Posts by Telesphoreo

dang youre quirky

Here are read/write speeds on the disk
Timing buffered disk reads: 3430 MB in 3.00 seconds = 1143.15 MB/sec

The connection from speedtest-cli was multi but here's a single one

Edit: I've decided to call this one a stalemate. I've tried with other VPNs on multiple devices and they all match with the speeds of my VPN. At this point I think it's just location more than anything else. Even the closest VPNs to me are still 5mbps down and 10mbps up.

Here's the new issue though. I decided to ditch OpenVPN entirely and switch to WireGuard. However, now I can't download the actual .conf files from https://vpn.telesphoreo.me

Basically you put the name and it'll give you the .ovpn (now it's .conf) file to import. However it's now returning error 403. I gave it full 777 perms and changed the owner from root:root to www-data:www-data

What gives with that???

This is from my Apache error.log

[Sun Nov 22 02:05:46.172092 2020] [:error] [pid 383217] [client 70.37.246.22:52161] [client 70.37.246.22] ModSecurity: Access denied with code 403 (phase 2). Operator GE matched 5 at TX:anomaly_score. [file "/usr/share/modsecurity-crs/rules/REQUEST-949-BLOCKING-EVALUATION.conf"] [line "93"] [id "949110"] [msg "Inbound Anomaly Score Exceeded (Total Score: 5)"] [severity "CRITICAL"] [ver "OWASP_CRS/3.3.0"] [tag "application-multi"] [tag "language-multi"] [tag "platform-multi"] [tag "attack-generic"] [hostname "vpn.telesphoreo.me"] [uri "/<redactedprofilename>.conf"] [unique_id "X7nHescczV8RVMPR5ZI2vAAAAAk"], referer: https://vpn.telesphoreo.me/

Edit: after digging through obscure posts on ancient forums I finally found the solution to unblock .conf on one vhost

        <LocationMatch "/">  
                SecRuleRemoveById 920440  
        </LocationMatch>  

in case anyone needs it and 920440 being the error ID from the Apache error logs

ProBoards isn't really a viable option anymore. It's half broken and it's really just an archive of everything at this point. Of course it's better than nothing but I'm just saying that switching back to our current ProBoards forum isn't really going to work well. I personally like Flarum and think we should stay on it.

I object. It's already implemented perfectly fine and there's no reason to rely on third party plugins that may not offer exactly what TF needs. The shop wasn't removed by Video in his own little world because it was bloated, but because he doesn't feel like it's necessary for all freedom servers. I think the shop is genuinely fine and if it ain't broke don't fix it

@wild1145#1544 I was using speedtest cli and just added the share flag. I'll see if it's single or multi. I can run a disk speed benchmark. If I actually recall the IO speeds on ovh were shit and the dedicated server was real NVMe SSD speeds

@StevenNL2000#1543 Yes both of them were on Ubuntu 20.04 but I upgraded to Ubuntu 20.10 which doesn't seem to have fixed the issue

So I have my own VPN. I primarily made it to bypass the ridiculous amount of blocked sites at school however I've properly adapted it into my server security. However one of the things that I don't understand is why my speeds are so slow.

Here's a speedtest from the server

It has a 1gig connection and it's a vps from ovh. But when I run a speedtest I only get 2mb download and 10mbps upload. My internet is 100mb download and 10mb upload. So the upload speed makes sense but is there any explanation for a such low download speed? I have a pretty much stock configuration. I have OpenVPN monitor installed on the server and I am the only one connected. Version is 2.5.0 I believe.

Any help would be appreciated because frankly can't pinpoint an issue. I had a VPN setup when I still used hetzner and I always maxed out and the actual internet speeds were the bottleneck not the VPN. The problem with hetzner is that the ping was pretty high at ~150ms. On ovh it's much more acceptable at ~60ms but that's explainable due to the location change. Is it just because hetzner was a dedicated server and ovh is a vps (not dedicated) and I'm just screwed? i don't really know how that works and if the network connection is shared between multiple vps's or not

@zeseryu#1502 Yes

Quote

@DragonSlayer2189#1491 Vouch, especially for the reddit, I had to make a reddit just for my tfm test server because otherwise it would flood my logs with errors. I think discord integration is fine, as long as you can enable or disable it, but there is still a lot of random shit in tfm that just doesn't need to be there (see /releaseparrots, /bird, /curse, etc.)

Yeah this was annoying. I was going to outright remove Reddit because I couldn't find a way to disable it. It kept requiring JRAW whether you used it or not. Seems to be a new bug that all the plugins have to be there or else it wont work ever since we went Aero-free.

