For people waxing about what "Syntax Control" is, and coming to some over the top descriptions of clever AI, machine learning and what have you.
You're overthinking it.
It's the most naive implementation you can think of. It's exactly this: you select "verbs" and the verbs are highlighted. You select "nouns" and the nouns are. Same for adjectives, adverbs, prepositions (as, from, in, etc) and conjuctions (and, or, etc).
Nothing earth shattering, nothing you can't do with simply tagging words, nothing marchine-learning or AI about it.
Beside the naiveness of the core functionality, notice also how naive the tagging is. As noted below by bosie, "programming" is marked as a verb in the phrase "in programming terms", syntactic context be damned. Same for "run-on" where run is highlighted as a verb too.
P.S Damn, it's even worse than that. All the hard work -- which was not that hard itself -- was already made by Apple in their CoreText API. According to the Verge:
">the heavy lifting is already baked into Apple’s developer platform. Since iOS 5 and OS X 10.7, Apple has provided a class called NSLinguisticTagger that segments natural-language text and labels the text with various bits of information, including parts of speech".
P.S 2 Now, since Apple itself added the capability to tag parts of speech in the iOS API, and since this is a very obvious example of using this (Apple themselves showed something similar in WWDC 2011 IIRC), isn't iA's patent application essentially like asking to restrict the use of Apple's very specific API? It would be interesting to see iA's lawyers fight Apple's lawyers.
P.S 3 This was part of a comment in a sub-thread, but thought it would be more useful posted at the top level to give everybody an idea of the feature.
Not to distract from the core of your argument, but "simply tagging words" is still a hard AI problem in general, with all those pesky "time flies like an arrow" vs. "fruit flies like an apple" cases. Of course you are not doing anything innovative in that area if you just call an API provided by someone who did the hard work.
The definition of "obviousness" in a patent is whether there exist two works of prior art that can be combined to create the invention. In this case, any prior art that stylizes text based on syntax could be combined with any prior art that determines the syntax of natural language to invalidate a later patent that stylizes natural language text based on syntax.
Whether the output feels different and non-obvious to a user is unimportant. Application to a different domain ("programming" versus "writing") is irrelevant. Patent law is very mechanical, and the rule is that it must not be any combination of prior art unless the means of combination is patentable on its own.
iA appears to be attempting to enforce their idea through pre-emptive public shaming of future competitors, which could be effective in a community of creative writers, but doesn't legally hold water.
Unless you're against all software patents on principle, I don't see an argument here for why patenting this feature deserves scorn.
If it behaves the way it's described, it sounds remarkably useful. Indeed it's obvious in retrospect, but that's true of many physical inventions worthy of patents.
Is there prior art? The example of Phraseology's "Inspect" doesn't seem similar at all. It's not a visualization tool; it's an analysis tool.
My understanding of the feature is that you can focus on a specific syntactical category and dim everything else out while you edit. This will let you focus solely on adverb use (for example) and make changes accordingly.
It's one for the lawyers I suppose. As far as I see it the only reason that I would buy writer pro is for the syntax control (and I want it). If an app came out with the feature for less money I would probably buy the cheaper version.
I think it's fair enough that they have attempted to file a patent here. Their business model relies on having no one compete with them on this feature.
That said, I think they could have handled the announcements with a little more grace. Maybe it's just the adrenaline of launching a 4 year long project?
Phraseology highlights selected words and is cumulative. Writer Pro dims the 'unselected' words and (I'm assuming) doesn't let you stack word classes.
I prefer Writer Pro's execution. I can see it as something that will actually speed my workflow up. Phraseology doesn't look like it will make things any faster for me.
It's not just about the nature of the feature, it's about how it affects my end-product as well.
Edit: double checked Phraseology and mixed it up with a Spanish-English app made by same dev team. Edited accordingly.
I started to write this a few years ago: http://www.mrspeaker.net/2012/03/24/syntax-highlighting-for-... - "syntax highlighting for writers", I called it. I was using an MIT library to get the semantic meaning for each word. Didn't get much interest from writers I asked, so dropped it... Though that blog post did get a bit of traction. I wonder if Information Architects ever read it ;)
Unless you're against all software patents on principle, I don't see an argument here for why patenting this feature deserves scorn.
If you aren't against software patents, you don't realize that any patent can more or less creep into any semi-related field ultimately producing something like a totalitarian regime where anyone producing any application will need multiple licenses.
And so, if you have already sworn allegiance to our future insect overlords, then indeed, no patent will receive scorn though they all will still deserve it.
If it behaves the way it's described, it sounds remarkably useful.
