Waffle is a weblog.
The author of Waffle, some guy in Sweden, also occasionally writes stmts.net.

Lately on Waffle

Who is NetNewsWire 4?

NetNewsWire 4.0, the Mac RSS reader, was released just the other day. As someone who was on the NetNewsWire 2.0–3.2 beta team and who has spent at least half an hour and more likely at least one and a half hour in NetNewsWire every day since 2004, I am interested in this. Here is the problem with what’s been going on.

NetNewsWire was originally made and maintained solely by Brent Simmons, although he was able to make it work only by relying on his wife Sheila handling support duties – they both make up Ranchero Software. NetNewsWire was huge both in terms of the app’s capability, feature roster and user base, and Brent on several occasions made large upgrades entirely free.

With Brent at the helm, features were added, considered, reconsidered, reimplemented, cut and wildly discussed. None of it happened in isolation. Brent may not have agreed with all of us on the beta list all the time, but he at least picked our brains. More than once, someone had a huge database which made a redesign of some corner of the app necessary. More than once, ideas were conceived, kicked around and implemented. (Brent and NNW was for a time in the hands of NewsGator, but it didn’t change how the software was developed beyond supporting their sync service such as it for a period existed. This tells me it’s more about the individual.)

NetNewsWire 4 started in Brent’s hands as NetNewsWire Lite. It’s not surprising, considering the organic way things grew in comparison with big advancements coming into being with every OS X version closer to the platform’s birth (and mostly free from the focus to have to poorly reimplement iPad features or metaphors), that the foundation of NNW or the interplay between so many features would eventually gunk up and could use a rethinking. NetNewsWire 3 was one, and this was to be the next. As NNW Lite was mostly finished, NNW was bought by Black Pixel in 2011.

After a long period of silence, they brought out the NetNewsWire 4.0 beta, which was a mildly souped up version of NetNewsWire Lite. After a bit of culture shock seeing so many features I depended on cut outright, I decided to brave it out and see where they’d take it. The place they’d take it was a bunch more betas. One beta was literally released to just extend an expiration timer.

And now, with the final release, they fixed a bunch of bugs that had been shockingly prevalent (like how the article view would stop redrawing except for during scrolling, or how changing from an article to a tab would sometimes not redraw, or how it would just stop reacting much to mouse click events at all) and added Sync. I empathize with them because they’d been working on a sync solution with iCloud Core Data which they just couldn’t get to work well enough and then switched to trying to do Google Reader sync before that product was mothballed to the Internet’s unilateral and universal uproar. Reimplementing sync was, I dare say, 40-60% of all the work Brent ever spent on NetNewsWire. It is time-consuming, hard to get right and has tons of error cases. And it probably requires a good long beta period to settle in.

So what’s my point? Is my point that Black Pixel has been mishandling NetNewsWire? A little bit. Have they been rude to me personally? No, of course not. Their support has been kind and service-oriented. I know I can’t expect the NetNewsWire beta discussion list to be continued by osmosis. NetNewsWire has to be produced by Black Pixel in the process they use and the way they work. Of course they had an internal beta for NNW Sync. But it is nuts to have a public beta process, let it languish terribly for months on end and then not even use it to test out the sync feature but just ship it directly into production. And I don’t expect to be treated like a sage old man just because I was part of a mailing list once. But now I feel like the grumpy man yelling at the TV. I feel like the curmudgeon whining about how things were better in the old days. I don’t want to be that guy, or feel like him.

But I also don’t want one of the five most important apps in my life to fade into something dumbed down. One interpretation of this post’s title is of course “who is NetNewsWire 4, who stands behind NetNewsWire?”. It used to be Brent and Sheila. It now is Black Pixel. Black Pixel has a more Apple-like approach and it’s not the way they work to think aloud, to eagerly solicit feedback, to herd cats and cat-like ideas of what should be for dinner tonight. I can deal with all of this. But I have to see progress.

Because the answer to “who is NetNewsWire for?” used to be “people who want to use an RSS reader that’s stable, full featured, regularly updated and fast to use”. And that’s not the case any longer. For example, NetNewsWire 4.1 could just be the reimplementation of the back/forward buttons for article item selection, such that if I went to an item and then went to another item and since the first one now was marked as read, I have to hunt to find my way back to it, I could just press the back button, and I would happily buy two extra licenses just for giving me some of the sophistication back.

With any luck, this is the end of a painful first implementation of a solid platform, and new features and updates will come more swiftly now. I hope so. I know that with the exception of syncing, which I don’t use, this is the point which I thought they’d have to get to before it was really theirs, and I’m willing to start fresh here and see what they’re up to. But if the next four years are like the previous four years, count me out.

Windows 10

  • I liked the Windows 8 Start screen for what it was. I thought Windows 8 was far too much about the experience of tablet/touch users subsuming everything several hundred million Windows desktop users had committed to muscle memory (or at least the apparently only 6% who know about alt+tab (the guy in that video is also unleashing the fury of Raymond Chen for referring to the “tray”)), but for the select few who had a Windows tablet and enjoyed Metro apps, I’m sure it worked.

    The initial Windows 10 Start menu looked like a great compromise between Windows 7 and Windows 8 as first announced — but the subsequent unexplained redesign dropped almost all of that. The subset of Windows 7 holdouts who want a vertical list of icons and text label items — very likely most of them — will find themselves yelling at live tiles. Or dead tiles, since the “live” aspect is not present for anything except Metro apps of Windows 8 or newer Universal pedigree.

  • With very few exceptions (like the taskbar and the excellent desktop picture) Windows 10 looks like a half-finished thought design-wise. Padding is awkwardly and uncomfortably narrow almost all over the place, and while this may be an effort to appease all the people who went to bits over “wasted space”, it does seem comically obtuse in the hodgepodge of buttonssorry, tiles at the bottom of the Action Center, with the small text edged up in a corner of a huge tap/click target.

  • The control box/icon buttoned window menu dropdown is dead again in Metro apps, meaning I can’t double click the top left to close those windows. (I did retrain. For several glorious months, just before getting my first PowerBook G4, I did go to the top right to find the close button; then I started using OS X where they are in the top left, and then I was back to seeking that corner instead.) Instead I’m activating the back button not once, but twice, winding up god knows where.

  • Oh yes, the back button in the top left (and indeed on the task bar in Tablet mode). Consider the following scenario: going to the Settings app (now finally, apparently, having absorbed a useful subset of the Control Panel’s functionality), finding my way to a setting, clicking a link to go to a related section where a related setting is, and then clicking back.

    The expected result, conditioned not only by common sense but by however many thousands of hours of web browsing is that I’m back at the screen with the link, and can fiddle with the rest of the settings. No, I am instead “back” at the screen from which I would have clicked into the screen that the related setting was found. Note also that the screen I’m referring to is not the section in itself, but the screen containing a sidebar to the left from which I can select sections that then show up to the right.

    To wit: I have a useful button using which I can go up in the hierarchy of settings screens for the current setting (skipping one level in the case of sidebar-equipped screens like the first level). Fair, but don’t call it “back”. Call it, possibly, “up” — anything but “back”. Also, provide an actual history trail. Don’t provide something that looks browser-like, with back arrow buttons and links and mostly plainly and dumbly flow-layouted contents, and then flake out on completing the metaphor.

  • With remarkable and uncharacteristic determination and consistency, Microsoft has taken “Universal apps” and really made them the undeniable “apps that run on Windows” platform. There’s no guessing which platform Microsoft will consider their Windows platform for the future. Where there is guessing is what everyone who’s written Windows desktop apps in C++ with MFC or other toolkits, or .NET with WPF are supposed to be doing. (.NET with Windows Forms is to go cry into a pint and then port to WPF.) It is not impossible that the free Windows 10 upgrade was offered solely to make “transitioning to a platform where people are supposed to write Metro apps under the expectation that more than, absolute tops, a fifth of the Windows user base will be able to run them” a brave and unproven strategy instead of just simply a staggeringly laughable strategy.

    Of the “bridges” to the Universal platform, the run-a-web site-as-an-app solution is brilliant in its simplicity and a keeper, while the bring-your-iOS-and-Android-code-bases solutions are as courageous as they will be Sisyphean efforts (and even if there’s not a lot of established native widgetry consistency to break with, people will not want to use phone apps as desktop apps).

Windows 10 is pretty good in many ways. It is not a mere whitewashing of the Windows 8 platform and endeavor. With it, Microsoft, Windows logo blue in the face, finally exhales after three years, admitting that as much as they need viable tablet/phone building blocks, the Desktop is where their bread is buttered.

In the same way as Windows 7 was the rescue boat that was needed to provide a viable Windows version after Vista, so is Windows 10 when Windows 7 is starting to age faster than XP, due to a world where it is likely to be the only six year old piece of software still in use.

But the new stuff is not new enough. Metro was introduced with a bold vision, applied ruthlessly; while Microsoft has managed to convincingly put the Desktop back in focus, it is being used to run “Metro 2.0” apps without personality, without ken, without oomph, without stretching for something in the distance. It is possible and maybe even probable that the potluck of mobile and desktop UI elements (hamburger menu, anyone?) cobbled together with some sort of ultimate familiarity in mind was a wiser choice than is plainly visible.

Vista tried to rethink the desktop and fail, Windows 7 saved it and Windows 8 and 8.1 went on a “let’s pretend we don’t have nearly a billion users who’d just like shit to keep working” misadventure. 20 years on from Windows 95, the time has come to redefine what a Desktop can be by embracing what it currently is, but by providing new ideas and new tools, by providing something that is as clearly as can be a follow-on to what has come before, which is familiar but better in every way. Windows 10 brings us a version of Metro, gutted of its originality, mixed out to a bland, featureless grey goo. It has extricated Microsoft from the Windows 8 debacle at last, but at the cost of the clean start that the Windows Desktop so desperately needed – especially since the future of the OS X Desktop increasingly seems to be “iPad” (except for when they steal Snap Assist and no one so much as bats an eyelash), and especially since Windows 10 now seems to be going into an “as a service” cadence with small iterations.

I will be installing Windows 10 everywhere I had Windows 7 and everywhere I had Windows 8. And as I do, I will endure parts of it, I will even straight-out no-joke like a lot of it, but I will weep for the reboot that never was, and I will seriously find myself wondering if the fabled Linux Desktop isn’t the last great hope of reinvention after all. (Until I remember that changing the window widgets is considered wildly extravagant, and that most of them consider reinventing K&R-era infrastructure and replacing stupid fucking X11 fool’s errands. We’re doomed. Dooooomed.)


How Microsoft Office’s codebase is being reworked to adapt to the new multi-platform requirements is the most interesting thing you’ll watch today, if you’re a programmer. Contains some nice breadcrumbs about how Mac Office development may be tilted on its head, and some brash admissions that the way forward will not be features that only work on one platform, because it’s hell going forward.

Via: Herb Sutter’s usual C++-will-solve-all-your-problems ditty from this week’s Microsoft developer hullabaloo, which contains a slideware admission that you can’t reach 95% of customers by doing Windows versions, but that you need Windows, OS X, iOS and Android to reach 95% of customers. It’s becoming a cliché to chalk all of this up to Nadella at this point, and it really doesn’t matter either way. The point is that the part of Microsoft that previously made damn sure you wrote to something somehow dependent on the Win32 API is now bending over backwards – relatively speaking – to support other things as well.

« Newer posts · Older posts »