Facebook may soon introduce you to M, a virtual assistant that is very different from Siri and Cortana and Google Now. For starters, you won't be able to talk with it and it will be able to do things that Siri and the rest cannot do at all.

You'll make requests of M by way of Messenger - that's right, you'll send it a text. If it needs clarification, it will text you back. It may handle your need the same way all the others do: use AI to look it up somewhere. But it also may refer the text to a human assistant, which means that M will be able to do things like argue with your cable company that your rate is too high!

I have no idea what Facebook intends to charge for that level of service, but it is interesting that they chose text instead of voice input. I think it was a smart move as voice requires configuration or perhaps downloading yet another app, which nobody wants to do. Text is also better for more complex interactions, especially if that invisible human has to become involved.

Should Apple and the rest be worried? Well, maybe. If I were Apple or Google and especially if I were Microsoft, I'd make my software M ready so that users could say "Ask M to.." or even just automatically fetch answers from M for comparison. That would help keep keep users in the fiold and could provide valuable competitive information. Of course nothing stops M from doing its own competitive research with questions it gets, so this could all get bogged down in a flurry of recursive questioning!

M is in Beta now in San Francisco - no clue when you and I will see it, but it should be interesting.

Facebook Launches M, Its Bold Answer to Siri and Cortana

Facebook M’s bet and the future of Web interfaces

Cortana vs. Google Now vs. Siri: Which Voice Assistant Wins?

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

For those just joining us, Ezoic is a "web site improvement platform". They take your site content, repackage it and spit it out to the world. The engine selects templates and tests results: what is the bounce rate, how much engagement, what's the ad CTR and so on. The algorithm then gives more face time to the templates that perform well, but templates don't get to live on their laurels - testing never stops.

There is nothing that Ezoic does you could not do yourself, but it would truly be a lot of work and would command your full attention constantly. My site, the site you are at right now, has been template driven for years. I experimented with layouts and ad placements and at one time made a decent amount of money. That all changed in 2011 when Google implemented new search algorithms which were intended to weed out junk sites. My content was not junk, but I got caught by collateral damage and my search traffic plunged. You can see that plainly in this graph from Semrush.com.

There was a second plunge you can see in August of 2013. That's the one that really crushed me. It was from template changes I made and nothing I did seemed to help. I could not get the traffic back.

Toward the end of April in this year (2015), I stumbled across Ezoic. As I explained at the time, I didn't have much confidence that they could do anything useful, but I really had nothing to lose: I was in the cellar and really could go nowhere but up.

And up it has gone. The Semrush chart shows that plainly and ad income has followed suit. Things are nowhere near what they were before 2010, but they are much better than they have been recently.

That's all good. What I DON'T like about Ezoic is the the templates that have accomplished this. I think they are often garish and that ad insertions are far too aggressive. You can control a lot of that with settings at their control panels, but even so automated layouts can never match what a human would do. It can get really ugly.

One particularly galling thing is that their templates see "<a href" as potential insertion points. That's so wrong: that's not a block level tag and inserting an ad ahead of it often breaks up a paragraph. I'm hoping they fix that soon.

But it works. By all metrics, this site is improving: more traffic, more money, more user engagement. I can't argue with that. I don't like the layouts, I cringe when I see how cluttered with ads some pages are, but IT WORKS.

That's what matters and that's why I'll be staying with Ezoic.

I'm going to let Ezoic mess this site up

Ezoic Day 2

Words can kill

Ezoic

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

I haven't yet switched back to Firefox, but it is getting more tempting every day. Chrome tabs are causing me more trouble than I'm willing to suffer.

Way back when Chrome first came out, one of its advantages was that each tab is a separate process. That means that one bad tab supposedly can't crash your entire browser. As it turns out, that's not entirely true: a screwed up tab can suck so much CPU that your only recourse is to force-quit Chrome entirely. This is annoying and seems to be totally unnecessary: the process can easily see how much CPU it is consuming and should throttle or even suspend itself if it is being greedy.

Suspending inactive tabs is available through an extension called The Great Suspender. It offers:

Unload, park, suspend tabs to reduce memory footprint of chrome. Tabs can auto-suspend after a configurable period of time or be suspended manually.
Tabs can be whitelisted to avoid automatic suspension.
Suspended tabs are retained after closing and reopening browser, preventing many tabs from all reloading after a restart.

Sounds great, but I'm not going to add any extension that can read and change data on every website I visit!

This sort of thing should be built in to Chrome. Each tab should also tell you how much ram and CPU it is using so you can spot troublesome pages more easily.

Update: Google explains how it will make Chrome suck less battery

Google's Masterplan To Make Chrome Suck Less

The Great Suspender

The great suspender (review)

Why I anm breaking up with Google Chrome.

What have you done for me lately, Firefox?

What Is Google Chrome Helper, and Why Is It Hogging My CPU Cycles?

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

The other day I had an email from Ziff Davis. Its subject was "Start a Blog with Toolbox for IT!". Ordinarily I would have just trashed that without looking at it, but I accidentally opened it and found it amusing enough to pursue. It started off with the typical schmoozing: "I came across your blog while looking for some insight on technology know-how, and was very impressed with the content you are providing.". Yeah, sure. Please stroke me some more while I sip my coffee.

It went on to say "I am always looking for talented bloggers to join the Toolbox for IT blogging community.". Talented? Well, gosh that sure fits me, doesn't it? I have to read more!

What came next just about knocked me off my chair. It said that if I joined their blogging community, I'd have the opportunity to "earn additional income based on your posts, with $50 per 300-word post (earning up to $500/month for ten posts)". Fifty bucks for 300 words? That's pretty high pay - I figure a 500 word tech post usually takes me about an hour, soup to nuts, so that's about $100 an hour and sometimes more. Not bad for tippity tappety on the keyboard, right? Especially not bad as the going rate for such articles on the rest of the internet is often much, much less. I've seen people offering $2 for 300 word articles - that's extreme, but we are competing with very low income talent, so rates are very low.

Of course they did say "opportunity". So what's the gotcha? I wrote back, expressing my amazement and requesting details. The response was quick and quite definite: "All the conditions are to blog about your field of expertise and meet the 300 word count minimum per post after signing a contract and returning at least a W9 to us."

A contract? Ahh, there's the rub. I asked for details on said contract and was refused: "I am sorry, but I do not have access to a contract. To see the contract, you would need to apply. Once your application is approved, legal would be touch."

Ahh. So I clicked through to apply. I found that in addition to providing the usual personal details, the application required me to submit a blog post. Huh? They expect me to give them a post and THEN tell me the financial details? I wrote back again, pointing out the incongruity of that. Within an hour they sent me a contract which they said was "part of the contract legal would send you".

Part of? So secretive, but I accepted it for what it was and took a look. It started off by explaining that if I gave them content, I was agreeing to give them "an exclusive, unlimited, perpetual U.S. and world-wide license to use, modify, and distribute" what I had submitted. My only rights to the content would be to use it in an "in the form of an e-book or hard copy"; other than that I could not reproduce it anywhere. Harsh, but there is that $50.00, right?

Not so fast. That was covered a bit later. It said "Author shall receive $50 per Qualifying Blog up to a maximum of ten (10) Qualifying Blogs per month." Of course it specified that the Blog must be in English, must be at least 300 words and so on. And then this: "ZD shall determine whether a blog is a Qualifying Blog or not in its sole discretion".

So, if I'm reading that contract right, anything I submit becomes theirs to use but they decide whether it is worth paying for. Curious, I took a look at the site and read a few articles others had submitted. They weren't awful, but it's hard to imagine anyone paying $50 for the ones I read. They weren't anything special, not really very technical and nothing you can't find all over the internet. In other words, the site is a content farm. I doubt they pay for much.

So I wrote back once more, asking if my interpretation was correct. If they decided that my article was not worth paying for, would I be able to withdraw it and use it myself?

So far, no answer to that question.. it's only been two weeks so I'm hoping they'll let me know soon :)

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

I added SolidOpinion Comments to this website recently. I did that because I'm wasn't happy with my homebrewed comment system. I wanted better authentication and the ability to nest replies. I could have added those features to my code, but I'm lazy and have too much else to do so I never got around to it.

So I took the lazy route. There are dozens and dozens of web based comment systems to choose from; all are easy to implement and their features are similar. My concern with any and all of them is that they may go out of business and take the comments with them. There's no good answer for that - it doesn't matter who I choose as any could go under.

Of course you might suggest finding open source code that I could work from. That's worse than adding features to my own code, because I have to understand their code before I can integrate it and if I had time and interest in that, I'd have the time and interest to modify my own! So that brings me back to the lazy route.

No good answers.

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

I am not a list maker. I have never liked lists. I had and have good reasons for that. Paper lists are clumsy when items are added over time as they can't be sorted without rewriting. Computer lists can be resorted, but in the past that wasn't much help: at some point that list always had to be printed and then it became clumsy again. There was also that I'd think of things to add to the list when I wasn't at my computer.

Then there was the sharing aspect. Our grocery list is a great example: we both add things to it. Our computers could do that, but only if both were on. I stayed with keeping things in my head and my wife put Post-it stickies everywhere.

Wunderlist changed that. It's on all our devices: computers, phones, tablets. We can share lists or keep them private. We can sort items or just drag them into the order we want.

It's not perfect. We don't use the "completed items" feature at all. Instead, we treat most of our lists as masters and we click the "Priority" star for what we need to buy this week and unclick it when done. Sorting by Priority makes that easy.

We still use our Amazon Echo lists because that's often more convenient. What would really be great would be if I could say "Alexa, add bananas to our Wunderlist Hannaford list" - or even better just "add bananas to Hannaford" and have it know to find that list at Wunderlist. I've suggested that to both Amazon and Wunderlist; maybe it will happen someday. Echo has recently added support for IFTT; if Wunderlist did the same we'd have at least part of this now.

Things I wish Wunderlist did differently:

• Option to show all priority items from all lists
  
  
• Option to show completed items as strikethrough rather than hidden
  
  
• Option to reset all completed items (weekly groceries, for example)
  
  
• Option to add Quantity column
  
  
• Option to add Sort column (grocery aisles, for example)
  
  

Wunderlist

Amazon Echo

IFTTT

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

Google seems to be obsessed with speed. Of course that makes sense: between search, Youtube, Google Plus and everything else their servers spit out, those bits and bytes represent an incredible amount of traffic. Any improvement they can make improves user satisfaction and cuts their own servers loads. Speeding things up translates into less equipment needed and that is money in the bank.

Google also nags website owners like me. Their Webmaster Tools site measures our pages speed and ranks it.

That's probably enlightened self interest too. Many of us run their Adsense ads; slow pages annoy users and that affects revenues. They also may eventually use speed as a ranking indicator. They say they don't now, but speed is plainly on their minds.

Enter QUIC (Quick UDP Internet Connections). QUIC adds reliability to UDP without the overhead that TCP imposes. It doesn't require changes to routers as they just see UDP traffic.

QUIC is already in use by Google since 2013. If you are using Chrome, you can see that with chrome://net-internals/#quic and by installing a Chrome extension that will show an indicator when a site is using QUIC. That's where it sits right now, but it may well spread to sites like this in time.

What will it take to implement this on your site? Maybe very little effort at all. For example, right now I use Cloudflare here. They provide SPDY (another Google "speed things up" protocol which will soom become HTTP/2) for me automatically and I assume that as QUIC becomes more popular they may do the same for that. That would mean I'd need to do nothing at all other than turn it on at Cloudflare. See Staying up to date with the latest protocols.

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.

I read at Firefox Fightback that Firefox wants to regain market share and a few highly placed folks have left. Market share? In this household, they have none at all. There was a time when I ran and recommended Firefox everywhere, but bloated and crash prone code caused me to move to Chrome in 2009. Since then, I've been happy wih Chrome until the very latest version, but now that has been freezing up on me at times so I thought I might take a fresh look at Firefox.

If the visitors to this site mean anything, I'd say Firefox is still a strong contender:

But Wikipedia says it is currently only "the third most popular web browser", and StatCounter says it's a poor third place at that (StatCounter Global Stats - Browser Market Share).

Firefox HAS been improving. They've been doing a lot since I dumped it in disgust six years ago. Some quick reactions:

• Fast again: Poor performance was one of my major reasons for defecting, but that seems to be fixed. I'm not interested in benchmarks; I don't know or care if it is faster or slower than Chrome or Safari - I just don't want a fat, wheezing Fox slowing me down. It isn't doing that to me now, so I'm happy.
  
  
• Crashing: That hasn't happened yet, but they do have a Firefox crashes - Troubleshoot, prevent and get help fixing crashes page.
  
  
• Customization:Firefox lets you easily (drag and drop) customize your menu bars. That's nice and it's easy to do.
  
  
• Yahoo for search? Are you kidding me? I understand that they don't like Google because of Chrome, but setting the default search to anything but Google is idiotic.
  
  
• IOS version coming soon? It's 2015 and they are just getting around to an IOS version? Wow.
  
  

Am I a believer again? I don't know. I'll try it out for a while and see how I feel.

-- This feed and its contents are the property of A.P. Lawrence, and use is subject to our terms. It may be used for personal consumption, but may not be distributed on a website.