Question thread #77

[pauamma]

It's time for another question thread!

The rules:

- You may ask any dev-related question you have in a comment. (It doesn't even need to be about Dreamwidth, although if it involves a language/library/framework/database Dreamwidth doesn't use, you will probably get answers pointing that out and suggesting a better place to ask.)
- You may also answer any question, using the guidelines given in To Answer, Or Not To Answer and in this comment thread.

Comments

[devilbear]

I think this would be the right place...? Sorry if it's not! I was directed here from the support board, so I'm copy/pasting what I wrote there:

So, my best friend and I have been kind of unhappy with our current roleplay home but we can't afford to buy premium DW features for all of our accounts... and that means making our own DW fork would be an interesting option for us. I was wondering if the following things are possible on a self-hosted DW fork:

- Give all accounts access to "paid" features by default.
- Have one permanent invite code for us to use to make as many accounts as we want while locking out strangers and spambots.
- Import content from insanejournal and dreamwidth.

I'd also like to know what kind of hosting would be best for such a project considering we have:

- Many entries, as we've been roleplaying in our private game for eight or nine years now.
- 200+ accounts we'd like to create so our characters can have journals for each au etc. instead of being lumped together. Probably 100+ icons would go onto each account.
- At least two communities we'd want to create.
- 500mb+ of rp icons I'd want to self-host and share through an icons community, as I currently have an IJ icons comm but I'm hoping to shift away from relying on that server

Any advice would be very welcome! We're trying to decide if the project is affordable and practical, so please don't sugarcoat it if the requirements would be high priced or require too much coding knowledge.

[denise]

The codebase does indeed support doing all of what you want, and you'd probably be able to run it on a single cloud server pretty easily given the minimal usage you're envisioning, but I'd advise against it unless you're really, really familiar with Unix and Perl. It's not really an install-and-forget system, and although we keep trying to simplify the install process, it's not at all what I'd call easy for a beginner. (I've been working on this codebase for coming up on 20 years now and I still have to consult cheat sheets for doing a lot of stuff.) Importing is especially going to be a pain in the fucking ass, because the system is pretty well-tested by now but when it fails, it fails hard.

I mean, I like to think that we have the documentation decent enough that you'd be able to make a stab at it if you wanted, and it's entirely possible that you'd be able to get things working without too much frustration! But it all depends on your tolerance for "oh my god everything is broken and I don't even know where to start".

[devilbear]

Yeah, that's a toughie, because I know precisely nothing about Unix or Perl and "install-and-forget" is definitely what we'd prefer. On the other hand, I've been known to pummel all manner of tech-like things into cooperation and learn through trial, error, and google when necessary. Just, y'know, css/html and php and changing a hard drive - not something as complex as this.

You've given me food for thought, thanks!

[pauamma]

To elaborate on what Denise said, you should expect a time investment of up to... I want to say 5-10 hours per week (assuming nothing goes wrong - if something does, all bets are off) monitoring your system and keeping it running smoothly (including the email part which you didn't discuss in your initial request and can be significant when a Big Player(TM) decides you are spamming them, or perhaps just doesn't trust you because you're a newcomer) unless you can invest a lot of up-front time automating that monitoring, which the Dreamwidth documentation doesn't cover at all. If you're still interested, let us know, and we'll point you to the documentation that does exist.

[denise]

I don't really think they'd have to worry too much about monitoring for a two-person install where they don't need 100% uptime or reliability, "notice that it's busted when we go to use it and fix it then" wouldn't be too awful for that use case. But yeah, ugh, the email thing.

[pauamma]

Fair enough.

[devilbear]

The time investment would be less than ideal for sure, but more importantly you (and denise's addition) have shed light on an important concern my friend and I had completely overlooked re: email. Given we'd want to avoid the headache of gmail deciding we're spam and don't have the time to dedicate to upkeep, this probably isn't the best option for us.

I appreciate the honesty of your advice! And your willingness to point me in the right direction, had I decided it'd still be a worthwhile undertaking for our project.

[denise]

Ughhhhh I didn't even think about email until pauamma brought it up, but yeah, running a mail server (which you need to do in order to have the site send you email) is a fucking pain in the fucking ass. I've been running my own for 20 years now because I vehemently distrust anyone else with my email, and I have come so close to saying fuck it and switching to a commercial provider so many times I've lost track. If you and your friend use gmail, forget ever being able to reliably recieve email from your install, ever. Gmail makes up its own mysterious standards and enforces them arbitrarily without any documentation and without any recourse when they just decide to stop accepting mail from you because you've violated some requirement they've entirely made up without any warning. I hate them to bits.

I, uh, may have some very strong opinions here.

[devilbear]

Your strong opinions have likely saved us a ton of headache, though!

We both exclusively use gmail and would definitely be attempting to use the mail server, since we'd want notifications and the like. Everything else, I think I'd be willing to try and tinker with, but that sounds like a nightmare I would never want to dig into for something like roleplay - which already takes up free time and doesn't generate any income to make up for it. I can't even begin to imagine the utter hell you've fought with to get any kind of important, time-sensitive help from the big G. One of my friends bought a gift card directly from them once and it took a month of being led in circles by auto-replies and people who just quoted FAQs before she could finally get it escalated to have someone look into why the site kept telling her the code was invalid.

[pauamma]

In case you were wondering, GMail is still at it.

[denise]

0% surprised. I did finally give up and migrate my domains' email handing to somebody else, because dealing with gmail just broke me.

[pauamma]

*nod*