Microsoft insider on why Microsoft sucks (and rocks)



  • Clickbait-y title out of the way, this is actually a pretty interesting series of reddit posts by someone claiming to be a Microsoft programmer.

    I'm quoting it here just in case it gets delete bombed (as it sometimes happens when stuff like this gets popular).


    15 year MS veteran, here.

    The problems here are mostly structural, in the sense that the problems come from the structure of teams, rather than the quality of individual people. Most of the developers at MS are actually quite good; many are top-notch. So why is so much of the product shit? It's because the goals are not set by the engineers. The final decisions are not made by engineers.

    For a long, long time, the "standard" arrangement in MS has been the "triad": Development, Test, and PM (Program Management). The thinking is that people in these three roles should focus on their strengths, and by their strengths combined, should make a great product.

    But here's the thing. It never, ever works that way. Historically, Windows NT was actually run by Dev. Office is (and always has been) run by PM. This is why you see some real technological advances in Windows, such hibernation, super-fast boot times, BitLocker, Hyper-V, ... (I could go on for a very long time on the good engineering done in Windows), but at the same time Windows has no real specific focus. Windows is a platform, so maybe that's OK -- a great deal of the value of Windows is the huge app ecosystem.

    But if you look at Office, it's a feature mill. Office hasn't created any new, interesting technology for well over a decade. Every release adds more eye candy, uses more memory, and is slower. That's because Office is not run by engineers, it's run by PMs. Office has never, ever figured out how to integrate with the web. They've tried, over and over, but every technology that they have tried has been chosen and pushed by PMs, not by engineers or anyone else who understands how people actually want to use Office these days. So, seriously, you could put the absolute best engineers in the world in Office, and you would still get more of the same -- more UI "features", more bloat, more useless stuff.

    For a long time this arrangement "worked", in the sense that people kept buying Office. But now, the world has shifted to the web, and Office really is not relevant in the web. It's relevant in IT, mainly because apps like Excel are still pretty solid at what they do. The core spreadsheet engine in Excel is a masterpiece of engineering, but Excel mostly hasn't changed in the last 10 years. The Ribbon was just a UI change. Nearly everything else has been a UI change, not any kind of advance in the core technology. Office is still file-based, not web-based.

    There have been numerous internal projects which attempted to branch out in new directions. I'll never forget seeing a demo of "NetDocs", more than 10 years ago, at a company meeting. Think Google Docs, long before Google was even a household word. An internal group had basically that working, but it was quickly killed. People here joke that the worst thing you can do is to demo a technology at a company-wide meeting, because it guarantees that someone else will get your project killed. It's not really a joke, it's just a grim reality.

    Things reached their absolute worst under Sinofsky. That man nearly wrecked Windows forever. He wanted to be the next Steve Jobs, so he was an absolute tyrant, who ignored everyone's criticism (everyone said Win8 was going to be disliked, but he ignored them). (Ignoring criticism is fine when you're right, but Sinofsky was wrong!) Thankfully, Sinofsky's reputation tanked when Win8 tanked.

    That man was so caustic that he prevented internal engineering groups from even talking to each other. For almost a year, the Windows Server team had trouble getting access to the Windows source code, because Sinofsky was so obsessed with secrecy for Win8. I'm not exaggerating, I was there, I saw this mess.

    Now that Sinofsky is gone, things have improved dramatically, but the culture is still nowhere near as healthy as it was during the Windows NT / 2000 / XP days. Satya Nadella is definitely setting a good tone, even though his direction is mostly the same as Ballmer's was. People see Nadella as an engineer first, which is a very good thing.


    Can you expand on how you view Microsofts prospects over the next few years?

    Actually, surprisingly good. The last decade has been punishing, and I think the senior leadership team has finally learned from that, to a certain degree. They actually understand some of the profound shifts that this industry is going through, but even better, they are actually responding to those shifts, rather than just (as they had been) ignoring them and hoping everything would stay the same.
    Here's the biggest shift of all: Windows and Office are obviously not going to be the main things paying the bills in the near future. They're not going away, but they are definitely changing. Windows will change the most -- Microsoft desperately wants to be selling Windows devices instead of Windows licenses to OEMs. They want to sell you an Xbox or a WinPhone, not another desktop license, for a variety of (mostly good) reasons.

