Last modified: 2013-05-23 (finished). Epistemic state: log.

Mostly just practice, which I report for beeminding purposes. Yay?

I’ve extended the cognitive tests with a simple questionnaire for some daily variables and now also track daily mood split into four variables (energy level, comfortableness, “everything makes sense”-ness, body load) on a scale of 1-51. I don’t expect that to yield useful correlations (with better than retrodictive powers), but it’s trivial to do, so whatever.

Also, first result: apparently, the more tired I am, the better I get at the Stroop test. Can’t rule out practice effects yet, of course, but I think ‘m getting a better feel for “cognitive functions are down, I’m not even bothering to process what this word says”, which makes the Stroop test much easier. Might be a useful thing to notice in general.

I’ve done some work on the next iteration of my language tools (now renamed to “mavothi”), might have some (unusable as-is unless you’re muflax, but maybe interesting nonetheless) results soon.

I’ve also done a lot of work on steno, which I’ve now started to actually practice. Actual practice, guys! I’m becoming functional.

I’ve transitioned some archives to git-annex. For my non-neckbearded readers, git-annex is basically a system to track archives over multiple locations (e.g. external drives, remote servers, different laptops) and automate most of the hassle of having those archives in the first place.

It has a Dropbox-y assistant which I found really awkward to use and which shredded my configs constantly while I tried to figure out how it works, so I’m not using it (yet) and do everything via shell commands, like a man. So my current setup looks like this:

  1. There are multiple independent archives, currently ~/books, ~/games/install (install files, mods etc. for vidya gaems) ~/テレビ (TV shows and movies) and ~/音 (podcasts, music, etc.), which contain a lot of big media. I have complete daily backups2 of most of this stuff, but I’d like to a) have more backups and b) be able to swap stuff out that I don’t use much. I regularly re-watch ‘Allo ‘Allo!, but there’s no point in having it on my main machine all the time3. Git-annex has this nice feature where you can drop files from one repository and it replaces them with symlinks so it looks like the file’s still there and you can move it around, but when you try to access it, it’ll tell you what other repository (which still has the file) you need to connect so it can get it back.

  2. I have a whole lot of external drives because I tend to update hardware faster than it breaks, and never throw out old drives (plus I have several sources who give me their old stuff too), so I’ve now set them up as “trusted”4 repositories for the archives and made sure all files have at least on external copy. I update these archives whenever my main machine is full and I want to drop files off it. That way, every file is always in at least two places and can’t be lost due to drive failure or basic stupidity.

  3. My two laptops (see above) used to just sync my whole home directory via unison, which is really wasteful and regularly annoys me because their drives are much smaller (and slower), and because unison isn’t too smart. git-annex fixes a lot of these issues because now don’t have to copy everything, but can still keep the directory structure synced.

  4. I currently manually sync the laptops with a simple shell script every time I need some data that’s not on the machine, but that should eventually become (semi-)automatic too.

What I’d like to do later:

  1. I’m currently syncing stuff on my phone through Dropbox (mostly images and audio logs), but once the git-annex android client is a bit more solid, I hope to use it there too so I don’t have to trust Dropbox anymore.

  2. Get the git-annex assistant running so I don’t have to manually add and sync files. That’s still rare enough that it doesn’t bother me much, but allegedly this should already be possible. It just constantly broke on my setup or was horribly slow, for some reason.

A minor belief update.

Back in December5, I wrote:

If one were to take a fictionalist approach to the [New Testament], that is to assume that, excluding maybe a few historical allusions here and there, they are works of fiction […], then it follows that the original scripture is forever lost to us. We may be able to reconstruct roughly back to Marcion’s work around 130AD, but no further. The evidence has simply been lost to time. One may speculate, as RMP and others frequently do, and all kinds of plausible explanations might be found, but closure is impossible.


