ewong or IanN_Away does mar have a mimetype?

Hendikins: long time no see

oh wow Hendikins is here

I've always been here?

I haven't seen you in years

Been floating around, just haven't had much to say.

eyes old :p

if you aren't up to date.. my unchecked escilating asshole thing is been severely curtailed since I was totally betrayed and purged so most here have been enjoying that.. mentioning it tho not so much :P

i am sure a-865 would agree

"I wonder how you use stuff like instant messaging", wait, since when did instant messaging require browsing?

frg_Away: do you know the mozilla rdf mimetype?

str8 xml will work but I want it proper

well the official one is application/rdf+xml but i dunno if Mozilla's varient had a vendor mime

Found only text/rdf: searchfox.org/mozilla-esr60/source/…ed/webconsole/network-helper.js#414

likely did that so it wouldn't get passed through the parser when they still had it

well if RDF is still being passed through the xml parser as just xml then devtools might end up with the no-style rendered version and not the plaintext

application/rdf+xml is also present

i couldn't find anything for mar tho

instead of shitting out http headers for content-type when I can just get/set ini setting default_mimetype which is provided if one hasn't already been specified .. getting the function to update the "registry" value means i can always know what it is but i don't have to actually directly output it..

Thrr is nothing in libmar so probably nothing elsewhere too.

should have done that years ago but at the time i just wanted a small selection of headers i'd need to send out a lot.. so html xml xpi bin text plus 404 and 501

well mar is a terrible format and the post-xpinstall application update service just haven't ever been 100% reliable despite patches and THAT VERY FACT is why I do not subscribe to the notion of a universal "maintaince" service

I love Mozilla code.. but if remote xul outside specific targeted applications is dangerous as fuck.. so is the mozilla platform running as NT AUTHORITY\SYSTEM (or one of its token varents)

idc if updater is not xpcom code (exactly) i just find it unwise

It solves a thing or two if the user is not an adminstrator but otherwise I hate these "services". Unlike malware authors who like them :)

a user who is not an administrator should not be allowed by any means to change system configuration including program updates

no this is all about FxaaS

a user who cannot update isn't on a rolling cloud enabled release

yeah. And the stub installers are something which should go away too. I dislike all the stuff going on in background and hitting the net regularly.

THAT is why .. anything administrator-y outside that is just bonus manipulation points to push the ill advised change

Stub Installers should be an OPTION

secondary

NOT primary

OR primary as a graphical button with text link directly under

just as long as it isn't hidden away

I do need to replace the update service tho.. i want it to use the archive files I already produce driven by an xpi file.. We already added xz for use in mar compression so xz is IN the platform.. and i can readd 7zip as well

first order of business tho on that front is to hack the add-ons manager and add-ons update service to accept xpinstall type 1 again which would be a modified install manifest that would contain the metadata for the platform to feed into the updater instead of having two COMPLETELY different AUS methods

frg_Away: how are you in treeview controller scripting?

also I'd love to know why DOMi can't traverse to treechildren which are loaded via a data source

rather than markup

The use. Can do patches and figure out things but overall not very good. rdf data sources are on the way out. I think no use left in our tree but the removal has not happened for add-on compatibility yet.

^the usual not the use