Templot5 - progress discussions

[Steve_Cornford]

The two remaining tgroupboxes now adjusted slightly:-

Steve

 

[Steve_Cornford]

Hi Martin,
Pull request created for map_clarity_unit
Steve

 

[Martin Wynne]

Hi Steve,

Looks good. Thanks for PR.

The red label on the right needs Transparent set to False to get its white background back -- assuming you think it needs it.

I'm hoping to do a release by tomorrow. Any last-minute requests?

cheers,

Martin.

 

[Steve_Cornford]

If you accept the PR for now, I will sort out the red label.
Steve

 

[Martin Wynne]

PR merged. Thanks.

Martin.

 

[Steve_Cornford]

Red text on white background fixed

Will push this time
Steve

 

[Steve_Cornford]

Well I think I have managed to update your main as I am now a collaborator & it allowed me to merge my pull request. I might have done it several times , but I am sure all is ok now.

No last requests..... I will save them for Templot5+.....

Good luck with the packaging
Steve

 

[Martin Wynne]

Well I think I have managed to update your main as I am now a collaborator & it allowed me to merge my pull request. I might have done it several times , but I am sure all is ok now.

No last requests..... I will save them for Templot5+.....

Good luck with the packaging
Steve

@Steve_Cornford

Hi Steve,

I'm hoping to have something on the server within the hour. Just made some strong coffee.

Martin.

 

[Martin Wynne]

@Steve_Cornford @Hayfield @James Walters @gavin @KHC1

Templot update version 245b is now on the server. Restart Templot to update.

This update is a milestone after more than 25 years of Templot. You now get 2 programs -- Templot2 and now open-source Templot5.

You can swap to and fro between them without losing templates or background shapes by clicking the big button at the top right:

It's been quite an effort to get this far in a short time and there are bound to be some bugs and quirks, but hopefully nothing terminal.

Templot2 is unchanged for track design and planning, but for plug track and COT track you now need to swap to Templot5.

I will write some more soon.

cheers,

Martin.

 

[Phil G]

Well done Martin very cool switch function
cheers
Phil

 

[Martin Wynne]

@Steve_Cornford @Alistair Ward

Hi Steve, Alistair,

I have now pushed/merged all the files as released in 555a.

Then created a new branch 556 for ongoing developments. Hopefully main 555a can be left locked as the initial release version. I can't find any way to lock it, so I have downloaded the Zip and put it in our resources here, should anyone need it outside of Git:

https://85a.uk/templot/club/index.php?resources/categories/templot5-code.22/

Going forward I think we can abandon the sub-build letters and just have version numbers. So the next version will be plain 556. The letter distinction no longer has any meaning, because I'm not planning to make any future changes to the binary part of the BOX file. Any future changes will be in a new XML trailing block.

I don't yet know how future versions of Templot5 will get updated and released independently of Templot2. For now the next version of Templot2 will jump to plain 256 to match the Templot5 556 contained within it. I'm hoping that it will switch to T5 by default on startup.

cheers,

Martin.

 

[Alistair Ward]

Hi Martin,

May I suggest a slightly different strategy for your branching.

In the first of my Git notes, I suggested that all development happen on branches (a branch per feature), and get merged to main after they're complete and tested. This way, main should always be in a releasable state.

When making a release from main, you put a tag on the commit used to build the release version. This allows you to always to back to that point if needed.

If you subsequently find a need to make a hotfix for a release (ie fix a bug without including any new development work), then at that point a would create a hotfix branch, make the release from that, and then merge it back to main

I hope to continue with my Git notes in the next couple of nights.

Cheers,
Alistair.

 

[KHC1]

Well done Martin et.al. Already updated and have done a switch from2 to 5 and no problems.

Keith

 

[Martin Wynne]

I suggested that all development happen on branches (a branch per feature)

Thanks Alistair.

I'm not too clear how a branch per feature would work. Features in Templot aren't neatly compartmented, everything is interwoven with everything else. That way could mean having multiple branches going on at the same time, which would get very confusing and almost guarantee mis-matches between them.

Also I don't know that a thing IS a feature until after I have tried it to find out. Creating a branch for it in advance seems to need a crystal ball?

A new branch after each release for all the work in progress to the next one seems more logical.

I know my way of working seems not to conform with anyone else's, but I'm not sure what to do about it.

cheers,

Martin.

 

[Steve_Cornford]

Is it possible for Martin to rename Templot5 to Templot555?

Then Martin could create a new repository called Templot556?

Steve

 

[Steve_Cornford]

Hi Martin,

Pseudo COT for 4mm?

Another of my light bulb moments daft ideas coming up.