It sounds like the syntax highlighting that every half-decent IDE and programmer's editor has had since forever. Which is odd, since there's no way that should take 4 years to implement.
Except this isn't a programming editor, and I've never seen one of those that uses dimming aggressively on all but a single syntactic unit. Maybe one exists. Screenshots welcome. And this is assuming I understand the feature correctly.
Unless the said invention carefully distinguishes a programming editor and a human language editor, the invention may still overlap on grounds of novelty. If the said invention does carefully distinguishes the two, the invention would still be questionable on grounds of obviousness unless it explains the key difficulties involved in migrating from programming languages to human languages and the invention claims a solution to that problem. I think the latter is unlikely.
Note: IANAL, and haven't read the claims of the said invention.
You obvious know next to nothing about English syntax. Have you never wondered why you can learn python in a day or two, but not English or Chinese?
First of, you need a huge dictionary of all words and rules for their conjugations.
Then you need a huge rule set for bound morphemes. Which adjectives can you add 'ly' to in order to make them adverbs? There are literally hundreds if not thousands of bound morphemes and they can sometimes only be applied to certain words within certain word-classes.
Then you need a huge rule set for when each word acts as a specific word class (is bank a noun or a verb? Is anime a noun or an adjective? Is like a verb or preposition? You know, "Time flies like an arrow, fruit flies like a banana", actually like can behave like several wordclasses)
These are not solved problems yet, but we are getting very good at them now. We weren't 4 years ago.
Failing that, you could also take and annotate word classes in thousands of pages of text, and then use a machine algorithm to apply similar annotations in the the text entered into the program. Unless you are expecting people to write articles for the Wall Street Journal, this might not be such a good route either.
They're just using NSLinguisticTagger. Correctly lexing English is harder than lexing a programming language, but that's hardly relevant when there's a pre-existing library to do it for you.
>You obvious know next to nothing about English syntax. Have you never wondered why you can learn python in a day or two, but not English or Chinese?
And you obviously know next to nothing about how this is implemented in Writer Pro.
We're not talking about AI-level intelligent understanding of english here (which not even MS Word has in its syntax check). We're talking about a ho-hum, "this word is a noun, this is verb" etc kind of tagging, which just needs some metadata (you can even scrap off of an online dictionary).
There is some context-awareness required, though. Verbs can almost always function as nouns, too. To determine which requires some reasoning about the sentence a word appears in.
"Drive to the store." vs. "I'm going for a drive."
"Don't punch Judy!" vs. "I gave Judy some punch."
You're right, though. It's not AI-level intelligence required for this.
You can't ignore the practicalities of what iA are dealing with, but it does suck to try to control what other people do by artificial means. If writing the "Syntax Control" feature once meant it was done forever, usable anywhere, untied to their product, I might sympathize more with the idea that nobody else should build & sell it because it wouldn't preclude using it. They'd be selling a software product in that case: the "syntax controller". But here they're basically requesting payment for the privilege of re-implementing their idea. Well, sorry, screw you. Information Architects doesn't really provide any service when they collect that patent money.
Well, I'm against all software patents (at least all that I can think of), so I suppose that describes me. If their editor, with its syntax control, is so sweet, they should sell a whole bunch of software and be happy.
Would you mind explaining how the feature described is novel and hasn't been done before? Because I wasn't able to discern anything I haven't seen before in one form or another.
You can point to a text authoring tool with a mode that will dim all but a selected syntactic unit of text for focused editing? I am not versed in the hundreds of text editors out there, but if one were commonly known surely a link to a screenshot would have made the candler argument much more convincing?
It sounds to me like the author doesn't like software patents regardless, and was offended by a misreading of the word "cheap": I don't think the original blog post implied that copies would be cheap-as-in-crummy but cheap-as-in-easy.
These are all linked to from other comments here. This is not new or unique. Sure, the design is different, and they are probably the first "big" name writing app to include it, but being more popular doesn't mean you invented anything.
I would distinguish the dimming aspect from the selecting syntactic units of text aspect. The former clearly isn't new[1], and adding an obvious technique to the equation doesn't make the entire method non-obvious.
So then the question should be: Is selecting syntactic units of text unique?
I don't know of any specific applications that do this, but I imagine that there are programs or libraries used in the field of machine learning that do exactly this or make doing it trivial and obvious.
[1] It's basically applying a dimming effect to an inverse selection. Off the top of my head, this is a common technique for highlighting passages from documents in video documentaries.
This is enough to make the concept of highlighting a syntactical unit a prior art (assuming the date on the said invention is not older), and the invention may now at best claim a method of solving this. The latter would barely be an invention on grounds of obviousness.
Note: IANAL, and haven't read the claims of the said invention.