Quote

@videogamesm12#1474 My aim wasn't to piss you off with that thread, but I get why you feel that. I don't exactly remember what I was thinking at the time (it was like 6 AM and I hadn't slept) but I do know I created the thread out of frustration mainly because I didn't know why Plex wasn't open source and requests to know why were denied.

As I already stated, that was none of your business. In the first place it was actually just because it was a fetus of a repository and didn't really have much. It was also because I firmly believe in keeping the people who criticize everything away from the people who are actually doing the work. This eventually changed as the idea of me resigning was becoming a reality. I wasn't going to open source it to set high expectations and then feel pressured when I resigned. Now that I have I don't really care saying that, but at the same time that was still none of your business. And by the way, 6 AM your time is 7 AM my time. If you go in my shoes you'll see how frustrating it is to get messages from you complaining about TFM and Plex as soon as I've woken up.

Quote

@videogamesm12#1474 The thread about how much TFM is bloated was going to be created as a way to highlight the bloat issue to the developers because I didn't think anyone really cared. This was mainly because nobody was being transparent about anything development related and hadn't been since the 2020 Ownership Crisis. I figured if I publicized the issue, it would hold the developers' feet to the fire and the issue would become a larger priority.

What I don't like here is how you're making it seem like the developers are a team and that's not true. I'm the only one who was a developer and, frankly, I had already told you all of this stuff. I told you TFM sucked and that was the point, I told you I would get rid of bloat, and I told you customizable messages and punishment history had already been tried to be implemented. I had already told you about this and then you go out to make a thread despite that you already knew this information. The reason there was no transparency was because 1. I didn't have access to the server and 2. nothing has been touched by me since Seth left. In terms of development, nothing was on the server and everything is open sourced so you could've seen the preparations anyways.

I'm sorry to sound rude but your post can be summed up in one word: entitlement.

I object. TFM should not be modified for a generic freedom server. That would, in it self, make TF a generic freedom server. If you want that to happen it should be in a separate fork and not affect the main TFM.

The shop can be disabled, not sure why it should be split out since it's 100% optional to use.

As I said about the useless features, that was going to be done. I already laid a roadmap out for you.

Customizable messages are a thing in Plex. I tried to make this in TFM but it failed. It was simply too late as the amount of messages in TFM was absolutely huge and would take an extraordinarily long time to make them all customizable. That seems like a waste of development time which is why it was going to be done right the first time in Plex.

What do I think?
I think you should think about what you're going to say before you post it.

@videogamesm12#1464 I understand your view point but TotalFreedomMod was made for TotalFreedom specifically. Not a generic freedom server. We try to make stuff generic enough, but that doesn't solve the core issue of TotalFreedomMod: complex code repeated in a bunch of spread out places. TFM is catered towards TotalFreedom and its up to other server owners to adapt it for their own needs. Plex was intending to keep the features geared towards TotalFreedom, but unify how everything was ran and make it much friendlier for anyone to jump in and work with it. To work on TFM, you kind of have to know where everything is and how everything works and when to add something where specifically you have to add it (often in multiple places, which can be daunting for a first time dev).

Right so now I'm at a computer so here's a more in depth write up:

This mainly has to do with the fact that I hate working in corporate environments. I already deal with it on a daily basis, and the fact of the matter is that TF is a nice break because it's not all corporatized. I like just hanging out and being able to have fun without worrying. Being a dev was fun and was actually rewarding. Even though every day I got a few new issues, it was still rewarding to solve an issue. I genuinely care about the server and did my best to keep everything fully maintained. I regularly kept everything up to date and sweeped out a bunch of redundant stuff. That was part of the fun. However, at this point, it's no longer rewarding and just feels like a corporatized chore. If we had a fuckup, it was the meme for the day and it was funny and we got over it.
This wasn't 100% determined by not having a panel. I was considering resigning before bringing it up with Wild in the first place, even over Discord DM's. Everyone from the community has been adapting to major changes. We went from ProBoards to MyBB to Flarum and lost the data 3 different times. We've all been adapting and I was perfectly fine to adapt to how Wild wanted to do it. At first everything was temporary and no promises were made or anything. However, the fact of the matter is, I'm simply burned out from being the only developer. I was burned out before, but I still made time for TF because it wasn't hard to just hop on and find a bug or do a quick patch. With the new system, it's just to inefficient for me. I'm sorry to sound like I'm complaining, and I've tried to adapt the best I can, but I simply won't settle for such inefficient methods on top of being burned out in the first place.
I started coding once again back in early 2018 when I helped TFM with the major 1.13 API changes. I spent almost every day over the summer helping migrate everything for the massive 1.13 changes. It felt rewarding, and even though I got treated like shit, I still knew that I helped a lot. I'm not going to do that again though. I'm not going to give TF everything I have. Not because I can't, but because I just simply don't feel like it. I can just feel a repeat of the whole TFM thing back in 2018. I'm not going to go through it again.