For COT track you have provided a small platform/web between the crossing timbers for the spacer blocks.
Here is an isolated v-switch in P4 printed as COT track as an example.
in 4mm scale, although the spacer blocks would be printed in resin, would there be any possibility/advantage in providing the platforms for the spacer blocks when FDM printing the timbering brick?
The platform could have pseudo sockets for the spacer blocks to plug into. This might be less fiddly than trying to install the spacer blocks after sliding in the point & splice rails.

ps I've just had scrambled eggs for lunch.

Steve

 

[Phil G]

Hi guys, have I missed something here?
I though you had tried all FDM printed 4mm track as the first cab off the rank idea, and it did not work. Mainly because the extrusion of the FDM filament is simply too big, for the required 4mm fidelity.
I get why its worth trying in 7 mm. Although having looked at all the photos, it still strikes me the fidelity of resin chairs is clearly obvious better even in 7mm scale.
cheers
Phil,

 

[Hayfield]

Phil

For me the quality is not bad and its not been refined. Yes resin chairs can be crisper and perhaps more detailed, but once painted would you notice ? especially from normal viewing distances. Plus its a lot less investment in £'s, time and effort.

For some added detail is a step too far, not saying COT track is the answer, but it may well attract a new untapped following

 

[Steve_Cornford]

Hi Phil,
I am not proposing printing COT track (combined chairs and timbers) for 4mm, merely that in 4mm we could pinch the idea that Martin has come up with for 7mm COT of the platforms/webs under the spacer blocks and use that in 4mm FDM printed timbers with the addition of a socket, so that resin printed spacer blocks be printed with plugs that fitted into the sockets on the platforms, rather than as at present where we have to add these spacer blocks after we have inserted the point & splice rails, as it is a bit fiddly.
4mm FDM timbering bricks, resin print chairs.
Steve

 

[Martin Wynne]

Hi I missed something here?
I though you had tried all FDM printed 4mm track as the first cab off the rank idea, and it did not work. Mainly because the extrusion of the FDM filament is simply too big, for the required 4mm fidelity.
I get why its worth trying in 7 mm. Although having looked at all the photos, it still strikes me the fidelity of resin chairs is clearly obvious better even in 7mm scale.
cheers
Phil,

@Phil G @Hayfield

Hi Phil,

It's been a long journey. I first tried FDM-printed timbering bases with integral chairs 6 years ago on the then new BIBO printer. After much tinkering about I managed about 4 timbers in 0 gauge which looked acceptable. But is was painfully slow. Threading the rail was difficult, and the chairs usually broke off in the process. It didn't seem very promising. When I tried it in 4mm scale it was hopeless. You could make some blobs which might look chair-shaped at a distance, but were very fragile and broke off easily.

But my main objection is and always has been to the concept of sliding the rails into position. It's just about feasible for a single turnout, although there is a problem with the knuckle bend. But it's impossible with complex formations such as a tandem turnout or outside slip. The only way to assemble those is to drop the rails into position vertically -- as in traditional soldered track building (and as on the prototype).

With solid one-piece chairs, that means separating the slide-on chairs from the timbers. Hence plug track, and the much better chair detail from resin printing. Subsequently I discovered that resin printing allows the outer jaw to be a separate component, possibly allowing the rest of the chair to be FDM printed integral with the timbers. But my experiments with that idea didn't get very far on the Marlin-based FDM printers, at least in 4mm scale. 7mm scale was better, but the chairs were no match for the resin-printed ones, getting the pin slot accurately sized was tricky or impossible, and the finished print needed a lot of cleaning up.

Then a few weeks ago @AndyB posted a topic on RMweb showing a few timbers of FDM-printed integral chairs in 0 gauge. I hadn't tried it on the new fast Klipper-based Neptune 4 printer, so I thought I would give it a try to see what happened. I could hardly believe the result:

That's straight off the printer. The chairs are not as good as resin printed, but not far behind and on a par with injection-moulded.

What's more they are strong and don't break when the rail is threaded. And no stringing or clean-up needed. For knuckle bends there is the option of having just the crossing chairs with loose resin-printed jaws.

There are two possible explanations:

a. the higher temperature and much faster Klipper-based printing somehow makes the difference, or

b. I didn't try hard enough 6 years ago to find the right printer settings. Had I done so the whole course of plug track development might have gone differently.

Whatever, this way of track-building seems a worthwhile option to include in Templot. At least in 7mm scale. For 4mm scale the results are not so promising -- the chairs don't look so good and break quite easily. I do have some extra tough PLA-ST filament to try, but I suspect the fine detail and stringing will be a problem.

And after that distraction it's back to Scaleforum -- which is all 4mm scale, and all plug-in resin-printed chairs. Not least because that method also allows for laser-cut timbers.

It's all about options and choices. Templot is a workshop tool, and the more ways of using it the better.

cheers,

Martin.

 

[Phil G]

Hi Martin,
Thanks for the clarifications
In summation it sounds as though I have not missed anything.
FDM in 4mm alone does not sound as practical or with the same level of detail as using resin chairs. But maybe still an option. (As you have said separate timbers, irrespective as how there made, and separate resin chairs does more truly follow the prototype methodology as well)

For 7mm it does make the FDM option more viable, and then its up to the user, to evaluate the pros and cons. The con is still less fidelity and also making more difficulty to manufacture the more complex trackwork. The pro is as John has pointed less capital outlay on equipment equipment, and also less tools to have to master to achieve a result.

I totally agree and endorse your comment, about the more options the better the overall concept has to be.
At the end of the day it is all about horses for courses.

There is also nothing wrong with a mix and match approach, if that is what somebody prefers to do.
Steve I did not fully get your point until you high lighted it. That too is quite valid.
cheers
Phil,

 

[Martin Wynne]

But my main objection is and always has been to the concept of sliding the rails into position. It's just about feasible for a single turnout, although there is a problem with the knuckle bend. But it's impossible with complex formations such as a tandem turnout or outside slip. The only way to assemble those is to drop the rails into position vertically -- as in traditional soldered track building (and as on the prototype).

With solid one-piece chairs, that means separating the slide-on chairs from the timbers. Hence plug track,

If the Neptune 4 can produce acceptable FDM chairs for COT track in 7mm, it can also make 7mm plug-in chairs. These are a raft of S1 chairs for code 125 rail and FDM printing (with snap-fit plugs):

They have printed fine (50 minutes for 16 chairs, say 3 minutes per chair), and when there is a bit more light outside I will post a photo.

The time could be significantly reduced by shortening the plugs and supports.

Martin.

 

[Martin Wynne]

.
Straight off the printer, with no cleaning up:

Martin.

 

[Hayfield]

.
Straight off the printer, with no cleaning up:

View attachment 11183

Martin.

Wow, photography is so cruel, especially when the item is enlarged by so much. On my print I doubt if its so crisp but at 1-1 ratio no steps in the printing is noticeable. I see there is a small amount of stringing, but nothing like the Kingroon

 

[Martin Wynne]

I see there is a small amount of stringing, but nothing like the Kingroon

@Hayfield

Hi John,

What little stringing there is, is very wispy, almost like cobweb, and falls off in normal handling.

See: https://85a.uk/templot/club/index.php?posts/13106

Why the Neptune 4 can do that, and the Kingroon can't, is subject to further investigation.

cheers,

Martin.

 

[Phil G]

Very much out of curiosity could the same 7mm chairs be made on the resin printer and then photo'd in the same way?
that would then be a very useful indicator of the 2 resultant processes.
cheers
phil.

 

[Steve_Cornford]

Hi @Phil G

Martin already has.

See here https://85a.uk/templot/club/index.php?threads/fdm-chairs-in-0-gauge.991/page-3#post-13106

Steve

 

[Martin Wynne]

.
If COT track becomes popular in 0 gauge, as I think it might, one welcome option would be the ability to change individual chairs to plug-in. This for areas of a track formation where sliding the rail into position isn't feasible.

At present clicking the customize button on a chair gets you this additional dialog:

Most of which isn't working yet. There is a not-working option to change the type of plug/socket fit, but not a COT option to switch plug/sockets on-off -- which is needed for COT track.

So that's my next WIP. In theory it shouldn't be too difficult -- but that doesn't always turn out to be the case! It's also yet another distraction from the 8-sided chair outlines and slab/bracket chairs. But as this is my hobby and not work, I reserve the right to be utterly distracted.

Martin.

 

[Steve_Cornford]

Hi Martin,
I think you have earned the right to be distracted.

Out of curiosity I have been investigating the fields in the 2D and 3D record defnitions in chairs_unit.pas and how they are used.

So far I have made a worksheet containing a separate sheet for each of the chair types, but not yet added a sheet for common fields (ie those note prefixed by a character based chair type) .

I have attached here a PDF print of that worksheet. E&OE.
Each sheet is named with what I believe is the chaircode+field prefix+"fields" ?
Each sheet has a column for the record (2D or 3D), a column for the field name, and then columns with counts of usage in each of the .pas files that reference that field.

The fields highlighted in pink are fields that have been defined in either the 2D or 3d record, but have not been given a value or referenced anywhere else.

I have produced this as it helps my understanding of what fields are used in producing which plugtrack component, so that I might make meaningful & helpful suggestions for future developments such as 8 sided chairs, chair families etc.

Ignore it for now if you feel it is a distraction from your distraction.

Steve

 

[Martin Wynne]

@Steve_Cornford

Hi Steve,

Wow. That must have taken many hours to compile. Thanks.

But I'm not too sure how useful it is?

The 2D stuff is all related to drawing the chair outlines on the trackpad (and eventually on the paper and PDF templates too). When coded for the DXF export it all becomes marks in the marks list with code numbers, and the DXF export is not concerned with the original dimensions.

The 3D stuff is used to create the DXF insertion blocks (jaws, keys, seats, bolts, etc.), and doesn't appear on the 2D screen.

Perhaps a Zoom meeting to discuss all this would be helpful?

cheers,

Martin.

 

[Steve_Cornford]

Hi Martin,
Didn't take too long using the "find in files" feature of notepad+.

I would welcome a zoom meeting to discuss, but perhaps after you have had some fun on your COT distraction, especially now that John (@Hayfield) has ordered a Neptune
You probably need a bit of relaxation having got Templot5 releaed to the wider commuity.

Steve

 

[Martin Wynne]

You probably need a bit of relaxation having got Templot5 released to the wider community.

@Steve_Cornford @James Walters

Hi Steve,

I was feeling under some pressure to get it released in good time before Scaleforum, so that everyone had a chance to discover the new functions beforehand. Also possibly to recruit a few more T5 volunteers.

Now taking it a bit easier for a few days. Although it's wet and grey as ditchwater here today, so might as well do a bit on Templot.

cheers,

Martin.

 

[Martin Wynne]

@Steve_Cornford @Alistair Ward

Hi Steve, Alistair,

I have pushed the first update/commit to the new T5-556 branch.

Can you see it? Can everyone else see it?

GitHub desperately wants me to merge it into the main branch with a big yellow banner -- why? I want to leave the main unchanged for now as the original 555 release.

cheers,

Martin.

 

[Martin Wynne]

.
Everyone loves another tickbox -- here are 4 more.

Sometimes you want some chairs without keys. For a stack of chairs outside a p.w. hut, or a wagon load perhaps.

That's easy to do with plug track -- just make a batch of chairs with keys switched off.

Sometimes you see a turnout which has been taken out of use. The unwanted rails are removed but the chairs are left in place on the long timbers until the next relaying. It would make an interesting model.

That's easy to do with plug track too -- just use some chairs without keys instead of the usual ones.

But for COT track it's not so easy. If you just leave the rail out the empty chairs will look a bit daft with keys in them. You can't just switch off the keys without losing them on all chairs of each type -- all S1, or all L1, or whatever.

The new tickboxes above let you omit keys on individual chairs instead.

Will be in 556 for the next update, and already pushed to GitHub.

cheers,

Martin.

 

[Steve_Cornford]

Hi Martin,

Can you see it? Can everyone else see it?

I can see your branches:-

also in my local github desktop, i managed to click on the current branch drop-down, & select your new branch T5-556, then fetch upstream, which gave me your "556 first update -omit chair key per heaved chair 10-09-2024 20:26"

and upon inspecting my local \github\templot5\ folder I can see that it now has the updated version of heave_chairs.lfm with DropDownCount = 50 on line 742.

Steve

 

[Martin Wynne]

@Steve_Cornford

Thanks Steve.

I now see that everyone can get to the branch by going to:

https://github.com/Martin-Wynne/Templot5/tree/T5-556

although there is no obvious indication that it exists on the first GitHub page.

Perhaps it would be better to use this as the front page for the project:

https://sourceforge.net/projects/opentemplot/

cheers,

Martin.

 

[Steve_Cornford]

Hi Martin,

I distracted myself today by having a little play adding chair labels to the PDF output, by cutting & pasting your math_unit code:-

not too bad if you switch off timber centre-lines and timber extension marks.
perhaps I should have tried adding the actual chair outlines first!

Steve

 

[Steve_Cornford]

Hi Martin,

Todays attempt at adding chair outlines to PDF has not gone to plan:-

Probably because shortly before this I had my hair cut!
Oh well back to the drawing board....
Might have some lunch first...
Does Coronation Chicken count as egg based?

Steve

 

[Steve_Cornford]

Hi Martin,

Does Coronation Chicken count as egg based?

The answer is... Not quite...

A small step forward...
Steve

 

[Steve_Cornford]

Hi Martin,
Sooty helped, so we have:-

Is the colour ok?

This is just the control template, and just in pdf_laz_unit so far, as is the chair labels.

I expect someone will be along soon asking for bolt-holes, & 8-sided GWR chairs

Steve