4 ms·

I use Pelican[1] (Python) SSG and they are adding (have added?) a feature to skip articles that have not been touched. That makes compilation really fast. [1]

by geekam 12y ago

I use Pelican[1] (Python) SSG and they are adding (have added?) a feature to skip articles that have not been touched. That makes compilation really fast. [1] https://github.com/getpelican/pelican https://github.com/getpelican/pelican

scott_karana 12y ago
I appreciate that they have that feature, but haven't makefiles had the same feature since the late 1970s? I love the idea of a static site generator, but they're really, really reinventing the wheel...
- pdonis 12y ago
  > haven't makefiles had the same feature since the late 1970s? Sure, if you're willing to use make as your SSG. :-)
  - scott_karana 12y ago
    Nothing prevents you from using make (or rake) to invoke an SSG... :-)
    - pdonis 12y ago
      Only if the SSG can operate at the single-file level. Most SSGs operate at the entire-site level; there's no way to invoke them and tell them "just rebuild this one page", which is what you'd have to do to use make or rake as your change-detector. Also, I'm not sure how well make or rake would work with version control, since make/rake, AFAIK, look at file modification times to detect changes, and DVCS pulls clobber those. In other words, suppose I make a change on machine A, re-generate the site, then push; then I pull from machine B. Make on machine B may not be able to tell that the site was re-generated, because the file timestamps are not preserved by the push/pull, so machine B is seeing different ones than machine A did.
      - scott_karana 12y ago
        It's true, yeah. Anyways, my point wasn't that make is sufficient (it isn't), but that a lot of SSGs have been pretty embarrassingly feature-sparse. Understandable, I suppose. :)
phloxicon 12y ago
I'm really interested in that feature. Do you know which issue that is on the tracker?
- geekam 12y ago
  Some work was done here : https://github.com/getpelican/pelican/issues/224#issuecomment-34637913 https://github.com/getpelican/pelican/issues/224#issuecommen...

by geekam 12y ago

scott_karana 12y ago

I appreciate that they have that feature, but haven't makefiles had the same feature since the late 1970s? I love the idea of a static site generator, but they're really, really reinventing the wheel...

pdonis 12y ago
> haven't makefiles had the same feature since the late 1970s? Sure, if you're willing to use make as your SSG. :-)
- scott_karana 12y ago
  Nothing prevents you from using make (or rake) to invoke an SSG... :-)
  - pdonis 12y ago
    Only if the SSG can operate at the single-file level. Most SSGs operate at the entire-site level; there's no way to invoke them and tell them "just rebuild this one page", which is what you'd have to do to use make or rake as your change-detector. Also, I'm not sure how well make or rake would work with version control, since make/rake, AFAIK, look at file modification times to detect changes, and DVCS pulls clobber those. In other words, suppose I make a change on machine A, re-generate the site, then push; then I pull from machine B. Make on machine B may not be able to tell that the site was re-generated, because the file timestamps are not preserved by the push/pull, so machine B is seeing different ones than machine A did.
    - scott_karana 12y ago
      It's true, yeah. Anyways, my point wasn't that make is sufficient (it isn't), but that a lot of SSGs have been pretty embarrassingly feature-sparse. Understandable, I suppose. :)

phloxicon 12y ago

I'm really interested in that feature. Do you know which issue that is on the tracker?

geekam 12y ago
Some work was done here : https://github.com/getpelican/pelican/issues/224#issuecomment-34637913 https://github.com/getpelican/pelican/issues/224#issuecommen...