1278730 - templates, gecko_minversion_inline and fx_minversion_inline show no contents

Status: RESOLVED FIXED

(Developer Documentation Graveyard :: Macros/Templates, defect)

Description

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

What did you do?
================
1. See after "For example:" of https://developer.mozilla.org/en-US/docs/MDN/Contribute/Structures/Macros/Commonly-used_macros#Version_information_macros

What happened?
==============
No contents are generated by the templates.

What should have happened?
==========================
Should show which version started to support the feature.

Is there anything else we should know?
======================================
This bug makes this table unusable:
https://developer.mozilla.org/en-US/docs/Web/API/KeyboardEvent/key#Key_values

Comment 1

[Tooru Fujisawa [:arai]]

These macros are not shown when it's lower than current gecko release.

https://developer.mozilla.org/en-US/docs/Template:fx_minversion_inline
https://developer.mozilla.org/en-US/docs/Template:FirefoxVersionIndicator
> var currentGecko = template("CurrentGecko");
> 
> if ((parseInt(currentGecko, 10) - parseInt($1, 10) < 9) || ($0==3)) {

https://developer.mozilla.org/en-US/docs/Template:gecko_minversion_inline
https://developer.mozilla.org/en-US/docs/Template:GeckoVersionIndicator
> var currentGecko = template("CurrentGecko");
> 
> if (parseInt(currentGecko, 10) - parseInt($1, 10) < 9) {

FirefoxVersionIndicator supports always showing by passing 3 to 1st argument, like {{FirefoxVersionIndicator(3,9.9)}}
We could apply this to GeckoVersionIndicator too, and use them for the place where min-version information should always be shown.
Also, it would be better adding dedicated macro for them, as the value '3' doesn't explain much when calling the macro directly from documents.


Then, updated Commonly-used_macros page to explain about when they're not shown, for now.
https://developer.mozilla.org/en-US/docs/MDN/Contribute/Structures/Macros/Commonly-used_macros

Comment 2

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

Thank you, arai-san.

At least I can say, hiding the information older than "current" release does not make sense because web developers may want to support ESR releases.

And the page documents the implementation of a living spec. So, the value will be changed if the spec will be changed. So, the version information is important. If it is the policy of MDN to hide old implementation's information, I need a way to hide old version's information completely rather than just dropping the version number.

Comment 3

[Tooru Fujisawa [:arai]]

We could fix this by adding CurrentGeckoESR template (or CurrentMinGeckoESR, as 2 ESR versions could be supported at the same time), and use it instead in GeckoVersionIndicator. so that the version information won't be hidden until it gets available on all branches.
it should need consensus on changing it (and maybe how to do so).  also, if the template is updated automatically, the automation should also be updated.

fscholz, can I have your opinion?

Comment 4

[Jean-Yves Perrier [:teoli]]

I wonder if we shouldn't just change the current macros to display the messages longer than 10 releases in the past: what are the number of regular releases that would cover 2 ESR?

That way we don't create new macros.

Comment 5

[Tooru Fujisawa [:arai]]

Thanks :)

for now, ESR is released after every each 7 cycles, and older one is still supported for 2 extra cycles.
so, there could be 'CurrentGecko - 9' version that is supported,
and thinking about transition after new release, 10 versions should be enough I think.

Comment 6

[Florian Scholz (Open Web Docs)]

10 versions sounds good to me. 

Keep in mind: The MDN documentation is browser-neutral, so banners should be avoided in most cases and the browser compat section (with the table and additional notes) is the place for how a browser implements a standard.

Comment 7

[Eric Shepherd [:sheppy]]

So we're agreed on 10 versions then? If so, this is an easy change and I'll do it.

Comment 8

[Tooru Fujisawa [:arai]]

Does showing the version number 10 more cycles solve your case?

Comment 9

[Eric Shepherd [:sheppy]]

I made this change last night -- so take a look and see if the concern is resolved. If not, we can revisit.

Comment 10

[Masayuki Nakano [:masayuki] (he/him)(JST, +0900)(away 4/27 - 5/6)]

(In reply to Tooru Fujisawa [:arai] from comment #8)
> Does showing the version number 10 more cycles solve your case?

Not for me, but I understand the strategy of MDN. I think that I should remove too old key value mapping. So, no problem.

Thank you, everyone!

Comment 11

[Eric Shepherd [:sheppy]]

I will classify this as FIXED then.