Thanks y'all. It was a genuinely fun time as being a developer and helping with everything. I'd still like to keep admin because that at least doesn't feel like it's a job. I can just come on and talk to people and manage the server if it gets chaotic. It's not some fancy job with fancy protocols and all that other crap. I'd also like to apologize for my rudeness yesterday and the day before. That's on me, but at the same time I feel like that thread was made out of pure spite of pissing me off. I know the server isn't perfect, but that's literally the whole point of why I'm here: to make things better. Making threads like TFM is bloated and open source Plex open source Plex don't help anything. I know TFM is bloated and making a big deal about it doesn't help. The November update was underwhelming (and ultimately never actually got deployed unfortunately). I was planning for a mid November update which fixed a bunch of bugs within TFM and would have made things better. The plan was that after the mid November update was done, to start getting rid of a bunch of obseleted and useless features in TFM. I was going to skip December for an update since that's holidays and time you should be spending with family (or, well, at least not hooked to a computer screen imo). TFM would have been cleaned up for the January update to tide us over until Plex which would have been done by mid 2021. Making threads to intentionally piss me off is a douche move, especially when I'm actually active and working on Plex and TFM at the same time as a one man job. If I was inactive or not doing anything, then I do think bitching about it would have been perfectly justified. But I was actively working on everything and making those kind of threads drains me and adds to how already burned out I am. I was planning on appointing (or having the community vote on) taah and super for being devs if they wanted when Plex was done. I don't know (or really care) who becomes the new Dev / Lead Dev and what their direction for the server is. Anyways even though I may have had the shortest run of an executive position in the history of TF, I'm fine being an admin because it's a fun way to relax and chill with people.

Thanks for understanding.

@DragonSlayer2189#1456 i saw that but it also gets rid of the shop and master builder stuff entirely. i appreciate the effort but I think a little too much of the wrong kind of bloat was removed. the whole point of Plex was to keep in the tfm exclusive features but have a much cleaner way of doing things that didn't require a bunch of repetitive code spread out all over the place like it is in TFM.

@Darth#1454 the plan was to do a major sweep through and get rid of a bunch of junk in TFM this weekend. hoped to get some of the junk cleaned up to tide us over until Plex was ready but that's not happening now. sorry that's whoever the new dev will be's problem

Quote

@Darth#1452 Does this also mean PLEX is going to be put on hold/cancelled?

Yes. If development continues, which is unlikely at this point, if will not be geared specifically towards TF and not be a complete drop in replacement for TFM which was the original plan.

title.

good luck.

@redeastwood#1417 no this deserves to be transparent

Quote

@wild1145#1414 But you wouldn't get anything since the server started either way because the console is cleared when you re-start the process, otherwise what you're really looking at is effectively a log viewer which is a very different thing to the issue you raised here.

No the logs will keep going as long as you have the page open. That is irregardless of how many restarts you do. There is no way to scroll up on screen. When you refresh the page it will bring you back a few hundred lines even, which is way better than like 60.

Quote

@wild1145#1414 Because then the servers would have even more different configuration. The point of the beta server is we can take a total clone of the live one from time to time and drop it in. The only config then that we have to change is to point at a reference core protect db rather than the ops one.

We agreed to keep the beta server and main server separate. We said that we couldn't move them over and would still keep them separated and would manually update them. Therefore this would still not be relevant. And also the even funnier thing is that the port management is done from Pterodactyl itself. So every start it will update the server.properties with a bind address of 0.0.0.0 and the port you set from panel if it is not that already. So even if we were to move the beta server in place to the main server and we were using Pterodactyl, it would still automatically do it for us.

Quote

@wild1145#1414 The panel itself, the fact you're exposing admin access to the world on a random ass panel... That's the attack vector. In the same way SSH is an attack vector. It's a way someone could gain unauthorised access to your server. You only open those up when the benefits outweigh the costs and risks...

Guess every time I want to login to my server I should travel to Canada where my server is and connect via Ethernet only. It's safer and one less attack vector! I really don't consider it a major attack vector. There are no costs, it's free. And there are essentially no risks if you use 2FA.

