lovetoxbecause we dont save disco results for different resources
lovetoxacutally i dont know what we do in this case
lovetoxhm never mind its no problem :D
fmhey guys!
a while ago i wrote a small syntax highlighting plugin for gajim. have been testing it with some friends since then and just fixed a minor bug. i think, it can be considered working now ;)
if you are interested, I would be happy to add it to the plugin repo
andrey.gThat's good. I've relied on the existing omemo discovery mechanism.
lovetoxyeah though its more difficult with chat messages, because i cant really care about what resource is online
lovetoxonly because a resource is online that doesnt support omemo, doesnt mean the user wants to send the next message unencrypted
lovetoxhey fm
lovetoxthanks :)
lovetoxcould you make a merge request on dev.gajim.org?
fmnp ;)
lovetoxcould you tell me your username
fmjust signed up ;)
lovetoxi have to add you to the member list of the project
lovetoxah found you
fmlol :D
fmthx
fmgitlab is complaining when i try forking:
Limit reached Personal project creation is not allowed. Please contact your administrator with questions
lovetoxah one moment
lovetoxdoes it work now?
fmnope
lovetoxnow
fmok, looks better
fmthx
fmyap, worked
andrey.glovetox: do you plan to allow forking freely for all (like on github)?
andrey.gAnd the merge request is ready.
lovetoxam yes but only for gajim projects
lovetoxso no private projects that are not related to gajim
lovetoxusers who register a own account on gajim, can already fork 3 projects
lovetoxbut there is some restriction on people who login with github accounts
andrey.gYes, of course. I'm asking, because Asterix had to allow me explicitly to fork gajim and now it seems the same issue with gajim-plugins by fm
lovetoxso we have to manually update these profiles
andrey.gAha, now I see.
lovetoxyeah we talked recently about this and we set a default of 3 projects for new members
lovetoxits not a problem to up that, but we fear spammers
lovetoxa bit so we see how it goes
andrey.gSince there are gajim, plugins and nbxmpp, it is exactly what one would need maximum.
lovetoxhey i give you a 100 if you are motivated :D
andrey.gthis is used often in Golang. If you don't need all the results, you emphasize this by using _ instead of variable names
lovetoxniicee i needed this a few days ago
lovetoxand thought what a waste to have to use the vars
lovetoxandrey.g, another inconveinience with the plugin repo
lovetoxif i merge this, its like
lovetoxif i merge this, its live
lovetoxpeople will not get a update message because you didnt modify the version number in manifest.ini
lovetoxbut if someone downloads the plugin he will get this
andrey.gOh, haven't thought about this. 1.0.1?
lovetoxquestion is
lovetoxdo you need it live?
lovetoxbecause i wouldnt want to bother the people with update notes if there is not something that actually changes for them
andrey.gNo, I don't need it right now as release, if you mean this.
lovetoxok, so we will merge it at a later point when you have all together so we could add it as a beta test
andrey.gIf I'd had already support for download resuming...
andrey.gUnfortunately I've received no feedback on https://mail.jabber.org/pipermail/standards/2016-December/031737.html yet. Implementing this would require also changes in Conversations. Much work.
fmlovetox, am done importing the plugin from github
fmmerge request is open ;)
lovetoxfm, looks great, i hope i have time tomorrow for a test
lovetoxcould you squash all the commits into a single one?
fmno hurry ;)
fmif you prefer ;)
lovetoxfor a initial commit i think its the best
fmok
lovetoxthanks
lovetoxhave to go, have a good night all :)
andrey.ggood night
fmgn8
lovetoxandrey.g i will write daniel tomorrow and ask him if he can comment again on the issue :)
andrey.glovetox: Thanks!
andrey.gfm: If you allow me to make a suggestion, a good git practice is for every big feature to start a new branch, commit there and push this branch (never commit over master).
fmwell, i developed on github, so, i just merged in the code from there
fmbut yeah, i agree, better to use branches
fm;)
andrey.gThe mentioned suggestion is what one can find in github git flow.
andrey.gOK ;)
fmlol, i forgot to push the branch i used to import from github :D
fmandrey.g, do you know whats prefered? should i just start the merge request from the branch or should i still squash the commits into one?
