Hacker News new | comments | show | ask | jobs | submit login
How QuarkXPress became a mere afterthought in publishing (arstechnica.com)
117 points by DanBC 1296 days ago | hide | past | web | 69 comments | favorite

It all boiled down to one thing for me as a print designer at the time: OS X.

Quark simply failed to release a version for the new OS, and like the article says, it treated Mac users with disdain when we asked for it. Anyone who worked in print during the '90s and Aughts knows that Windows was disliked by designers for its poor handling of type and PostScript. We weren't about to switch our entire computing platforms, thousands or even tens of thousands of dollars worth of equipment, just so we could use a layout program from a company that seemed to want us gone.

As far as the print community was concerned, Quark had stopped releasing software. We didn't "switch" so much as we were forced to look elsewhere when the product disappeared.

It was actually a bit strange turning to Adobe at first, because they had picked up Aldus, and Pagemaker with it, and it didn't have the best reputation. Discovering that InDesign was actually superior to Quark was just the cherry on top. Think of it like being dumped by someone you really loved, only to find out there was a much cuter, smarter, funnier person who'd been crushing on you the whole time, just waiting for you to see that they were there. Quark became the ex, and InDesign got the ring.

Yeah, Quark made a lot of mistakes, but I agree that this was the Big One. Quark's audience were Mac users, Mac users were moving to OS X, Quark refused to go with them. Even at the time it was an utterly baffling decision; it hasn't gotten any easier to understand with time.

I would love to read an article that interviews Quark insiders about why they made that choice, what the internal factors were that drove them in that direction. Seeing as how it was basically corporate suicide and all.

What portion of their users were Mac users?

I vaguely recall Quark being in the category of things where the software cost more than the hardware. At some level it can make sense to tell the user to switch their hardware and OS (since it's the smaller portion of the cost), but you still need the customers to listen.

There's more to the pain of switching your hardware than just the material cost. As a designer, when I finally got away from Quark, it was good riddance to bad rubbish. Their entitlement problem (No, you switch your hardware, ecosystem, technology culture) went beyond their marketing acumen into their UX design.

It's easy to forget how expensive computers were in this time period. A new high end mac (or PC) could easily set you back $10,000 or more. A Mac IIfx was $8,970-10,970, monitor not included.

You're talking 10-15 years earlier than the time at hand (early 2000s). By then computers were already pretty cheap.

During lots of this, I don't remember their being a non-Mac version available.

(checking WP) yeah, it was released in 1987, and the Windows 3.1 version didn't come out till 1992.

VMWare could do well to remember this now they have killed VMware Service Manager.

The article mentions Quark lost because it became rubbish but does not give the more interesting reason as to why that happened, basically that "The original owner sold the company and the new owners fired the entire development staff and outsourced all development and customer service to India"


For what its worth, I worked on the InDesign team during the InDesign 2.0 - CS timeframe, and my boss at Adobe had worked for a long time at Quark. He was very knowledgable about Quark's code.

During that time period, he repeatedly asserted that Quark would never be able to produce an OS X/Intel version. He believed that there were too many MacOS/PPC specific optimizations, and that the code was so difficult to modify that it would never happen.

There were also a non-trivial number of top people from Quark working on the InDesign team at that time. My impression was that Quark was a pretty unpleasant place to work, and that many of the best people left long before InDesign became dominant.

I think the article is right that Adobe tried harder to win over customers during this period. From a technical perspective, however, I think this was a classic story of an old product, architected with older technologies and for slower machines, being outpaced by a product architected with newer technologies and for faster machines.

InDesign's internal architecture is the most robust, extensible desktop application architecture I ever worked with. But it was quite heavy compared to quark's, and initially that caused performance problems. But in the end, faster machines overcame the weight of the architecture (as well as optimizations which were facilitated by cleaner code), and enabled Adobe to innovate faster.

I think you see this same kind of phenomenon with something like the Nokia/iOS transition. At first, iOS seemed to have fewer features, be slower, and require excessive hardware. But pretty quickly the hardware outpaced the architecture, and the power of the underlying architecture enabled vastly more innovation.

I suspect Quark was doomed whether they outsourced their engineers or not.

As a side note, I always wondered if the Creative Suite product offering (which bundled ID with photoshop) was just hype or also really got people (especially students) to just forget about Quark. That seemed to happen around the same time, give or take some years of overlapp.

The article makes that point explicitly:

  And things got even worse for Quark when, in 2003, you
  could get InDesign as part of the newly introduced
  Creative Suite Bundle, which as a publisher or graphic
  designer, you would have probably bought anyway for
  Photoshop and Illustrator. You were basically getting
  InDesign for free—it was a Trojan horse in the bundle box,
  and I think it really sealed the deal for people on the
  fence about the switch.

Quark India had a policy of hiring only fresh graduates and paying them very little. After 6 months of training and 1 year of work, they would fire the whole lot and recruit fresh graduates again[0][1].

When this cycle was repeated every year[2], word got around that hundreds of well-trained but desperate Quark employees would hit the job market in March[3]. Other software companies would wait in anticipation and snap them up.

[0] http://www.coderanch.com/t/29639/Jobs/careers/Perot-Quark

[1] http://www.jatland.com/forums/showthread.php?7090-Walkin-in-...

[2] http://timesofindia.indiatimes.com/city/chandigarh/Quark-Cit...

[3] http://www.tribuneindia.com/2005/20050307/biz.htm#1

Reminds me of the story about Excel and Lotus notes: http://www.joelonsoftware.com/articles/fog0000000052.html

I love this:

There's nothing wrong with being in a market that has established competition. In fact, even if your product is radically new, like eBay, you probably have competition: garage sales! Don't stress too much. If your product is better in some way, you actually have a pretty good chance of getting people to switch. But you have to think strategically about it, and thinking strategically means thinking one step beyond the obvious.

The only strategy in getting people to switch to your product is to eliminate barriers. Imagine that it's 1991. The dominant spreadsheet, with 100% market share, is Lotus 123. You're the product manager for Microsoft Excel. Ask yourself: what are the barriers to switching? What keeps users from becoming Excel customers tomorrow?

On episode 54 of Accidental Tech Podcast [1], they mentioned a recent flair-up about the screen writing software Final Draft.

It seems the CEO went on a podcast about script writing called ScriptNotes two times [2] and it sounds like he was caught totally flat by listener feedback about how far behind his product was. Film director Kent Tessman wrote up a post [3] about how foreseeable much of this was and just how bad the situation is.

It reminded me quite a bit of this article. An old & dominate piece of software that is built on old technologies and has obvious deficiencies that doesn't seem to be progressing and seems ripe for disruption.

1. http://atp.fm/episodes/54-goto-fail

2a . http://johnaugust.com/2014/the-one-with-the-guys-from-final-...

2b. http://johnaugust.com/2014/period-space

3. http://www.kenttessman.com/2014/02/notes-on-scriptnotes/

Fascinating. There's been lots of recent discussion about problems with the Wii U launch by Nintendo. One of the interesting things to come out of it:

Pre-launch American third party companies kept asking Nintendo about how the networking system would work, referencing XBOX Live and Sony's offerings as examples. Would it be on-par, or similar in features, or totally different?

Nintendo responded by asking the third party developers to please stop referencing the other company's on-line systems because nobody at Nintendo had any experience with them.


"This was surprising to hear, as we would have thought that they had plenty of time to work on these features as it had been announced months before, so we probed a little deeper and asked how certain scenarios might work with the Mii friends and networking, all the time referencing how Xbox Live and PSN achieve the same thing. At some point in this conversation we were informed that it was no good referencing Live and PSN as nobody in their development teams used those systems (!) so could we provide more detailed explanations for them? My only thought after this call was that they were struggling - badly - with the networking side as it was far more complicated than they anticipated. They were trying to play catch-up with the rival systems, but without the years of experience to back it up."

I remember reading that. I love Nintendo, but not only was that line incredibly sad, but it spoke volumes about my experience with Nintendo's software in recent years (outside single player games). It rang too true.

The timeline the CEO gives for Cocoa is full of pooh. It has been a lot longer than 4 years ago that Apple said to use Cocoa over Carbon. 64-bit just forced the issue. Unless you are Microsoft or Adobe, Carbon was a very temporary waypoint as far as Apple was concerned.

As a Graphic Design major at the end of the 90's, I had to learn to eat, sleep and breathe Quark. Dealing with the company was HELL on a student. For context sake, this was right around the time that Apple released their new iMac with NO ADB ports. (USB only) Quark used a hardware dongle even for their educational version (which cost over a thousand dollars if I remember correctly) The company was extremely obtuse when it came to solutions. They suggested I buy an ADB to USB adapter. It flaked out constantly. I tried a few different models. I turned to Quark for help. Almost every conversation with support seemed to be "you don't like it? try formatting your text in Microsoft Publisher". All I was asking was to assist me in using their software - which I had paid for. At that time, they OWNED the market, so anyone that they helped was almost guaranteed to be a future customer. Adobe on the other hand had educational versions of Illustrator and Photoshop for 99 bucks a piece. I graduated when InDesign 1.0 had just come out. No one took it seriously at that point as it was brand new. But Quark's "we don't care, we don't have to" attitude is definitely part of what unseated them.

I have used both inDesign and Quark in a professional print-design setting.

Another thing that the author left out here was the to be a power-user of nearly _any_ software package requires you to learn keyboard shortcuts.

I can remember day 1 I made up a big list of all Quark shortcuts, printed & laminated it and used that thing all day every day. This was a major hurdle to new Quarks users - they're too slow until they memorize the shortcuts.

inDesign was smart about this, not only could you _change_ your user shortcuts, but they were malleable enough you could even make the same as Quark's shortcuts if you so desired. You had the freedom to mold the software to your way of working.

Ultimately it became very very trivial for Quark users to switch _to_ inDesign, but going from inDesign _to_ Quark was like a step back in time.

I worked at a prepress & design company (in prepress) that used Quark. I had time to kill every now and then so decided to start reading and working through the InDesign 2 instruction manual (actual printed manual). I'd learn something and small and think "holy shit, this might actually be a wonderful application". It gave those tear to eye of joy type of moments constantly (something I don't get much with just about anything anymore). So I worked on getting the designers to switch. They were skeptical and at first mildly unwilling because it would interrupt their workflow but were eventually forced to by our studio manager. Their attitudes changed very quickly despite the pains of learning a new app and dropping something that they were masters of. Anyway, while I had a rep for being quite technical and quick to learn stuff, they all had quite a few years experience more than me and were faster with the keyboard (I was no slouch). Switching to InDesign leveled the playing field dramatically for me, it was at that company where I think I went from juniorish/intermediate to senior level mac op within about a year or less. I decided to learn the native shortcuts because it was obvious that this was the app of the future. The designers used the Quark shortcuts but probably learned the native ones eventually. Well, one hopes anyway. Doing web application development now of course, can't stand prepress which has quite a low glass ceiling and is quite depressing really.