Quote

@wild1145#1414 Then as I say, you don't have to use it. You're more than welcome to send me config and plugins and I'll do it... I've made my decision and the suggestion has been declined. I've only left the thread unlocked because I don't want to come across as an asshole for locking it mid way through this discussion.

Is it fair to decline it when discussion is still going on though?

Quote

@wild1145#1414 Yet it somehow hasn't changed my view or position on this... As I said in DM's when we had these conversations.

I didn't bring any of those points up in your DM's though. You only told me why it wouldn't work and I said "ok"

Quote

@wild1145#1414 But we're not... Nothing you've highlighted here are things we use. So far you've just moaned that it doesn't work how it used to work and gone out your way to be awkward about it.

That's not true. Read the thread and just above. We use most of the features that Pterodactyl has to offer and it just makes everything easier.

Quote

@wild1145#1414 Given we have full stack traces every time it happens, and it didn't happen on the temp server, the first thing I'd do is roll back everything to then and validate that's still the case. I've got some ideas as to why it's crashing and hanging as I've said before and I personally think we should be focused on fixing that rather than masking the problem with a panel... If admins and the community aren't feeling the pain of the problem they aren't going to hold either of us to account on why it's still a problem.

I don't have the stacktraces and it would be really hard for me to even get them because you can't scroll up in screen. And there are multiple log files for each day. So you'd have to go through each one until you find the right server session and then find the NPE. Do you not understand how inefficient this is?

@Windows#1386 That's not the problem. It's extracting the actual logfile itself since they all look nearly identical. 2020-11-11-1 and 2020-11-11-2 for example. Tab completion won't work on this and it hurts my eyes to look at all that.

Quote

@wild1145#1388 No... You don't... If you are needing to unzip, then console wouldn't help anyway given it wouldn't have the history in console from previous executions...

It doesn't need every single execution. The latest one is usually fine. It's not often that I test an NPE and then immediately restart the server.

Quote

@wild1145#1388 Because the NAT Port-forwarding re-directs external port 35565 to the LAN IP on port 25565. As I said on Teams... As long as the VM is listening on 25565 and you connect to the IP on 35565 it will get routed appropriately. Same for everything else. That's ultimately how the main server works, the difference being they both listen on 25565 and the ports are exposed 1 --> 1 because that's how I knew what ports needed to be exposed and didn't want to confuse things.

That sounds extremely complicated. Just allow 35565 through the firewall and change it in the server.properties. I genuinely don't see any reason to overcomplicate such a dead simple thing.

Quote

@wild1145#1388 And yet it's almost like that's still not the point, but hey ho.

That's what the main attack vector is, so it is the point.

Quote

@wild1145#1388 That I appreciate, I can understand you feel this is best for the server, and as I've told you multiple times I strongly disagree. I've told you that in DM, and yet you then get shitty when I tell you on the thread that it's a no as well... I don't think this proposal is a good answer for TF and I don't think the risks outweigh the rewards.

Well, because I hate working directly using screen and I'm sure many other community members would appreciate having a panel as well. I made this thread because I'm disgruntled and I want the communities opinion as well.

Quote

@wild1145#1388 And I'm not suggesting you don't, but given we discussed this multiple times in DMs and I've explained my reasoning multiple times, it feels like a waste of both our time keeping re-visiting this nearly every day now... That's why I take issue with this and the fact we're still having this discussion.

I brought up valid rebuttals to all of your reasons for why it won't work / it's a waste of time. That I did not say on the DM's.

Quote

@wild1145#1402 I think I've fairly well summed up why I don't want a panel... Especially at the moment, it's an unnecessary distraction and something we don't need... The last little bit of work we needed to do to get the scripts finished so I'm hoping that should highlight that we've done what we needed to get the bot working, and hopefully that should prove if the issue is a lack of admins in right timezones, or the server scripts not working properly.

I think it's pretty necessary. It's missing some pretty damn important things that we previously had in Pterodactyl and even AMP.

Quote

@wild1145#1403 So it's certainly possible to detect the hangs without a panel if we can't get to the bottom of what's actually causing it. Given the nature of the crashes (And lack there of proper crashing) I'm not sure any real panel is going to reliably detect this either way because it's not really crashing.

That's the whole point. It's able to detect when it's in a hung up crash and can kill and restart the server automatically. I high;y doubt anyone will be able to pinpoint one exact reason to why the server keeps hanging up.