It's only words... and assumptions; and priorities; and... ouch!

February 6th, 2018 by Diana Coman

On the very first of December 2017, a man whose knowledge I admire started publishing the FFA (Finite Field Arithmetic) series on his blog. Having previously struggled to actually follow in the immense log the scattered bits and pieces and discussions and rants and what-not on the topic, I was absolutely delighted that he finally took the time and committed to this weekly series. I was (and am) also grateful that he decided to put in the significant effort for this last but crucial step of writing software: documenting it essentially, providing at the very least a single place for the interested reader to start from.

As my delight is not of the mere words type, I proceeded to adjust my schedule to fit in a weekly dose of FFA. So I looked around at my madness of a weekly schedule, cut and put away various things that were not quite as important as I considered FFA to be, cleared up Saturday afternoons and set that time aside specifically for FFA reading and practice. As the FFA chapters were meant to be published every Friday as per their author's own statement of intent, I considered Saturdays to be perfect fit: fresh chapter ready to be read and enough time to digest it properly until the following week and the following installment. In a word, perfect, right? Sadly, it turned out I was wrong.

It really took only three weeks for the Friday FFA to become... the Saturday FFA. And this was indeed a change half-sort-of-announced in the logs1 - I recall it as a statement from Stanislav along the lines that he *might need to* delay the FFA chapter because of work load. This statement in any case didn't sound AT ALL like a loud and clear "from now on FFA will be published on Saturdays because workload far exceeds expectations and I'm only human, sorry". As I consider public complaints to be rather futile, I did not even complain about it - in hindsight I suppose I really should complain more, if not for any effect (I still doubt any effect) then at least for the record. For MY future record especially.

Going back to this sliding of Fridays into Saturdays, the very unpleasant effect for me was of course that my painfully cleared time on Saturday afternoons was all of a sudden not fit for purpose anymore. And not just because "no FFA this week" but consider this: one can slide, sure; this week I'll read last week's FFA chapter and so hopefully I can at least still use the whole setup; except... certainly not fresh anymore, certainly not exactly keeping up to date, certainly not solving the puzzles as intended before potentially seeing other people's attempts in the chan if nowhere else. And to top it all, this week a monster chapter2 further stretched the schedule anyway: by the looks of it one week slide behind with the reading will still not be enough to avoid being left stranded. So what am I supposed to do exactly to be able to follow FFA? Honestly, my current approach - and I really don't see another - is simply to say: fine, stuff this "follow" idea, it's *way too costly* simply, not going to happen. Let it run its course and when it's finished and all ready, there will hopefully still be time to go and take it chapter by chapter, as possible. At least once it's there, one can make a plan on eating and digesting it, without further surprises to blow up one's plan.

Note that this has nothing to do with the FFA itself. In particular, this is NOT in any way saying that Stanislav "should have" done more work or different work or whatever other work. The whole sticking point really is on the planning and communications side and this came to the forefront even more visibly only yesterday through the "hardware debacle" for I can't call it in any other way. As far as I can grasp from the log itself, the work there behind the scenes was again done and done well as Stanislav always does it. However, the last step, the 20% that kills the whole other 80% was botched beyond belief and the saddest part of it is that Stan himself does not even see how or why exactly he ended up in the hot soup.

From previous practical experience I can say that it almost never helps to tell a man what he can't see for himself. There is a reason for that particular blind spot of his eye and that reason is not addressed through someone else's explanation. Nevertheless, I'll say - for my future records mainly - only this: assumptions made but not explicitly stated for the other party to accept or challenge are a way of shooting yourself in the head; conflicting deadlines and/or requests are to be shouted from the rooftop as soon as spotted; in this horribly limited reality of human life, practical costs can and do quite often trump ~everything else. The unexpected in particular tends to have a horrible cost, so the best thing you can do for those you work with is to make sure you don't add to their share of the unexpected. Otherwise put, please, please make sure that they really can trust your exact words at any moment as much as they can trust your finished product: any assumptions clearly stated and defended if needed, any limitations made clear from the start, any change to previously promised issues clearly argued, any priorities made explicit, any implications of new developments (be it new requests or anything else) stated as soon as identified. This being said, communication is of course a not-exactly-precise activity and one in which I certainly do not presume to give advice to anyone. I simply listed there the main parts that I found myself adding to my own "communication manual" over the years - it's been of use to me and that is enough reason for it to be here but if you, reader, find anything of use for yourself in there too, that's even better really.

