Reducing the code size

[Sa'ar Zac Elias]

A part of a discussion with Fred:

Clearly the code is constantly getting bigger. So how about reviewing the code and try to reduce it: use 1 and 0 for true and false when possible, remove the "defined" configuration when possible (leaving just comments about them of course) etc.

[Alfonso Martínez de Lizarrondo]

I think that all the true/false -> 1/0 conversions should be done at package time. That way the code remains as is and it will be future-proof.

[Frederico Caldeira Knabben]

@alfonsoml, I think we're using the strict equality operator in some places, which would break things. Maybe better to have the 1/0 thing is a coding style best practice, so we have better control of it.

[Sa'ar Zac Elias]

These changes should become routine for the core developers as well as the contributers, so adding documentation regarding flags and initializing of config settings is recommended IMHO.

[Garry Yao]

Isn't this suppose to be a [MicroChanges micro-change]?

[Sa'ar Zac Elias]

That was my initial thought, but Fred asked me to open a ticket for that and put the patch on review.

[Garry Yao]

In which sense does the lazy config option initialization style contribute to performance? Anway some of the styles configs are also used by 'pastefromword' plugin thus get broken. Besides, some of the changes in the patch are CKPackager's task instead, e.g. quotes transformation, simple block braces, check ​this out.

[Garry Yao]

For me this's pretty much a packager ticket really, the 1st feature I would vote would be an invariant variable feature that caches reference like editor.lang.common.generalTab will contribute a lot to the reduction.

[Sa'ar Zac Elias]

Replying to garry.yao:

Besides, some of the changes in the patch are CKPackager's task instead, e.g. quotes transformation, simple block braces, check ​this out.

These are just some coding style fixes I added.

the 1st feature I would vote would be an invariant variable feature that caches reference like editor.lang.common.generalTab will contribute a lot to the reduction.

I agree, that could be great, but let's do it ourselves in the meanwhile.

[Garry Yao]

the 1st feature I would vote would be an invariant variable feature that caches reference like editor.lang.common.generalTab will contribute a lot to the reduction.

I agree, that could be great, but let's do it ourselves in the meanwhile.

Moved to #6475, so don't worry about it now.

[Garry Yao]

1kb of footprint reduction is produced after the patch.

[Sa'ar Zac Elias]

Fixed with [5949]. Additional minor fixes should be commited as micro changes.
These small things should become routine, let's have a documentation for them.

[Frederico Caldeira Knabben]

A note... it was wrong to have such big change in a micro release. That's why it has been targeted to the 3.5 earlier. More attention is required in the future.

Another note... just to remember: the "this" trick is not needed. CKPackager automatically replaces "this" if there are enough occurrences of it to compensate it. Anyway, there is no need to change what has already been changed.