Winners by categories.

[SirWumpus]

(Post IOCCC28) It would be useful for the web site to list winners by category or tag. All the one-liners, all the mazes, X11, utilities, languages, layouts, etc. Some might be in multiple categories. A colleague of mine was looking at ideas and sometimes it would be easier to find stuff and compare what has gone before if it were some simple classifications.

[xexyl]

I like the idea. I also am glad you mentioned that some might be in multiple categories (I would have suggested it otherwise).

It'd be ideal if it could be done programmatically but I guess we'll have to wait and see.

[lcn2]

We believe we have addressed all of the current questions that still need answering at this time. If we've missed something or something else needs to be clarified, please ask again.

[SirWumpus]

Think this should be raised again (after IOCCC29?) when there is not a rush of pre-contest work. Yesterday I pointed out on Discord entries that are "quines"or related. Not sure we could create categories for everything, but maybe a small (SQLite?) database of keywords, some sort of cross-reference or "index" derived from winner README and documentation. Maybe add tags to .md files, eg. {utility} {game} {quine} {oneliner} etc.

For developers, its just a matter of cloning the winner repository and find/grep the documentation, but other curious visitors might not go to such lengths.

[xexyl]

Update 0

FYI for an example tag cloud see my Nanozen blog on the RHS.

Okay but the second post there is so sad. Anyway I'd have to look at it another time - after a rest. I'm not sure I can tell what you're getting at (unless you mean the tags which maybe you do?).

[SirWumpus]

You are missing the point. Why add hand written JSON (or modify existing JSON). Have a simple tool that auot-gens the cloud; have a small dropin JavaScript that fetches the current cloud and displays. Even then you really don't need the JavaScript; it could still be generated statically a cloud and included into the two pages; maybe some CSS for font controls.

[SirWumpus]

Update 0

FYI for an example tag cloud see my Nanozen blog on the RHS.

Okay but the second post there is so sad. Anyway I'd have to look at it another time - after a rest. I'm not sure I can tell what you're getting at (unless you mean the tags which maybe you do?).

I mean the TAG section on the right-hand-side of the blog which is the cloud; there are fancy interactive graphical word clouds but I think its overkill for IOCCC website design choice.

[xexyl]

Update 0

FYI for an example tag cloud see my Nanozen blog on the RHS.

Okay but the second post there is so sad. Anyway I'd have to look at it another time - after a rest. I'm not sure I can tell what you're getting at (unless you mean the tags which maybe you do?).

I mean the TAG section on the right-hand-side of the blog which is the cloud; there are fancy interactive graphical word clouds but I think its overkill for IOCCC website design choice.

Right. I thought maybe that (when looking a second time). Thanks for confirming.

[xexyl]

You are missing the point. Why add hand written JSON (or modify existing JSON). Have a simple tool that auot-gens the cloud; have a small dropin JavaScript that fetches the current cloud and displays. Even then you really don't need the JavaScript; it could still be generated statically a cloud and included into the two pages; maybe some CSS for font controls.

Because it doesn't fit our design. And as for JavaScript - right. We went out of our way to avoid JavaScript. As for why JSON it's because much of the website uses it and it would be easy to do this. It in theory would be easy to adapt one of the scripts to do make a page with categories. A static page of course.

[lcn2]

We do NOT have backend support on the servers for IOCCC, therefore we cannot run something like a database or any backend code of any kind. All of the web site magic has to be done in the generated static HTML. Adding a small database would not be useful to someone browsing the web site.

One could add .keyword.json files to all of the entries. This would allow someone, for example, to create an IOCCC search web page elsewhere that linked into the main IOCCC web sit, for example.

However we have no plans to do any sort of keyword design before the end of IOCCC29. Right now we are focused on replacing the current IOCCC28 manifest code with a new table drive walk code. Right now are focused on code completion by the end of the month.

UPDATE 0

The MAIN priority NOW is getting ready for IOCCC29 by 2025 December.

[SirWumpus]

You wouldn't need a running service. This would be generated when the HTMl is generated and it would probably be create in-memory (or temporary SQLite file thrown away when done); the result would be static.

And yes IOCCC29 is priority, but so far I have no assignment (other than wikipedia tweaking). This is not something you or @xexyl need concern yourselves with just now. Now that I have an idea, I'll build something, test it, and demo it for you if I think I have something viable.

[xexyl]

However we have no plans to do any sort of keyword design before the end of IOCCC29. Right now we are focused on replacing the current IOCCC28 manifest code with a new table drive walk code. Right now are focused on code completion by the end of the month.

Nice to hear this. I'm still curious what you have in mind. You taking care of that simplifies it as I was not entirely sure what you were after. From the sounds of it it seems like this might quite simple to put in the current code that I wrote, updating it so that it uses the tables instead. Given you also intend on working on the semantics tables issue I would also guess there will be a pull request in jparse - and jsemtblgen would be updated too I presume.

I'm happy (and not happy if you understand me) that IOCCC29 will be started this year.

[xexyl]

You wouldn't need a running service. This would be generated when the HTMl is generated and it would probably be create in-memory (or temporary SQLite file thrown away when done); the result would be static.

And yes IOCCC29 is priority, but so far I have no assignment (other than wikipedia tweaking). This is not something you or @xexyl need concern yourselves with just now. Now that I have an idea, I'll build something, test it, and demo it for you if I think I have something viable.

Good luck with it! Interested to see what you have in mind. Even if it's beyond my faculties at this time (or who knows, maybe it'll always be beyond my faculties :-) ).

[lcn2]

You wouldn't need a running service. This would be generated when the HTMl is generated and it would probably be create in-memory (or temporary SQLite file thrown away when done); the result would be static.

And yes IOCCC29 is priority, but so far I have no assignment (other than wikipedia tweaking). This is not something you or @xexyl need concern yourselves with just now. Now that I have an idea, I'll build something, test it, and demo it for you if I think I have something viable.

We need to focus on soooo many other things right now: We won't be able to consider anything related to tags until sometime in middle 2026, at best.

As we don't have the bandwidth to consider any keyword demo now, you might as well consider such a demo next year.