oracular ramblings

Hopefully this will help somebody else avoid the ‘fun’ of trying to track down why, exactly, LaTeX has suddenly gotten cranky, and (under OSX) the TeXShop ‘typeset’ command suddenly started dying miserably with:

pathsea: Running mktexmf ptmr7t
! I can't find file `ptmr7t'.
<*> ...:=ljfour; mag:=1; nonstopmode; input ptmr7t

Please type another input file name
! Emergency stop.
<*> ...:=ljfour; mag:=1; nonstopmode; input ptmr7t

A few quick basics — I’m running TeXShop as my front end, with the handy BasicTeX package (discussed halfway down the page) for a lightweight TeX backend on OS X 10.4.11.

It’s a bit embarrassing to admit that I hadn’t noticed that there was an issue with generating pdfs from TeXShop until some well intentioned directory cleanup. It appears that TeXShop will cheerfully show you an existing pdf by the appropriate name if one exists, and it can’t actually typeset the new version… sans error!

Once I’d discovered that there was a problem, the yak shaving began. Skipping over the chasing faint hints and vague suggestions, I found the following:

• IEEEtran clearly included something problematic
• This debian bug report which pointed out that the ‘ptmr7t’ font wasn’t included in the base TeX Live install.
• An exceedingly simple document resulted in the failure:

      \documentclass{article}
        \usepackage{times}
        \title{Brief Article}
        \author{The Author}
        \begin{document}
        \maketitle
        This is a test!
      \end{document} 

• It’s bloody hard to find what’s actually missing.

As it turns out, the problem seems to be that the PSNFSS macros only included the font description (.fd) and style (.sty) files … which wouldn’t matter if the psnfss definitions weren’t used for a number of basic things, including various math functions.

You can download the whole PSNFSS package — but the missing fonts for ptmr7t et al were in lw35nfss.zip.

From there, I extracted the fonts, and:

• Copied the extracted fonts from ./psnfss/fonts/[vf|tfm] over to $texlive/…/texmf-dist/fonts/[vf|tfm]
• Ran ‘texhash’ and ‘mktexlsr’ (it’s possible that I only needed one or the other, but I wasn’t about to try and break things again, just to see)

Hope that helps… or at least reduces your yak shaving!

After I found out that my LISA ‘07 talk about The Security Butterfly Effect had been scheduled for 2pm on Friday – the last day of the conference, and just before closing ceremonies – I decided my job was edutainment, more than education.

I’m prone to involving my audience in the talk at the best of times – and this was no exception. Heckling was welcome and encouraged, and audience participation was as mandatory as possible with your average herd of sysadmins.

That aside, the general gist of the talk is this:

“Small variations in the initial condition of a system may produce large variations in the long term behaviour of the system”

Most of us don’t actually know what the initial condition of our system (environment, if you prefer) is. As such, we make countless assumptions every day about the initial state of our system. With the increase in automation and automated processing of information, the assumptions that we’re functioning under are often neither ours, nor visible to us. The rise of the machines and vast increases in complexity make it easy to miss that first small misstep that leads to later, catastrophic events.

A much more interesting topic that I didn’t really explore is the interaction of complex systems, where singly innocuous changes become serious vulnerabilities in combination.

I always have a dreadful time setting out to write, whether it’s writing a blog post (as you can see!), writing a paper, or writing an article. The only thing that seems to flow freely is pithy responses to mailing lists.

Given that, I was happy to find the following set of links, starting at the excellent Light Blue Touchpaper with How (not) to write an abstract, following a comment link to Microsoft’s Simon Peyton Jones delightfully pithy presentation about writing a paper, and finally ending up at a lovely collection of links put together by Mark Leone, and referred to at the end of spj’s presentation.

The beginning of wisdom is to call things by their right names.
— Chinese Proverb

“Everybody knows what reputation is — why bother to define it?”

The OED has four pages worth of definitions and examples of reputation – even pithy and lightweight dictionaries provide several options. Here I’ve defined a starting point and a common set of descriptions to build from. If everybody knows what reputation is, clearly this starting point will be background information. If not – or if you feel that reputation is something other than described, this can be a starting point for discussion.