A product with 95% market share needs engineering a lot less than it needs sales/finance. As engineers, we see this as "business guys screwing it all up." But these companies are acting rationally give the environment they are in.

Why the focus on sales? Because at that size of a market share the features of the product matter less. Adding feature X is very unlikely to get you more share. Additionally, the product has so much inertia that it can become bad, and you won't tank over night.

The only way a company in that market can continue to grow revenue and profits is to reduce costs and increase sales. Hence the focus on sales guys who can talk the talk, bundle software and discount, and do everything to eck out more share. Meanwhile finance guys come in and look to see what they can trim, and often that's engineering.

Now we can argue about the net effects that the enormous pressure to keep growing can cause. Or about how the stock market and C-level executive comp structure has promoted short term thinking over long term strategy (Just look at Michael Dell. He had to literally take his company out of the public markets to do the long term work that needed to be done to save Dell). These companies are acting rationally give the environment they are in. The environment is just screwed up.

I don't fault Quark for being stupid. Instead I simple recognize that they were not exceptional enough to free themselves from this cycle. Very few do.

Focus on selling what, and to whom?

Remember, they were selling one-time licenses, not subscriptions, so 95% market share means 95% of users had their product, not that they sold something to 95% of them that year. One of the biggest challenges about being a dominant player is actually that you mostly compete against the previous version of your products. "Why should I upgrade? This isn't any better than the edition I bought last year." So the business isn't fighting to get that last 5% of the share; that only allows it to grow a measly 5%. Instead, it has to focus very strongly on finding ways to make its existing customer base give it more money. And you do that by improving your products and services so that you can get existing customers to buy more of them. It's the only lever you really have, so doing almost anything else is crazy.

And here's a company failing to even address existing concerns with its products, so it's hard to construct a story on how it was acting rationally. It's not even clear how much money it was even making; how well did those new versions sell, given that they didn't seem to improve much? So I submit that even on a short-term basis, they weren't acting rationally at all.

Not to mention the long-term basis. You should note that Quark was privately held, so the stuff about the stock market doesn't apply here. One of the big advantages of being private is that you can play the long game. But they weren't! If you're not making your customers happy, someone else will. It's easy to blame the environment, but the environment was actually giving them all kinds of feedback about what they should be doing and it's the environment that allowed a competitor to come in and eat their lunch.

Isn't that a textbook example of the dangers of short term thinking? In any engineering field, it's irrational to think that the industry won't shift underneath you in a way that makes your 95% market share irrelevant. And when the market does shift, the incumbent that fired all its engineers will be easy pickings for the new breed of competitors.

It's not just a tech industry issue either. When the U.S. auto industry was ascendent, Japan had a reputation for poor quality products and South Korea wasn't even on the scene yet. Well, Japan kept improving the technology and with that the quality, and not only caught up but blew past Detroit and the automakers did not even see it coming. U.S. automakers have spent the last few decades investing in the improvements Japanese companies had the foresight to be investing in all along.

It's not "rational" for companies to mortgage their futures to goose profits in the near term. It might be "rational" for executives to do so, but not for corporations as a whole.

Yeah, this is what happened. I supported Quark users at the time, and everyone was just surprised and confused. In contemporary terms, it would be as if Zynga didn't ship Farmville for iOS 7 and everyone had to switch to something else. Maybe they made a weird bet that customers wouldn't adopt OSX, or migrate at a decades long WinXP-pace. Since Apple binds the OS to hardware, it happened in (only) about five years.

Running Quark 4 and 5 under Classic was ugly and problematic. When Quark 6 finally shipped it was so buggy that it couldn't be deployed. By the time Quark 7 arrived, most people had switched.

You could hire more sales people to try to eek out a few more sales, or you could spend your engineering money on diversifying your product line. XPress is essentially Quark. So much so that when people say Quark they mean XPress.

The company should have looked to grow into new markets, rather than eek out a few more sales in a saturated one.

In addition, while not sexy, at some point you need to modernize your software, and money should have been spent there instead of on sales. I switched from Quark to InDesign because of OS X. It's the OS I used, and Quark was only one program that I used, and I was never going to use OS 9 again.

We're not talking about writing a native Cocoa app here, but just porting over a Carbon app. Quark couldn't even be bothered to give us a crappy Carbon app like Microsoft did with Word and Adobe did with Photoshop.

> The only way a company in that market can continue to grow revenue and profits is to reduce costs and increase sales. Hence the focus on sales guys who can talk the talk, bundle software and discount, and do everything to eck out more share. Meanwhile finance guys come in and look to see what they can trim, and often that's engineering.

Honest question from a non-businessperson: Why do you have to keep growing revenue and profits? Like, at some point, if you're making fucktons of money, can't you just focus on guaranteeing you continue making fucktons of money? Isn't it inherently unsustainable to have to be making even larger fucktons?

If I'm traveling someplace, I accelerate until I'm going as fast as I want to be going, and then I don't have to keep accelerating anymore more.

At that size you need to innovate to even stand still as the environment is changing around you, and this flux only gets magnified by your market share.

The first question here is why there's a maximum speed you would like to be going. If you keep accelerating, you'll get there even faster! It's also not inherently unsustainable to make ever increasing amounts of money, because the economy can grow with you; it's not a zero sum game.

That said, you're right that a lot of companies totally bone themselves trying to grow unsustainably, letting the stability of their company (like customer happiness, employee retention, that sort of thing) fall apart in the meantime, and then collapse under their own weight. While I think the parent is wrong about how it applies to Quark, they're quite right that the incentives for short-term growth are driven by how the stock market works. It may have even tilted our whole business culture to worship growth regardless of context.

Here's how it works. Let's assume a simplified version of the stock market in which stocks are priced based on the current performance of the company [1]. So let's say I buy stock at IBM today for $188/share. IBM makes fucktons of money, which is great for IBM but useless to me. It's useless to me because that's what determined the price in the first place, so if it keeps on making the same number of fucktons then the price is going to stay at $188 and my money is just sitting out there doing nothing for me. What I need to happen is for the stock to be worth, say $200 so I can sell it and make some money. So I need IBM to grow. As much as possible. In fact, it's the only thing I care about, because after I sell the stock I don't care what happens to the company. So I vote to put people on the board who want to grow the company and they hire executives with that mandate (who are anyway paid in part with stock and thus have the same incentive) and the whole corporate culture aspires to growth at all costs. If they said "well guys, this year growth isn't really going to work and will probably just sink the ship, so we're going to spend our time making our customers happy and pocketing our billions in profits" everyone would sell the stock, the board would be fired, etc. So that's why you have to grow. [2]

[1] That's not really how it works; expectations of future growth are baked into the current price. Sort of. The relationship between a company's performance, growth, dividends, and the overall economy is complicated, but it doesn't really change the point much, which is that as an investor, you want the stock price to go up, and to do that, you need as much growth as possible.

[2] Following on from [1], there's the argument that in a world with perfect information and rational actors, if the expected future growth is priced in, then sacrificing long term sustainability for short term growth wouldn't raise the price. But we don't live in that world.

Interesting new-boss-same-as-the-old-boss comment on the OP:


> I remember the exact moment that Quark pssed me off so badly that it forced me to install Indesign for the first time. I mistakenly ordered another Quark seat for apps $800-ish, and mistakenly pressed "PC" instead of "Mac" on PCZone's website. Zone wouldn't fix it so I called Quark to return the unopened PC version for a Mac version but they would do nothing. No transfer, no refund, I was S.O.L. of $800. That pushed me over the edge to try Indesign, and I never looked back. I now owe Quark thanks, because as a printer and early adopter of Indesign, I personally helped dozens of my designer customers discover Indesign and switch from Quark. Adobe was a true partner company and we all made great businesses together….*

> …. until Adobe Creative Cloud. It's dejaVu all over again, where Quark and now Adobe customers feedback can be ignored. Adobe with CC has succeeded in pssing off former advocates so badly that it's a distasteful just to say Adobe. In one swoop they destroyed tens of thousands of users paid value in anyone's full paid CS6. (except for a slight discount for first year towards forever payments). This CC Adobe bundle of 20+ crap apps to create value, much like Corel tried to do unsuccessfully a decade earlier, is an insult.*


I'm still using my legal copy of CS5, though I recently upgraded from Lightroom 4 to Lightroom 5. There was some kind of deal to use the Creative Cloud but I ignored it because even though I was only upgrading because LR4 doesn't yet support my new camera, the $99 upgrade is a pittance compared to the time LR saves me in photo editing. Second to Sublime Text, Lightroom is my favorite "work" purchase I've made as an individual customer.

So, what do people think of the Cloud? I didn't really read the terms, just that it sounded cheaper, in the way that "Just 30 monthly installments of $10!" sounds cheaper on certain TV advertisements.

I needed Premiere for a short project, so the Creative Cloud business model was great for me. I was prepared to spend a couple of bucks to use this software for a month, but they actually give you an automatic first month free.

I don't normally do any video editing, so this saved me a lot of money. But I certainly wouldn't buy my PCB layout software this way; I use it all the time and it makes a lot more sense for me to own a license with the ability to upgrade when a new version comes out rather than pay in perpetuity.

It seems to me that Adobe's business model here is amazing for people outside of their target market. A casual user who rarely needs their tools will love this, but their core user base is losing out. That's very short-sighted on their part.

I pretty much only use Illustrator. I'm still on CS6. It didn't help much that CC1 was one of those releases where I look at the list of new features and say "there is nothing here for me" that's happened every 3-4 years since Adobe moved to the marketing-driven "new release every year" model instead of the developer-driven "new release when it's done" one.

I've done the math and the Cloud model is not worth it for me. It starts to make sense if you regularly use like 4 of their big apps, I think.

Edit. Oh shit I just looked and they've quietly added some new features that I actually would like to have. Now I'm running the numbers. The existing pre-CC user pricing works for me - $360/y for my one big app every day, plus being fully legit when I need to briefly use InDesign to lay out a book or Premiere to edit a Kickstarter video - but once that trial year is over I think I'll be ditching it for the single app subscription plus briefly subscribing to other apps and canceling them.

Our shop is still using CS6 and we have no plans to upgrade to CC. There is no compelling reason to do so, and financially it doesn't make sense for us to join. For the rare times when a customer submits artwork created in CC, we simply ask them to re-save it in CS6 format.

In my previous job where we developped our own PDF interpreter (neither based on Adobe owns PDF lib or Print Engine) we only needed InDesign in short occasional burst, mainly to produce test files when new PDF features were introduced, or to check some issues reported by customers. Switching to the CC model for us made sense as it saved us money.

And in my current job we only need to use Adobe applications to generate demo files for tradeshows and sometimes private demos. So, we never require a full-time license. And we can use it as an alternative to a floating license as any of us can log in to use one of the application before logging out once finished. However now that I think about it I am not sure whether our marketing team switched to the CC model or have been using CS 5 or 5.5 the whole time. Considering they use it much more often than us it would make sense if they did not use CC.

I dug this out and submitted it to HN: https://news.ycombinator.com/item?id=7194751

I witnessed this turnover first hand when I was hired on as a customer service representative at the launch of QuarkXPress 5. This was a major release given that they hadn't released a major version for 5 years! Hilariously enough, we were trained to tout OS X support... via Rosetta. Native support didn't come until QXP6.

While I do think the major reason why Quark failed was due to a lack of a native OSX app, I think there's another major reason... Creative Suite.

The vast majority of desktop publishing users required Photoshop and the pricing difference between Photoshop and Creative Suite was much less than adding (or even upgrading) QuarkXPress. There's also the integration between CS apps that Quark couldn't compete with.

On a sidenote, people have mentioned how Quark was sold to Fred Ebrahimi and he outsourced the vast majority of the engineering staff to India, but did you know he also created Quark City? http://www.quarkcity.com/

I think this is a good example of what lots of people call the "MBA business model". It's a bit unfair to the MBA, but the idea is this:

1. End up in control of a thriving business. Doesn't matter how, flashy credentials and a nice suit, mixed with lots of buzz words and business speak usually work.

2. Fire everybody responsible for driving the product forward, frame this as "increasing profitability" or "cutting costs" or similar.

3. Increase sales staff and squeeze as much blood from this stone as possible.

4. When the market for your product inevitably shrinks because you've made your product irrelevant or because of competition or other reasons, blame shifting market winds for making your product irrelevant.

5. Sell the product or company off to somebody else, collect cash out your stock, rinse and repeat...or spend some time on your yacht.

An anecdote also related to the article. I grew up in family business...not very successful family business, but nevertheless. One of the businesses my family ran for a number of years was a small (very small) print house. Most months it didn't generate enough to pay my parent's paycheck, but they stuck with it till retirement.

One of the interesting things I had a chance to see was the invention and growth of the electronic desktop publishing industry. At the start layout was a very long process of getting blocks of text in a chosen font and size printed on small dedicated machines not too far removed from what Gutenberg would recognize. Then a specialist would, on a light table, manually cut the text to size, and glue the pieces of the layout to a master sheet. When all the artwork was finally in place, they'd go and produce a photographic negative of the final page, and then using something like lithography burn the image to a metal sheet which was then used on a press to produce the final page. As you can guess this took forever, but one of the most laborious parts was getting the original little blocks of text made up.

Enter the computer age and a small business like my parents could drop $20k for a pre-press dedicated layout system. I don't remember too many details of it, but it looked basically like your typical early 80s green screen computer. Everything was typed up by a specialist in some kind of meta-LaTeX-like language. There was a secondary very high res and large for the time black and white graphics preview monitor you could render the page on to get a look and proof (if you had enough memory). Then you could "print" the result to a photoprinter that worked quite a bit like a mini-darkroom. You had to keep various reservoirs stocked with the appropriate photographic chemicals, but the output was state of the art of the time. The rest of the process was virtually identical to the above -- even then it cut turnaround time on a job in-half. I wish I could remember the name of the machine, because I've never seen anything else like it. On occasion, when it broke down, a specialist would be called in to come repair it.

Over time, and as small printing business printing got more complex and this system aged, they replaced it a couple of times until eventually they hired a guy who also had his own smaller printing business on the side. They bought him out and he brought in-house a very early full-pre press desktop publishing system with him. Which was Windows 2.0 and a very old version of Aldus PageMaker, I think it was 3.something and a very expensive laser printer. They pretty much stuck with this setup until they sold the company many years later. Basically what this bought was that it essentially removed all the manual gluing and carefully measured layout such that, once a page was setup, you printed it off, took it to the dark room, made your negative and then the plate and the printing was off. It saved days of work.

Again, as the business grew more complex, the demand for color and complex layouts increased they started to need to do full color proofs. In printing, one of the principle problems is that, once a design gets through all of the darkroom, lithograph, ink color matching, screens etc. it rarely looks much like the original design. This magnifies when you start working with 2 colors, then gets exponentially more complicated up to 4 (CMYK). So the need for rapid proofs of what the design would probably look like became more and more important -- even then the proof only had a passing resemblance to the final product. Unfortunately, the equipment that could produce these proofs was prohibitively expensive. So my parent's graphics guy would outsource his designs to a graphics bureau to make a proof. I think a few pages might take a day or two to turn around and cost a couple hundred dollars.

I used to deliver the disks over to the bureau and remember the first time I walked in. A gleaming open room, packed full of Macintoshes running QuarkXPress and PageMaker, and some very expensive printer looking devices all along a back wall. I'd give them the disk and the job order and inevitably get a groan when I pointed out that it was a PC formatted disk (this was back in the days that this mattered). "Oh, it's you guys, when are you ever going to get a Mac and get Quark? It'll make this much easier." They'd wheel out, on a movable desk, an old PC, blow the dust off and fire it up and make sure the disk was readable. "Okay, we'll sigh call you when it's done."

Various incompatibilities with some of their proof making equipment and their PC caused lots of time and cost overruns, but it had to be done. I'm pretty sure, in the local market they worked in, they may have been the only PC graphics shop around. It sucked interfacing with the rest of the market, but at the same time, everybody who brought in their own artwork did that work on a PC, not a Mac. All the churches and parents getting wedding invitations, all the small businesses getting tri-fold fliers, all of them. And they got lots of business because other companies refused to deal with their homespun PC formatted artwork.

I don't have a point, just thought I'd add a little first hand history.

I worked at the Auto Trader in the late 80's/ early 90's. First as an assembly worker, then as an ad layout artist.

We had the typesetting printers and a small team of dedicated typists who entered the hand-written car descriptions into the standard ad size box (paid by the word!) and then matched the black and white photos of the car to the ad copy.

When you say glue, I'm assuming you actually mean WAX. Using a very light coloured wax offered the same adhesion as glue but allowed you to move the photos and type around, and was invisible to the stat camera.

The wax machine was half the fun - two rollers, one rolling through hot wax, would grab the paper or photo from the back, roll hot wax onto the bottom surface, and deposit it on the other side ready for placement.

The second most fun tool was the stat camera, which allowed you to resize photo or type by physically moving the photo or type from the camera lens, or allow you to put a halftone filter over a black and white photo to allow for single colour printing.

What used to require a team of 40 people - ad layout designers, line workers, typesetters - to handle six weekly magazine could probably be done by one person now.

> When you say glue, I'm assuming you actually mean WAX.

We used to use a rubbery glue called Cow Gum. It too was repositionable and invisible to the camera. I haven't seen it for years, gone the way of Letraset probably, but it was good stuff. The raging solvent-fume headaches after a day of pasting up were not so good though.

Wax was great because it was non-toxic, quick drying and smelled good.

You just had to make sure not to wax things twice because it gunked up the machine!

You've jogged my memory and are correct. I remember a hot wax roller that sat in a reservoir of some kind, and little wax sticks about the size or 3 sticks of chewing gum or so.

I think there was also several kinds of adhesive tapes used for some of the work with different patterns on them as well.

> I wish I could remember the name of the machine, because I've never seen anything else like it

Sounds like a phototypesetter:


(I'm probably going to get some terminology wrong here, but...) My mum worked at a regional daily newspaper. They had 2 or 3 phototypesetters in the corner of their computer room. Each one was about the size of 3 full height racks. They'd print on to photographic paper and were designed to be automatically fed in to a machine (about 1.5 full height racks high) from the same manufacturer that would develop the paper, ready for a Compositor (ie, a surly unionized employee) to cover it with hot wax, slice it with a scalpel and physically lay it out on the page. If the printer crashed, it would have to be rebooted with a series of switches to bootstrap it far enough that it could load it's bootloader (?) from paper tape.

Where she worked decided to use a different manufacturer for the printer and the developer such that it wouldn't automatically feed from one to the other. So the printer dumped it's outputs in to boxes mounted on the wall of a darkroom and would wait for someone to go in and manually feed it in to the developer.

It was glorious.

If anyone ever gets the chance to go on an open-day or tour of the noisy end of a newspaper or large print shop, i'd totally recommend it.

"Sounds like a phototypesetter:"

You are correct. I owned one type - Itek Quadritek (another was Compugraphic). The fonts were essentially plastic disks. At the time (early 80's) each font was $40 (80's) dollars.

I later (about '86 iirc) replaced this with a Linotronic which was driven by not only a green screen front end (as described) but later the first Mac 128k (I think Pmaker 1.0 required 512k mac). The lino was $80,000 in 80's dollars approx iirc.

"Sounds like a phototypesetter:"

Wanted to add that to this day I can still remember the sound of the stepper motors spinning and manipulating the plastic font disks. A really cool sound. (Similar in a way to modem negotiation but lasting much longer obviously).

On those machines for anyone not familiar you actually had to write code (mark point, return point) in order to draw a simple box. Then more code for the actual type. Everything was placed by coordinates.

So when (even) mac paint was released and you could draw a box and put type inside of it the potential seemed quite obviously. It also added greatly to creativity because a designer could work in real time coming up with a design.

Even more fun was the Compugraphic Junior. It didn't use disks, but film strips with the font. Then your had to switch out some gears to change the font size.

Seemed pretty futuristic at the time though.

Some of this sounds familiar. The Berthold unit in your link looks like the same general family of technology...everything is recognizable (the Linotype photo doesn't look familiar in style at all). But from what I remember their setup was maybe a couple steps more modern. Nothing so big to be rack sized.

The green screen bit was your typical 1980s CRT green on black terminal with two 5 1/4" floppy drives vertically mounted next to the screen. The connected rendering monitor was a vertical format...the screen was maybe 17"? And the photoprinter was maybe the size of a large color laser printer today. I think all the reservoirs were semi-internally mounted and I think the width of the printout wasn't even 8 1/2". It was really suited just for blocks of text to be later hand pasted up.

There was no place for the printouts to go, so for a while they just dumped on the floor until somebody nailed an extra empty paper shipping box underneath to catch them.

I don't remember there being any other pieces of hardware, but again, I'm going back through the haze of a few decades here and I didn't personally work on the equipment.

after a little looking

The closest I can find is this http://macpro.freeshell.org/quadritek/Digitek.html

I know that they went through at least 2 or 3 generations of these before switching to digital desktop publishing, I think my memories are mostly of their last system. So the previous ones could have been more like the other ones being described here.

Your comments brought back a lot of memories for me.

I worked as a production artist (someone who takes a designer's idea and turns it into somethng that can actually be produced, hence the name) off and on for 15 years, starting in the early 1970s and ending about 1991. I worked for an ad agency, typesetter, publisher, and finally in the art department of a flexo plate maker. When I first started, we used drafting boards, T-squares, X-acto knives, and rubber cement. Type was of the hot variety - photo type was only for display type, not body copy. In the late 80s I was trained on two SOTA production machines: a DuPont VASTER system (used both vector and raster graphics), and later a Context system that ran on Sun workstations.

The typesetter your parents bought was likely either a Compugraphic or A/M (Adressograph Multigraph) - in the pre-DTP days, they seemed to have most of the market for phototypsetting systems (at least for smaller businesses). I worked with both, and your description would fit either of them.

I guess I would somewhat dispute your assertion that layouts (or paste-ups/mechanicals as they're more properly called) took "forever" or that it was a "very long" process. Using the "Thin and Pretty" layout in the article as an example, the last place I worked we probably would have budgeted an hour for the job - typesetting, mechanical, camera work, and (maybe) printing plates (all done in-house). I don't think that's terribly out of line with the time it would take using DTP. (That doesn't include design time!)

When we were first making the switch from the old manual operation to digital, we often found it took longer to do the work using DTP; it wasn't unusual to hear an artist say "Screw it" and go back to his light table to finish a job. But that was to be expected - unfortunately, a number of my co-workers had little or no computer experience. Over time that ended, although the company did keep two full-time artists on staff who did nearly everything the "old" way until about 2000. The huge advantage of DTP, especially early on, was that changes/additions/corrections were so much easier. And in our case, it was considerably easier to actually find artwork on the computer. All our paste-ups were done on #20 illustration board (for dimensional stability) and they were filed in row after row after row of architectural file cabinets. Unfortunately, no one had ever devised a good filing system, so it would often take 2 or 3 or 4 times as long to find the freakin' art you needed to make a change on as it took to actually make the change...

Thanks for providing some clarification and color to my very hazy brain dump.

I do remember my parent's "climate controlled" filing room and filing system for client artwork and other sundry. It was basically a big storage room filled with racks of old paper shipping boxes and large grey envelopes. The customer order would get written up on triplicate carbon free paper (customer, contact info, type of job, ink, references to artwork, due date etc), one color for the front office, one for the "lookup" catalog (organized by customer name) and one would go on the front of the envelope. All the artwork, disks, negatives, proofs, a few final samples and other assets for the job would go in the envelope and get organized onto a shelf.

Every job generated a new job number that was something like yearmonth-sequential# so a job coming in today might be 201403-38 and jobs folders were roughly organized by the 201403 part.

If a customer came in and said "I want another 10,000 of that job I had you print lat Oct" we'd go to the catalog, lookup their name, find the 201310-# job and go pull it from the shelf to review. It usually took just a few minutes to find something. I think every 7 or 8 years we'd go through and purge old files just because there simply wasn't the physical space.

They'd also use the job # part of the serial, when writing up jobs, to keep a rough estimate on business volume. Since every month the job # started back over at 1, and you knew that you'd on average do 600 jobs a month, you could get a feel if the month was slow or fast just be seeing what the next job id was going to be on a given date.

The job id's were then used in scheduling. For a long time my father drew up the schedules by hand. He had a grid printed up on large format paper, and I remember him coming home every day with a stack of the front office carbon-free copies for the day and drawing up the production schedule on the grid.

I think the hours of the work day were along the top of the grid and the equipment operator was along the side. He used some kind of color coding system (with the full job # and a legend) to trace the job through the work-day as it moved through various pieces of equipment and such, and he'd use some known metrics for how long it'd take to do that kind of job on this or that equipment, and budget it out. There was quite an art to it. Especially for big jobs that spanned multiple days that needed to be kept track of.

It was kind of a Gantt chart, but a little different. And it provided a clear visualization of all the production going on at a glance. He'd make a couple copies and one would go in the production room so any operator could go check it for their next job, or to see if they were on schedule (sometimes they didn't agree and they'd go make their case for a schedule adjustment).

Sometime in the late 90s he finally got a computer and reproduced the entire thing in Excel, which meant he could do all the scheduling at work instead of at home. When they finally sold their company, the new owners were impressed enough with the system that I think they adopted it for their business.

This is all too true. The interesting thing is when companies get into this sort of death spiral you can usually spot it if you look carefully. If you're a competitor with an "800 lb gorilla" which has just entered this territory you have a good shot at killing them if you execute well.

I speculate the reason this sort of technique persists though is that the only people who get hurt here are the lower level employees and the customers. Smart investors make money on the short selling, the top management makes money/reputation on the original upside.

I've always been interested in the particular song and dance that's performed during the board meetings to convince them the CEO's blood squeezing strategy is not what's causing the company to noticeably fail.

All of your history is essentially correct. I owned a company in this business in the early 80's (started it right out of college btw.) We bought the first linotronic ($80k, driven by a Mac running Pagemaker 1.0) in our city so essentially we had the first service bureau you might say. Taking and printing disks (even from a Mac) was a complete mess and generally super aggravating. Side fact: This was all before fax machines were in widespread use for proofing. That in itself was a major leap in getting customer approval.

Thanks for adding any color or detail you can remember. It's very hazy in my memory, I was pretty young and didn't ever work too much in the artwork department. I mostly just made money in the summers doing odd jobs and running errands.

One thing I do remember was the endless difficulty in receiving artwork from customers who produced everything in RGB instead of CMYK and IIR converting it for color separation was a huge PIA.

I also do remember when FAX machines started to get used more. I'd say a good 30-50% of my parent's business relied on Fax. I think they kept a couple of spare machines in a cabinet in case it failed it was so critical. The first few machines printed on thermal paper, and once spamming fax machines caught on, simply became too expensive. They'd come back after a weekend and have an entire spool of paper all over the floor full of advertisements and other junk.

I think once the web and e-mail started to catch on, and inkjet printers became cheap enough, they definitely felt it and business dropped tremendously. There's still no cheaper substitute for printing off a few thousand copies of something that's available for home users. But lots of people just get by with e-mail or some alternative to physical copies.

Also the state of expectations started to catch up to them. You really couldn't own a 1 or 2 color offset print shop anymore. You need to handle 4 color, and do it without a huge cost premium, because everybody wanted full-color brochures and wedding invitations and whatever. Having lots of sunk cost in previous technologies and staff made the cost of switching simply too difficult.

It was a lot of fun though.

I used to deal with this issue of PC PageMaker disks in a Mac & Quark world at a service bureau back in the day. The problem was that PageMaker had unpredictable results when printing to an imagesetter.

I would open the file in Mac PageMaker, delete all the graphics, save the pages as EPS, then rebuild in Quark. Quark had this great Xtension that let you place the output with great precision on the imagesetter film roll.

> I don't have a point, just thought I'd add a little first hand history.

Your anecdote is much better than the article I submitted.

Having gone through all this I definitely appreciated both the article posted as well as the parent comment.

I will add my own anecdote about Quark from that time period. They had a total cocky shitty attitude. This was back at the time when by comparison Apple (iirc) seemed downright friendly to customers. Also users of Quark definitely looked down upon anyone using Pagemaker. Quark (as the posted article mentioned) had a firm lock on the market and was seen as the choice of true artists. (Once again, from my memory).

Not at all! The article dug up some old memories I thought folks might be interested in reading a brain dump of.

actually, i felt the submitted article was the best thing i've read on HN all week. thanks for posting it.

I remember seeing PageMaker around that vintage. It came with a runtime-only version of Windows since everyone was still using DOS.

This provided a lot of insight into an area that I knew absolutely nothing about. Thanks for taking the time to type it all out.

Thanks for taking the time to write that up, very interesting and well-written.

Quark's postmortem comes up a lot, having been in the publishing trenches I see it differently: Quark bet heavily on interactive CD-ROMS as the future, not the internet. They developed cross-compilation from print to interactive CD-ROMs. In a betamax vs. VHS war, the interactivity of CD-ROMs lost to the horrible and graphically disastrous web 0.9beta. Somehow this is forgotten when gurus want to classify every business story according to pet-theories of success and failure. Could you be betting on the wrong future?

Having been around when Quark started to fall behind and then switching to InDesign myself, I don't agree with you. QuarkXPress was crap, and Quark had no interest in modernizing their software. It took forever just to get native OS X support, and the software lacked a lot of InDesign's print publishing features. Desktop publishing has always been a Mac-centric affair, and ignoring Mac users was a pretty poor decision.

Quark tried to milk its customers dry. It did and then they moved on. They never diversified their company, and the one product they made was becoming despised.

Final Draft is a good modern-day version of QuarkXPress.

A quark user could smoke an InDesign user, mintues vs hours. High pressure, high speed design is still not matched by InDesign today. But there is very little high speed design left. Of course, InDesign is wonderful now.

Had it gone that way, they would have been slaughtered by Macromedia's offerings (Authorware and Director). And Macromedia somehow managed to "shock" their tech onto the web when the little disks mostly went away.

This reads a bit like something from


albeit with more focus on the features and less on what led Quark to screw up.

I would emphasize the utter hatred QuarkXPress had for its customers[1], and given Adobe's track record, they were some of the worst in the industry.

1) dongle issues

It was interesting that the article used the word bo{2}b in the section about regular expressions.

I'm not particularly bothered by it, although I recognize that it's not neutral with everybody and thus best avoided. I'm just fascinated since it it's very rare in mainstream publications, especially from the US.

Well it's ArsTechnica, not Disney Channel.

A bit immature but nothing really offensive.

"listen to your users and respect them"


Applications are open for YC Winter 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact