a secret EDF hidden post when toxic dump are flood in sewers. After understanding that EDF is a little guilty in that invasion due to the toxic dump (which cause pig mutation) and due to an"alien" traffic that EDF manage to operate with "secured" conteners, you enter a huge dump destructor where you need to make all you can do to put a boss in. And then, like in combustion chamber from HL, you have to switch on the dump combustion to kill the boss with a few tries and escape. Issue : wall limits in mapster32!!

[Striker]

Unless you can summon Plagman, it'll probably stay as it is, lucky to even get a few fixes. It's not anyone's fault, really, it's just people have moved on. Some fixes do make it in now and then, like a crash that was discovered, but that's about it. As the engine advances, old things with nobody to maintain them are likely to break. Best thing you can do, is find a build that works best for your project, and stick to it until such a time that either someone has the time and knowledge to work on polymer, or replace it with something better.

At some point, when I'm not busy with NetDuke32 and Ion Fury, I'd like to try to at the very least get some things like Polymer's highpal system working in Polymost, so the sting of Polymer being in the state its in won't be quite as painful.

[Mark]

What makes it frustrating is when bug fixes/features are added in the last year or so that supposedly were not Polymer related but they had the consequence of glitching Polymer and they don't get fixed. Most of my modding here over the years has been using older revisions. I'd like to be able to take advantage of newer eduke features but stuff like this ends up happening far too often. Without knowing anything about what caused the bugs I have no way of knowing if a recent bug-causing change can be reverted or if its an integral part of the new direction being taken and will NOT get fixed. I'm starting to think its the latter because of how long some of these bugs have been an issue. I had hoped that by testing new revisions in my projects quite often and catching a bug early and report it that there was a better chance of a fix. Sadly thats happening less these days.

Bottom line is I can still get by using old revisions and just have to go without new features or other fixes.

[Radar]

Assume any issue you find in Polymer will NOT get fixed. It is unsupported. You have probably noticed yourself that when issues do get fixed, it is the exception rather than the norm.

[Mark]

Polymer specific updates I'm not expecting. But I am hoping the things that broke as a side effect of a non polymer change get fixed.

[Paul B]

Mark, on 30 August 2023 - 08:38 AM, said:

Polymer specific updates I'm not expecting. But I am hoping the things that broke as a side effect of a non polymer change get fixed.

I can relate with you Mark. I too heavily relied on the Polymer renderer by choice as nothing quite measured up and still there isn't an equivalent. It's a shame to see it being slowly phased out or no longer supported. I can't imagine how difficult it must be to maintain it's no easy feat.

[Mark]

After testing new revisions, finding new glitches introduced, posting them, and getting no results I am no longer going to bother testing new revisions. It sucks that its got to this point but I've reached the end of my patience.

[Phredreeke]

You can always become a tester for Polygraf

[Mark]

[oasiz]

Mark, on 29 August 2023 - 05:37 PM, said:

Without knowing anything about what caused the bugs I have no way of knowing if a recent bug-causing change can be reverted

This is usually the case.
There just hasn't been a proper rendering programmer on the team so whatever gets fixed is often done based on what can be seen as a "before/after" and learning from that as things go.

A lot of the devstuff in the last few years are mostly the result of fixing/improving stuff for Fury, Netduke and/or TCs like AMC. (Loads of recent TROR fixes, interpolation, tons of perf improvements, voxel improvements, etc.. and yes... the very popular clipping stuff)
People who develop the engine do not primarily use polymer so it naturally gets little attention. Same reason possibly mapster's feature set has largely stagnated after helix's stuff... aside from maybe few minor patches per year.
In case of Fury, polymer lacks so many features that even jokingly testing with it has always looked outright broken so you just don't get to use it.

There is no secret new direction or anything going on, it's just the nature when things get changed/fixed/broken as features get added, improved or changed.
Polymer is 10+ year old abandoned code with no active maintainers. Just like any piece of software, they go rotten over the years.

I understand it's frustrating but there is just no good answer to give either.
You've done what you could and hopefully once people free up, they can take a look.

Real way to the future is to either get a new renderer for eduke or another source port.

[NY00123]

In addition to what's been written by oasiz above, I hope that terminx won't mind me quoting this recent text from the Duke4.net Discord (posted at Sep 9 2023, 0:59 AM UTC time):

terminx said:

polymer needs a bunch of things rewritten to be performant

I had a local branch I was working on that rewrote stuff like the use of occlusion queries to squeeze more perf out of it but it never made it to a point where it was worth committing and the experimental nature of the branch made it as much of an exercise in what not to do as an exercise in what to do

Here's an earlier one from Nov 30 2019 10:16-17 AM UTC time, which I recall to be quoted in the forums beforehand:

terminx said:

Polymer is in kind of an unfortunate state
when we were exploring rendering options for the console ports of IF we talked about maybe reworking it into something usable but even Plagman said he'd probably just throw it out and redo it before trying to actually do something with it at this point

[Mark]

As I mentioned before, I don't expect any optimizations. I would just like to see recent breaks get fixed or at the very least a reply about why nothing is going to be done. Being left wondering sucks.

One example of many is the breaking of the flashlight feature. TX said he did some light interpolation magic and that was likely the cause. He asked for and received the flashlight code and was going to either change his code if possible or come up with a tweak of the flashlight code to work along with his interpolation changes. That was quite a while ago and then silence. If its not going to be addressed, just post it so we know.

[oasiz]

For TX... probably once aftershock is done?
Loads of things beyond just engine stuff keeping him busy on that.