Disqus - Latest Comments for mcherm

Re: Jay's Blog

Michael Chermside — Sat, 17 Jul 2021 07:58:48 -0000

Thank you!!

I find that to be a much more satisfactory "conclusion". Listing some particular results with practical applications, along with the acknowledgement that (much as with the continuum of real numbers) for practical problems one could always (in theory... it might sometimes be to hard) construct a choice function explicitly but it is far easier to have a general rule allowing it -- that wraps it up more completely.

This was a really useful and well-written essay.

Re: Jay's Blog

Michael Chermside — Fri, 16 Jul 2021 14:25:20 -0000

This was a WONDERFUL introduction to this topic -- very accessible writeup. I am putting this in my list of interesting writing on the topic.

Unfortunately, while I really liked the way you expressed everything else, I felt like the end was unsatisfactory.

You wrap it up by saying the following (I'll attempt to paraphrase):

The Axiom of Choice (AC) is independent of our axioms for set theory, so we could take it to be true OR false without contradiction. We choose axioms because they are USEFUL. Dealing with infinite quantities is weird and unintuitive (even without considering AC), but it is useful because it allows us to do things like approximate the physical world as continuous. The unsavory implications of AC all occur with truly infinite sizes so they never come up in practice. Therefore we should be comfortable accepting AC.

What you set out to convince us was that we should be comfortable accepting AC. But what you demonstrated is that we should be comfortable accepting infinite sets. If you ALSO showed that there are useful parts of mathematics (like approximating the world as continuous) which work with AC and don't work with not-AC, then this would be convincing. But you didn't show that (and, while I'm certainly not an expert, I don't know of any practical ways in which ZF + not-AC gives different results for problems that DON'T involve actual infinities).

Any thoughts?

Re: Trolls Target Female Scientist

Michael Chermside — Mon, 15 Apr 2019 13:13:44 -0000

The title is completely incorrect. The trolls may disparage her contribution, but they do not diminish it.

Re: Never accept a counter-offer – Vardan Torosyan's Blog – Vardan Torosyan's Blog

Michael Chermside — Mon, 03 Oct 2016 09:18:29 -0000

> I think if you’re happy with your current job and the salary is the only
thing you’re unhappy with, you shouldn’t be looking for a new job in
the first place. You need to communicate with your manager and ask for a pay raise.

Actually, I think looking for a new position and bringing information about what THEY are willing to pay is one reasonable way of "asking for a pay raise". Not, to be clear, a technique I have used myself, but I think it is a very reasonable approach.

Re: Schlock Mercenary

Michael Chermside — Fri, 12 Jun 2015 15:28:17 -0000

Schlock Mercenary has been, ever since you got your feet just very shortly after the strip began, rich in character, scene, and plot, and one of most consistent pieces of writing I have seen anywhere. 15 years of DAILY output is impressive, and output of this quality is just astounding.

I have a list of best online comics: it includes fantastic things like Girl Genius, things that have grown with age like Questionable Content, simple but venerable pieces like Freefall, and some with rich, complex plots like The Order of the Stick. There are even a few newer items that have risen near the top like Gaia or Quantum Vibe (recently ended). But Schlock Mercenary is the first one on my (sorted) list, and given the way you work, I don't expect it to be displaced anytime soon.

Thank you so much for your fine work. I hope it continues to be as profitable and rewarding to you as it is enjoyable to your fans.

Re: Girls With Slingshots - GWS #2008

Michael Chermside — Mon, 16 Mar 2015 11:37:04 -0000

I just wanted to say, Thank You. You have managed to create a Fine Thing... which is something not everyone accomplishes anytime in their life.

Here's hoping you get some rest then go out and create another Fine Thing.

Re: Yoink

Michael Chermside — Fri, 16 Jan 2015 08:36:04 -0000

Too bad, I would have loved the Quark Express joke.

Re: Schlock Mercenary - Touching on the Topical

Michael Chermside — Sun, 14 Dec 2014 11:18:12 -0000

I spent a good 10 minutes searching, and didn't find this blog, so if someone DOES know where it is, please post. Howard... if you included a link in the original post that would be great.

I did not see it mentioned in Howard's twitter feed (https://twitter.com/howardt.... I did not see it in his (outdated) livejournal (http://howardtayler.livejou.... I did not find it in the primary blog on the SchlockMercenary site (http://www.schlockmercenary.... I didn't see it mentioned in his regular publication "Writing Excuses" (http://www.writingexcuses.c....

I'm sure there is some obvious place that I forgot to look, and I'm hoping that Howard (or someone else) will point it out. And if not, then some of those links above are fun things to read instead.

