docs: image base reference #545
Conversation
.tool/genheader.go
Outdated
| func main() { | ||
| version := specs.Version | ||
| if strings.HasSuffix(specs.Version, "-dev") { | ||
| version = "master" |
There was a problem hiding this comment.
Why not just keep the *-dev version?
There was a problem hiding this comment.
Why not just keep the *-dev version?
Ah, because it won't work for <base>. Maybe use specs.Version directly in <title>, and then use the HEAD commit hash in <base> when specs.Version ends with -dev.
There was a problem hiding this comment.
lemme take a stab at that.
Makefile
Outdated
| $(OUTPUT_DIRNAME)/$(DOC_FILENAME).pdf: $(DOC_FILES) $(FIGURE_FILES) | ||
| @mkdir -p $(OUTPUT_DIRNAME)/ && \ | ||
| $(PANDOC) -f markdown_github -t latex -o $(PANDOC_DST)$@ $(patsubst %,$(PANDOC_SRC)%,$(DOC_FILES)) | ||
| $(PANDOC) -f markdown_github -t latex -o $(PANDOC_DST)$@ $(patsubst %,$(PANDOC_SRC)%,$(DOC_FILES)) && \ |
There was a problem hiding this comment.
No need for && chaining. Make will die if any individual command fails.
There was a problem hiding this comment.
i don't recall that to be the case. It's not like there is an implied set -e inside the target
There was a problem hiding this comment.
i don't recall that to be the case. It's not like there is an implied
set -einside the target
$ cat Makefile
foo:
false
echo "still going?"
$ make
false
Makefile:2: recipe for target 'foo' failed
make: *** [foo] Error 1
If that's not convincing enough, I can try to hunt down some docs for you, just let me know.
There was a problem hiding this comment.
heh. fair. my memory fails me on the need for painful amounts of new-line escapes for long one-liners. No biggie.
Makefile
Outdated
| @mkdir -p $(OUTPUT_DIRNAME)/ && \ | ||
| cp -ap img/ $(shell pwd)/$(OUTPUT_DIRNAME)/&& \ | ||
| $(PANDOC) -f markdown_github -t html5 -o $(PANDOC_DST)$@ $(patsubst %,$(PANDOC_SRC)%,$(DOC_FILES)) | ||
| $(PANDOC) -f markdown_github -t html5 -H $(patsubst %,$(PANDOC_SRC)%,header.html) --standalone -o $(PANDOC_DST)$@ $(patsubst %,$(PANDOC_SRC)%,$(DOC_FILES)) && \ |
There was a problem hiding this comment.
$(PANDOC_SRC)header.html should work fine, and it reads more clearly (to me, anyway).
vbatts
force-pushed
the
html_img
branch
3 times, most recently
from
5cf770b to
cec7700
Compare
|
updated. PTAL. |
|
On Wed, Feb 01, 2017 at 02:39:02PM -0800, Vincent Batts wrote:
updated. PTAL.
cec7700 is still using ‘master’ for *-dev versions. I'm happy to take
a stab at my HEAD approach [1] in a PR against this one (or a
follow-up PR?) if that would be helpful.
[1]: #545 (comment)
|
|
oh oh. You're saying to get the commit and reference the precise remote head commit. hrm. |
|
fixed. Now a |
|
On Thu, Feb 02, 2017 at 08:11:47AM -0800, Vincent Batts wrote:
… which pre-supposes that the commit exists on github.com…
Yeah, but that seems more likely than assuming a similar-enough image
will still be in master.
Anyhow, 7692fe4 looks good to me.
|
|
LGTM |
|
needs rebase |
|
@vbatts Please rebase! |
|
rebased, but will need to rebase again on once master is fixed ... |
Fixes opencontainers#247 this makes an ephemeral header that gets included into the html doc that bases the images path to pull from the same tag on github. Signed-off-by: Vincent Batts <[email protected]>
|
rebased |
|
LGTM |
|
LGTM, lol, whatever |
|
@jonboulle you mad |
|
@vbatts dat html/pdf generation |
5 participants
Fixes #247
Fixes #435
this makes an ephemeral header that gets included into the html doc that
bases the images path to pull from the same tag on github.
CC @RobDolinMS @wking
Signed-off-by: Vincent Batts [email protected]