Ok, I keep thinking about the $6k figure that @Puttitout mentioned in his thread.
For some reason I'm not seeing the costs here.
1. Bandwidth
2. Hosting
3. Cloud flare
4. Dev time costs
5. Legal ???
But we don't know because he didn't come out and say what it is in reality.
So let's say that we have enough people here who could help with programming. Depends on the language.
We have people who could help with server management.
Those who can't or are afraid of don't can and are probably willing to click a few ads per day.
Member management can also be done by a committee but ... Umm ... Yeah ... I get it. The point is that we can source this too.
But !!! I'm gonna say that he wants out of his day job and make this go big like Reddit. We are that close.
So, let's be a bit pragmatic here ...
Nobody likes .net so we go to another language maybe one that has the most goats able to work in; screw the ideological argument of what language is best -- just get the most guns on board.
Once the language is chosen, decide if a framework fit our needs.
Then we move to project manages who lay out a time line and just keep an eye out for areas that aren't staying on track.
I don't see why voat can't be converted to another language within weeks. I do laravel work and could probably have the majority in place within a month if I didn't have to go to work. The database structure is set, the core functions are all there so it's just rewriting those functions in the new target language. Some parts tend to JavaScript / node really well and others need a back end for api endpoints.
Once could starts getting submitted we get the DB guys to evaluate queries. And we get the security guys to hammer the could for all possible entry vectors.
How many people are we technically looking at? Max 20, more likely 10.
The biggest issue is starting paralysis and arguing about what language to use. The project managers can focus on functionality and ranking them in order which helps the coders to stay on target.
Once a beta is in place, we start looking at the marketing group to push for advertisements and social media spamming to pull in more people to the site. Ads can work and I think many of us can live with that ... Maybe have a on going outside link to an Amazon affliate that we go to before shopping at amazon. This could and probably should be a separate site with no voat branding to keep it low key.
Finding a hosting company can't be that hard and I can't see why a $250/month machine run everything. Bandwidth and content legality is the big problem.
I have no idea about cloudflare but I'm sure somebody here knows about this and we try to keep those costs down and still stay secure.
Writing all that out it makes me wonder why the community hasn't been called upon unless it's a business thing which is cool by me because all the work put into this site does deserve some kind of payout.
What do you all think?
view the rest of the comments →
caprep ago
What is the biggest cost right now? Bandwidth? Memory? CPU? Disk?
If we knew the biggest issue we could focus on that first.
Is Putt running the same sized instances all the time, or scaling up / down instances as traffic demands it?
Suggestions:
Bandwidth - Make all images WebP and save ~25% of image bandwidth right away.
Bandwidth idea 2 - Minify the HTML. There are a lot of long class names in the code and extraneous spaces. I'm sure we could easily get a 20% bandwidth savings from the main HTML.
Disk - Take old posts, put them into a torrent and delete them from voat. If people want them they can host them.
CPU - Most of the visitors probably aren't logged in. Why not compile a static HTML page every 30 seconds and save it to disk. When a non-logged in visitor views the main page give them the stored one. Same with comments on popular articles. The page doesn't need to be up to the second
Memory - Most all of the suggestions above should help, especially the pre-compute option.
There are a number of quick things we can do to relive the pressure, we just need to know where to aim our autism.
WhiteRonin ago
Yep, yep and more yeps ...