fmok, tomorrow then ;)
fmgood night
andrey.gI'd say, if you already have splitted commits, it is very nice and in a separate branch they look as something distinct. For me personally splitted commits are useful, if changes are made at many places to have a better overview, what is related to what. Since your plugin is new, it might be not such relevant, as lovetox mentioned. (For history preservation and learn purposes I'd like the original.)
andrey.gGood night.
fmthat was my intention, to preserve the history of the github repo
fmeven though the history in this case is realy not that important ;)
fmanyway, tomorrow...
linusFwiw I find the practice of squashing commits with nontrivial changes together terrible :p
fmagreed
linusLike at work, I worked on porting our huge application from django 1.6 to 1.7. All of that work got squashed into one commit when it was merged in. That was later regretted.
SaltyBonesAnybody know what happened to this conversations muc?
SaltyBoness/this/the/
linusIt's still there...
linusxmpp:conversations@conference.siacs.eu?join
SaltyBonesGajim tells me "group chat does not exist"
SaltyBonesMaybe it's an s2s problem.
linusOr a gajim problem. Those aren't unheard of either :p
SaltyBonestrue but I have the same problem with conversations itself
lovetoxfm no you should not merge your PR yourself as it would be defeat the purpose of a PR so that people can test it before it goes live
lovetoxand every push to the plugin repo is instantly available to all users, there are no releases
lovetoxalso i would say to squash trivial commits, which are almost all your commits
lovetoxthat you moved files into a subdir, or how you build up the Readme file over months i would say has no real value for us later
fmlovetox, I will open a new merge request later with all the commits sqashed and the fixes for the instructions proposed by Asterix (they were still for my github version...)
fm"you should not merge your PR yourself" did not intent to do so ;)
fmgot a little late yesterday....
lovetoxthanks, i tested the plugin
lovetoxand its an amazing contribution, love it :)
fmcool :)
lovetoxi said this about merge yourself, because you could do it, gitlab doesnt give us much choice with permissions
fmwill see if i can find time to port it to gtk3, bt probably not before new year....
lovetoxgtk3 is in no hurry
fmok
fmcan I update prs in gitlab? or should i close and reopen it?
lovetoxno you can update
lovetoxjust do your changes, afterwards i push with git push -f
lovetoxand overwrite my repo
lovetoxthe mergerequest will always only show the difference between your branch and master
fmok, thougth it kind of stores the commits/diffs
lovetoxit works a bit of different in gitlab than on github
lovetoxi like it more on github
lovetoxwe have gitlab not for long
lovetoxso just do it, and see how it goes, we can close and reopen later anyway if we want
fmyap, gitlab seems to be even to new for google ;)
fmok, i will do it when i'm back home later
fmgot to go, later!
lovetoxfm i looked into your issue with message correction
lovetoxwith the way you modify now the textview its probably never going to work
lovetoxin gajim we have extension points
lovetoxmeans, gajim detects a pattern in a message
lovetoxlike a http link or some other formatting
lovetoxand before it prints this special text to the textview, it hands it off to an extension point
lovetoxyou can hook this extension point, and get from it the iter where we at in the textview, and the special text that was matched
lovetoxfrom there you just have to hand the special text to your encoding method and then insert the whole text at the iter which was also handed by the extension point
lovetoxi think we can save much code here in your plugin, and addtionel make all the other features like message correction work
lovetoxthe problem why you couldnt do this from the start is, that gajim doesnt matches your chosen pattern @@ as special text of course
lovetoxbut i think we will add this and with the next release of gajim we can do this more efficient
lovetoxbut message correction for now not working with the syntax highlight i consider a minor issue, so lets get this released :)
fmhi again
fmlovetox, I have to confess, I did not dig into gajims plugin system very much. I wanted to get it to work asap, so I just took the latex plugin, which obviously also is the foundation of the @@-wrapping ^^
fmthe latex plugin probably also requires this fix then
arunefm: which plugin have you developed?
fmsyntax highlighting for source code
fmnot yet in the official repo, got to fix the pr first ;)
fmhm... for some reason I can't squash the initial commit (a56f5f7). anybody knows why?
arunefm: right, nice
lovetoxhm in which direction do we squash
lovetoxolder into newer
lovetoxor newer into older
lovetoxprobably everything has to go into the inital commit
lovetoxyes newer into older
lovetoxso you can squash your first commit, because there is no older comit
lovetoxso you cant squash your first commit, because there is no older comit
lovetoxfm
fmyeah, trying it that way
lovetoxso set squash on everything that is not the initial
fmarune, thx, I am happy, people like it
fmerror: could not apply a56f5f7... Initial commit
fm-.-
lovetoxhm i will try
fmlovetox, I could remove all the commits and start afresh
lovetoxwait a moment
fmjust playing around locally so far ;)
lovetoxno i think its because
lovetoxof this merge commit
lovetoxthe last from github
lovetoxthat is weird
lovetoxdo we need this commit?
lovetoxdoes it add something?
lovetoxor better do we need all the other ones before?
lovetoxah no thats not it
lovetoxhm
fmas i said, guess its best I drop the whole thing (last 11 commits, 10 on the branch, one merge) and add it again
fmi think its because i added a unrelated history
fmso git *maybe* treats the first commit in this bunch different from the rest?!
lovetoxi think it has to do with the folders maybe
lovetoxbecause you didnt originally clone the repo
lovetoxyou had a own
fmany reasons to research it further? will removing all my commits brake something in gitlab?
lovetoxno, but try this
lovetoxi think i have it
lovetoxdrop all commits except the merge one from github
lovetoxthe last one
lovetoxand check if everything is here
fmwith "the merge one from github" you refere to a56f5f7 or to 5c125d6 (just to be sure ;) )
lovetox5c125d6
lovetoxbut this doesnt show up
lovetoxnah do it as you wish
lovetoxi just thought there is something to learn here
lovetox:D
lovetoxyou cant break anything if you operate only with merge requests
lovetoxi think the problem is definitly the folder structure
lovetoxyou had in the first commits no syntax highligte folder
fmwell, think I learned not to start this in the night after some hot wine ;)
fmi can squash the commits before the creation of the folder with the one that created it
fmjust not with the very first one in the history
lovetoxso whats now
lovetoxdid you lose your commits?
lovetoxbecause there are not anymore in your repo
lovetoxi think i know what was needed
lovetoxskip all commits in the rebase
lovetoxthen at the end it will pull this merge github commit into it
lovetoxwhich is not listed as a normal commit in my log
fmi cleaned up (aka deleted) everything, now I will just cleanly add everything again and also already fix what Asterix and you proposed
fmits still weird, never had an issue like that with git
lovetoxits defintily because of the folder
lovetoxyou made the correct folder structure in the last commit
fmmaybe you're right...
lovetoxah damn, i wanted to play around with that
lovetoxbut now your repo is empty :(
fmwell, i still have a copy of the repo ^^
fmhttps://xkcd.com/1597/
fmcan put it in an archive for you to play with it ;)
lovetoxhaha
lovetox:D
lovetoxthats so true :D
fmdefinitely
fmhm, i had a look into gitlabs pr edit page. seems i cant change the source branch of the request. should i push everything to master to reuse the pr or just open a new one?
fmlovetox, question about your comment to add self.activatable: is there a reason you used __import__ in the "test" if axolotl/protobuf are available?
fmor just because you don't realy need the import in this place?
lovetoxthere is something horrible called demandimport.py in gajim
lovetoxit delays imports of modules
lovetoxwhich breaks als most modules
lovetoxwith __import__ demandimport cant delay the import
fmlol.
lovetoxbut i would only use it at the start for the external module
fmyeah, i think i will just stick with "normal" imports and surround them with a try-except block
lovetoxbut test it at the end
lovetoxuninstall the pygments
lovetoxand try if the plugin loads
fmi will.
lovetoxand only to say it again, you need this only for pygements, we can safely assume the rest of the libs are there
fmsure
fmdamn, removing the commits broke the comments in gitlab
lovetoxyou have it as email or
fmwhat was the unused import you have been talking about?
fmyeah, but the context is missing (if there was any)