But going from a *ly grep to a syntactic unit is merely a matter of using a context-free grammar instead of regular expressions. The latter extension is obvious.
Imagine a patent on something that's basically UNIX shell pipes. Filed after the UNIX shell had been invented. And it turns out that the program written by the patent filer just calls out to /bin/sh to do its pipe stuff.
That analogy doesn't make sense, as I understand the feature. They're not patenting "basically" NSLinguisticTagger. They're patenting the use of linguistic analysis to provide visual focus in a text editor. Again, yes, this is obvious in retrospect, but AFAIK, no one else was using linguistic analysis in this way before.
Cotton gins are obvious in retrospect - you basically strain the cotton through a mesh that's smaller than the cotton seeds and have an automatic brush that cleans the mesh to prevent jams. Put the straining process on a wheel instead of a straight belt and mess with some gears to make everything work off of turning one crank, and you're basically done. In retrospect it's straightforward, but we went centuries or millennia harvesting cotton without figuring that out.
Most things are obvious in retrospect - you can look at a traffic light and say "yeah, I'd obviously think of that - you need a bright and clear signal positioned where all can see it, and you need to have a warning light in addition to stop/go since cars have a braking distance" but it took years for someone to come up with it.
I agree that just because it's obvious doesn't mean it's not valuable, but my position is that it does not deserve the various government-granted legal protections of a patent. If it is obvious then it is something that you run into, not something that you create. By all means exploit your first-mover advantage as much as you can (be it a genre, a market, or a technique), but don't ask the government to grant you a time-limited monopoly. Just because it didn't exist for centuries doesn't mean it took centuries to create.
(note: "obvious", "straightforward" and "easy to understand" have various degrees of subjectivity attached so it's never easy to argument in these terms)
If I can program the feature from scratch in less than a day with only a simple explanation of the feature, I don't really think it should be patentable...
Time-to-reimplement is not and should not be a measure of patent validity, because a patent isn't about difficulty of writing the implementation, it's about the cleverness of the solution. What needs to be measured is "given the same problem and existing tools but no knowledge of this solution, how long does it take to come up with this idea". A clever solution that takes 20 minutes to implement but that nobody would think of in a dozen years should be patentable while a solution that takes months to implement but seconds to imagine shouldn't be patentable (but is copyrightable).
In this case though, I'm not sure "we took a first-party API and applied it like other products, but we used dimming instead of highlighting" qualifies.
I was a happy user of iA Writer: loads fast, does what I want, prevents me from messing with settings for hours, etc.
Writer Pro is just BS. A half-assed product to make money off of people who already bought Writer.
The "workflow" involves just showing you the same text with a different font, depending on which step you selected (Notes, Write, Edit, Read). When you step into "write" mode, you're editing what you wrote in "notes" mode. If you step back to "notes" mode, you just see what you were editing in "write" mode -- that is, you cannot keep notes that are not part of the final output, something every writer does. Only the font chances, you're working with one document all the time.
As for the show syntax "innovation", it's just highlighting different parts of speech (verbs, nouns, etc). Nothing extremely worth your while.
Those 2 half-baked features (with several MIA from iA Writer) are not worthy of the "Pro" moniker.
Opinionated software I'm fine with (I like iA Writer). This is more: "software that has the opinion that it should scam you out of your money" (especially with the only way to check it out being buying it from the App Store).
Since Apple provides the metadata tagging of the words via NSLinguisticTagger, then is this the web equivalent:
//dim non-anchor tags
$('*').not('a').addClass('dim');
//hightlight anchor tags
$('a').addClass('highlight');
If the browser provided the syntax metadata, then I don't see how this example with anchor tags is logically different. This appears to be obvious and trivial, where a patent requires non-obviousness and non-trivialness given the state of the art.
I'd like to see iA being hit with their own medicine, to taste what it feels like.
Let some patent holders hit with other text editing patterns that they violated in iA Writer and Writer Pro.
The "font adjusting size to the size of the window" would be a good start. Or how about "word count based on text selection". Or "auto formatting markup syntax". Or "modal writing workflow".
This raises a larger question about patents in software. If I can anonymously create a program or feature that violates a patent, and release it to the wild as free and open...the patent holders have nobody to go after.
The end-game for this scenario is for the whole package to provide enough polish and other features that you'd be stupid to cobble together the OSS equivalents after you reached a certain level of affluence. And then it's the business' job to make sure they can profit off that audience division. Good companies understand that and don't waste their time with needless software patents. Poorly run companies insist that "it's mine, you can't have it" and wind up stirring up trouble from amongst their user base.
Of course, it would seem iA is in the later category. Hopefully they've done some market analysis that they still have a critical audience mass who are either ignorant/apathetic to such tactics and buy it anyway.
Haha, and then you wind up looking like the RIAA, suing your hoped-for audience. Only tone-deaf execs and greedy lawyers would sue their audience members.
And before you say, "but they were stealing and so weren't going to be part of the audience." I'm well aware of that, but the reality is that you're drawing out the ire of folks from amongst your hoped for audience. Just not a good plan.
the fact that it's not a good plan hasn't stopped others from doing it. It's not much of a comfort to know that the person ergo
who sued you made a bad business move.
It's not too hard to find what seems to be prior art for this idea, even if it's not a feature you actually see in many word processors. For example this blog post from 2007 shows a screenshot of a text editor syntax-highlighting English:
1. Until we see what they're trying to claim, we don't know what the patent covers. At this point it seems it's just a provisional, which means they might not even have drafted the claims yet.
2. The use of Apple's existing APIs to do the "heavy lifting" has no bearing on the "quality" of the patent or the novelty (if any) of the work. All inventions are built on prior work.
3. As a corollary, the triviality of implementation is orthogonal to obviousness of an idea. As an extreme example, RSA is pretty easy to implement, but it was a breakthrough nonetheless. (Of course, patents very rarely cover anything as groundbreaking -- the bar is and always has been much lower.)
4. If you are concerned that this leads to an unnecessarily broad patent, submit it to askpatents.com, along with any relevant prior art. This will increase the chances an examiner will find it, especially since it's highly unlikely that they're going to look at WWDC videos.
Unfortunately, as it's not even a non-provisional application at this point, you would need to wait for one to be filed and published. This could be as much as two and a half years from now, and I doubt anyone will even remember by then.
http://en.wikipedia.org/wiki/Part-of-speech_tagging
So the "new" "invention" they want to patent is "we can display the result" which basically is nothing more than syntax highlighting? Or did I get something wrong?
I don't get what syntax control is. Is it just a count of different grammatical categories? Does it do some kind of sentiment analysis and tell you if your sentence structure is aggressive or something?
As I understood from the original blog post, it identifies bad syntax more easily, enabling you to correct them. In programming terms, it would be a macro that finds code smells and lets you refactor easily. The example he gave was identifying excessive adjective use.
Imaging writing a 100+ page document and doing Menu > Syntax Control > Adjectives and seeing all the places where you use unnecessary adjectives. Or really long, run-on sentences. Or extra semicolons. These may be correct grammatically but they might lower readability so you want to minimize their use.
I could be wrong but I think this is what they meant by Syntax Control - ability to detect syntax use in English and not highlight/fading text color. If it is the former, I can absolutely see why it took four years to create.
Actually you're overthinking it. It's nothing of the short (I've bought Writer Pro).
It's exactly this: you select "verbs" and the verbs are highlighted. You select "nouns" and the nouns are. Same for adjectives, adverbs, prepositions (as, from, in) and conjuctions (and, or).
Nothing earth shattering, nothing you can't do with simply tagging words, nothing marchine-learning or AI about it.
P.S Damn, it's even worse than that. All the hard work -- which was not that hard itself -- was already made by Apple in their CoreText API. According to the Verge:
"the heavy lifting is already baked into Apple’s developer platform. Since iOS 5 and OS X 10.7, Apple has provided a class called NSLinguisticTagger that segments natural-language text and labels the text with various bits of information, including parts of speech".
thank you for the screenshot. though i do wonder, what exactly is it highlighting?
"in programming terms," - why is programming a verb and what good does it do to highlight it?
Please put the update at the top of the article, felt a little mislead after reading the full drama then only being told at the end that it was no longer current.
I doubt somebody that gets upset about a small company protecting their idea with a patent (that he hasn't even seen) would actually buy software from them anyway.
His example ("prior art") wasn't remotely relevant.
So many upset about something you have no information on. But go ahead, grab your pitchforks and go hurt another small developer.
Hindsight is always 20/20, especially when it involves obvious things that somehow nobody managed to combine in a certain very useful way before. You can slap a piece of fried squid on a bowl of risotto, but that doesn't make sushi an obvious derived work.
This is an area where the downvotes are just waiting to crush you...but as per usual, I disagree with the premise that these guys have done something wrong.
Patents and copyright provide small businesses or inventors the means to monetize their innovation for a time. I don't see what's wrong with that. It could stand being reformed, like almost everything in American civil life...institutions are broken all over...but holding this small group of ira glass look-a-likes responsible for the larger system being broken is kind of lame.
They made something wonderful, they should be able to be rewarded under the terms of the law. Don't like the law? Take that up in the political process.
At this point usually everyone says that the political process is irreparably broken and the Republic is lost. If you think that's the case...I don't know...just kill yourself. There's nothing you can do and everything, everywhere, is bad. Woe is me. Oh. Oh, woe is me.
Patents and copyrights are more different than similar. Copyrights provide immediate protection with almost no up-front cost, and apply almost invariably to cases where people have deliberately and overtly derived value from your work.
A small business attempting to make use of the patent system can expect to spend many tens of thousands of dollars for the privilege of waiting 5-7 years for the issuance of a document that entitles them to wait 2-5 years for the privilege of a hearing in court which only has a chance of being remunerative if the counterparty is a gigantic tech company that can afford to bury you in legal costs.
Aside from all this, flip a coin: heads you're in the "right" and have a "case", tails your patent claims are insufficient to keep anyone from realizing value from the idea underlying your patent: you lose, and we all breathe a sigh of relief.
Patents make sense for some industries, where inventions are clear (or at least justiciable) and product lifecycles are measured in half-decades. But they are decidedly unuseful for software entrepreneurs.
Patents are a pox on the software industry.
(I have a few, from earlier in my career; I've never been the one paying for them).
Things that are perfectly legal and acceptable in some areas may be heavily frowned upon in others. The law can't take into account best practices for each group, it has to be broadly applicable for the entirety of the population. In that regard, it doesn't sound like IA has done anything illegal or even anything that should be illegal.
What can be done, however, if you disapprove of their actions is to do exactly what the OP is doing. Calling them out, shaming them to a group that might feel similarly. Likewise, if one of my friends is being a jerk, I can choose not to be friends with him and also tell my other friends what he did that irritated me. Hitting on my girlfriend is not illegal, but it's not acceptable in my social circle either.
Not everything needs to be illegal, sometimes it can just be immoral to a certain audience.
> Patents and copyright provide small businesses or inventors the means to monetize their innovation for a time.
No they don't. Quite the opposite: they prevent small businesses from innovating, as they're hampered by the legal minefields of patent and copyright infringement. They only benefit large corporations that already have a legal team and their own patent portfolio to bargain with.
Two lies that can be ignored; a rich man telling you what's best for the poor and a large corporation telling you what's best for small business :-)
Considering the number of posts and links in the threads here on HN showing prior art, making the assumption that just because you saw it on IA first doesn't mean they invented it.
While information architecture is important, having a dedicated information architect leads to an "ivory tower" effect. This generates massive amounts of paperwork and unnecessary design artifacts and also decouples the resulting architecture from the needs of the application. These effects more than negate the benefit of having a specialized resource fill the information architect role.
Don't support information architects... hire multi-skilled developers who can do the information architecture themselves.
Yes I did (some nut want to patent syntax highlighting). And then I thought it would be amusing to try to see what would have led to the headline meaning what I had expected it to mean. Apparently you people have no sense of humor. :)
As you may have realized by now, this is HN, where you expect sense of humor at your own peril. In general, humor is not (much) appreciated here, except when used in an intensely witty context, as a left-handed compliment or couched in ten layers of pseudo-wisdom.
You're overthinking it.
It's the most naive implementation you can think of. It's exactly this: you select "verbs" and the verbs are highlighted. You select "nouns" and the nouns are. Same for adjectives, adverbs, prepositions (as, from, in, etc) and conjuctions (and, or, etc).
Nothing earth shattering, nothing you can't do with simply tagging words, nothing marchine-learning or AI about it.
Here's an example of Syntax Control in action: http://imgur.com/NrseuZV
Beside the naiveness of the core functionality, notice also how naive the tagging is. As noted below by bosie, "programming" is marked as a verb in the phrase "in programming terms", syntactic context be damned. Same for "run-on" where run is highlighted as a verb too.
P.S Damn, it's even worse than that. All the hard work -- which was not that hard itself -- was already made by Apple in their CoreText API. According to the Verge:
">the heavy lifting is already baked into Apple’s developer platform. Since iOS 5 and OS X 10.7, Apple has provided a class called NSLinguisticTagger that segments natural-language text and labels the text with various bits of information, including parts of speech".
P.S 2 Now, since Apple itself added the capability to tag parts of speech in the iOS API, and since this is a very obvious example of using this (Apple themselves showed something similar in WWDC 2011 IIRC), isn't iA's patent application essentially like asking to restrict the use of Apple's very specific API? It would be interesting to see iA's lawyers fight Apple's lawyers.
P.S 3 This was part of a comment in a sub-thread, but thought it would be more useful posted at the top level to give everybody an idea of the feature.