138172 - mozilla.org documents modified for moz.zope.org should be checked in gila

Status: RESOLVED DUPLICATE of bug 151557

(Developer Documentation Graveyard :: General, defect)

Description

[Fabian Guisset]

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.9) Gecko/00200203
BuildID:    20020417

While migrating some content from the existing mozilla.org website to the new
moz.zope.org website in order to get a demo up and running, I had to correct the
markup of most of the documents to accomodate the stricter markup/style rules of
the new website. I also did some very minor modifications.

Fantasai told me that all modifications should be checked into the current
mozilla.org website so that they don't "wordrot" (the documentation counterpart
of the code "bitrot").

I am ok to check those modifications in gila.mozilla.org if people want them.

Modifications include:
- HTML4.01 strict markup
- No inline style (only using the persistent-style.css stylesheet)
- Semantic instead of presentational markup
- Relative instead of absolute links
- Better markup formatting

Reproducible: Always
Steps to Reproduce:
1. For example, go to http://mozilla.org/hacking/nutshell.html
2. Then compare with http://66.123.57.58/mozilla/contribute/hacking

Actual Results:  the moz.zope.org version is better

Expected Results:  the mozilla.org version should be of the same quality

Note that this is quite a bit of work on my side, so I will only do it if
mozilla.org is interested in those modifications. I can produce diffs if people
want them. Feedback is welcome.

N.B: should anybody else be CC'ed?

Comment 1

[Gervase Markham [:gerv]]

If no content is changed, and if (if you make major changes or use advanced
styling) you check that it still works in Mozilla (current), NS 4.x, IE 4 and
up, and Opera, then I would be inclined to say to just check them in on top.

endico knows the procedure for getting CVS access to gila.

Gerv

Comment 2

[Fabian Guisset]

I already have cvs checkin access to gila.
Of course I would only checkin documents if they render well in the different
browsers. Thanks for the feedback.

Comment 3

[Gervase Markham [:gerv]]

So if you aren't after checkin access, what's this bug about? If you have better
versions of the docs, check them in. As far as I'm concerned you don't need to
ask permission to do that :-)

Gerv

Comment 4

[Dawn Endico]

yes, you should ask permission.

switching documents over to the strict doctype means you're signing up
the document's maintainer up for a bunch of extra work to make sure that
later updates to the page continue to validate. We don't want documents
that claim to be a certain doctype but are not.

which documents are you talking about?

Comment 5

[Gervase Markham [:gerv]]

I don't see anything in the original post about changing the doctype on the
documents - just about changing the markup so that it does validate with that
doctype. Most of our documents don't have a doctype, because they use the
wrapper, which doesn't have one. 

Or have I misunderstood Fabian?

Gerv

Comment 6

[Fabian Guisset]

Gerv is correct, I'm not changing the doctype, just the markup, which would
validate _if_ we changed the doctype to a strict one. Gerv (via Henri Sivonen)
has done such modifications in the past (see bug 93742)
I guess I just filed this bug to make sure it was ok to check in those
modifications.
Docs I have modified are mainly in hacking/:
nutshell.html, reviewers.html, portable-cpp.html, ...
As I migrate more content to moz.zope.org, I will modify more documents, and the
current mozilla.org website would probably benefit those changes.

If neither of you think this bug is useful, we can close it, it's fine with me :-)

Comment 7

[James Cox]

we should only make changes if those changes do not require a doctype to be
defined. It's always good to ask the owner for permission to update (ie, for
active documents there may be a new one coming), otherwise it should be ok.

i also think it would be good if you posted a sample for general review.

Comment 8

[Daniel Wang]

what this bug asks for is essentially bug 151557

*** This bug has been marked as a duplicate of 151557 ***