And so I realized, I needed not a new proclamation of the message, but an entirely different work, one capable of reflecting the true author of confusion. I considered merging all traditions into one gospel, or using elaborate annotation (even footnotes to footnotes), but then I thought, ideally the text would be a dialog between all those strands, would give the Marcionite, the Simonian and the Gnostic their first fair hearing, and so we really need more gospels written from these perspectives, including the Devil’s own version that they read down in Hell, but even that would not be sufficient, not even to replicate the delightful state of interwoven connections I see in the text, for nothing short of all of the tradition would be capable of representing the tradition.


But can the text be made simpler, without breaking it? After all, the canonical NT is already reduced, lacking several gospels, epistles and so on. Yet, it seems to me that these reductions don’t necessarily simplify the content of the text. All ideas are still present, they just aren’t as close to the surface. And this does not have to interfere with the interpretation of scripture. After all, the 19th century Dutch and German radicals managed to reconstruct many aspects before most of these alternative text had been found! They predicted the Gnostic texts in great detail and were entirely right.

I still expected that there wasn’t a single source of origin (or linear sequence of origins) for the core NT, that like modern mediums, the authors were telling stories of their own invention with no real connections between them, until decades (and sometimes, centuries) later various churches would try to integrate them into one tradition (with the well-known result that for every assertion in the NT, there is also a counter-assertion that contradicts it). This would make a successful reconstruction essentially impossible because there simple isn’t a coherent shared thread.

I no longer think that’s necessarily true. I’m not sure that there is a single source of origin (and if so, that it is knowable) either, but due to a lot of recent reading I’ve done, particularly on the Gospel of John and its possible (more-or-less) Marcionite origin, there now is a budding live hypothesis in my head that would make sense of all the texts and their origin, that answers virtually all important questions about the NT, and that postulates just such a single origin (of Simon -> Marcionite hermeneutics -> Catholic hijacking). Funnily enough, that is still in line with my judgment that whatever scripture may have existed before Marcion is lost to us; I just didn’t expect Marcionites to be that influential for our canon.

I will discuss all of that in detail in due time, but I’m only about 20% done with all the necessary skills I’ll have to learn (including several languages) and books I’ll have to read (thankfully all in languages I know), so this is obviously speculative play still, but the mere possibility of a such a hypothesis and the fact that it can be directly verified with the evidence we do have, is dazzling.

There are two vectors that I still find deeply puzzling and possibly beyond our reach - where did the Buddhists get their ideas from, and what’s up with Simon?

Christian Lindtner claims (and unfortunately I don’t have the skills yet to even fully comprehend his theory) that the core NT texts are based on Buddhist sources (which Christian scribes could easily have had access to thanks to Ashoka), but then we have to trace the underlying ideas back to India, and there we will likely lose them. Recent research makes me increasingly skeptical that any of the Jewish texts meaningfully predates the Hellenistic Period (and many not even Christianity!). The case for Greek priority of the Pentateuch (and a Hebrew translation second!) seems stronger every year, and even if we “merely” have to declare pre-Septuagint developments as forever unknowable, that should equally make us skeptical of Buddhist stories before, say, the Visuddhimagga.

It would be much easier for my ideas if instead Buddhism took at least its mythological narratives from Christianity (and then historicized them way in the past), but that’s not a good criterion for truth. :) And if it is the other way around (which is certainly plausible), then we have to explain, independently, how the Buddhists got started. The connection still deeply confuses me, and I follow Lindtner’s suggestion that I’ll have to learn Sanskrit to adequately understand this eventually.

The other origin, Simon, is equally mysterious because we simply don’t have many crucial sources anymore. RMP makes a decent case that at least his Great Proclamation has been more-or-less completely preserved, but much of it reads like typical Gnostic gibberish. I feel increasingly comfortable with various Marcionite writings and their interpretation, but what’s the deal with Simon? I currently have no idea.

I feel sorry for anyone who wants to start a Post-Simonian Church. Luckily, I’m merely trying to start the Post-Marcionite one, and that seems very much doable. Marcion, Appeles, Valentinus, Basilides and others will be slandered no longer, and one face of “Paul” will be reconstructed once again. But Simon? Who knows. We might have to fall back on channeling after all.

Luther, in his Commentary on Genesis, writes:

Therefore it is far the safest not to be too curious and inquiring in these subjects [i.e. hermeneutics], because they are placed above our human capacity. For how can we understand that order which God himself establishes and approves? Yes, reason must here be put to shame, for what is order in the eyes of God we judge to be confusion of order. Thus the stars seem to us to be arranged thoughtlessly in wild disorder in that the bright ones are scattered among those more obscure, and the lesser among the greater. Who would judge this to be order? And yet it is the most perfect harmony, so constituted by the all-wise mind itself. In like manner we judge other matters. It seems confusing that our Elbe and all rivers flow to the sea in an irregular winding course. Such disorder there seems to be also among trees, yes, between man and wife, where it appears there is no order. But all this only proves that God is a God of order and that his judgment as to order is quite different than ours.

We might find ourselves one day, discovering Eris legible all along, in the setup for Creation’s final pun…

  1. Meaning of the scale for the variables, with 2-4 as the typical range:

    • energy level:
      1. “can’t get up”
      2. “ugh”
      3. “functional”
      4. “movement is light, doing stuff is easy”
      5. “mania”
    • comfortableness:
      1. “paranoia”
      2. “anxiety”
      3. “people are nice”
      4. “feelings are awesome”
      5. Overly Attached Girlfriend
    • “everything makes sense”-ness:
      1. “nihilism”
      2. “knowing stuff is too hard”
      3. “the world makes sense”
      4. “everything according to plan”
      5. horse_ebooks
    • body load:
      1. “Nurgle”
      2. “nausea”
      3. “some stomach cramps”
      4. “Ibuprofen”
      5. “dissociation”

  2. I use a custom rsync script to copy everything to a RAID-mirrored backup drive. Daily backups go 1 week back, plus there’s an additional monthly version for the last 3 months. I used to have significantly more past versions (monthly backups back to 2010, 2 weeks of daily backups). The main idea was that I wanted to have a safeguard against deleting a file and then, a week later, noticing I actually still wanted it (which happened way too often). Nowadays, I have most of these things in git repositories and spread over multiple locations, so restoring arbitrary past versions is no problem, and I’m far less trigger-happy with deletions in general. The old backups used up a lot of space, so I reduced my backup horizon to 1 week (daily) / 3 months (monthly), although I still have one offsite drive with all the old versions (because I only update that every few months).

    I might eventually replace some of that with git-annex too, or something like it, but I’m nowhere near comfortable enough with it to trust it with my core backups.

  3. Unfortunately, the .git directory itself tends to use up quite a bit of space. For example, my ~/books directory, which has basically all my read-only documents in it, is \~70GB large and has \~35k files. The .git right after adding the files was already \~400M (after doing a git gc). However, size tends to depend almost exclusively on the number of files, so I did some restructuring and packed some stuff up, and reduced my number of files to \~8k, and now got only \~160M of .git and adding new files is reasonably fast (<20 seconds on the slowest machine). (And to be fair, part of that is git’s fault, not git-annex’.)

    So my heuristic for now is that git-annex adds \~1% overhead of the total archive size to every repository, which is still worth it most of the time (because you can distribute stuff and space is reasonably cheap), but it sucks a little.

  4. git-annex has 4 trust levels:

    • trusted: files are assumed unchanged since the last time we saw the repository
    • semitrusted: files count towards required redundancy, but we have to check first if they’re still there
    • untrusted: location is tracked, but we never expect data to still exist
    • dead: data is gone

    “trusted” makes sense for backup drives that never change on their own and are only updated in one direction, “semitrusted” is the level for my different machines that might have independent changes that need to be merged, and “untrusted” covers drives that are known to be failing and aren’t dead yet.

  5. Old logs are currently unavailable due to some restructuring. Might take some days/weeks/lives.

blog comments powered by Disqus
dlog » daily log » powerdrills and bonesaws