thank you. but i would prefer not to host stuff on google drive. but some other provider could work (as i said, i am willing to pay a reasonable amount).

So keep a backup. Torrents can be messy because they can be broken if there are no seeders.

If the content is static, then I’d recommend some older P2P filesharing like eD2K to keep one big zip/rar file backup shared among peers.

yes. content would be partially static. most files will not update, only new stuff will be added. and some files be updated.

i did not know syncting could have multiple masters. afaik, syncthing had a master-slave architecture, where a folder on a device is master, and another folder is slave (both can true simultaneously, a folder can be used both as source and sync). if there is another folder, it can be slave of prior 2, but not master, because then you can have conflicting results (which master to pick). do you possibly mean something like a pyramid/tree architecture, where a father nodes has 2 daughter node, and each daughter has 2 and so on. if so, that is even harder to setup (getting people to ask others if they will be their father/daughter cell. this also has problem if some node is out of sync (because of being offline or something), daughters and grand daughter will also not sync. A cyclic link list is also possible, but again chain can be broken. and this can not be a doubly linked list either (2 masters). or is there some other way?

cheating aside

that was just one case. I was stupid. Now I am not.

I did appreciate such archives and made use of them.

to share notes, PDFs, try old exams to prepare…

exactly why i never said no to anyone. it is the age old saying - we stand on shoulder of giants. I remember falling ill, not being able to attend classes, and missing stuff. but then some friend would lend their notes. drive imo is just for that.

some day you’ll graduate as well and life will move on for you.

I am graduating. That is why when i leave, i want to leave stuff in a functional state so they do not have to start a fresh. I did mention this in post, but i wrote a whole lot more than i should have, and i do not expect anyone to read all this.

ou’ll move far away, get a full-time job, maybe have new hobbies or a family and time will come and you’ll stop supporting it as well. I’ve seen that all the time and most privately run things vanish sooner than later.

absolutely. as i said, individuals work for selfish reasons, and once i leave, i would not have a selfish reason anymore.

And I’d say if you’re the main/sole contributor of content, it’s questionable if this even survives long term. Unless people upload recent exams and material, the content will become obsolete after a few years.

yes. it does get obsolete. but our department is still relatively new (5th or 6th year since establishment) and hence, most course have not been taught by 2 or more profs. hence, much of it will stay relevant as long as professors stay.

My juniors have started bugging me again to get drive working again (new sem has started).

So you kind of need some community anyways.

I would have to pull some shit to form a sub division of department society. then i can get budget to either buy some drive subscription, or set something local, but set it behind some proxy, so it would appear not to be hosted in college (reverse vpn if you will)

i do not think syncting would scale well with ~200 people during exam seasons. Also, that would require everyone to download syncthing (practically impossible task)(i think syncthing android app is depreceated or something, and only forks are alive, no idea about ioss clients). Also, that would actually download stuff, all from one server, that would be expensive (fetching 1 file or 1 course worth of file is relatively cheaper as compared to fetching all course files. At that point, i might as well implement a private torrent.

there is another resource run by dev club, but not many of our department folks are in dev club. Also, their solution imo is worse. they techinically do have a index with search, but not a good one. also - very slooooooooooooow. Ans this is when their solution is smaller than mine (and their solution gets contribution from other departments as well).

I understand it would be better be done by a entity instead of a person, but problem is, entities would have to abide by institute rules. and then whole lot of problems from “why not possible” point 2 applies.

I could get some of my juniors to form a small group within department, but i do not think many of them do anything unless they are given a motive for it. and there is no real motive to maintain a good database other than helping others. entities can do “good things” but individuals often do it for selfish reasons. that was partially the case with me. I used to make notes, and then tons of messages from classmates to share notes, and got fed of sending stuff individually. then i started sending stuff in group chat, but not having a good way to search chat history meant people would not find it, and ask me again. so i made a drive. a course happens where people have to install software, but actual instructions are very hard - i get messages - i make scripts to install, or compile the end product and just ship it. You might think these are good deeds, but they are still selfish acts. I used to maintain a good directory structure anyway, might as well upload it.

it is meant to be editted by others. I do not think many others will be able to use cryptomator (afaik you would have to download the whole cryptomator folder, and then update). Still, thanks!

I would seriously consider this, and try to implement this weekend.

Like this your stuff is compliant and you allow folks to access a google drive for example that you or someone else can manage.

people would not differentiate (in future, i would not be the one to upload stuff, but newer gens would), so i would set everything to be uploaded to google. I would still try to maintain 2 offline backups in case google gets angry.

