Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)LE
Posts
0
Comments
142
Joined
2 yr. ago

  • I'm curious, how would you do this in such a way that it wouldn't come at the expense of effecting your high availability?

    If the server were on-prem or in the cloud... and the system crashed/rebooted, how would you decrypt (or add the passphrase) to the encrypted drive?... cause the likehood of the kernel crashing or a reboot after and update is higher than an FBI raid... and it would get tiresome to have the site being down, while we wait for Bob to wake up, log in, and type the passphrase to mount the encrypted hdd.

    You could use something like HashiCorp Vault, but it isn't perfect either. If the server were rebooted, it could talk to Vault and request the passphrase (automatically) , but this also means that the FBI could also "plug in" the server (at their leisure) and have it re-request the passphrase. ... and if Vault were restarted there's quite a process to unseal (unlock) a vault - so, it would be as cumbersome as needing to type in the passphrase on reboot.

    My point / question is: yes, encryption (conceptually) is easy, but if you look at "the whole life cycle / workflow" - it's much more complicated and you (as an administrator) might ask yourself "does this complexity improve anything or actually protect my users?"

  • I really think the "simple" approach of categorizing bot VS non-bot and federate vs defederate are only masking the underlying problem : all posts do not have the same amount of "value".

    However, with Lemmy they do. And I think this is what's broken. If you or anyone in the community has time or interest, I think focusing on rewriting the "what's hot" algorithm would reduce/remove many of these "workarounds" (like the one you're suggesting).

    (I'm just thinking out loud) but a better "what's hot" would have each post weighted:

    1. Against the number of people subscribed to a channel (more subscribers == more relevance)
    2. Against the average number of comments by different users/ post / community. (many comments from different users == more relevant) This would implicitly address the issue of bot spam, that you mentioned.
    3. An upper limit on new topics / community. This would avoid the meme community from hijacking all of "what's hot".

    Of course this cannot all be done in real time. Things like "average number of comments per post" could be precalculated daily, but I think it'll be "good enough" and a radical improvement to what Lemmy currently offers.

  • ... another option: you use the web based Teams.

    If you want more isolation, you could have a dedicated web browser for it.

    Of course, the web version of Teams has a few annoying limitations (you can only see 4 people at the same time, opening multiple tabs to Teams kinda breaks it, etc), but it is endurable.

  • As others have mentioned use a credit card instead of debit.

    But if you need/want to use a debit card, then take a look at services like Revolut or Wise (non-referal links included).

    Both provide you with debit cards that you can enable/disable instantly within their app. Revolut gives you "virtual cards" which can be used for online subscription, so you can create a dedicated virtual card for each subscription (minimizing the impact if/when one of your cards is leaked). Revolut also has "one time use cards", so a new debit card number for a single purchase. In practice, more and more vendors are disallowing "one time use cards", but you can create a similar effect with the virtual cards.

    Both platforms also allow you to set up dedicated (monthly) spending limits on either the physical or virtual cards. So you can limit your exposure that way too.

  • Hey congrats, that's quite an accomplishment. I bought the game from Steam and spent 2 hours struggling to get through the tutorial... which I failed.

    I ended up returning the game. I still have it on my wishlist, but I'm going to wait for it to reach some epic all time low price, before I subject myself to that level of suffering (and growth) again.

  • I did a bit more homework and you're right.

    "Back in the day" running Javascript increased your attack area. But now-a-days I guess it's consider "safe".

    I did find this old (7 years ago) posting which talked about concerns. Today, I guess the rule of thumb is to avoid (or limit) browser plugins.

    Thank you clarify that.

  • But the nice thing about email is it's decentralized, and everyone already has it.

    That is true, but in the case of email as an issue tracker: only the people who have received it will know of its existence (unless it's mirrored on public facing websites - like Debian does with their issue tracking).

    The thing we'd lose is the "ease of access". Tbh, I'd see Usenet being a better distribution medium than email for OSS apps... but I really appreciate the intention behind solutions like git-issues: move the issue tracking into the same tools used to track code changes. It, in my opinion, is more in line with K. I. S. S.

  • You make a valid point regarding losing important contextual information like PR and open bugs.

    However, I don't think email offers the same level of visibility as we currently have with github workflows.

    There is an creative Git based issue tracker, I used called git-issue. Basically, the entire bug/issue/pr process is captured as yaml (I think) files, which are kept in a dedicated branch.

    When I used it (as I wanted a self-hosted bug tracker), I found it functional but a bit cumbersome. However, I could see someone creating a very nice github like web interface for it.

