Re: Math mode wackiness in memoir class.

blmblm@xxxxxxxxxxxxx wrote:
In article <1152433526.038894.104720@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
Mark Mephinasony <twisted0n3@xxxxxxxxx> wrote:
blmblm@xxxxxxxxxxxxx wrote:
biased in favor of those who ask questions politely

My initial posts to threads here have all been quite civil, thank you
very much.

Can you say the same about follow-ups?

That depends on the civility of the post being followed up.

Keep in mind that if no-one throws a first punch, then fights never
start. If someone sometimes throws a second punch, that's called
"provocation" and perhaps "self-defense". Also "human nature".

Maybe I'm misunderstanding, but I infer from the above that you
consider your time more valuable than that of the people who have
replied to your posts.

No, I consider it equally valuable. I reserve the right to skim or
ignore posts when they are part of a large batch of mostly diatribes
and invective. So, I expect, do they; goodness knows that some of the
responses to my posts give ample evidence that the responder didn't
read my post very carefully first. So how come people who start
flamewars can skim, but people who get attacked but never threw any
first punches must read everything carefully, in triplicate, before
they dare pipe up? That's quite a double standard, and it penalizes
exactly the wrong people with extra work.

And then there was the exchange in which someone suggested that
you add "\nonstopmode" (or whatever it was) to the beginning of
your source file, and you thought this was meant as a command-line
option, and much confusion ensued, all of which could have been
avoided if you'd read more carefully.

To recap:
There were a dozen new posts to threads I'd posted to in c.t.t. that
day (not to mention maybe 200 additional, unrelated new posts). Most of
them were long, and long on invective and short on useful information.
Needless to say, rather than spend a good eight hours reading and
rereading everything carefully, given this sizable load of material I
skimmed it all instead. Perfectly reasonable, and expecting me to
behave otherwise is quite unreasonable.

Next, since I'm asking for info on how to modify error-reporting
behavior (c.f. -wall -werror for gcc, to make compilation of a C source
stop on anything suspicious such as if (x = y) rather than if (x ==
y)), I'm asking (implicitly) for a command line switch (where such
options traditionally go -- not the source file!), and therefore
expecting the response to be a command line switch. Also perfectly
reasonable, particularly for someone with a C/C++/Java background.

Finally, I reach that particular post and see something starting with a
slash that's supposed to make it halt on error. I figure oh goody,
there's the command line option I asked about earlier, and stick it in
my batch file. Nothing wrong with this behavior, either.

What happened was what is often called a "misunderstanding". There are
several people where part of the blame can be apportioned, reasonably;
but I am not one of them. We have:
* Everyone who's posted lengthy, largely offtopic, generally flaming
responses to simple questions I'd asked; without them I might be
reading responses more carefully because there'd be fewer, shorter,
more useful-information-rich ones to promote such a strategy. The low
SNR in responses I've gotten, OTOH, promotes the strategy I've actually
used: skim and extract likely information opportunistically.
* Moreover, there's the particular batch of useless or mostly-useless,
lengthy replies on that particular day, which are especially
significant since if there'd been only the one concise reply with
useful information that day I'd have read it carefully.
* In any event, it wouldn't have mattered if the poster had given me
what I'd implicitly asked for -- a command-line switch, rather than
something else cleverly disguised as one. When someone wants to modify
logging/reporting/interactivity behavior of a job, they are asking for
a command line switch. Remember that.
* On top of that, some can once again be laid at the feet of the
TeX/LaTeX coders/developers, on two counts: one, putting in the
misfeature to read error reporting/logging controls from the source*;
and two, making these such that they'll look like command line switches
if they appear out of context in a usenet post.

*I myself earlier suggested warnings for literal " and ... in source
files that can be temporarily disabled for a block of source using an
in-source command, but there's a good reason for that, namely that
these occurring literally may not deserve a warning in all instances,
and in particular commands/environments that typeset code or other
stuff verbatim. Control of this sort of thing on a fine grained level
obviously requires source code commands. But interactivity/batch mode
and similar stuff really do logically belong in command line parameters
-- which do, in fact exist in (La)TeX for these purposes, making it
even more baffling that the poster didn't mention one of those instead,
given it's the "natural" way to alter job interactivity. (After all,
whether you want interactivity depends not on the source file, but on
the time and circumstances, as a rule! Why do you think there's -werror
but no #pragma werror in C compilers?)


People sometimes get defensive when their favorite tools are attacked.
It's not optimal behavior, but it's reality.

An attempt to cause a tool to be improved constitutes attacking it?
Well, I guess the "its perfect as it is" crowd might feel so, but the
"its perfect as it is" crowd are clearly irrational. If they don't like
or want a change because they're used to and proficient with the tool
as is, they are free to not upgrade to a hypothetical changed, newer
version, after all. If LaTeX3 actually one day appears and, to boot,
contains some of the suggestions that so ticked them off, they could
always stick to LaTeX2e, for instance -- or even create their own
hybrid of the two without the objectionable features. It is, after all,
open source.

Right. Your contribution is to point out where the tools are,
in your opinion (*), lacking. It's someone else's job to actually
make the desired changes, after figuring out all the consequences
and corner cases.

Yes, of course. I am a user, albeit with some programming knowledge in
general. I have no special knowledge of the tools' internals. This
makes me unqualified to do the actual implementation. No one person can
know all the corner cases, because no one person speaks fluently every
language on the planet, and most of the corner cases seem to be obscure
and specific to particular localized usages.

On the other hand, users are exactly those best qualified to point out
flaws or room for improvement. In fact, users that are not developers
are better for that job than users that are, because a lot of problems
with a lot of software stem from an unclean separation of the user
interface from some implementation details, and developers, who know
the latter by heart, simply tend not to notice these or, if they do,
don't find them problematic. Also, developers may base their assessment
of putative flaws/nonflaws on esoteric test cases; ordinary users base
theirs on how it actually performs in production circumstances in the
real world, and on what things bother them or cause them
confusion/trouble/awkwardness/don't work when they attempt production
use.

This makes it essential that, with ANY software whatever:

a) Users are able to report difficulties, either with the usability or
the actual functionality, without difficulty. -- This means without
prejudice, and without hoop-jumping of any sort, such as (as commonly
encountered, esp. with commercial software) having to complete any sort
of registration form, having to divulge personal information that is
irrelevant to the goal of solving the user's problem or making the
developers aware that users are having the problem, having to login to
anything (other than their Internet account itself of course), or
outright having to pay money.

Here, the "without prejudice" appears to be being violated.

b) User feedback, possibly with duplicate similar incidents aggregated,
reaches the awareness of developers.

I am not confident that this is happening in this case. People that
speak with authority about the internals, as if they are developers,
are piping up, but they appear generally to be hostile to feedback.

c) Developers are not hostile to the above feedback; also, developers
consider behavior users don't like to be a bug, whether or not it was
intended behavior. If it was intended, but in the real world it turns
out not to be all it was cracked up to be, then it is a problem, as
surely as if it was unintended. Also, developers consider user feedback
about the usability of the interface to the software to be as important
and worthy of heeding as feedback about the actual functionality. If
users complain it's not obvious how to do X or that it's a pain to do
Y, this suggests that how to do X and Y may need some tweaking and/or
better documentation.

I am not confident that this is happening, either.

d) Development is active, preferably on a release early, release often
basis, with separate stable and unstable branches.

It's pretty bloody obvious that this isn't happening. We have
third-party mods that basically are LaTeX2e with bells on, rather than
major fundamental fixes to deeper flaws that have been identified. And
the recurring vaporware promise of LaTeX3 RSN -- which has been
promised for years, without anything actually appearing that ordinary
users can download and use. If development is active at all, then, it's
clearly not on a release early, release often basis, which means users
with gripes have little hope of seeing their concerns addressed before
the sun goes nova.

what you consider an enhancement someone else might regard as a
waste of time.

Often what one user considers an enhancement others consider a waste of
time. As long as the proposed feature can be turned off/avoided by
those uninterested in it, this does not support any argument against
its inclusion.

I don't want to argue who's right here, just to point out that opinions differ.

And software can sometimes support surprisingly many different opinions
without bloat, especially if there are parallel forks/distros oriented
at different target audiences (consider Ubuntu Linux vs. Red Hat!)
instead of, say, a Preferences dialog with sixty-seven tabs a whole
three of which are devoted to preferences-saving-behavior preferences.

Oh, and notice the extra ">" in front of the "From"? That is, as
far as I can tell, part of your posting as provided to me by my NNTP
server. (Anyone else get the same thing? or not? artifact of
posting through Google Groups?)

A Google Groups employee has recently been interfering with my postings
through GG. Initially, they disrupted my postings to a specific thread
in another group -- first, a bogus "posting limit exceeded" message (if
real, whatever "limit" was claimed was set ridiculously low); followed
by one followup appearing as nothing but "- Show quoted text -" when I
saw it listed the next time I checked the group after posting it. When
I examined it (by clicking the link), I saw that a single > character
had been added. At this point I began to suspect intentional
interference rather than glitches. When my followup asking what had
happened was rejected with the bogus "posting limit exceeded" error
message, my suspicions were confirmed: someone with privileged access
to my postings before they propagated, ergo a Google employee, was
intentionally diddling my postings to that thread. (Once is a glitch.
Twice is coincidence. Three times, especially twice in a row, is
intent.)

But it appears that whoever it is has now attempted the same thing in
this froup, trying to alter my posts so they seem to contain no (or
less) original material, perhaps in the hopes of hiding them from
people or making people more likely to miss some of what I wrote. That,
and the spurious error messages, indicate a more general aim of
censoring some of my posts, apparently unofficially (if I had done
something wrong as far as they were officially concerned they'd
probably just cancel my account, forcing me to make a new one) and
apparently in multiple froups.

Now I ask: is it one of my detractors here that put the employee up to
this? As far as I am concerned, disrupting my postings (either by a)
altering some of them from what I wrote before sending them out or b)
outright preemptively suppressing me from posting some of them)
constitutes hacking, i.e. computer intrusion, fraud, and abuse, and (in
the unauthorized-modification case) arguably copyright infringement or
plagiarism of some sort. As such, the employee and whomever passed them
the bribe are both liable for civil penalties and possibly criminal
prosecution.

Hope you're now feeling scared, whoever you are. To avoid legal
troubles, here's some advice: leave me alone. Should be simple enough
-- even a three-year old can understand and follow that one, simple
instruction. (Even if three-year olds often don't, they can.)

If you're looking for a justification (a "why this is, and it's
good") for all of your "why?" questions, you will probably be
disappointed. Sometimes the only answer to "why?" is "because
that's how it is" (possibly with a historical explanation).

I'll accept "for hysterical reasons" as an explanation, but not flamage
for simply having asked the question. Goodness knows there's a lot of
legacy cruft in TeX and descendants.

Anyone who has worked on a large piece of software with a longish
lifespan will probably recognize this phenomenon; there's probably
something similar in other areas of life as well. Sometimes design
decisions are made that in retrospect are not optimal. Revisiting
them is often considered to be more trouble than it's worth. You
may know this.

Yes -- sooner or later, a rewrite/ground-up new system with the same
philosophy and functionality is needed.

Could be. My impression is that you got crosswise with the regulars
almost immediately, for reasons not entirely your fault

Not "entirely"? Try not "at all".

What were the very first objections?
a) I didn't read the FAQ. It wasn't in the obvious place (recently
reposted to the froup).
b) When someone suggested I do so, I still didn't. Their "suggestion"
was the telegraphic phrase "Minimal example? someURL Ask questions?
someotherURL.". Looking at that, you can see how it seems to be
off-topic and is mainly URL by weight, which is commonly a hallmark of
spam. People tend to glance at a posting that looks like that and then
move on, ignoring it. Sometimes they even complain to the provider that
originated the message. Now, apparently, it was actually a link to a
Web copy of the FAQ (why two links though?) and asking me to provide a
minimal example to reproduce a problem. This should have been clear; it
would not be difficult to modify that text to be far more clear as to
what it is trying to convey.
c) When I suggested as much, I was told nobody has time to write
something more clear and in plain, full-sentences English prose. Well,
if someone doesn't have the time to write a proper response in plain,
easily-understandable English, then they don't have the time to write a
response and should shut the *** up, IMO. A cryptic, telegraphic post
is worse than useless: it may be misconstrued (as it was in that
instance) and worsens the SNR. (Later, I discovered that the cryptic
text in question resulted from someone spamming a macro key. That means
there's no excuse for its unclear nature; it could be a 3 kb
English-language essay and still be posted at the press of a key in
that instance. The work of composing a clearer, more legible version
would only have to be performed once, rather than once a posting.)
d) Any attempts to explain why I acted the way I did were responded to
with flamage, rather than a) reading and understanding how they had
failed to communicate clearly, producing what they perceived to be a
problem and b) resolving to avoid such miscommunications in the future.
So when I pointed out that the linky posting looked like spam and
exactly why it did, instead of the poster saying "Oops, sorry, I
thought it was clear" and altering their macro, they attacked me.

As far as I can tell, at no time did I commit anything that might be
considered to be a mistake, let alone act in intentional bad faith.
(Lack of information leading to a hindsight-poor choice is not a
mistake, it's having not had enough information and having done the
best you could with the information you did have. A mistake is "could
have done better with reasonable effort", as I'd define it.

that any further discussion is apt to be colored by the history.

That is wrong. History is irrelevant. If I ask a new question here,
anything that's passed under the bridge is irrelevant. (Note the word
"new"; a question I'd asked before that had been answered before may be
another matter.)

Do I need to change my posting preferences to change name after every
question then?

.