Re: An Overwhelming Majority Thinks We Shouldn't Convict Based Solely on a Rape Victim's Statement

Michael Chermside — Mon, 27 Oct 2014 12:11:46 -0000

I would vote for option 2 as well... given the scenario you described.

Suppose a woman goes to the police and reports that she has been raped. The rapist, she says, used a condom and she spent the 24 hrs since the incident sobbing with the support of a close friend and showering repeatedly -- thus a rape kit will show no evidence. She looks through store video footage and claims to recognize her attacker. Credit card records lead to a person from a town 80 miles away. In court she identifies him as the rapist, but he testifies that he just drove there to pick up a pack of cigarettes at the store and walk around, and that he has never seen her before. (Whether he has an existing record or police had previous suspicions of him are facts not admissible in court.)

In THIS case there is a good chance I would convict. But in this case I have background facts (like the reasonableness of his being in the location) and the plausibility of each person's story. I don't have "just" their statements.

In a different case, a woman comes to the stand and says "that man raped me". The man comes to the stand and says "No, I didn't." Without question, I would find "not enough evidence to prove him guilty beyond a reasonable doubt." And this seems to be the situation you described in your poll question.

Re: Using github pages on github enterprise

Michael Chermside — Wed, 03 Sep 2014 15:00:46 -0000

Thank you! This was exactly what I needed to know.

Re: Ferguson fallout: The daunting odds of a meaningful dialogue about race

Michael Chermside — Tue, 26 Aug 2014 08:37:41 -0000

I think you are correct... that we need to use Ferguson as a stepping stone to spark two conversations: one about the relationship between police and the community (willingness to shoot a suspect, "stop and frisk", over militarization of the police), and another about race, particularly as it relates to the first topic.

But how do we make this conversation happen? These are issues that I have been aware of for some time... the events in Ferguson, while extreme and unacceptable, were not surprising to me. But I have done nothing to change them. How CAN we make something productive out of the Ferguson affair?

Re: N.J. woman, 98, now seeks redress in Rosenberg spy case conviction

Michael Chermside — Tue, 26 Aug 2014 08:25:38 -0000

tomblair:

What on earth is your problem?

It sounds to me as if your mind is closed. You say "I don't believe her.", and yet you have not read the government documents that she says were just released and prove her innocence. You are making up your mind without the facts -- so I am forced to conclude that you are being driven by prejudices and political preferences rather than actual knowledge or understanding.

For what it is worth, I would say the same if you had loudly claimed she was innocent without any evidence.

Re: O’Reilly purchases Pearson’s stake in Safari

Michael Chermside — Mon, 04 Aug 2014 16:26:06 -0000

Will Safari continue to carry books by Peterson and other publishers, or will it slowly revert to a mostly single-publisher offering?

Re: PyPy.js: First Steps

Michael Chermside — Tue, 06 May 2014 15:33:33 -0000

So, you managed to achieve your "lofty, crazy, good-for-motivation-but-likely-futile goal" within the first few months of 2014, less than 12 months after you posted this. I am very impressed; I would never have thought it possible.

Re: The Glitch That Will Kill Bitcoin

Michael Chermside — Fri, 14 Feb 2014 10:21:38 -0000

I own bitcoin: partly as an investment and I also use it for occasional purchases (more frequently as venders start supporting it). However, I would NOT be one of those 140,000 "accounts" you list because my bitcoins are stored at Coinbase (in several large accounts managed by them).

I am not saying you are absolutely wrong -- to the contrary, I really appreciate your fact-based approach -- but I wanted to point out that there is one possible way in which your are undercounting usage.

Re: maori.geek

Michael Chermside — Fri, 31 Jan 2014 12:52:06 -0000

Great article. I just wanted to drop this here:
http://shop.oreilly.com/pro...

It's newly published a book on jQuery deferreds. It was written by a friend of mine, so perhaps I am a bit biased, but I think the book does a really great job of explaining things with examples and exercises. Someone who was confused by this article or who understood it all and wanted more might want to check out the book.

Re: Amended Reply Brief in United States v. Auernheimer

Michael Chermside — Mon, 30 Dec 2013 11:45:43 -0000

Really very nicely written and concise. As a non-lawyer, I appreciate good, clear legal writing that explains itself well enough to be understood by any party, not merely a legal expert. By that standard, I found this brief to be quite clear and persuasive.

Re: How to use the Python Imaging Library

Michael Chermside — Thu, 07 Nov 2013 13:30:22 -0000

Instead of using PIL, try using PILLOW instead. (http://pillow.readthedocs.o... PIL had been only lightly maintained so PILLOW is a fork that was created and which is more up-to-date (supports Python 3 for instance) and with newer bugfixes.

Re: Real-time payments. No interchange. An idea is now reality.

Michael Chermside — Fri, 25 Oct 2013 08:59:53 -0000

Dwolla:

I'm not sure what is up with all the Bitcoin comments. I am a strong supporter of Bitcoin, but I think this real-time payment implementation that you (Dwolla) are rolling out looks wonderful. The Bitcoin advocates need to realize, this is a PAYMENT mechanism, while Bitcoin is a CURRENCY. Hey, in the future there is no reason that Dwolla couldn't support Bitcoins, Pesos, and a multitude of other currencies on the same platform.

I am much more worried for the company that I work for (a major US credit card provider). If we can't manage to innovate this way, we (the credit card companies) are at risk of becoming dinosaurs: huge (and profitable) but doomed to extinction. My own hope is that my company can step up to the plate and give Dwolla a run for their money, but with features like this I'd say they're a few points ahead of us right now and we need to bat in a few runs to catch up.

Re: Why “Innovation Experts” are Selling Snake Oil

Michael Chermside — Wed, 03 Apr 2013 10:14:25 -0000

You point out that employees not understanding what the comany really needs to accomplish (what the important problems are) is a common impediment to inovation and that not having an "innovation room" in which to work is nearly never the impediment. I would like to point out one other common impediment: a corporate culture that does not encourage actually acting on innovative suggestions.
People at your company like to stand around the water cooler and gripe about things that annoy them. (I know this because people do this at EVERY company, although the "water cooler" may be a different location in your office.) If those conversations often end with "Hey, that's not bad: I'll bet that would really help. Let's try making the change and we'll see if it works.", then you're doing great. If they almost NEVER end up that way it might be that your employees are lazy bums who can't be bothered to fix the stuff that bothers them. Or just perhaps it's not the lazy bums, but the fact that the institution is set up to discourage that sort of change. Can employees start a (small) project without having to paperwork documenting the ROI of the change and getting it approved by the projects committee? Do people other than management ever make proposals? If not, then can YOU change that somehow (even within your own group)?

Re: What to Do When Your Non-Technical Boss is Just Plain Wrong

Michael Chermside — Fri, 01 Mar 2013 11:35:13 -0000

This article really does an EXCELLENT job of pointing out some specific techniques that are effective for addressing this problem. I'd like to mention one additional approach that makes a huge difference over that long run.

It helps a lot when the non-technical manager really trusts their technical expert. If you have a long working relationship and excellent track record then they are more likely to consider your advice and defer to it. But how do you establish this record in the first place? Often non-technical managers feel frustrated because of their inability to understand some of the things they are trying to control. This only gets worse when "that article I read" seems to contradict "the tech guy that works for me". And it is worst of all in managers who USED to be technical and really don't like to think of themselves as losing mastery of the technical details.

So as the technical expert, you can cut through all this by working extra hard to communicate the reasons behind your recommendations. Never suggest that they should "just trust you", instead go out of your way to explain, without excessive technical details, the reasons for your suggestions. For instance, "Multi-threading is an excellent solution for many servers, because it lets the computer work on several things at once. But our app is 'IO bound', which means that it usually waiting for a connection, and doing other work in the background -- we are ALREADY doing several things at once -- multi-threading won't help." There is a real art to describing technical things in an elementary fashion, but it pays off in situations like this.

Re: Through sickness and health … and joint accounts?

Michael Chermside — Tue, 04 Dec 2012 14:14:14 -0000

What I really want is BOTH--joint account and individual too.

Re: Simpler Twisted deferred code via decorated callbacks

Michael Chermside — Mon, 15 Oct 2012 10:14:41 -0000

I'm curious whether the timing of this insight is in any way related to Guido's recent call for input from people experienced in using deferred and other asynchronous APIs.

Re: What My Pacifism is NOT

Michael Chermside — Mon, 01 Oct 2012 19:21:11 -0000

> My pacifism does not reject violence as a solution in some situations.

[...]

Hmm... Somehow I never considered this. Years ago I stopped calling myself a pacifist (but did not stop practicing the same behaviors) because of just such situations. Perhaps instead I should have used a richer, more nuanced definition of pacifism.

Re: Always Go Home with the Lady Who Brought you to the Dance

Michael Chermside — Sun, 04 Mar 2012 02:58:02 -0000

Ah, yes: that's a reasonable distinction to make.