    I think WinPhone is actually a decent product, and the Nokia acquisition was a good idea. I really want to see some "flagship" phones, though, and MS has been slow in producing those. NOTE: I do not have any kind of insider info on phones, phone releases, etc. I legitimately like WinPhone, and would like to see it succeed, that's all. But the truth is, it has only seen significant adoption in the low-end and middle markets. Which is good -- you have to start somewhere -- but I think WinPhone hasn't broken into the high-end market mainly because of brand perception, not because of the quality of the product itself.

    Office is successfully migrating to the cloud. O365 has been relatively popular, and is competing pretty well. I use it and I like it. I love OneDrive, and I think the teams behind these products are actually doing fairly good work. Mainly because they have a clear customer profile -- they know who they want to sell to, and the customer actually wants the product (unlike with phones).

    Azure is a pretty solid service. But more importantly, it's the kind of thing that only a large company could provide -- AWS, Microsoft, and Google are really the only competitors in the space for providing large, scalable cloud services. It's hard to imagine small companies breaking into a business which requires billions of dollars of capital outlay, just to build the datacenters alone. I think that will play to Microsoft's strengths.

    What has kept you at Microsoft for so long, considering you've been with the company through what is probably the most painful period in its existence?

    Ok, this is going to sound really corny, but here's why. Microsoft made its name by establishing the PC era. More than any other company, the PC era owes its existence to Microsoft. That's not to say that they did everything perfectly, or did it for noble reasons, or that others weren't involved. But personal computing really owes a lot to Microsoft. And to me, the personal part of it really matters -- PCs were a thing that you owned, that you controlled, and this was a huge shift away from the very centralized computing vision of the 1970s.

    For the last 15 years, I've wanted to contribute to that -- to make PCs better for people. I've shipped code in Windows and in other products, and our goal was (honestly) to improve life for our users. I remember I worked on a network feature in Windows 2000, and in the first version of it (long before Windows 2000 shipped), using this feature required that the user go to Control Panel and enter some obscure network configuration info, just so that they could use some voice/video programs (mainly NetMeeting) in a certain situation (crossing firewalls). We were pretty happy with that -- yayyy, NetMeeting can cross firewalls! But we looked at it and said, You know, configuring this is a pain in the ass. So we worked at making the firewall-crossing logic automatic. In the next milestone, the user only had to enter one piece of information. I worked at it more, and found that I could automate 100% of the configuration-detection logic, so that NetMeeting just worked, even when there were firewalls involved.

    The end result was that no one knew that this code even existed. Windows and NetMeeting just worked, seamlessly. There wasn't even a dialog box left, by the time we were done with it. I was really happy with that. No one knew, but one product reviewer commented on the fact that NetMeeting just seamlessly crossed his firewall, and he was surprised with that. That made a couple of us very happy.
    The chance to do that kind of work, to deliver a polished, reliable, useful thing to people, is why so many of us work here. It's not a perfect environment -- none are -- but every time we shipped something good to people, you could see it.

    Which is why I'm kind of sad about the state of personal computing. A lot of personal computing has shifted to devices that are not PCs -- not in form factor, and not in spirit. Most people don't "own" their phones -- the physical phone is subsidized by the carrier, and the code running on the phone is also locked down by the carrier and the phone OS provider. By contrast, the PC is a completely open platform -- you own it, top to bottom. At the other end of the spectrum, so much of people's data is moving into the cloud -- which is also emphatically not an open platform. The cloud is far more centralized than even the mainframe vision of the 1970s! There will only be a handful of cloud providers (AWS, Google, Azure), while in the 1970s things were centralized per organization. We're moving to nearly planet-wide centralization.

    I hope, I deeply hope, that the computing power of mobile devices increases to the point that we see another PC-like shift toward decentralization. When every mobile phone can hold terabytes of storage, when networking links are > 1Gb/s between phones, will we move away from carriers running the show, and move back toward giving users control again?


    Ok! Story time. True story.

    I worked for a guy who at one point was an Office PM. We'll call him "K". He was the second PM to work on the Office Assistant -- you know, Clippy and such. This was back when BillG was still running the company.

    BillG basically did non-stop reviews of feature teams in the company. So your feature would periodically come up for a "BillG review". These were always a big deal. Bill had laser vision, and could see through bullshit in demos and talks. BillG ran the reviews -- you answered questions when asked, etc.

    Anyway. It was time for the Office Assistant to have a BillG review. K, the guy who told me this story, is sitting in Bill's meeting room, along with the dev manager for the Office Assistant, waiting for BillG. BillG walks in, and asks his scheduling assistant what this meeting is about, and the scheduling assistant tells him.

    BillG says, "Oh, that fucking character?" K's heart sank. This was his feature, what he spent every working hour on, and BillG had just called it "that fucking character". ...

