It seems like every time #pixelfed eeks closer to a v1 release and that (comparatively) boring polish stage where notifications are fixed, remote profile avatars are loaded.. discover pages correctly calculate posts (including individual images from albums/multi-image posts), discover federates..see more replies link is a link and not just underlined text.. a properly developed and documented api..

… a flurry of new features that subjectively should be post v1.0 are rapidly advertised..

.. or the MicroUI is teased (yet again)

Mr Dan S Up.. I love your project and support your work, I frequently contribute (via feedback and issues raised) and I get that introducing new features is much more exciting than the grind of functional fixes.. but 1.0 isn’t going to arrive unless the fundamentals all work as they should (across platform, DB flavour etc.)

Maybe you could concentrate in competing with instagram on a feature basis when you a solid, reasonably bug free and perhaps a slightly feature-lesser form




Raised an issue (https://github.com/microformats/h-entry/issues/19) to introduce a microformats property that could translate to content warnings if syndicating to #fediverse sites.

Hopefully, a bit down the road it can be implemented as I’m currently apprehensive to post things like photos of food to my site because I know that when syndicated to my pleroma account they wouldn’t currently be collapsed.

Being able to flag sensitive content at source and have it reflected as such at the destination would, I think, be a nice addition.




Just configured and enabled ekster as my endpoint which should, I hope be quicker than aperture.. however it seems that the initial posts that it pulled from the various feeds have vanished..

Testing, yet again.




Took a few minutes this morning before starting work to troubleshoot the issue I was having with instagram import on my pixelfed instance .. and I got it working.. so 137 new old photos uploaded.. very glad they are unlisted during this process!

Almost all of the photos are alcohol or vegan food\cooking related with a few stray snaps of early parenthood and a dug.

Spotted another directory with even older instagram photos from 2010-2012 but without the json files so not sure if I will manually import those or not.. I might cherry-pick some.




Last week I raised a couple of tickets with the brid.gy dev, one of which related to truncation length for fediverse using the mastodon api but with larger character allowances than the default 500.

Ticket here: https://github.com/snarfed/bridgy/issues/953

In my case its for pleroma which has an additional max_toot_chars field and on my instance this is set to the default of 5000 characters.. I think perhaps misskey uses this field too..

Anyway, this post serves as a test of the implementation sitting as it does at 541 characters.




I just raised a couple of feature request tickets for brid.gy.

The first is to check a page for a specific element and if extant pass on as a content warning to the mastodon api.

The second is to read the server’s character count limit at the time of connecting an account and using it as a basis for truncation rather than apply a 500 character limit across the board.

I’m not sure how doable these are but seem reasonably simple and would be handy.




OK, I think I’m reasonably happy with my site now..

I finally found out how to deal with hyphens in frontmatter in hugo (the documentation of which is a lesson in frustration).. Now to see when I hit post if reverting my micropub endpoint hacks have broken anything..

I’m now using a webmentions hook to deliver interactions to the site.. though, I want to fiddle a bit with nginx to move some of the php files to a non-public directory..

I’ve given the html a semantic overhaul which will hopefully make it reasonably accessible though there will be more work to do there (particularly around captions and alt tags for images).

In this vein I also removed all floats (I think) and am using grid and flex for layout. Have dropped a bunch of excess html and css too..

Have hidden my switch to 日本語 button until such time as I actually have a reasonable amount of Japanese or bilingual content.

Still a work in progress but much further down the line than previously.


Categories:

Giving webmention.io hooks a go for dealing with mentions as I seem to have lost some with the current process.


Categories: