Shally's Alley Live

Friday, September 9th. It is nearly 1 PM and you're on Shelly's Alley. So, um, let's welcome fall, shall we? I mean, this is the beginning of fall, isn't it, officially, Glenn?

Or, uh, spring in Australia, I guess.

Sure, spring in Australia, that's right. Um, but is fall— like, September actually in fall, or we still have a little bit of summer left?

If you live in the South, it's still summer.

Yeah, has it started snowing up there yet?

No, here in Boston it's just pleasant. We're good.

Nice. Well, welcome to Shelly's Alley. Welcome to fall. Today we are going to riff off of what might appear on the surface to be a little basic, but I'm reminded how important it is to really know the basics and how often we forget that the, the real fundamental stuff is very necessary and can help you. You just don't remember using it. So we're going to talk a little bit about that basic stuff. And I'm actually going to go ahead and open the chat overlay to see if we have any questions coming in from people out there. What I want to start with though, is just the fundamentals of Boolean. Just straight up, what is Boolean? Boolean is quite simply just 3 commands in Boolean logic, which is a type of math that doesn't necessarily require numbers. If you were to think of it in terms of English, if you are you know, a language connoisseur, Boolean would represent a logical disjunction or a logical conjunction, whether you're separating terms or joining terms. And all there is to Boolean is this: it's AND, OR, NOT. Bring that up on the screen there. See if that showed up. There it is. Those are the three commands: AND, OR, NOT. And these work in most databases and in most search engines. Databases like your applicant tracking system or your CRM usually support Boolean. In some, like I've seen perhaps, uh, Beemery, if I'm not mistaken, you have to choose from the dropdown when you have keywords in the search box and it asks you, do you want these to be AND or OR terms? And you can select that way. It may not allow you to use the command line. I'm not sure if that's changed or not, but that happens occasionally. So read the instructions or the help manual. But the bottom line is these three things are the basics or the concept behind Boolean. Think of AND as a way of joining, or a basically a logical conjunction of joining words that you want to appear in your search results. If you are in an ATS and you want somebody to have both a CPA and tax, you would simply put CPA AND tax. In most systems, the booleans are required to be fully capitalized, every letter, so capital A, capital N, capital D. The same thing, capital O and R, and capital N, O, T. In some systems, you are required to encompass the disjunctions in parentheses if you are going to have alternative terms that are synonymous. So let's say, for example, you're looking for somebody that's an expert in either tax or transfer pricing. That would be something that you would include in parentheses. Inside of the parentheses, you would have the 'or.' So in that case, it would be more like this. It would be tax or transfer pricing. The Boolean here is the OR. The parentheses on the outside are not Boolean commands. They're just an envelope that is not required in some places. For example, on Google, if you left out the parentheses, Google would still understand that you're looking for the word tax or the word transfer pricing. This is very different than if you actually made that an AND because then you would be looking for both of these. In this case here, the results would require that the word tax and the phrase transfer pricing be present somewhere on the page. That's the difference between AND and OR. Parentheses is simply a package or an envelope. Now, the other thing that you see here are these quotation marks. Be careful, this is something that I learned from Glen probably, oh, I don't know, 25 years ago. Be careful when you use quotation marks that you don't store these in your Word documents, like a Word or even PowerPoint, because these applications are used to correcting quotation marks to be typographically accurate, which is when you see them in print, they curl into the phrase on either side, and those are different characters to a search engine than these vertical straight brackets. So if you're going to use these brackets and store them, make sure that you do that in an application like Notepad or TextEdit or Evernote or somewhere where they're not automatically corrected, which is incorrect, into the curly brackets. Glenn, you gave me a tip on how to turn that off. Is that still relevant today?

It still works. So if you use Microsoft Office and you don't want to have curly quotes change or take— well, if you don't want to have any quote change to anything other than the straight, which is what you want, you can go under File menu, Options, and then under there, there is something called proofing. So the proofing tab within Word options and also other Office programs will let you turn that off. And it also lets you turn off one other related thing. I guess it's not technically a Boolean, but the dash, which we know is the not command on a lot of search engines, including Google. If you type a minus sign, sometimes that will get expanded into a long dash. So the hyphen becomes a long dash, and that will not be recognized by the search engine either. So if you want to turn that off, again, Word Options, Proofing, you can turn off both the curly quotes and the long dashes.

So it won't autocorrect the quotes, and it won't add the long dash. Now, to exemplify what Glen just described with the long dash, let me show you what I mean here. In the case of this example, let's say that we're looking for tax, but we're not looking for a CPA. So we're going to do tax not CPA. That, that minus sign is the same as the not symbol, so it would essentially be just the same as if you did tax-not-CPA, like that. Both the top and the bottom are identical examples and would work the same in most search engines. However, there's another dash, there's the N- and the M-. I know it's confusing and it probably doesn't even sound that different, um, to most of you, But the n-dash is essentially the hyphen that is also kind of like a minus sign. But the m-dash is the longer of the two that is sometimes used in— you might see it in dictionaries or something. And let me show you an example of what that looks like so you can see how easily confusing this gets. So there is tax and then CPA with the other dash, the no-no dash. So the one on top is the minus CPA, and the one on bottom doesn't mean anything. It won't work. It will actually just search for the word CPA and ignore that longer drawn-out, um, I don't even— it's an em dash. It's— I don't even know what the proper is. Glenn, is there an actual You've got a good education. Is there an actual name for that thing?

I always call them hyphen and long dash.

So I don't—

okay, there may be an official term, I don't know.

There must be. I think, I think the em dash is the long one, but, um, in any case, the minus is the one that we want and not the other one. So if you put all of this together What you would end up with is, let's say combining all of it, you would, you would do tax and CPA or accountant and not, let's say, state, so you don't get state tax. So that includes the 3 examples that we just talked about. I'm not being very fair, let me actually edit that other one to be not as well.

So here we go. What's interesting about these is that Word and PowerPoint will automatically autocorrect those the wrong way, but not Excel. Microsoft Excel leaves it the way you type it, which is great, and that's probably a reason you should create your search strings in Excel.

In Excel, yeah, yeah. My sourcing template is in Excel and it doesn't autocorrect because it's, you know, mathematics, right? And with mathematics you don't want to be messing around with the, with the characters and changing them without telling the author. So these, this Boolean that you see here can also be the same as— I'm going to write the other way. This would be the same as saying tax CPA and minus state. These two are identical. They have the same result. That's supposed to be a colon, by the way. Tax and CPA or accountant not state is the same as colon tax, and you can see the and is left out because it's implicit in most searches. You don't really need to state it, and then the minus is the other one. Now to get really fancy, in some search engines like on Bing and on Google, There is another character that can replace the OR, which is a vertical dash, sometimes called the pipe symbol. If you're looking for it on your keyboard, it's below the backspace key or above the enter key. Another way to think of it is Shift+Backslash. The Shift+Backslash will draw that vertical pipe, which is the same as the OR. And of course, the mathematical symbol for AND would be the plus sign, as you can imagine. So if we were to spell everything out and put it all together, it would look like this: Tax AND CPA OR accountant NOT state is the same as Tax PLUS CPA PIPE accountant MINUS state.

And what's also interesting about that is the pipe symbol is actually used in programming for representing or. So if you create a regular expression, for those of you who are a little technical, you will know that if you want to say or in a regex statement, you need the pipe.

So it actually does come from math, and that's, that's why it's there. I mean, George Boole, the, the the person that's credited for coming up with the concept of these non-numerical mathematical symbols, was looking for a way to express mathematically what was being expressed in English so that it wouldn't have to be written out. So one of the things that I wanted to clarify though is, because this is something that is asked of me all the time, well, Shelly, why do you want to use the plus and the pipe and the minus when you've got the and and the or? Aren't you just giving us more stuff and confusing us on purpose? So that you can make more money. Um, no, that's not the case. There's actually a very practical reason for that, and quite frankly, it's one of economy. In economy, I mean in a couple of different senses. First of all, it takes a lot less keystrokes to type a plus, a pipe, and a hyphen than the words themselves. There's also a lot less probability of error. And why, when I say error, I mean even the same kind of error that Glen and I commit all the time, which is to accidentally not capitalize the D or the OR, or the R of the OR, or to instead of saying NOT, we just put in no because we're typing too fast or whatever. All of these are very simple errors that when you start building a search can really trip you up, and you remove those almost completely when you use plus, minus, and pipe. And finally, the last economic advantage, if you will, to this is that these characters do not count towards your character limit in search engines that have character or word limits. On Google, there's a 32-word search limit. If you were to type in a 33rd word, it would say that we found all the results up until excluding this because It's outside of the limits. However, if you include AND, OR, and NOT spelled out, they are counted as words. But if you put plus, minus, and pipe, they are not counted at all. And so you saved yourself quite a few. You can actually build searches that have more variables in them without running into that limit as early because of the characters actually counting. So it's really just a speed and efficiency and economics argument.

And I get a Boolean amen about that.

That's right. Or, or whatever the equivalent is. So you've got AND, OR, NOT. Beside that, some people consider the parentheses to be a Boolean even though it's just an envelope, and the quotation marks also just an envelope that surrounds what is referred to as a forced phrase or a phrase that must appear in that order. Because if you search for words that are in quotation marks, they only will match the words in that directionality. So software engineer in quotation marks will only find the word software to the left of the word engineer. If you don't use that, you can find the word software and the word engineer anywhere on the page. And that may mean that somebody's talking about software and someone else on that page is talking about a civil engineer, right? And it's not a match. So that's the forced phrase that brings it together. You can think of forced phrases as natural language, the words that we use when we put the words together. So for example, "I was at" or "I went to" in quotation marks would be a natural language search. So there's— oh, go ahead, Glenn.

I mean, I want to make sure you're done with this segment because there is another Boolean operator you haven't mentioned, and it does come up, but not very often. Go for it.

All yours.

It's called the exclusive or, or XOR. So the only places that I find as a sourcer that I come across the use of XOR is on patent databases. So if you go to some of the major country patent databases, you can do this Boolean operator in between called XOR or exclusive or. And what it does is it's basically saying you can find one of them, Or you can find the other thing, but not both. So where you normally use OR, it's good for more sourcers in general. Yeah, there's no X and no E in front of the X. It's just XOR. But the thing about sourcers is when you typically look for something or something else, you're also happy to see a page that has both terms. And that again will usually work because you're looking for a term or a synonym of that term. And then add both terms, you'd be good. But in the patent world, I can see the use case for having it where I want to see one of them. I might want to see the other one, but don't show me both because that gets me different results that I don't want to see both together.

Yeah. And again, read the help file, the help menu, because some search engines will allow you to do that and some databases will as well. There's one last— moving on to another topic, but there's one last item, I guess, that is very frequently confused with a Boolean operator, and that is the asterisk, which is sometimes called the wildcard. Now, the reason it's confused with Boolean is because people used to use this on some search engines, some older ones like AltaVista, as a way to find variations of a word by rooting or stemming a word. The most common example would be engineer. If you searched for the word engine and put an asterisk at the end, that would find in those search engines all variations of the word, including engineers, engineering, engineered in the past, and so on. However, that's no longer supported by search engines today. So if you try to use the asterisk on Google, It will not truncate or stem your word. It does work on some of the private searches or tools like SeekOut, for example, does support the stemming of a word. And Exalead, which is not kind of a beaten-off-the-path search engine, as well as Gigablast, both support truncation or stemming. And in that case, the asterisk takes the place of everything else to the right of it. As a wildcard. In Google though, the asterisk is also a wildcard, but for its own self, a word, meaning that it's not a wildcard for a piece of a word like engine asterisk. It's a wildcard for a word that may be located between two others, like software asterisk engineer will find all the variations of three-word phrases where the first word is software and the last word is engineer but the word in the middle could be a variable such as software development engineer, software test engineer, software quality engineer, and so on. It's still a wildcard, but not for the part of the word, but rather the entire word itself. However, it is not a Boolean. A few other things that are not Booleans: filetype and site command and inurl and intitle. Those are syntax commands that are very specific to some search engines. In Bing and in Google, they support a lot of the same ones like filetype and site, but Bing has others that Google doesn't use. So these are not booleans. These are essentially, uh, think of them as line item commands, like similar to what you would use if you were programming or something like that. Now they all have help files, most of them do anyway, that show you what those commands are. And some search engines like Gigablast have a very long list of these commands, others a very short one. The trend nowadays, Glenn, I don't know if you've seen the same thing, is to essentially remove these. So I've seen search engines whittling them down and eliminating the ones that are no longer supported because they're frequently not used or infrequently used. So we're sort of trending backwards instead of making it easier to customize your search with these commands. We're essentially removing the support for some of these commands. Are you finding that to be the case?

I would say that's generally true, except in the cases of things like SeekOut where they're using the Lucene platform behind the scenes, and that supports so many different options. And it's just a question of whether the host of that website decides to turn those options on or off. And if they do turn them on, then they have a lot of capability, but Again, not all those commands are going to be obvious to the user without some documentation.

Right. Again, read the help file. Yeah. Most of the Lucene-based ones will have some of those activated and there are way more even than the ones that are on SeekOut. So if you're curious, one of the best ways to hack these databases is to learn the Lucene commands because most of them are based on Lucene. If they're not based on Lucene, they might be based on Elasticsearch. The other one would be Algolia. Those are the top 3 search, at least in my experience, search algorithms. If you read the documentation, you'll see what their commands are and they'll usually be the same.

You might want to type those 3 platforms in the chat because I don't think anyone knows how to spell those.

It would be Lucene, Elasticsearch, and Algolia are the main search engine software programs. And these are the ones that power, if you will, many of the searches that you're familiar with. Um, Glen, what do you think about around?

I think Some of the use cases I've seen crop up recently by some sourcers have been really interesting. The good news is Google, which used to inconsistently support the around command, lately seems to be supporting it very well. So that's great for sourcers who know what the around command does, and then you can start using it in some fairly complex queries, including Google Custom Search engines, if you want.

And so the syntax would be around— although technically it's not a Boolean— as a fully capitalized word. So all of the words capitalized, and then there's a number that goes after it in parentheses. What does that number represent?

Yeah, so that's the number of words that could be in between the word on either side of that around command. So you might—

So let's say 3.

Yeah, so if you want to do an example that has 3, that's good because now you're saying that word should be close to another word, but not a phrase. I mean, it could be a phrase, because that would be around with zero. But around 3 means one word, like, let's think about, you know, configuration and server. Those two words mean something different if they're close to each other. So it's really good when you know that if you imagine someone's profile or a paragraph of text, And if a sentence had those two words pretty close to each other, that would imply a different meaning and tell you something specific about what's being talked about in that paragraph.

Whereas if the two words were on totally different parts of the page, it's almost like using that quotation mark but, um, and the wildcard, but in a more extended way. And also it's bilateral. So remember the example I gave of software blank Engineer could be software QA engineer or software test engineer or software development engineer. Well, the around, let's say you did software around engineer and you had a 3, would look for the word software and the word engineer to be within 3 words of each other, but it could actually be in either direction, software first or engineer first, and that is a little bit more expansive than the other. You could imitate that with the quotation marks and an asterisk by saying open quotation marks software asterisk asterisk asterisk engineer close quotations, but that would only be one direction, meaning the software would be the first word. Around changes that so that it's the two. One of my favorite applications, and the reason I brought this up, is that the around command is very useful in peer regression searches. If you're wondering what that means, a peer regression search is essentially finding people that are like the people that you already know. If you know that someone is, let's say in this case, Shelly and Glen are both sourcers. If you already know Shelly and you know Shelly's a sourcer, well, can you find people that are around virtually and, you know, figuratively Shelly? And for example, Shelly around Glen now will bring back results where on the same page the words Shelly and Glen appear together within 3 words or 4 words apart, most likely indicating that that's a page that includes other sourcers listed in either a conference or an event or pictured in a photograph, in captions, etc. So it's a very useful technique for finding, um, people that are connected or somewhat related. Not just for people though, also companies. If you have a company like, say, Shopify and Grubhub, you'll find a list of other similar companies. So it'd be Shopify around 2 or 4 or whatever, Grubhub. So that is a different command. It's not technically a Boolean. It's one that belongs to an old-school set of commands called fuzzy logic. There used to be a few others like it. WITHIN was one that was used for a while. Glen, are there any other fuzzy logic commands that are still around and practically used today?

I'm trying to think of a fuzzy logic command that I use. Boy, you're putting me on the spot. There might be. I'm not thinking of one. I'm sure you have one though.

So fuzzy logic, essentially to define it, is kind of a cutesy way of saying that there's a variable or several possibilities, as opposed to the very much linear logic that doesn't leave a whole lot of room for error that is used with AND and OR and NOT. Those are essentially absolute Booleans. If you want an apple and an orange, the result has to have both apple and orange. There's no interpretation, there's no variability to that. But a fuzzy command would allow for that variability. So it would be similar to what we sometimes call semantic search. In that case, if there was a fuzzy command that was similar to the AND, like for example, like, um, apple like orange would bring back pages that have apple and pages that have other round yellow-colored, orange-colored citrusy fruits, would not necessarily limit it to just orange. The around command is a fuzzy logic term because it's looking at a variation of various possibilities. 3 in the parentheses means 1 word apart, 2 words apart, or 3 words apart, but it means a maximum of 3. In other logic, fuzzy search commands that used to exist could, for example, replicate what we sometimes do with regular expressions or what Glen was talking about before with regex. For example, you could say any number from here to there. There is one other fuzzy search command that exists on Google and it is the numrange command, which is not a Boolean, but it is a fuzzy search command. You don't actually need to type it out though. You don't actually need to type numrange. All you need to do is separate numbers from each other using two dots, not the ellipsis, not three dots, but two dots. So if you were looking, for example, people who graduated anytime between 2017 and 2019, you would use 2017, the lower number, separated by the higher number with these two dots. And this is a fuzzy search command that would look for 2017 or '18 or '19. That's not a lot of choices, but you can imagine if that search was, for example, 2010 to 2020, that's 10 different years. And as Alan Flora puts it, it's really good for zip codes because zip codes are numbers that are very serial in nature and they happen to have a directionality to them. So as they go up in number, so as the, the number value is going higher, usually the value of the area is going lower. So you can search an area this way or that way by using NumRange with zip codes. Also other numbers that might be similar like phone extensions or street addresses and things like that. That's another type of fuzzy search command that's still in use.

Well, I got a couple of things to say about that one. First, I'm kicking myself for not mentioning numrange when you asked for other fuzzy commands. But of course I use it. The interesting thing about the zip code example though is you need to watch out if you're dealing with the Northeast of the United States because those zip codes—

Very true.

Best guess. And that will totally screw up Google if you put a NumRange starting with a 0.

Yes, now the way around that is not a very comfortable way, but Glen has a very good point, which is that the zip codes that start with 0 to us are still zip codes with the exact same number of digits than any other zip code has. But as soon as you put that into a calculator, which is Google, it eliminates the trailing— no, sorry, leading, that's what I meant, the leading zeros. There is no zip code such as, you know, 00001, but if there was, that would essentially be 1 to Google, although to us it's that number. So you have to put that in quotation marks, but then that means you also have to search for each one of those zip codes as a text, each one at a time, like, you know, 0892 or 0893 or 0894 and so on because of that. So this is why in New England we're not using zip code search that much unless there's logical expressions which would allow you to say I need a number that is 5 digits long and it starts from 000000 to 999— actually, I don't know if there's a 99999, but you see what I mean. So a logical expression would be able to tell you that you would be able to search for 5-digit long numbers even if they started with 0. However, Google does not have that, so this is something that would be more like instead of a SeekOut, for example, where fuzzy— not fuzzy, where logical regular expressions are possible or exist.

So another great use of NumRange is when you're looking for people who manage teams because they will talk about the size of their team. If you look for manage and then put a num range in there and then put something else after, I guess, I mean, you could put team. You could even probably work a wildcard in there somewhere, but that will get a wide range of natural language results like I managed a 30-person team of developers, or I'm a member of a 200-size team of something. So yeah, there you go. That's a nice kind of search.

In this search, we're looking for 3 different words in a logical disjunction. We don't need the parentheses because we use the pipe. Again, another reason to use the pipe instead of spelling it out is you don't really need the parentheses even for your own visual determination. It's looking for either managed, supervised, or led, and any range from 5 to 20, and also people, staff, or engineers. So it could be managed 18 people, supervised 12 people, or led 7 engineers, or any other combination of those. And it will find all of those. That is unfortunately something that only works on Google. Bing does not have, to my knowledge, a NumRange equivalent. Does it?

Well, Bing has a number of problems, but that certainly is one of them.

Yeah, I don't think there's actually a NumRange equivalent. Yeah. The only other one that I can think of that is still in use is with Twitter, there is the within command, and that applies to geographic search. So if you go to search.twitter.com, you can still search by geography by specifying a number of either kilometers or miles. So let's say that you want tweets that were posted within 10 miles of Atlanta. Or within 10 miles of zip code 30072, for example. Um, so that within colon command is actually a fuzzy search because it's anywhere from 0 to, uh, which, which makes it a fuzzy search. That's all the ones that I can think of. I don't know if there's others out there.

Some with EXALEAD, but sorry, I think that's worth typing in the chat too because that's a great example, uh, and another way that makes Twitter a more useful sourcing tool because now you're doing geographic ranges.

So it'd be within, let's say, 10 km for kilometers and mi for miles. Um, It would be—

and that is nice that Twitter does support either km or mi as your part of your string if you're more familiar with measuring distances one way versus the other.

It's like that. Yeah, you just switch it from mi to km, essentially. I'm still used to miles. So that's— yeah.

So I promised I would tell a short story that shows that your influence has gone decades now. So I was talking yesterday with one of my NVIDIA colleagues, and he's, he's now a full-cycle recruiter, but he used to be a sourcer earlier in his career. And he was telling me shortly after, um, I don't know when he told me, I think it was yesterday actually when we chatted, and he said, you know, I've still got the stuff that you guys shared way back when you were doing full-time training. And I said, what, what thing are you talking about? He said, it's this 12-step thing. I'm thinking, wait, this is not an alcoholism program. I know what you're talking about. This is Shelly's list of 12 different sources to tap when you're looking into finding, um, you know, the biggest universe of relevant talent. And it's in a specific order of most likely to yield something to least likely, so you won't necessarily have you go through all 12 steps to be useful. But if you do go through all 12, then you really exhaust the pool, but you might actually get what you need from steps 1, 2, or maybe just 1. So that slide, I think, has probably been duplicated thousands of times over the years. And this guy remembered it and said in this early sourcer job interview that he had to get, because he had that sheet. They asked him a question about his sourcing process. He was able to rattle off the 12 steps, and they were blown away, and he got the job.

All right, that's great. That's really good, man. That makes me feel good. I get the warm and fuzzies when I hear that. There was that cheat sheet also that we had.

You had the first cheat sheet, so Everyone should know that the first Google cheat sheet that you probably saw was the one Shelly created in, I think, '05.

Um, it was earlier than that. Oh really? Yeah, it was 2001, I think it was.

Okay.

Yeah, it was like Google was just a baby. It was really early on. There was a, there was a, uh, one of my mentors in the industry who is— he's passed on, but he was probably one of the foremost thought leaders in what's called reverse engineering, which is not just necessarily hacking. It's also— it's hacking, but it's also solving puzzles and cracking things from, you know, starting from the end and working backwards. And, um, you know, I was, I was a, an apprentice of his, and there was a challenge where he was, he was asking the people in this, in this group about, um, ways to use Google to find, um, pictures, photos of a particular— it was basically a, like, not a contest, but a, not a test, more like a fun puzzle, right? So he posted a He posted this, this painting. It was actually a photograph of a painting that was a legit masterpiece from a museum from his home city. I can't tell you any of the, uh, other stuff because it's private, but he posted this picture and he said, I bet you can't find out what this is, because it was, it was a masterpiece that had been recently discovered and put in a museum that hadn't been digitized. And he wanted to test us to see if we could figure out what it was. So a lot of us got pretty creative with how we were using Google and other searches, AltaVista and a few others as well, to try to figure out what the image was by describing the image with language because you didn't have— there was no image search back then and it hadn't been digitized anyways. So, you know, oil, pastel, ocean, mountain, cloud, you know, that kind of stuff, and we were working with that. And that is when I discovered the asterisk. And I realized, I came to realization because I was trying to use stemming like you could do with AltaVista. I realized that stemming didn't work. It wasn't producing any results. So I was getting something, but it wasn't results as I expected. And then I started playing with it and I realized that it actually functions as a bridge between two words. Ironically, I just happened to use the asterisk, but there are 7 other characters that are equally functional that would do the same thing. So you could put these not or stop characters in the break between the words and they would function the same as the asterisk. It wasn't just the asterisk. But anyway, the point is that's how I stumbled on the asterisk, and then I went back and, you know, defeated in my quest and posted to the group, I, you know, I give up. I'm unable to find what this picture is. And it was an impossible problem to solve really from the very beginning. I mean, he was testing us because it was really impossible to do. But I said, along the way I discovered this, and it was the very first documented, uh, I don't know, bug, uh, from Google, because Google had never published anything related to their commands or anything like that. It was the very first time somebody had found a command that was functional on Google that was a result of the base code that they used. Later on, I went to find a dozen more, and that's how the cheat sheet got created. And then when I was working at Google in 2003, that's why I said it was before 2005, and I was there, I, I was presenting to, um, a group of recruiters there, uh, and I had my Google cheat sheet up. And Larry Page walks by, takes a look at that Google cheat sheet. The next day it's published on Google when you click on the help icon where it's help, and then you can see my cheat sheet. No credit to me, but hey, I know who started that. He just thought it was a good idea because nobody had ever posted a help file on Google to say, here's the commands that are available.

Well, that is a cool story. And I guess to toot our own horn slightly more, people may know that I used to work with Shelly full-time as a trainer at another company, and that was between 2008 and 2010, just after we had worked together at Microsoft. And one of the things I worked on there was, because I really liked the Google cheat sheet, I said, can we create more of these?

More cheat sheets? Yeah, that's right.

And so Shelly said, well, I've, I've actually been thinking about doing one for LinkedIn, and maybe we should do one for the other major search engine, which at the time was not called Bing, was called MSN, and then got renamed Live. And we created cheat sheets for MSN Live, Bing, and LinkedIn, and later one for Twitter as well, which was pretty cool. And those were all very popular cheat sheets in their own right.

Yeah, so lots of people got benefit from that. I'm very, very happy. Well, thanks for joining me, Glenn. This was fun. Good retrospective, and also I think a really good basic tutorial for folks that maybe just have forgotten about some things like parentheses and AND and OR and quotation marks. You know, use them. They're, they're very, very useful in helping you hone in on your searches.

Indeed. I think it was a great hour to spend. Thank you for inviting me.

Thanks. All right, everybody, see you next Friday for Shelly's Alley. Have a good weekend. Take care, everybody.