Fixpoint

2022-07-02

#jwrd Logs for Jul 2022

Filed under: #jwrd logs, Logs — Jacob Welsh @ 16:05
Day changed to 2022-07-02
[16:05] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jun-2022/#4596 - Can you please give me another example of the syntax needed for an excluded directory in the EXCLUDE file? I have tried muliple variations to no avail. On the other hand, I have successfully written the operation to exclude directories using --exclude='filename'. Nevertheless, I have now spent so much time trying to write a
[16:05] sourcerer: 2022-06-17 17:48:24 (#jwrd) jfw: you might think you could give it in absolute form i.e. '/dev' to match it exactly - and this *would* work except that for (misguided) security reasons or something it strips any leading slashes from the filenames when building the archive.
[16:05] caai: successful tar command with the operation '-X' that I would like to accomplish that task as well.
[16:13] caai: jfw: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jun-2022/#4652 - in addition, it is difficult to reach the top of a mountain that you worked your entire life to climb, only to then realize that there is another mountain still yet to climb.
[16:13] sourcerer: 2022-06-29 19:57:09 (#jwrd) jfw: caai, dorion: perhaps an obvious point but it occured to me, regarding why the training is so hard to sell: we're going for people who've worked hard for their money, i.e. been quite successfully distracted from what mattered in the world by their saltmine overlords and/or own stupidity, or else started out rich or otherwise lucked into riches, yet got themselves distracted just the same; so by
[16:22] caai: to my understanding, the first stage of grief (this is a flawed analogy) is denial. it is far easier to deny the importance/urgency of the matter than to start at the foot of another mountain. why not let people save face and simply pay for in-house service at a premium price? as time passes, some will desire to learn.
[16:23] caai: pasito a pasito
[17:22] jfw: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4659 - ok. the first point was that the exclude file needs to be a simple list, one path per line. the second more subtle one which applied for either --exclude or -X variant was the need to start everything from "./" in order to match the full path rather than subcomponents (for example a file in the home dir that happened to be
[17:22] sourcerer: 2022-07-02 16:05:20 (#jwrd) caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jun-2022/#4596 - Can you please give me another example of the syntax needed for an excluded directory in the EXCLUDE file? I have tried muliple variations to no avail. On the other hand, I have successfully written the operation to exclude directories using --exclude='filename'. Nevertheless, I have now spent so much time trying to write a
[17:22] jfw: named "mnt").
[17:23] jfw: thus, your previous exclude file becomes: http://welshcomputing.com/paste/qzemvybugs
[17:25] jfw: and your option 4 becomes: http://welshcomputing.com/paste/kewfj9beeb
[17:35] jfw: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4662 - this sounds like a more charitable way to put it indeed; at the same time, that's exactly what children are asked to do all the time, and the grownups don't tend to be swayed by protestations that "but I graduated fourth grade art class and that's enough for me, to hell with math"
[17:35] sourcerer: 2022-07-02 16:13:13 (#jwrd) caai: jfw: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jun-2022/#4652 - in addition, it is difficult to reach the top of a mountain that you worked your entire life to climb, only to then realize that there is another mountain still yet to climb.
[17:39] jfw: and to be clear, I'm not arguing against the "full service" plan if it helps make money; not at all.
[17:52] jfw: it's like, I dunno, I'd rather teach people to eat well & exercise but if everyone around already suffers from advanced cardiovascular illness then it'd be foolish not to set up instead for treating heart attacks.
[17:53] jfw: in that vein, "lost coin recovery" is another service that could be built.
[17:55] jfw: (for things like corrupted files, failing storage media, "software upgrades" rendering old wallets unusable etc.)
[19:12] dorion: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4674 -- yeah, and it's from all sides too. it's not just the legal asset protection instruments are gradually deteriorated, but also it's hard to understand just how cancerous phones have been to otherwise intelligent people's minds.
[19:12] sourcerer: 2022-07-02 17:52:19 (#jwrd) jfw: it's like, I dunno, I'd rather teach people to eat well & exercise but if everyone around already suffers from advanced cardiovascular illness then it'd be foolish not to set up instead for treating heart attacks.
Day changed to 2022-07-03
[13:38] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4669 - received. thank you. i see that there were erros not only in the EXCLUDE file but also the script. part of my confusion was not understanding that i should start by changing directories right at the beginning 'cd /', as well as not understanding that you were using a semicolon ';' as part of english syntax in the /mnt example,
[13:38] sourcerer: 2022-07-02 17:23:15 (#jwrd) jfw: thus, your previous exclude file becomes: http://welshcomputing.com/paste/qzemvybugs
[13:38] caai: rather than part of the script itself. hahahaha
[13:55] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4674 - i understand. it is a sad case indeed. a bigger concern i have is how few people use bitcoind. even if we do everything correctly, there will be no one to transact with because they are all on exchanges.
[13:55] sourcerer: 2022-07-02 17:52:19 (#jwrd) jfw: it's like, I dunno, I'd rather teach people to eat well & exercise but if everyone around already suffers from advanced cardiovascular illness then it'd be foolish not to set up instead for treating heart attacks.
[13:57] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4675 - this is an excellent idea
[13:57] sourcerer: 2022-07-02 17:53:54 (#jwrd) jfw: in that vein, "lost coin recovery" is another service that could be built.
[14:01] caai: jfw: when i mount flashdrives on the thinkpad, i notice that i receive messages saying: couldn't mount as ext2, ext3 due to feature incompatibilities. i formmated the flashdrive to ext4. should i have formatted it to ext2 or ext3?
[16:32] jfw: caai: the semicolon works as a command separator for the shell much as in English; when writing an actual script though I'm more likely to use newlines for legibility rather than mushing everything into a one-liner.
[16:37] jfw: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4686 - no you're fine, that message is just the linux filesystem developers being sloppy/confusing.
[16:37] sourcerer: 2022-07-03 14:01:18 (#jwrd) caai: jfw: when i mount flashdrives on the thinkpad, i notice that i receive messages saying: couldn't mount as ext2, ext3 due to feature incompatibilities. i formmated the flashdrive to ext4. should i have formatted it to ext2 or ext3?
[16:38] jfw: you could indeed use ext3 or even ext2 for more compatibility with ancient systems, though I doubt that's much of a consideration here.
[16:40] jfw: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4682 - bitcoind, or an actual bitcoin wallet at any rate rather than whatever subversions du jour.
[16:40] sourcerer: 2022-07-03 13:55:12 (#jwrd) caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4674 - i understand. it is a sad case indeed. a bigger concern i have is how few people use bitcoind. even if we do everything correctly, there will be no one to transact with because they are all on exchanges.
Day changed to 2022-07-05
[21:11] jfw: https://seclists.org/oss-sec/2022/q2/216 - out-of-bounds read in gnupg / gpg can cause higher-level programs that parse its --status-fd output to be tricked into accepting signatures from the wrong key.
[21:12] jfw: the flaw was reported in gpg2 but I've confirmed the broken code exists in 1.4.21 too (write_status_text_and_buffer in g10/status.c)
[21:17] jfw: for completeness, users of --status-file could be affected too (I don't know what those might be)
[21:18] jfw: v.pl and fetch-bitcoind are unaffected because they don't use those "status" options, simply getting verification status from the return code of the program, and moreover only include the trusted keys in the temporary keyring
[21:23] jfw: I suppose people could be tricked directly rather than just programs, if they use --status-fd for some reason.
[21:40] jfw: we'll probably want to do a patch for Gales in any case.
Day changed to 2022-07-10
[15:06] caai: jfw: update - I have begun reviewing lesson 10. the last few weeks, i went down a rabbit hole of reading the GNU tar manual, as well that of chmod and chown (some videos as well). as a novice user, when trying new things, my inability to effectively do what i want (and know can be done) launches me into investigation mode.
[15:30] caai: these investigations lead to testings, which are necessary to confirm one's understanding. unfortunately, this causes me to advance through the lessons slower. however, my goal is to be a competent computer user and not only finish the lessons.
[15:33] caai: today, i reviewed the fetch-bitcoind-0001.sh just to get a basic idea of what operations it is running
[18:03] jfw: caai: any interesting finds from the 'tar' reading?
[18:07] jfw: GNU programs tend to be on the overbuilt side, with knobs for all kinds of dubiously useful things, which is why I don't tend to refer to their manuals when talking about the essential features of the standard utilities. On the other hand, the Busybox "docs" are threadbare to nonexistent in some cases, so it can be necessary.
[18:11] jfw: as far as wandering off the lesson plan, it sounds quite fine really since you're satisfying hungers that arose internally. I believe learning works much better by "pull" rather than "push".
[18:12] jfw: at least when one has the luxury.
Day changed to 2022-07-11
[10:12] caai: i see what you mean, the tar manual is extremely exhaustive. i didn't find any
[10:12] caai: extraordinary operation that represented a 'silver bullet', rather
[10:12] caai: many minor details about the operations and syntax that i found useful. i
[10:12] caai: suppose i could have found these essential features faster via another means, but now
[10:12] caai: i know! would you then see GNU manuals as a last resort when other manuals fail
[10:12] caai: to cover a subject matter?
[10:19] caai: to answer your question, i learned that one can use the tar long form or short form (they may also be intermixed in a single tar command) - I can see the utility in both, though i will probably stick with using the short form most of time; the necessity to place the '--file' option before the 'archive-name'; the usefulness of '--verbose', especially as a beginner;
[10:20] caai: how to list (--list) contents of an archived directory; how to extract (--extract) contents of an archived directory; how to check what tar version i am using (tar --version); how to obtain a list of tar defaults (tar --show-defaults); details with appending and updating archives. and as you mentioned, there are knobs for all types of dubiously useful things and it is a deep well.
[10:29] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4704 - thank you. i am sure i will be seeking your guidance soon as i work on lesson 10. i estimate by august i will be due for the next lesson. veremos ....
[10:29] sourcerer: 2022-07-10 18:11:37 (#jwrd) jfw: as far as wandering off the lesson plan, it sounds quite fine really since you're satisfying hungers that arose internally. I believe learning works much better by "pull" rather than "push".
[20:46] jfw: https://blog.chrishowie.com/2019/09/19/hot-swapping-virtio-disks-on-qemu/#comment-767 - in minor outreach for credit where it's due; archived in case the blog is as dead as it looks and the comment never gets through moderation.
[20:48] jfw: caai: "faster" will come with experience I imagine, as usual.
[20:50] jfw: "long options" are a characteristic GNU feature indeed, perhaps adopted elsewhere over time but not the historical norm; single letters indeed don't suffice once you've piled on as many options as they do.
[20:51] jfw: sometimes long forms are helpful just for clarity though, to be fair.
Day changed to 2022-07-18
[17:27] dorion: welcome back caai.
Day changed to 2022-07-19
[12:05] caai: dorion: thank you!
[12:11] caai: minor troubleshooting report: yesterday, my node stopped syncing, therefore, i ran 'svc -kd /service/bitcoind'. then, i ran 'svc -kd /service/bitcoind' to reinitiate the node while observing '~/.bitcoind/debug.log'. i noticed 2 messages in particular that stated the same thing: ***Warning: Disk space is low
[12:19] caai: i then ran 'df -h' and saw '/dev/sda3 Size: 896.5G Used: 850.8G'. upon a little investigation, a saw that the culprit was a backup directory that i had created from a few months back '/bcrepairdir/old-blocks', the backup of the database from a few months back. i then proceded to delete the old blocks and now i am back on track: 'df -h' output is now '/dev/sda3 Size: 896.6G Used: 555.7G'.
[12:23] caai: at the moment, i am reviewing lesson 10. i just finished reviewing: Node (bitcoind + gbw-node). i am now reviewing Wallet (gbw-signer)
[12:25] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4717 - yes, i imagine so
[12:25] sourcerer: 2022-07-11 20:48:31 (#jwrd) jfw: caai: "faster" will come with experience I imagine, as usual.
[12:30] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4719 - thanks for the explanation. they certainly have piled on a myriad of options
[12:30] sourcerer: 2022-07-11 20:51:15 (#jwrd) jfw: sometimes long forms are helpful just for clarity though, to be fair.
[17:46] jfw: caai: great to hear you were able to track down and resolve the problem.
[17:48] jfw: minor correction, it'd be "svc -u /service/bitcoind" to resume the node; perhaps that's what you meant since it seems to have worked.
[17:49] jfw: (u as in "up")
[21:16] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4730 - yes, you are correct. it was 'svc -u /service/bitcoind'. thank you for the correction.
[21:16] sourcerer: 2022-07-19 17:48:52 (#jwrd) jfw: minor correction, it'd be "svc -u /service/bitcoind" to resume the node; perhaps that's what you meant since it seems to have worked.
[21:27] caai: jfw: could you please share what permission settings you suggest on a personal computer for directories and files (.txt. .jpg, .pdf, mp3, etc.)? i will most likely not share access to this computer with anybody, and will mainly operate as a 'user' and not 'root'.
Day changed to 2022-07-20
[03:33] jfw: caai: for a single-user machine there isn't normally a need to change anything from the defaults
[03:36] jfw: though habit wise it may be best to think as if it were a multi-user machine; thus you definitely don't want write permission for "others", possibly not for "group" either; so for a regular file that'd be 644 (rw-r--r--) or 664 (rw-rw-r--)
[03:37] jfw: and particularly sensitive things might be set 600 i.e. not even read access other than for the owner
[03:39] jfw: directories always need 'execute' permission in order to be traversed, so the equivalent of the above examples would be 755, 775 and 700 respectively
[03:40] jfw: if you can share more about what motivates the question I might be able to shed more relevant light.
[13:34] caai: jfw: perfect. this is exactly what i was looking for; a general guideline for permission settings. danke schon. i don't have a particular motivation other than to habituate myself to operate with best practices
[13:41] caai: in addition, most of my directories and files were copied from my past machine; in turn, the settings were (rwxrwxrwx) for almost everything. it didn't seem best to leave these permission settings
[13:51] caai: jfw/dorion: as of today, my node is fully up-to-date (block height 745,624) - mission accomplished
[15:47] dorion: hey, congrats caai !!
[15:47] dorion: now to keep up with the regular backups to mitigate any future issues.
[15:48] jfw: ah, yes the rwxrwxrwx tends to happen when copying from a fat32 or ntfs (DOS/Windows) filesystem since they don't track permissions (or at least not in the same model as unix)
[15:49] jfw: so 'chmod -R go-w $directory' can clean that up.
[19:31] jfw: https://www.palemoon.org/roadmap.shtml << nice to see these guys seem to be still alive. "full ECMAScript 6, ES2016, ES2017 and ES2019 standard support, and additional support for other ES additions (ES2018 and ES2020+) to bring the browser's scripting and DOM capabilities mainly level with what is requested by the current web. It was expected that this scripting storm that has been raging for a
[19:31] jfw: few years would die down, allowing us to keep pace with any new developments without being in any way lacking in our support, however this does prove problematic as big players are continuing to try and fundamentally redefine JavaScript and CSS into amalgamations of programming paradigms that they were never designed for." No kidding.