  1. I can't even find it now, all I could come up with was Mircea Popescu's later surprise at this development...  

  2. Similarly to previous schedule stretching, this was also announced as a mere "thinking of" rather than "previous planning was totally inadequate for these here reasons; new plan as follows with x, y and z measures taken to avoid repeat of same mistake; sorry and thank you for sticking around." This time however, I decoded the message better given previous experience so I was at least *expecting* no FFA this Saturday basically, what can I say. 

Comments feed: RSS 2.0

26 Responses to “It's only words... and assumptions; and priorities; and... ouch!”

  1. The (not really secret) revelation is that the time I have available to work on the FFA material each week, is a ~random number: specifically what remains after subtracting the time taken to work in order to eat.

    I was able to maintain a semblance of regularity at the expense of sleep. This has a limit. After a certain point it comes at the expense of proofreading, and then of thinking, and eventually the product is not what it should be.

    Chapter 9 in fact was weak. And the one regular reader I have remaining, pointed this out to me in the comments.

    What I will not do is maintain any kind of timetable whatsoever at the expense of correctness. I would rather write correct material that no one wants to read or otherwise use, than to risk producing a well-received but subtly broken FFA that eventually gets good people killed. We have enough kochware as it is.

    So if I must lose readers, I would rather lose them like this, than like-that.

  2. Anonimosu says:

    As far as I can grasp from the log itself, the work there behind the scenes was again done and done well as Stanislav always does it.

    Nu prea se arata ce zici tu acolo. Tipul da aiurea din top zilnic, ca mai nimereste din cand in cand cate-o jumatate de adevar/banalitate nici nu e de-a mirarii in conditiile alea. E plin in log de exemple, uite de ieri Si de-acum trei zile La cate prostii barboteaza nu-i de-a mirarii ca mai nimereste si cat de cat in zona din cand in cand.

  3. Diana Coman says:

    Anonimosu Sa nu confundam treaba cu vorbitul zic. Sigur ca bate campii destul, da, da' nu despre aia era vorba in propozitie ci despre *lucrul* omului care e in cazul asta anume spec pt servere (facut din cate am inteles in amanunt in fapt pana la proiectie re cat s-or mai gasi fix procesoarele+placile cu pricina). Altfel zis omul e foarte, foarte competent tehnic. Dar are o problema cu vorbitul, dupa cum remarci si tu, da.

    Stanislav You are fighting there a strawman though. This is NOT EVEN about losing readers, wtf! I get what you are saying - it's not that the problem though, not AT ALL. The problem is promising weekly when you can't guarantee it. So don't promise *that* to start with, promise then every month or something - something you CAN however guarantee! Even better, make the stuff in advance so as to have buffer (and before you say it: I know it *changes* while you go through it again, of course it does). Anyways, more words thrown after the others I guess. For my curiosity: do you/can you see at least my problem from this side?

  4. Anonimosu says:

    O fi, daca zici. Eu nu prea ma am cu astea, dar nici nu cunosc multi care stiu lucra dar nu stiu vorbi. Mai mult ciobani, salahori, treburi care se invata manual decit oral. In fine.

  5. Diana Coman says:

    Corect ar fi sa zic interactiona mai mult decat vorbi in sine. Da' cum in cazul dat interactiunea e aproape exclusiv vorbita, deh. Ca aia de zici tu au posibil o problema cu vorbele da, da' e alta chestiune, drept.

    Altfel ca idee, cum era aia cu "ochi inteligenti da' nu stie sa se exprime?" Ca nici nu-i nevoie s-o inventez eu cum ar veni si clar despre injineri e vorba, nu ciobani, manual, alea. Si treaba cam de acolo vine, ca problema nu-i la gandire abstracta sau vocabular sau ceva, ci la lucrul cu ceilalti.

