Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
-
@frankot Yeah, keep repeating that OSS mantra like the good fanboy. Have you ever tried it yourself? For just about every problem I've had I found an issue from a few years ago, and most of them were closed with nothing but a comment saying something like "Blocked by X" (with X done long ago) or "See [dead link]" or "Planned for [a version released 2 years ago]". I've even found (and tried) a bunch of stale / dismissed merge requests which solved some problems after merging & building them manually, yet they've been there for a long time without being accepted. Besides, some projects only have a locked SVN nobody can contribute to.
Not to mention complex environments and lack of build instruction on a project with 3000+ files in 10+ languages, yeah right I'm going to spend weeks setting it up and figuring everything out just so that my merge request gets ignored like so many before me. -
@spongessuck Among many, also quite a few things I've tested use it under the hood and just add some management UI.
-
@melezorus34 Fair point, though I didn't even bother with unmaintained projects, everything I tried has been actively contributed to in the last couple of months.
-
@hitko well, most such oss is free to use, as in you do not pay for it.
That means that unless you contribute to it you actually cannot make demands.
They offer their work for free and on shit on them for it, not really the best way to motivate them.
Many projects are started as a hobby and maintained in spare time, often its external contributions that add the extra nice to have features and that help the project to spread.
So if you cannot find any good oss video conference, pick one and help them be a little better, your help can be what makes the next one select it and do their contribution and once more get involved you get payback in a better product.
Problem is that project that is more niche often have a hard time to get support as many hobbyist do not have a need for them and therefore do not spend time helping them. -
@Voxera I've explained a part of the problem in a previous comment, i.e. projects being too complex / lacking directions for someone from the outside to contribute to, and project owners ignoring contribution. I'm always willing to contribute to a project if it doesn't take me days just to set up the build environment and figure out where my code should go.
To address the other part, almost every solution I've tested (well at least the main ones, I tried a bunch of wildcards before finally giving up) has a hefty backing from some company which provides commercial support & features. -
@sleek or too cheap to actually work on it
-
@frankot I totally agree on that part; I'm complaining more about the mentality of OSS projects than the software, e.g. "We demand 100% test coverage on merge, but we don't need to automate build because Joe has it all set on his machine already and we can just message him on Slack. Also we need to test each commit manually on our machines, so we won't merge it until we feel like building it for half an hour." Or maybe it's "We never bothered decoupling UI from the rest of our code, so we'll just keep that crappy look until people forget this product even exists. Yay time well spent!"
-
@hitko I'm still stuck on manual review.
-
@hitko ok. If there is a company backing it does change a bit of cause.
But still, as long as its free its still no use complaining ;) -
@hitko It seems without a decent initiative from capital owners, a whole lot of OSS projects just don't realise how much more they could focus on actual software by organising and automating their processes.
-
@Voxera I know, I know... The paradox here is how
- so many people advocate it over paid software, yet
- they acknowledge you get what you pay, yet
- they say it only takes a bit more effort to set it up, yet
- they don't account for that extra time in the final costs, yet
- they say how OSS developers take their valuable time to offer something for free, yet
- they don't recognise how a decent project structure and processes would allow everyone to spend more time on valuable contributions, yet
- they enforce "good practices" like test coverage, yet
- their code still fails at some basic real-world use cases, yet...
I wouldn't really care too much if I only had a bunch of crappy OSS alternatives to choose from; at worst I could say "This paid thing has these important features OSS doesn't." Now I have to make it work with OSS because it clearly offers all crucial features and there's no way I can explain how fucked up maintaining this thing is to a non-tech person. -
There are some proprietary solutions that are Bullcrap as well. Does that make proprietary solutions bullcrap?
Does a bullcrap software written in Python makes all software written in Python bullcrap?
Please stop generalising.
Yes there are many OSS that are terrible - but a substantial number of them are as robust as one can be. -
@TheKalpit I think it's clear that many different projects were trialed and none were satisfactory.
-
Normally I'd say "Contribute if you don't like the current state".
But I won't, because Linux development has been so much fucked over the last years, mainly be GNU and systemd.
Everything now depends on logind, Glibc, Bash and systemd-units.
Especially new developers won't follow standards for compatibility, because these projects don't follow them too.
The big distros also don't get tired spreading that bullshit, infecting newbies before they get the chance to learn it properly. -
Finding the perfect publishing model is like finding the perfect programming language. In some areas and situations it works really well in others it does not. As you may have noticed. Creating and maintaining a project alone takes a lot of effort. If there are a lot of beneficiaries. Even corporate contributions because either open source is required for security, adaptation or no viable/affordable closed source solution available it tends to work really well. It is where all the prominent successes come from. Another reason that is under utilised in my opinion is when a problem/field is to complex it should be a collaborative standardized effort that benefits all and moves it along.
The Linux kernel is an example that bit all the spots.
There are obviously a lot of situations where open source projects are just enough to satisfy the creators needs.
There are a lot of situations where keeping software proprietary is in the best interest to the company and thus to everyone (else the well working solution would not exist). That still hurts as an open source fanboy but can't see it any other way. -
@hitko Sounds like you don't maintain open source software?
I do, it's hard and sometimes super annoying and boring.
Most of the time, I don't have the motivation required to make sure the quality is maintained.
Related Rants
This project must have taken a toll on this guy
Why the fuck open source solutions need to be such a load of bullcrap? I've spent a week trying to set up every single self-hosted video conference software, and the only thing I've got is a shorter lifespan.
How the fuck does your (judging by GitHub, well maintained) software only support Ubuntu 16.04? And I mean ONLY, there's no support for docker, or any other distro either, and we're only weeks from getting the second LTS since 16.04. And why the fuck does documentation tell me to manually go through 20 different config files just to enable SSL?
Why the fuck doesn't your official AWS cloudformation template include VPC or other required parameters? I've had to rewrite the whole thing just to get a valid stack you dipshit!
And how fucking hard is to make your software look decent, I can't expect clients to chat with me using something that looks like an incest child of 2003 MSN and eDonkey?
Oh, and it'd be fucking dandy if your documentation wouldn't return 404, maybe I'd be even able to test what your product has to offer?
I guess after everything I've tried I'll go with Jitsi; it seems the most decent, although it lacks some pretty basic features like limiting chat features for guests.
rant
video conference
oss