[19:33] jfw: "While we do aim to implement new additions and enhancements to existing standards, we may not necessarily follow the Google Chrome-focused implementation-first barrage of features that attempt to redefine the web in a much more centralized, client-computing focused format. As such we are more critical than ever about which additional complexity to add to the various engines in GRE and the browser
[19:33] jfw: and to strike a much-needed balance there to keep Pale Moon and GRE agile and easier on resources for use in other applications and environments than a generic web browser."
Day changed to 2022-07-21
[13:14] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4744 - yes, i will probably do so once a month
[13:14] sourcerer: 2022-07-20 15:47:53 (#jwrd) dorion: now to keep up with the regular backups to mitigate any future issues.
[13:21] caai: jfw: in regards to the system backup script, i included another line in the 'EXCLUDE' text file: './home/user/.bitcoin' because it seems superflous to backup the bitcoin database via the system backup script when it is being backedup via its own dedicated script. would you agree or am i missing something?
[13:27] caai: http://fixpoint.welshcomputing.com/2022/jwrd-logs-for-Jul-2022/#4746 - ahhh, i see. 'chmod -R $directory' will come in handy. gracias
[13:27] sourcerer: 2022-07-20 15:49:50 (#jwrd) jfw: so 'chmod -R go-w $directory' can clean that up.
[16:09] jfw: caai: I agree, it's superfluous indeed and a good example of the use of excludes.
[16:10] jfw: and with that, it should be less painful to run the system backup more frequently.
Day changed to 2022-07-24
[15:17] caai: yes, it is far less painful
[15:26] caai: as you know, i don't have sqlite3 on the thinkpad. in turn, you only assigned me the gbw-signer portion of the homework. i created a directory named: 'demo-wallet' with a sub-directory named 'keys'. i then created 3 text files: 'change', 'fee', 'outputs'
[15:33] caai: i then ran 'gbw-signer gen-key'. what i am little less clear about is: 'Create a contrived outputs table from public data and test addresses and practice creating a signed transaction'. should i input completely ficticious information? does this information orginate from the database when sqlite is working? please advise
[17:45] jfw: caai: normally it would originate from a database, yes; but the idea here is both to practice without that prerequisite and to get a better idea of what exactly that data means (the fields in the outputs table described in the README). So yes, it can be completely fictitious; it just has to be correctly formed. for instance, for addresses you can use the one you just generated. The transaction ID
[17:45] jfw: might be the trickiest one: it's a 256-bit hex string, thus 64 digits (4 bits per hex digit). You might get a sample from a block explorer website, or maybe just running 'sha256sum' on some arbitrary input
[17:50] jfw: (indeed, for address you'll have to use one that's in the wallet, since the private key is needed to be able to sign a transaction from it.)
[17:53] jfw: because of the contrived data, the transaction obviously won't confirm with an actual node; but it's not the wallet's place to make such judgements (quite explicitly, since the point of a secure machine is that it doesn't know about the outside world)
Day changed to 2022-07-29
[22:51] caai: i understand. i created the contrived outputs table. I generated 3 keys/addresses and then inserted them into the table with corresponding amounts of btc (i inserted 1 of the addresses in the change file). Then, as suggested, I ran sha256sum to generate a hex-string, which a inserted. After that, I added #notes. finally, i ran 'gbw-signer send 'address' 'amount'
[22:57] caai: i believe it was done succesfully since the outputs reflected the transaction as commanded and the transactions file was created. i believe i will implement the gpg part after i have completed the rebuild
[22:59] caai: in the #notes portion i bascially replicated what you put, but with my own ficticious information*
[23:00] caai: now, i will start to focus on the rebuild
Day changed to 2022-07-30
[00:22] jfw: caai: congrats.
Day changed to 2022-07-31
[14:53] caai: jfw: thanks
[14:59] caai: i am half-way through reading '/var/build/gales-linux-20191201/BUILD'. once completed, i will read Bernstein's writings. then, i will review the last training video.

No Comments »

No comments yet.

RSS feed for comments on this post. TrackBack URL

Leave a comment

Powered by MP-WP. Copyright Jacob Welsh.