  6. Weekly series appeared practical precisely when it still consisted of "make the stuff in advance so as to have buffer." When during the rewrite (which is what this is, compare the current material with the summer's paste-drafts) it diverged, fundamentally, from the year+ of "buffered" content -- from that point, no longer.

    Reader is more than welcome to visit my site monthly and pretend that it's a monthly series. (Or not to read entirely. Nobody forces anyone to read my articles, most of the world very happily gets by without doing it, believe it.)

  7. Diana Coman says:

    From what I see, you are still turning around the point and ultimately avoiding it. This was NOT about the reader - here the reader's trouble is but an example of some impact of the issue that you say you don't see. The issue is simply that unreliable promises and unreliable/unclear communication of changes (regardless of reasons; perfectly valid reasons included) mean unfortunately large costs for anyone wanting to work with you (as opposed to using your output if/when available). So no matter how much one values your knowledge and your work and your output, they are forced by those (avoidable) costs to avoid working with you - the most they can hope for is to *use* what you produce. If that's what you want it's fine, sure. Hopefully it's clear though that it's your choice (perfectly valid reasons for it!) to isolate (or perhaps to filter others??) in this way.

    You are precisely saying (and yes, it fits the facts as far as I can see it) "fuck the readers/collaborators, it's up to them to mold themselves around what I put out there if/when I put it" - which is at the end of the day a perfectly valid choice, of course, but it comes in very weird opposition to your previously stated wish for feedback/discussion and even to giving any exercises/homework. Why bother with that anyway, readers can devise their own or not do anything anyway, they'll get by happily too, right?

    If the answer to that is "oh, but *anyway* I didn't get any useful feedback/discussion/anything even when keeping to the schedule" or "oh, but look, apeloyee is the most/only one useful *anyway* and he doesn't complain/still hangs around", take a break and re-read it at a different time.

  8. Anonimosu says:

    @Diana Pe de o parte e asta cu cainele inginer. Pe de cealalta parte insa, ai vazut ce-a zis Popescul cu memory hole effect si alea. Faptul ca omuleanul debiteaza tampenii cat casa la un nivel si nu la altul nu inseamna neaparat ca el chiar are doua trepte de viteza. Poate fi si ca el debiteaza uniform dar tu le detectezi pe unele mai usor decat pe celelalte si ramai cu impresia gresita.

  9. Diana Coman says:

    Anonimosu De posibil e posibil, sigur. Nu-s de parere ca-i anume cazul aici, dar recunosc ca-i adanca si ramificata baza acestei pareri asa ca nu ma chiar simt in stare sa o expun anume in clar si amanunt. Oricum si oricand ramane insa ca ... e posibil, da.

  10. @Anon This theory of yours is theoretically not bad, but not more than that. Looky that I know, "astea" as well as "alea", an' I can tell you the guy does have his moments of lucidity. His problem isn't that his brain doesn't work, his problem is that he's carrying a monkey on his back, and as is overwhelmingly often the case with autodidacts never to date encountered anyone or anything that demanded he get rid of it ; but rather was perpetually able to slide through on spurious alternative considerations and other such make-believe.

    @Stan Hanbot told you the same thing, namely : that if you regularily have a problem with delivery, and especially on self-declared deadlines, it's absolutely not the task nor whatever alt-explanation you may readily conjure ; but absolutely always your own issues with time management and generally executive function.

    Go to your local community college and take the remedial time management class, alongside the confused sophomore girlies overwhelmed by the banal requirements of Chemistry 202 or Drama 330 or whatever the fuck triteness their self-hopes and auto-dreams crashed on the jagged shores of. And do it like one trying to learn something, not like one trying to persuade the minimum wage earning teacher that he really aughtn't be there in the first place (which, with any luck, will be absolutely all of them). Yes it'll be mildly humiliating, but in your present sorry sad state that is absolutely the best you could possibly do for yourself.

  11. Mircea Popescu et al: don't you think that it is even a little bit odd to complain re delivery schedules of an unfunded, spare time project ?

  12. Diana Coman says:

    @Stan Nobody is complaining. If you really know what you are doing and are happy doing it, carry on, sure.

  13. Diana Coman: actually I'm not particularly happy doing it -- would rather do it full-time, and in a tenth of the time, and without the angry mail from disappointed readers. But I'm not independently wealthy, so all I have is this.

  14. Diana Coman says:

    @Stanislav Datskovskiy I meant the unreliable communication. Whether your communications are reliable or unreliable does not depend on time or on being independently wealthy or on anything other than you really. Hopefully the difference between communication of facts and facts is clear enough.

    Fwiw people are trying to help. If you perceive it simply as "angry readers" or some such, there isn't much anyone else can do really. And indeed, helping someone when they don't want it is quite a huge fight so it does look a lot like people being very angry, granted.

  15. hanbot says:

    @Stan Why doesn't meeting a deadline fall under the heading of your otherwise very respectable notion of "correctness"? Perhaps you're looking at time as a separate component of work? It is not, and neither is it a kind of trough to be scooped from when various details start to snag. Stop abusing time, and you'll stop being surprised.

    Also, ben_vulpes' fuckup post might be worth a re-read here --I know I've made several passes, myself.

  16. hanbot: Correctness is an attribute of the algo. Strictly. And incidentally if you disagree (as I understand MP, for instance, does) you are welcome to not read my material, not use, warn others to avoid, negrate, etc. Because I cannot give you the "good-enough, cheap, quick, predictable" type of correctness that you seem to want. Try Walmart?

    Incidentally there is a whole world of software written under "good management", by people who "communicate well", under strictly kept plans, schedules. Why would you want to use mine instead of this ? Go, instead use e.g. IBM's. I promise not to cry.

    For completeness, I will point out that everything you see on my WWW, since day 1, is "an abuse of time" : FFA and every other creation of mine exist at all strictly out of scraps of time which I bloodily carved away from saecular scutwork.

    That being said, you and every other unhappy customer are entitled to a full refund of every penny you paid me for reading my site. Yes, even the retarded Ro anon commenter. (How much do I owe ~him~ ?)

  17. Diana Coman says:

    Stan, where did anyone say you owed them anything? Simmer down and if you can't yet not deflect at least let it be for now, it'll still be there, the same, tomorrow.

  18. Certainly a useful re-read two years later.

    My take-away being reinforcement of the theory : the post-human split is between the sub-humans who WILL NOT WORK, no matter what, but gladly communicate nothing in particular, from the "world health organisation" to whoever ; and the sub-humans who WILL NOT COMMUNICATE, no matter what, but gladly work nothing at all until they drop.

    And you're more than welcome to sort something out of this. Somehow.


  19. Diana Coman says:

    Huh, some different sort of ultra-specialisation (like that madness observed in computer graphics) ie some specialise in talking and some in working but there's no bridge between the two at all?

    Alternatively, looking at it closer to how you state aka based on refusal of one part rather than narrow focus on another, it brings to mind the amount of relentless hedging that I keep hearing and/or the anxiety that they have in fact, hm.

    While I have some ideas, I can't say that I have yet a clear full model that I'm also satisfied with.

  20. -----BEGIN PGP MESSAGE-----
    Version: GnuPG v1.4.10 (GNU/Linux)

    -----END PGP MESSAGE-----
  21. Diana Coman says:
    -----BEGIN PGP MESSAGE-----

    -----END PGP MESSAGE-----
  22. […] projects will eat into the log reading time. Certainly, all of this is a continuing lesson in priorities and pacing over […]

  23. [...] more than half a year since last time I really looked at it, FFA (Finite Field Arithmetic) finally made its way back up on my list of tasks. Given the rather [...]

  24. [...] of - I'm through the article itself but reckon I should also include the (English) comments and log [...]

  25. [...] line is, tell people what you want to do, what you think is important and do it well. If your words don't match your actions, write about what you're doing to fix it. A few example plans are Lucian [...]

  26. [...] of fact, that making roasted cabbage out of one's priorities, words and assumptions inevitably leads to pain, which, when improperly digested1, itself leads to a feedback loop of further pain, spilled guts [...]

Leave a Reply