-
Notifications
You must be signed in to change notification settings - Fork 3
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Suggestions from gtmetrix.com #11
Comments
Are you open towards replacing Apache with this transition as well? With Caddy you get common mime types by default when enabling gzip/brotli, all you need is one line:
You could also change the static assets location to be I understand if you're more comfortable with Apache and rather manage such config more explicitly/verbosely though. The caching concerns can generally be handled by a build tool importing your assets and using a content hash for the filename on the build output, thus the name itself always matches the intended content and can be cached indefinitely (or some other long duration) without concern. This doesn't appear to be something Hugo supports at a quick glance? (or the documentation isn't particularly clear about it) If that's not an option, a post process could probably be run to generate the content hashes and replace the original file names from the build output as a workaround if this approach was desirable. I think Hugo is offering something similar via the |
@polarathene Hello again. And thanks for all your pointer here.
Unfortunately, given your efforts here on the Caddy front, No. For a variety of reasons Apache is our go-to currently and as-is we have everything working, if not optimised.
This is definitely an element of our continued use as it servers all our needs so I'm keen on the not fixing what isn't broken approach. Given that all non trivial software is at least broken in some way or another. But in this case Apache is an at least semi known quantity and we can leverage the the time/effort we have already invested in it's config. Your pointers/investigations re Hugo are interesting. We arn't really a major traffic site so I'm mostly interested in keeping things simple but not slow if we can help it. Currently we are way under-optimised and I copied this issue in from ages ago when this repo was private and post a bunch of clean-ups we did prior to the publishing. Such as the removal of some early google analytics etc. Another element here is that our images in this repo have been somewhat scaled down from the currently in-play ones on the production website but bar that it's pretty much what is 'live' currently. I'm not that familiar with Hugo myself just yet but it's intended use was primarily to ease contributions and translations and dig us out of the mess that we currently have; while maintaining our existing urls if possible. Thanks again for your input here, much appreciated. We have to try not to be too ambitious re optimization but I did want to pop in what we have already. Plus these tests were carried out prior to our now complete https transition so at least we have that one (mostly) sorted. Bar the mixed content issue of course: |
Note the following was based on a test performed prior to our move to https. Some factors may have changed but copying in here from our prior closed repo to keep a note of this performance orentated test facility.
Note also that we are likely to gain some sanity / speed from completing the following issue in this repo:
Migrate to Hugo #1
So we may be better off returning to this issue once we are Hugo generated.
Apparently we are a little behind the times: From a quick test we have have a couple of easy speed-ups:
https://gtmetrix.com/enable-gzip-compression.html
In .htaccess
Potential issues, we also serve Rock-ons and older CentOS base repos from this domain.
And:
https://gtmetrix.com/leverage-browser-caching.html
again in .htaccess
and:
The text was updated successfully, but these errors were encountered: