No Food for Thought

Food is something you should provide to your brain long before coming to this blog. You will find no food recipes here, only raw, serious, non-fake news for mature minds.

The European Union's Interference Equation: Inaction + competition = over-reaction

admin Saturday July 21, 2018

In 1996, yearly worldwide PC sales went beyond 70 million units. It was obvious that personal computers would become ubiquitous and that the World would crucially need operating systems and commodity software.

The network effect on computers was already known in the early 1980s. In fact, in 1996, software vendor lock-in was already very much a reality.

In 1996, it was obvious what would happen if the world didn't make such software and let private companies and individuals tackle the problem. In fact, in 1996, private companies had already started creating operating systems and software with intentional vendor lock-in, and individuals had already started creating badly underfunded free software. Efforts were duplicated and allocation was highly inefficient. In 1996, the main PC operating system was Microsoft Windows 95; the software world was already plenty messy.

And yet, in 1996, neither the United States nor the European Union, nor any other union decided to offer its citizens a "universal operating system". A few years later, the situation had unsurprisingly worsened, and Microsoft's dominance was even greater.

At that point, the World could have learned from its errors and decided to avoid doing the same errors with the next big innovations. But rather than offering a public Web search engine or starting an operating system for mobile devices, the United States dug up old legislation and recycled it to pretend it was doing something about the problem. In 2001, the United States government sued the largest software enterprise which its inaction had forced to fill in the gap, claiming the Evil Microsoft had broken antitrust legislation. A settlement stopped the government short of dealing a grave blow to entrepreneurship and the free market.

The European Union, having done nothing more than the USA, was plagued by the same problems. In 2004-2007, it used the same pretext found by the USA to suck a €497 million fine from the dangerous Microsoft, which was evil enough to include a media player in its operating system. The EU too would have done something about the problems.

Smaller computers, bigger interference

With all that energy spent blaming the Bad private sector, the World had no energy left to prevent repeating its errors, with the advent of the Internet and handheld computers. The World let the private sector provision web search engines and operating systems for handheld devices. We let Google build a web search engine, and when it created an operating system for handheld devices, surprise surprise - the Evil Google made Android's default search engine Google Search.

Will these new failures be enough for us to learn and tackle the next problems before it's too late? At least in Europe, the answer seems to be no. The EU has chosen instead to stick to its pattern and slap a €5 billion fine on Google, that Evil innovator which provides the world with the open source Android operating system. As for the USA, its current political situation sweeps away any chance of seeing the Federal government anticipating any future problem in the next years.

Intervention in moderation

If further state intervention is to be expected, is our world doomed to have neither the public nor the private sector provisioning solutions to coming challenges? Or can these interventions be turned into something positive?

I believe governments can indeed intervene without undue interference. There is no better way to illustrate how than to use a real-life example, so let's take the first Google practice the EU blames Google for:
Google has required manufacturers to pre-install the Google Search app and browser app (Chrome), as a condition for licensing Google's app store (the Play Store);

If such a practice is deemed problematic, here are 2 alternative interventions I suggest:

  1. Forbid sale of the relevant devices to minors without agreement from their tutors
  2. Inform customers buying the relevant devices about the practice and require them to confirm their understanding of that practice in order to complete the purchase.

By using such moderate interventions instead, we would let transactions without negative externalities occur, but we would also make citizens aware of problematic practices, and - perhaps - make consumers wonder why such practices have come into existence.

AV1: A Victory for Open Video

admin Saturday July 14, 2018

Rejoice. One major component of modernity - video - can now be compressed using a stable, specified and fully open royalty-free video codec, with AV1, which is also most efficient.
I did not play any role in this achievement, so I have to thank all contributors, notably Google, Cisco, Mozilla, VideoLAN, IBM, Intel, AMD, ARM, Microsoft, Netflix and NVIDIA, BBC, Amazon and Realtek.

Impatiently waiting for an AV1-ready environment

"Cryptocurrency": the pseudo-currency which no one understood or accepted, but which many bought

admin Tuesday June 26, 2018


This article was written when "cryptocurrency" hype was peaking. While the situation changed, its conclusions remain relevant.

On the cover of the May issue of Canadian magazine L'Actualité, I read "LE BITCOIN POUR LES NULS" ("Bitcoin for Dummies"). One of my first reactions was "true". In fact, it announced an article trying to explain what "cryptocurrencies" are, titled "Bitcoin, la monnaie que personne ne comprend" ("Bitcoin, the currency which no one understands").

I find that title interesting in that those who "own" Bitcoins surely do not "understand" it. But it seems to suggest Bitcoin can be understood. As I expected, the article didn't directly tell readers to "buy" "cryptocurrencies". It warned about weaknesses. But as I expected, it overall suggested that "cryptocurrency" had value, just perhaps not as much as some claim. The article even quotes a guy trying to make a living by selling "cryptocurrency" claiming that we can't miss out on "this type of asset[sic]".

It seems that with the current hype about "cryptocurrency", the critical view no longer has as much prominence as the view of scammers/enthusiasts. In fact, it seems the critical view doesn't have any room left. So I thought it was time for me to intervene.

What is "cryptocurrency"

I couldn't decide which definition of "cryptocurrency" was best, so here are a few.

cryptocurrency (plural cryptocurrencies)

  1. a clever term coined to suggest that worthless bit sequences sold by clever terminological scammers have value, like actual currencies
  2. an invention which is so abstract and seems so complicated that it must be worth something
  3. a unit of nothingness whose value is so cryptic that it currently manages to trade above nullity

What is "cryptocurrency", really

I guess I should explain more clearly for those who didn't understand the humorous definitions above and intend to "buy" "cryptocurrency". Basically, "cryptocurrencies" have no value. Not because there's hype about them. Not because they degrade the environment. Not because their hypothetical value is highly variable. Not because they are involved in an economic bubble (if that term can be applied to something with null value). Not because they are highly vulnerable and easily compromised. Not because they are used by criminals. Not because they are inefficient. Not because their price is manipulated. Cryptographic "currencies" have no value because, like other immaterial "alternative currencies", they do not store value.

While this may seem obvious to economists, this has gotten less intuitive to the average developed world citizen, with the growth in types of intangible assets. Many intangible things have value. Patents, texts and other intellectual property are intangible, but can have value. And many without financial knowledge would even consider all stocks as intangible. Moreover, most real currencies no longer have intrinsic value. But even though paper money has negligible physical value, representative money, in the broad sense, has value, relative to the issuing government(s).

The problem with alternative "currencies" is that, by definition, they are not fiat money (money backed by governments). Alternative "currencies" are issued by random entities. Usually, even these entities do not guarantee anything in exchange for a unit of these "currencies".


To be fair, some immaterial alternative currencies have tangible backing. One of the most notable cases is the Toronto dollar, created in 1998, which has fixed exchange rates with the Canadian dollar, and which is backed by the Family Life Foundation of Willowdale.

Except that in fact, the Family Life Foundation of Willowdale (now just "Family Life Foundation") is a small religious organization. When the Toronto dollar's hype fell and honoring the exchange rate became more costly than sales generated, the Canadian "charity" obviously stopped honoring its guarantee, and 15 years after its creation, the currency lost its only backing. Therefore, those who had bought or accepted Toronto dollars are now stuck with a "currency" which - of course - nobody wants anymore.

While the exchange rate being lower for sale than for purchase made it obvious from the start that the currency was more targeted at faithful souls than at investors with the slightest hint of judgment, it is now obvious to even the wishfulest thinkers that Toronto dollars have no value. Has the issuer apologized for the 100% loss of its customers? Apparently not, since the website of the Family Life Foundation no longer even refers to Toronto dollars. It even gave up (if it didn't sell) its domain to an even more doubtful entity using it to boost its sales.

Will those who created Bitcoin apologize any more when the Bitcoin crashes? No. And there are several reasons for that:

  1. Unlike classic traditional alternative "currencies", the Bitcoin's creators did not sell all bitcoins, just the first bitcoins, making perhaps a few million euros at most.
  2. The Bitcoin's issuer has never backed it by any guarantee.
  3. The Bitcoin's creators are anonymous.
  4. The Bitcoin's creators are smarter, and the "currency" they created is complicated, so that "nobody understands it". Which is why many honest people have joined the scam, despite #2, and despite #3. The Bitcoin never had any value, even hypothetical/temporary. It never needed any backing, and will never lose such backing, so the Bitcoin's actual "value" will never be more obvious than it currently is (unless citizens get less geeky / more educated).

But, theoretically, reliable entities could strongly back a cryptocurrency, giving that currency some value. I insist on "theoretically" and "reliable". A corporation which creates a "cryptocurrency", claims some guarantee and then tries to sell the first units is not a reliable entity (or, in other words, the value given is proportional to the reliability; such an unreliable corporation would not provide significant value).


If you still think a "cryptocurrency" might have value, the main question you should ask is "What is that value?" And the simplest way to answer that is to ask yourself not for how much you could re-sell it, but what you could buy with a unit of that "currency". And - if you want a more precise answer - what you could buy with that same unit in 1 month, 2 months, 4 months, etc. If the answer to all of these questions is "Nothing", then the value of that "currency" is null (unless you are confident that there will be irrational market participants willing to pay something).

A quick search turned up surprisingly little material debunking "cryptocurrencies" (although perhaps I am not alone feeling a bit bad treating the subject, fearing to give it even more unwarranted publicity). The main document I found is David Golumbia's Cryptocurrencies Aren’t Currencies. They Aren’t Stocks, Either, which does a good job debunking propaganda used by many "cryptocurrency" merchants.

National currencies

This post is not meant as an apology of national currencies. The ideal currency would have a constant value, would be ubiquitous, would cost nothing to produce and would be secure. National currencies have a variable value, and even their rough stability is reliant on the good will of their governments. National currencies are numerous. They use coin which require material to produce. Verifying the authenticity of coin can be difficult, and even authentic coin can be stolen. Undoubtedly, national currencies are not ideal.

If my understanding of "cryptocurrency" believers is correct, the error they make is to think that since national currencies are imperfect, it's OK for "cryptocurrencies" to be imperfect. Imperfection is comparable to imperfection, isn't it? The simple appearance of a second "cryptocurrency", and then of a third, and then... of countless more, and the fact that anyone can create its own, should have made it obvious that "cryptocurrencies" were worthless, but it didn't for many. In good part because there had already been numerous currencies for centuries, legitimately.

If I decide that my house's land secedes from Canada to form a new "Filipstan" country, on a 2500 m² surface, against the will of all Canadians but my own family, that I have no army nor natural resources, even if I claim that Filipstan taxes will be collected in "Filipoin" and Filipstan emits 1010 units of its own "Filipoin" currency on the first day of independence, you should probably not pay much more than its melt value to buy a Filipoin. But, if there's any chance that Canada does not annex Filipstan on the next day, Filipoin will still be worth more than any cryptocoin (theoretically, not marketingly). At least, it will have some backing.

National currencies are problematic. But trying to create a better currency won't be possible without backing that currency, whether or not that currency is humanitarian, multicultural, multilingual, non-religious, non-partisan, cryptographic, fractal, recursive or whatnot (apologies to those who were about to launch a fracurrency - better luck next time!).


If you've "invested" all your fortune into buying (or "mining") "cryptocurrencies", don't despair. Peak Crypto is hopefully behind us, but you most likely still have plenty of time to sell to a better believer than you are, per the Greater fool theory.

If you were considering to "buy" "cryptocurrencies" though, realize that not losing means you will have to find someone as foolish as you are. It's a zero-sum game in which entry requires actual money or energy, so if you want to avoid losses, you'd better excel in technical analysis.

"Cryptocurrencies" may be an ecological disaster and worthless, but I think the phenomenon they created teaches a valuable lesson to learn when they can claim more than 300 billion euros in "market capitalization", even though it has been nearly 10 years since they were introduced and several exist.

When I was a kid and my mother refused to buy me some toy, I countered her argument about lacking money saying that she could just use her plastic card. Eventually, like most people, I understood that for representative money to be valuable, it needs scarcity.

What few ever understand though, is that a currency stores value. And having followed a university class introducing macroeconomics, I must say this is no surprise. The class treated central banks, and explained how private banks "create" money. But it never explained how backing affected a currency's value.

Our reaction to the cryptocurrency phenomenon should not be to prohibit alternative currencies. While we should try preventing sales of alternative currency to minors, if adults want to hope that the "currency" they believe in will become the future's global currency, we should not prevent them to put their wealth into getting a share of that dream any more than we prevent people from financing any religion. We could mandate cryptocurrency sellers to warn each buyer that any currency's value is proportional to its backing, but we live in a free market where mentally mature and sane individuals should be free to pay what they want for things they want, no matter how worthful - or worthless - they are. People buying alternative currencies are not hurting others, and not even literally hurting themselves. People "mining" are not necessarily hurting anyone, and if they do by consuming dirty energy, the negative externalities should be compensated by regular tax disincentives.

We should not prohibit Bitcoins, Litecoins, foocoins or the next clever scheme alternative currency makers come up with. In part because that would be pointless; there is an infinity of worthless things we can create. And in part because some complementary currencies could be valuable if they had significant backing. What we should definitely ensure is that people who study macroeconomics, and ideally not only these, get an understanding of how actual money stores value.

2018-08-19 Update

This article's publication marked the end of Peak Crypto. Now that the hype is well over, it is time to "sell" your "cryptocurrency", if you still haven't done so.

2018-09-04 Update

The hype is really over, now that The Economist acknowledges "cryptocurrencies" are useless. Hurry!

2018-09-16 Update

While "cryptocurrencies" keep crashing, nuance about their worthlessness is getting more important. 2 "stablecoins" have appeared, "stablecoin" being another clever marketing term to suggest that a cryptocurrency's value will be stable since it is pegged to a stable asset - in these cases, the USD. The guarantee of convertibility is simply provided by the private company which issues each one of these "currencies", but the Gemini dollar is regulated by the USA and seems to have a guaranteed value of 1 USD.

Another cryptocurrency with a guarantee from a state was launched in February - Venezuela's petro, which is directly backed by natural resources of the country. However, the petro cannot be "mined", and appears to be a scam from the government to help alleviate Venezuela's financial crisis, nearly as worthless as traditional "cryptocurrencies".

2018-10-12 Update

Rush! The implosion is coming sooner than expected!

2018-11-16 Update

With the Bitcoin fever over, Axios has published graphs putting Peak Crypto in the first half of 2018.

2018-11-24 Update

Bitcoin and its brothers are getting more prominence in shops. Will "cryptocurrency" owners soon be able to use their "currency" as such? No such luck... the shops won't accept it, but they'll "sell" it. Next to other "goods" with a clear value, such as tobacco... at least, tobacco buyers can feel a little less guilty about wasting their money now, by simultaneously "investing" in another poison.

Ah, the Ultimate Irony...
"Cryptocurrency": the pseudo-currency which no one understood and which stores never accepted... but sold!
Yep, even with its contradiction now exposed in broad daylight

Le poids (ou la masse?) des mots

admin Monday April 30, 2018

Il y a quelques mois, ce blogue faisait pour une première fois référence à un article de la série "Des gars, des filles" du magazine L'Actualité. En lisant « Le poids des mots », que signe Noémi Mercier dans le numéro de mai 2018, force est de constater que cette série, qui n'est certainement pas à veille d'être à court de pertinence, n'est toujours pas non plus à court de contenu. J'ai trouvé cette courte lecture (1 page) très enrichissante.

Voici l'étude de Armand Chatard auquel le sixième paragraphe réfère, et selon lequel les garçons sont plus confiants de pouvoir exercer un métier traditionnellement féminin lorsqu'il est présenté en utilisant autant le masculin et le féminin que lorsqu'il est présenté en utilisant seulement le masculin.

La belle langue? Si ce fût autrefois le français, les critères de beauté et la société ont bien évolué depuis. Lorsque de simples mots amènent des maux aussi complexes, il faut plutôt catégoriser le français parmi les langues sexistes (et peut-être réattribuer cette épithète obsolète à son descendant sans sexisme linguistique, l'Ido).

Goodbye Flash, Hello Complete Freedom!

admin Sunday April 22, 2018

Although Debian has been my main home PC's OS for more than a decade, I've always used proprietary software on it. Usually various drivers or firmwares, and sometimes applications. But always, the 2 most popular browser plugins: Adobe Flash and Sun's Java plugin.

That was until approximately 2013, when I realized Sun's Java was no longer needed. And until now. Today, it had been a while since I had remembered to update my main PC's Flash plugin. And as happened last time, Debian's buggy flashplugin-nonfree package failed to perform the update. I thought it was time to check how badly my plugin was outdated, so I checked the version I had. And, after a while wondering what was going on, I realized I didn't have the plugin installed.

And indeed, my shell history shows I tried updating the plugin in September, and when I realized that was broken, I decided to uninstall it. It's coming back to me now. I uninstalled it thinking "Let's see what happens without". And I believe I realized playing videos got different. I do have considerable performance problems playing some videos now, and the lack of Flash may be a key factor in that (but it's a 5-year old PC, and perhaps there's a driver problem).

If you had asked me in 2006 if I could browse for 7 months without Flash without failing to do anything, I would have answered a most confident No. Yet, I am now using a free browser with 0 proprietary plugin/extension, and I didn't even realize it! And indeed, less than 5% of websites now use Flash! In fact, on this PC, the only proprietary software I still use is Citrix Receiver (for remote connection to work), AMD CPU and GPU microcode, and some proprietary scanning plugin for my HP LaserJet Pro all-in-one printer.

Thanks VP9, thanks Opus, thanks WebM. Thanks HTML5, ECMAScript and the rest
Congratulations Google, congratulations free codec developers. Congratulations web standard developers, web browser developers and webmasters
We made it!

Modestly Moving Away from a Monstruously Mad Mozilla

admin Sunday April 22, 2018

In 2003, I was using Internet Explorer, Hotmail and Microsoft Windows when I discovered the Mozilla Suite. I could certainly count the number of open source applications I used on a single hand at that time. I installed Mozilla 1.4, a friend told me about Firebird, and then I switched to Firebird and never went back to MSIE as my main browser. Impressed by Firebird (now Firefox), and curious about free software, I migrated from Windows to GNU/Linux later that year. Since then, Firefox has always been the main browser on at least one of my operating systems. For most of these years, Firefox (or its close relative Iceweasel) was also my main browser.

Soon after discovering Mozilla's browser, I tried Thunderbird, which I started using instead of webmail. For most of these years too, Thunderbird (or its close relative Icedove) was my MUA. So it has been nearly 15 years since Mozilla became one of the projects with the most influence on the software I use.

For years, I was evangelizing friends and family about Mozilla software. I was also a developer of projects upstream of and downstream to Mozilla. When Firefox was released, I was helping early adopters on IRC channels. For more than a decade, I also sporadically contributed to Mozilla's wikis. I wrote and helped triage issue reports. For some time, I even edited Wikipedia's page about the Mozilla-Debian licensing/branding issue. When I discovered Firebug thanks to a friend, I immediately installed it. Over time, I would report many of its issues.

Despite the licensing issue and XUL's reliance on GTK+ instead of Qt (my favorite desktop environment's toolkit), I never hesitated before contributing to Mozilla. In 2010, I was thrilled to realize the milestone we had reached when Microsoft's Internet Explorer dropped below 50% in browser market share. Years later, Mozilla finally agreed to relicense its logos, and the old Mozilla-Debian conflict ended.

For years, we wrote history. And then, in 2016, I realized a major issue using Gmail via Thunderbird had been affecting me for years. That issue had been slowing me down and putting the security of my customers at risk. Following that initial realization, I spent numerous person-days dealing with the damage (cleaning up my mailbox, directly and indirectly). But in the course of that long process, I also realized more issues in Thunderbird/Gmail, as well as in the Mozilla project itself. Since these are numerous, I will not list them here, but instead ask those who volunteer in Mozilla to read the overwhelming report I sent to Mozilla's governance and bugmasters mailing lists.

This mail appears on the archives of Mozilla's governance mailing list, but careful readers will notice it does not show in the archives of the bugmasters mailing list. Indeed, although I did send the mail there (using Thunderbird!), the mail apparently never reached the list, for reasons unknown to me.

That additional problem is one of those I reported in a follow-up to "Issues, meta-issues and transparency" which I sent to the governance list on 2018-01-13. Weeks later, I hadn't received any response to my follow-up, and realized that mail had apparently not reached the governance list, so I resent it on 2018-03-10. Unfortunately, I can only link to a local version of this mail, since the second sending visibly also didn't reach the list.

I have to conclude these events did not just reveal issues in Gmail and Thunderbird, but huge issues in the Mozilla project - not just in its issue tracking, but also in its mailing lists (and yes, Mozilla also knows about these). And so, more than a year after I reported what I personally witnessed, the persistence of this situation shows not only that Mozilla is broken, but also that users are far from its priority (if it's not simply unwilling to fix itself).

Going forward

All of this is not to say that Mozilla products are worthless, nor that no part of Mozilla can be salvaged. Many large free software projects struggle with separation of duties, decision-making and prioritization. No matter how one looks at these problems, ultimately they "just" indicate a governance issue.

So in fact, if there were no issues other than those I reported, Mozilla - the open source project which has more resources than any other - could fix these quickly. Unfortunately, beyond the governance issue(s) exposed, there seems to be an extra issue at Mozilla. What this debacle and the aftermath shows is not just that Mozilla needs to review its governance. It also suggests that Mozilla does not want to solve the issue(s). My report was not met with silence; in fact, several people replied. Some contributors genuinely tried to help a bit, and I must thank at least Svetlana Tkachenko for offering a significant and credible contribution. But no one has offered thanks for the report, or even acknowledged the issues. In fact, I believe most replies were - most ironically - claims that the report was off-topic on the only forum which it managed to reach (the governance mailing list).

In my opinion, this reaction is evidence that Mozilla is denying its governance issues. Mozilla is in its 15th year and would not be the first once thriving organization trying to hide or deny its difficulties now that it is struggling. So unfortunately, even if all of this "just" comes down to governance problems which could be solved with reasonable resources relative to Mozilla's size, "just" solving these issues from within an organization which denies them seems like a challenge beyond what the old and busy man I have grown into can reasonably tackle.

In any case, what it certainly means is that this will remain my last contribution to Mozilla on a volunteer basis, and I will no longer endorse any Mozilla product. My departure is in no way because Mozilla's mission is complete; despite all the progress accomplished in the last decade, the web and its standards have evolved a lot too and much remains to be done. I hope that other contributors to Mozilla or other people interested in Mozilla's objectives can either fix it or get involved in other projects which share some of these objectives.

As for Mozilla, if it wants to remain a project where its goals can be accomplished, it will have to review its priorities and put users, quality and transparency first. If the current situation is just the result of a lack of resources, Mozilla should focus on its own issues. Proving contributors that Mozilla products and processes could be trusted will be a lot cheaper than "rewarding" contributors, and much more effective in winning the loyalty of remaining and potential contributors. Mozilla, just make our work effective, so we will become proud of our accomplishments and willing to pay for Mozilla clothes ourselves. Stop offering us to pay for traveling to your HQ, start valuing our work, and those who want to meet will be willing to pay their tickets themselves. Stop putting the resources you have into producing an Internet health report; people expect Internet health reports from the Internet Society, not from Mozilla; what people expect from Mozilla is reliable software. If you want to help the Internet, focus on your core mission and heal your own systems before worrying about the health of the rest of the Internet.

It's unfortunate to have to make this decision right after Firefox regained its relevance, with Firefox Quantum and the "integration" of my favorite extension (Firebug), both in the previous 12 months. I haven't decided yet what this resignation will mean for myself. I am redacting this resignation in Mozilla Firefox. I already mostly migrated to Google Chrome, even though the product is not strictly superior to Firefox, but even though this is recent, I already have doubts about the Chrome project's management. As for my MUA, I really don't see any free software alternative at this point (I concluded that KMail was way too buggy a decade ago, and I am under the impression this has not improved since).

Farewell, Mozilla

Timex T128 alarm clock: Imperfect, but Smooth

admin Saturday April 21, 2018

Sometime in the second millennium, Realistic manufactured a Chronomatic radio alarm clock, which my parents bought for me. Over 2 decades of use later, my reliable alarm clock had only started dying. It had lost its radio capability, but it could still be set to beep loud enough to wake up an entire room, no matter its size. Unfortunately, if the volume control could possibly set the alarm's sound in the past, that feature had broken long ago, so when I realized the prospect of being awakened by such a heart-threatening sound had perhaps become scary enough to make me lose sleep, I decided my oldest electrical device had earned the right to its ultimate sleep, while I had earned the right to buy a new, smooth alarm clock.

After a careful on-line search, I decided the Chronomatic's successor would be a Timex T128BC3, sold by Walmart for 25 CAD, part of Timex's T128 series. While this remains a fairly cheap device, after a few weeks of usage, I can say this purchase will have been worth it if the clock is still working as well in 20 years. The alarm can be set to be gradual, very smooth. The display is OK. The battery-based Sure-Alarm feature is well-designed, although it's disappointing to be unable to see the time when the power is out (there is no button to display temporarily). The nap feature is nice.

The weaknesses I found are:

  1. The DST control is a good idea, but it would save a lot more time to add a button for each digit to set times quickly, rather than simply time-consuming plus/minus buttons.
  2. Enabling and disabling the alarm cause a confirmation sound which is not incredibly loud, but still easily loud enough to wake up anyone else sleeping in the same room.
  3. The "24 Hour Set & Forget Alarm with auto repeat and auto shutoff" feature Timex brags about is worthless in practice. Unless you wake up at the same time every day (including weekend days), you'll still have to manually disable and re-enable alarms. If you wake up before the alarm time, there's no way to prevent just the next alarm. You have to disable the alarm completely (then re-enable it at night). Which makes weakness #2 considerably worst.
  4. Warranty of merely 90 days
  5. The feedback of buttons (ALM 1, "-", "+", ALM 2... all of them) is really bad. Very often, buttons do not actuate even when they have traveled down and are depressed. I did not investigate this deeply, but this seems to be related to button width. To ensure buttons actuate, you need to press in the center.
  6. The alarm seems to be buggy. Even though my clock is set to start smooth, twice its first 2 beeps were loud. Then it went back to quiet beeps and started its gradual increase.

Update 1: A few months after replacing my Chronomatic, I was shocked to notice that a woman my age was using the same alarm clock as I did. Hers had only started dying too. Wow, these Chronomatic clocks sure stand the test of time!

2018-10-13 Update: For the first time, I was awakened by a Sure Alarm during a power outage this week. I did not manage to stop the alarm without removing the battery, which is non-trivial (supposedly requires a tool, although I managed without). Either the button feedback issue combined with the lack of light and my lack of sleep that night explains my inability, or the clock failed to disable the alarm. I will try to update this next time Sure Alarm triggers.

Questrade's "Margin Account": A marginally considered nomenclature for a regular account

admin Saturday April 14, 2018

So you've joined Questrade and now want to open a main investment account to invest most of your money. It's not for an RRSP, or any other registered account. At the Account Selection stage, Questrade asks you which account type you want, and you can only see specialized account types. Questrade offers you a margin account, a managed account, but you fail to see a cash account. Could Questrade be stupid enough to not offer the simplest and most common account type?

Turns out the answer is negative. Questrade simply wanted to facilitate (and probably encourage) margin trading, so it made margin trading part of its regular accounts. So they named that account type "Margin account", only failing to specify, either in the name or in the account type's description, that using margin is completely optional, so that the majority which is just looking for a regular unregistered account should also use that type.

Unfortunately, that's not the only problem. Questrade also automatically allows you margin if you open a margin account. For example, although I just put the minimum 1000 CAD in my "margin account", I still have a 3330 CAD buying power. And that is not only the default setting; you cannot disable borrowing. So when you bid in your margin account, you'll have to consider "Cash" minus your open orders, not "Buying power".

Twitter's Bootstrap shall now bootstrap itself

admin Sunday April 1, 2018

Last Summer, as I was working on the Tiki project, which uses Twitter's Bootstrap framework a lot, I realized that Bootstrap is quite hard to discover organically and I decided to bite the bullet and go through Bootstrap's documentation.

The OK

Doing so, I hit quite a few issues, some of which I reported:

From these 4 initial reports, 2 were reportedly solved and closed. #23754 was a wide report which could be considered as reporting several problems. Several of these were fixed, and when the ticket was closed, I opened more specific/clear reports, #24705 Grid system documentation refers to undefined columns and #24706 Inappropriate paragraph in grid system documentation's "How it works" section ("Sounds good?"). As I could not verify the fix for #23769, I filed #24288 Indicate getbootstrap.com website freshness (documentation version). The only worrying case was #23772, which was closed without solution nor explanation, and which remains closed as of ½ year later. Since I couldn't reopen it myself, and since no one saw my request to reopen it, I filed a new ticket, #25225 Blockquotes example and others use cryptic spacing classes before their introduction.

The Bad

At that point, I had 4 new issue reports : #24705, #24706, #24288 and #25225. All of these were closed. Worryingly though, only #24706 was actually solved. The 3 other issues still persist, to my knowledge. Even though I asked to reopen 2 of them months ago, the 3 remain closed.

That's when I figured it was time to ask the mailing list whether Bootstrap contributors can expect serious treatment. And that's when things got ugly.

The Ugly

The next problem I hit was I count not find any official mailing list about Bootstrap development. When I realized one could not find any reference to a discussion forum in Bootstrap's documentation, I had to give up, and instead reported #25824 Documentation does not refer to discussion forums.

I suggest anyone who would consider contributing to or using Bootstrap to read that last report. The report was closed in less than a day, without any result being mentioned­. Not only did Bootstrap not get any official forum, but the documentation still fails to acknowledge that there is none.

Given the level of interest the Bootstrap project seems to have in getting forward, I am afraid the most helpful thing I can do is to stop wasting my time on it. This resignation shall be my last contribution to Bootstrap.

HP LaserJet Pro M227fdn (Debian GNU/Linux, Microsoft Windows)

admin Saturday March 31, 2018

After giving up on my Epson WF-3620 all-in-one, I had gone through a sixth consecutive defective inkjet printer. 6 printers of 3 different models, from 2 manufacturers, in 2½ years. Fearing loss of even more of my time and mental sanity by getting a seventh printer which would also fail near-instantly, I reviewed our needs and choices. Although I had always avoided laser because we print little (most of our usage is non-business), I decided to drop inkjet printers altogether and decided to replace with a laser all-in-one. After realizing that color laser printers were too massive, I decided to go with a monochrome HP all-in-one. After I gave up trying to add a M227fdw to my cart, I called HP and was told by an agent that was because HP had no such device in stock...

I then decided to go with a slightly less flexible but immediately available and considerably cheaper M227fdn (no touchscreen, no front USB port, no wireless). I thought I was lucky, since Debian stretch offers the very first hplip version which supports the M227 series. Setting up the printer was indeed easy after I remembered I needed to install the hplip package. I was impressed by printing speed.

Things got considerably worst with the scanner, which I connected via USB. I tried launching Skanlite, which refused to open with an error message. When I tried launching from the command line, it whined:

skanlite wrote:
sane_open(" "hpaio:/usb/HP_LaserJet_MFP_M227-M231?serial=VNB3J02222" ", &handle) failed! status = Error during device I/O

With such a helpful error message, I decided to try Simple Scan, which opened, but failed when asked to scan, complaining it was "unable to connect to scanner". After searching the web significantly, I was lucky enough to find the solution, given by another poor LaserJet user. The solution is to install hplip-gui, launch HP Toolbox, and install a proprietary plugin for that scanner. While that solution does fix the problem, if the M227fdn does require a proprietary plugin to scan, it's easy to consider HP Linux Imaging and Printing's table of supported printer models, which lists the M227fdn as having "Full" support and as having "Scan to PC", as fraudulently misleading.

The LaserJet has a relatively powerful web interface, but reports cannot be viewed, only printed. Inexcusably, the 10 reports offered, including the usage report, can only be directly printed on the LaserJet, even with firmware 20180510. If you want to see how many pages you actually printed before buying a new cartridge, you can thankfully use Supplies Status in the Home tab. That will also give you the total number of pages printed (in fact, the number printed with your drum, which should support about 18400).

I thought setting up the scanner on Windows would be easy, but I was overestimating HP once again. After installing HP's software, I was somewhat able to scan using Microsoft's built-in utility, but I could not see a user-friendly application from HP. I tried the menu entry "HP TWAIN Administration". It is not clear whether that application is purely administrative or if it's supposed to allow scanning directly. The home screen says to click on "Numériser" ("Scan" in French), but there is no such button. The name suggests it's just an administrative tool, but I could not find where the presets controlled there show up.

So I went to the manual and found that after installing the software on Windows, there should be an "HP Scan" shortcut in the Start menu. Unfortunately, even though I installed 2 different versions of the software on 2 Windows 10 installs, none had such a shortcut.

Thankfully, I then noticed that in one session, HP did install an application called "HP Smart", which does allow easy scanning (among other things), and which works. Yet, it wasn't installed on the PC on which I used the latest installer, and it apparently only installs for the current session (you have to repeat for each user!).
A big applause for HP for such a failure... how Smart is that? It's HP Smart.

But, after all this time wasted, it seems to do the job. So here's hoping I didn't get a seventh consecutive defective HP all-in-one, and that this will be my last post about printers for a very long time...

Update 2019-01-02: After trying to copy a document, I must say this printer's interface is crap. I knew it would be less intuitive than a touchscreen, but having to use 4 different buttons to launch a copy, only 1 of which is labeled, caused 2 adults together to fail to copy a document until we gave up and consulted the user guide.

Update 2019-03-02: We printed 1979 pages on the original cartridge, but the output was starting to get gray rather than black. The web interface didn't show any fill percentage anymore (although a bar suggested 1% left). It had been several weeks since the printer started whining about low toner though, so I just adjusted the low toner threshold from the factory default 7% to 4% in the web interface's System tab (Supply Settings), since it's apparently quite annoying on Microsoft Windows. Changing the cartridge is not as easy as it was with previous printers, but it's straightforward enough once you know you just need to follow the pictograms, letting the printer figure out automatically it needs to initialize the cartridge.