Reputation is only a candle, of wavering and uncertain flame, and easily blown out, but it is the light by which the world looks for and finds merit. — James Russell Lowell

Reputation[0] is a poorly described and understood area with strong links to identity and trust. The following is a summary of my thoughts (with a little help from my friends[1]) on a definition of reputation.

Formally, we can say that:

Reputation is a shared subjective valuation by other entities about a property of an entity.

Thus, no entity can create reputations on its own, either for itself or for others. While a reputation may be based on proveable facts, there is no inherent requirement that a reputation be proveable.

Less formally, reputation is another subject that could easily be described by the infamous quote:

I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description; and perhaps I could never succeed in intelligibly doing so. But I know it when I see it… — Mr Justice Stewart

More practically, although there are certainly a variety of areas where fact, rumour, opinion and reputation overlap, reputations do have some clear differentiating properties.

Using the scurillous statement “Robert’s mother wears army boots”, it’s initially hard to decide if we’re looking at a rumour, a fact, an opinion or a reputation.

Breaking down our definition of reputation, we can explore the murky depths of how to easily describe whether this statement is a reputation – or what additional information is needed in order to be able to describe it as a reputation, rather than a fact, an opinion or a rumour.

• A reputation is a statement by others about a property of some entity
  Our entity here is Robert’s mother, and we’ve definitely got a property here, specifically that she “wears army boots”. The important condition that this is a statement by others isn’t met though (although it is implied by saying “Robert’s mother”).
• A reputation is intrinsically unproved
  There is no expectation that a reputation is in any respect based on fact; reputation can dance on the border of slander and libel, or be based on hard statistical data[2].
  “Your mother wears army boots” is a classic grade-school taunt, and in this case, we don’t know if Robert’s mother does or doesn’t wear army boots. We also don’t know if the hypothetical boots that Robert’s mother wears are actually army boots, or might be rubber galoshes!

  Obviously, if we have empirical proof that Robert’s mother does wear boots – and those boots are army boots, we have a fact on our hands. This doesn’t mean that we don’t have a reputation – Robert’s mother can still be known for wearing army boots whether there’s any factual basis for her reputation.

• A reputation is communicated between and by multiple entities
  A reputation must be believed to be ‘factual’[3] by multiple entities, and be communicated between multiple entities; a valuation that isn’t shared by multiple entities is merely opinion, while a statement that has a single source (anonymous or not) is either an opinion or rumour.

  We’re definitely falling down here – we don’t know how we came by the statement that Robert’s mother wears army boots! This could be a rumour, since rumours are generally described as unverified statements of uncertain origin – and this is certainly an unverified statement – but rumours are typically also of uncertain origin. If we’d been told “I hear that Robert’s mother wears army boots”, that would clearly be an uncertain origin, and thus a rumour, but all we have to work with is the bald statement “Robert’s mother wears army boots”, which could easily be of a very certain origin.

In the end, we’ve got a statement that’s neither fish nor foul. Without empirical evidence that Robert’s mother does, indeed, wear army boots, we don’t have a fact.

If we knew more about the source of the statement, we might have an opinion – the source might be declaring that Robert’s mother is wearing something on her feet that it interprets as army boots (or that it holds the belief that Robert’s mother wears army boots).

We probably aren’t talking about a rumour – “Robert’s mother wears army boots” is a statement which while it doesn’t have a defined origin is in a format that typically implies a very certain origin (namely the source of the statement).

On the other hand, we don’t clearly have a reputation here either – we don’t know how many entities hold this particular valuation about Robert’s mother.

If we change our murky statement slightly, we can separately meet fact, opinion, rumour and reputation.

• Fact: Robert’s mother wears army boots as can be seen in these photographs of Robert’s mother..
• Rumour: Somebody said that Robert’s mother wears army boots.
• Opinion: I think that the boots Robert’s mother is wearing are army boots.
• Reputation: Alice, Bob and Edward have said that Robert’s mother wears army boots.

[0] Why reputation? It’s my area of interest and research – how social and computer networks interact and impact trust, identity and privacy.
[1] Many thanks to Nick, Ben, Adam and the folks I should remember but have blanked on.
[2] This leads to interesting implications about trust and identity, since determining the apparent value of a reputation also depends on the source(s) of the reputation, and whether it’s possible to verify the content(s) of the reputation.
[3] Factual in this case being a subjective not an objective understanding