  • I like the idea of improving the quality of "what's hot".

    At the moment, the current implementation is pretty weak. Even in this thread, as I'm reading it: Your post is top... even though it's 25 minutes old and has only 3 upvotes, compared to the second thread which is an hour old and has 39 upvotes.

    I can see how Lemmy would benefit by modularizing the "hot" algorithm. This would allow each Lemmy server to install/test their own (or shared) "hotness" algorithm. Eventually, I think, everyone would converge but in the meanwhile it would allow for a rapid exploration of different possibilities.

  • You've got a good point regarding Photoshop. Gimp exists on Linux, but I find it immensely powerful but hard to wield.

    Gaming with the Steam Desk has gotten better for Linux with the introduction of Proton and I imagine this'll only improve.

    You can see if your favorite game is supported with Proton here.

  • I'd say, let's wait for a catastrophic event at github before we jump ship.

    Git , by its nature, is distributed. If, worse case, github.com went down (without warning). Life would move on, people will have local checkouts of the "important/popular" repos that would be pushed "somewhere else".

    Yeah, github actions wouldn't work, build that pull from github repos would need to be refactored, but life would move on.

  • The end user's ip is hidden in the onion network. The server will get the ip address of the "last node" your client routed it's request through (and that node only has the ip address of the previous node, etc).

    However, the clients ip can be leaked if a server creates some Javascript which makes an Ajax call (basically, an additional http request). A malicious Ajax call will not go through the onion network and thus expose the clients real ip. Hence, it's recommended to disable Javascript and other features while using tor.

  • Yeah, I had joined Beehaw shortly before they defederated. I knew this happened, but I thought "meh, it'll be alright". I tried to make the best of it... but at the end of two weeks I was asking myself "Is this all there is to the fediverse? It's pretty disappointing".

    So before I gave up on Lemmy and the fediverse, I looked for a new Lemmy server that wasn't defederating nor defederated from the fediverse. Eventually, I settled on Lemm.ee and I see know just how much of the fediverse was being filtered out for me.

    disclaimer: I don't fault Beehaw for their decision to defederate. It is their choice to make, and I greatly admire and respect their transparency in the matter. However, for myself, I don't need, want or appreciate these extra guard rails "to keep me safe". I'm an adult and are willing to act and be treated like one.

  • Windows (and most other operating systems) have a "user land" and a "kernel space".

    "user land" is where all your applications run. A "user land" application can only see other applications and files owned by the same user. Eventually, a user land app will want to do "something". This can be something like read a file from disk, make a network connection, draw a picture on the screen. To accomplish this, the user space app need to "talk" to the kernel.

    If user space apps were instruments being played in an orchestra, the kernel would be the conductor. The kernel is responsible for making sure the user land apps can only see their respective users files/apps/etc.

    The kernel "can see and do everything", it reports to no one. It has complete access to all the applications and every file. Your device drivers for your printer, video card, ect all run in "kernel space".

    Basically, the OPs link: they've ported Doom to run effectively like a device driver. This means that if doom crashes, your PC will blue screen.

    This has no practical purpose, other than saying "yeah, we did it" :)

  • "The Fediverse promise is one where the user has the power today." ftfy

    The concern people in the fediverse have with companies like Meta joining, is that:

    1. Embrace: they will "start off" by making the fediverse easy to access for the masses. There will actually be great growth in the fediverse. People will flock to Meta as their choice platform because they will be faster and more reliable than self-hosted fediverse servers.
    2. Extend: Big companies will begin to introduce new features, some of it will be added for the open source community to use. Eventually, there will be new proprietary features added (integration with WhatsApp for example) . This means that Meta's Fediverse will be different from the open source fediverse. It'll probably start out as something innocent like "needing a Facebook account to post a message / comment in their channels." Then it gets worse...
    3. Extinguish: Now the masses have flocked to Meta because it's fast and stable. This results in many/most of the Self-hosted services to become extinct. Then Meta starts to add more" security", like a fediverse "reputation". Meaning, if your self-hosted service submits "enough" posts/comments that are not spam, then your allowed to read/post on their platform. This means if your self-hosted and/or a smaller member you will be barred from accessing/posting content. Thus, the fediverse is now owned by big corps and you need to use their platforms (and watch their ads or subscribe) to access content.

    Source: Compare the history of e-mail (the original fediverse) before Gmail and Hotmail compared to what we have today. I (as an individual) can run my own mail server, but most of my messages will be marked as "spam", if I send it to a friend who has a Gmail address, because my reputation is too low. This forces me to "pay" for email.