I use it to share with family that is not tech literate.

that is definitely reinforcing.

I would still try to find if i can host directly in campus, and not google or any 3rd party at all.

this is actually very close to pseudocylindrical projections (an example is the orange peal like sliced maps, wiki says its called Goode homolosine). they look wierd, are dicontinous, but still largely preserve area and angles.

this one is very similar, except sizes are still wierd, and uses different radii spheres, which makes it both non angle and size preserving.

thanks for the suggesion, but syncing is not the problem currently. I can use stuff like rsync, or rclone for drive providers. problem is where to sync to.

I could try these, but the problem is, I am graduating. I could set it up once, and maybe even give someone else (or myself) remote access to the hosting infra, but I would likely be less available to manage stuff.

Good dsolutions though. I could possibly try to make the latter solution work (managing nexcloud is relatively harder imo, and i have no idea how would i mount a encrypted google drive to it). It still feels like something only i would have to maintain, but if I can get it in a setup and forget stage (or like a annual maintainence), then I could consider it.

the only problem now is money. I would have to use vps for this kinda stuff. the sbc + ssd idea was something i had proposed to a junior. but hosting anything in college premises with college internet would have to “techinically comply with copyright rules”. if it was a git like solution or torrent like, even administrators would not be able to access the stuff (comparable in tech literacy)(i am not talking about people who manage or internet infra, but copyright stuff, like our library department). With a drive like setup, they would be able to use it too.

Can i setup a password to access the stuff, like a simple password, common for all? then only student who have password be able to use it. maybe i can setup http authentication. but then again i would fall back to - it is getting too tough for them to use it.

i have access through university. no fancy stuff. my uni subscribes to must research publications (most unis do)

I am sending it to you in a dm (or to anyone else too if you ask)

you have encountered the same bug as me (https://lemmings.world/post/28452192) the problem is that arch (or cachyos) now defaults to using wine-wow64 builds (no 32 bit libraries required) and that has something to do with it. I have not beeen able to fix it, and do not know where to report (wine team will not consider bugs concerning pirated installers)

I recently did that and my account was banned. It was not expensive (~$1), but still felt bad

I did not know that. Thanks!

i have not tried any others (just don’t “pillage” enough). But my wild guess would be no, mostly because others do not compress that much or use freearc. fitgirl compresses a bit more than others. I prefer their repacks because of their great repuatation, large amount of seeders, and small size (i have terrible internet).

This is not a huge problem, you can still get wine builds with older lib32 implementations (for example on arch, you can check aur (and by extension chaotic aur), most other distros would have not changed yet), so you do not have to install on windows and move over.

If only I could

i will check it out, but eventually they will also drop lib32-* libs and will get the same problem. I want to solve the root problem

WoW64 is a Windows subsystem for running 32-bit stuff on 64-bit Windows. You’re talking about Wine’s implementation of WoW64 - there’s the old one which needs 32-bit Linux libraries and the new one which doesn’t.

I do know about this. to be specific, wine team since 9.0 stablised using there newer wow64 stuff (iirc, this was one of the blockers to use wine on non x86-64 architechtures), and since then, there has been aur package to play around it. I used to only use it, but receently i wanted to download a game, and faced this bug. then I reverted back to normal arch package which is in multilib repo and depends on lib32-*. Now the main arch package is using the newer wow64 implementation.

I did hint at the the fact that I did not find much in logs. I have not checked the verbose logging (which shows each dll involved, each bit of memory read, etc), but with regular logging,

0130:fixme:win:WINNLSEnableIME hwnd 00010108 enable -1: stub!
0158:err:environ:init_peb starting L"C:\\users\\sg\\AppData\\Local\\Temp\\is-SE85P.tmp\\FlushFileCache.exe" in experimental wow64 mode
0158:fixme:ntdll:NtSetSystemInformation (0x00000015,0x40bb84,0x00000024) stub
0158:fixme:ntdll:NtSetSystemInformation (0x00000050,0x40bba8,0x00000004) stub
0170:fixme:ntdll:NtQuerySystemInformation info_class SYSTEM_PERFORMANCE_INFORMATION

these are the only lines that come when actual installation starts. The installer stays stuck at at some percentage (varies depending on total game size, and size of compressed binaries)

I think it is related to freearc decompression, because at this point in installer, the various .bin files (the compressed files are named like fg-*.bin, where * is some number, they are FreeArc archive files) are decompressed when using older wine version, with this, they are just stuck. I can not C^c the proceess, I have to pkill all the running wine processes. There is something related to wine written in fitgirl faq, which also hints at replacing the dlls for this, but that does not work for me