    So after the meeting K and the dev manager had a good laugh. They decided to use "TFC" (That Fucking Character) as an internal acronym for the project from that point on. So header files were called <tfc.h>, #define names were called TFC_FOO, etc.


    Wow, thank you for writing this. I've been devouring every Windows 8 developer blog post and Sinofsky speech to try and figure out what went wrong, internally. Stories of internal strife among smart people fascinate me. It makes me wonder how the heck Windows 7 came out as well as it did.

    Here's why Win7 was so successful: It was what Vista should have been, to begin with, and fixing Vista's problems was an obvious right thing to do.

    Vista had a few serious problems, but at the same time, Vista also brought a lot of value to customers. There were many, many new developments in Vista -- so many that I have a hard time listing them all.

    • New pluggable filesystem architecture. This fixed a lot of the problems that anti-virus packages ran into, and it also made it possible to do things like perfect live backups, without taking a system offline.

    • A completely new display driver model (WDDM). The marketing name for this was DirectX 10, but that doesn't tell you how much work went into the driver model. The driver model is utterly different than the DX9 model -- it was far cleaner, much more efficient, etc.

    • A new "composited" desktop, based on DX10. The "Aero" (glass) look was just icing on that cake -- the underlying compositor is much more important than Aero.

    • Lots of work on IPv6

    • Massive improvements in networking, such as Wifi

    • Massive improvements in storage, such as iSCSI, etc.

    • 64-bit kernel, supporting both 32-bit user space and 64-bit user-space

    Like any really big new release, Vista had its share of problems. In retrospect, it's amazing that Vista worked as well as it did, when you look at the huge jump in tech from XP to Vista.

    What really killed Vista, though, was two things: 1) the OEMs weren't ready with enough device drivers, and 2) memory usage increased enough that perf suffered on laptops.

    So, with Win7, the fix was obvious: 1) Get solid device driver support, 2) Drive memory usage down, 3) Fix bugs. That's exactly what the team did, and the end result was a product that really shined. Win7 is what Vista should have been on day one, but Vista was already so late that they felt compelled to ship it before it was really ready. (And that -- Vista being literally years behind schedule -- is a whole different story!)


    One thing I've wanted to ask for several years... Linux kernel with all the required drivers has size of about 5 MB, or at least about that order of magnitude. Why is my c:\Windows folder (Win 7 Professional x64, legal, activated, up to date, everything proper) about 50 GB - 10 000 times bigger? I know there's (a lot) more to Windows than just kernel and drivers, but still that seems unreasonably huge. What do you need 30 000 folders and 130 000 files for?

    Several reasons.

    First, c:\windows contains the entire installation image. This contains every single device driver that is shipped with Windows, and every optional feature of Windows itself. Drivers for every printer, every NIC, every sound card, etc. -- everything that Windows supports "out of the box". I think this started with Vista or Win7, I forget which. The OEMs wanted this because Joe User could never find the installation CD when he connected some new USB device. Since hard drives were basically infinitely huge, it was (on the whole) a rational decision.

    Second, on a 64-bit system, c:\windows contains binaries for both 32-bit Windows and 64-bit Windows. The 32-bit binaries are stored at c:\windows\syswow64. For example, if you really, really need to run the 32-bit version of Explorer.exe (because your customer has a 32-bit shell extension that must work in order for you to get paid), then -- there you go. It's the exact same as as running apt-get install multilibs or whatever on Debian. Most Windows binaries (but not all) are built for both 32-bit and 64-bit, and are installed in their respective subdirectories (system32 for 64-bit, syswow64 for 32-bit (yes I know that seems backwards)).

    Third, keep in mind that there is some virtualization going on here. If you right-click on c:\windows and look at the "total size" figure, the tool which computes the total size is going to count some files twice. This is because they are hard-linked to each other. This is true for most of your active device drivers. For example, if you run dir winusb.sys /s /b from c:\windows, you'll see something like this (taken from my dev box) :

    C:\Windows\System32\drivers\winusb.sys
    
    C:\Windows\System32\DriverStore\FileRepository\winusb.inf_amd64_ed914f5101cd8e33\winusb.sys
    
    C:\Windows\WinSxS\amd64_winusb.inf_31bf3856ad364e35_6.3.9600.16384_none_e8b34490e99860c5\winusb.sys
    

    These are all hardlinks of the same physical file. When the Device Manager installs a device driver, it checks for the driver in System32\DriverStore. If it finds the driver there, it just hard-links the binaries into their final position. So if you run the command fsutil hardlink list C:\Windows\System32\drivers\winusb.sys, you'll see this:

    ] fsutil hardlink list C:\Windows\System32\drivers\winusb.sys
    \Windows\System32\drivers\winusb.sys
    \Windows\WinSxS\amd64_winusb.inf_31bf3856ad364e35_6.3.9600.16384_none_e8b34490e99860c5\winusb.sys
    \Windows\System32\DriverStore\FileRepository\winusb.inf_amd64_ed914f5101cd8e33\winusb.sys
    

    This file (winusb.sys) will usually be counted three times if you right-click on c:\windows and look at the file size. That's just because the directory property page code was not updated when hard links were added to Windows, and most users never deal with hard links.

    You can verify this yourself by creating an empty directory, copying a file into it (let's say foo.zip), and then doing this:

    fsutil hardlink create foo-1.zip foo.zip
    fsutil hardlink create foo-2.zip foo.zip
    fsutil hardlink create foo-3.zip foo.zip
    

    Now right-click on this directory in Explorer and check its size. It will show 4x the real size of the file, even though there is only a single file.

    Another reason is SxS -- "Side by Side". Windows contains many different versions of some libraries / packages, such as the C runtime library. This is similar to having /usr/lib/libc.so.4.0, libc.so.5.0, etc.


    This implies that you think engineers would do a better job of understanding user needs that PMs do, despite that being the single core competency a PM is supposed to have. So, either Microsoft PMs are all so weak that someone whose skillset is an entirely different area are better than they are, or it's harder than it looks. I'm going with the latter.

    Seems to me that an engineering-led organization would give you open source projects or Google, neither of which is particularly good at producing user-facing awesomeness.

    It's more like this. A good PM is someone who understands the engineering deeply enough that they can make informed decisions about what will work for users. A bad PM is someone who says "build me a Star Trek transporter", without knowing that "Star Trek" was a TV show, not a tech demo.

    Let me give you an example. Why did Chrome become so popular, so quickly?

    What feature did it have that Firefox didn't? The #1 feature was performance.

    Chrome is fast. No Office PM ever tried to sell a faster Word. Obviously -- every release has gotten slower. But I still run Office 2003 on my home machine (which is a monstrous machine). Why? Because it's fucking fast, and because Office 2013 doesn't do anything (for me) that Office 2003 doesn't do.

    To an engineer, performance is a feature. To a PM, performance is invisible. The same is true for reliability.

    So, yes. I think engineers would do a better job than PMs, any PMs, on certain aspects of design and prioritization.



  • The "Triad" explanation sounds like a just so story told by worker bees with a poor view of what's really going on. Not that there isn't some truth to it.


  • Discourse touched me in a no-no place

    @cartman82 said:

    This implies that you think engineers would do a better job of understanding user needs that PMs do, despite that being the single core competency a PM is supposed to have.

    That cracks me up. The admittedly small amount of experience I have with PMs suggests that their primary skill is creating endless meetings and viewgraphs. I understand user needs better than just about every PM I've ever worked with (and again, that's a small set, I admit.)



  • @cartman82 said:

    understanding user needs that PMs do, despite that being the single core competency a PM is supposed to have.

    This is a different sort of PM function than I familiar with. Admittedly, I work in hardware, not IT, but IME, understanding user needs is driven by marketing through system architects. From where I sit, the main function of PMs is driving (arbitrary, unrealistic) schedules.


  • Discourse touched me in a no-no place

    @HardwareGeek said:

    From where I sit, the main function of PMs is driving (arbitrary, unrealistic) schedules.

    Oh, they have uses. My team once had a meeting with a vendor, who dropped the bombshell on us that they wanted to do data exchange via email instead of something sane like FTP[1]. The PM whipped his hand out and hit the Polycom's mute button before the rest of us erupted.

    [1] or SFTP, etc. FTP was just a starting point for discussion.



  • @FrostCat said:

    Oh, they have uses.

    I don't deny that. Around here, quarterly bonuses depend on whether your project is behind, on or ahead of schedule. In theory, it's good to be ahead of schedule, but beyond a certain point, the bonus money goes into a pot to be paid at the end of the project, rather than the current quarter, and there are a lot of things that can go wrong and make that money evaporate.

    The good (from the employees' POV) PMs are good at manipulating the schedules to get as close as possible to the sweet spot of the maximum immediate payout (assuming you're on or ahead of schedule) and keeping any ahead-of-schedule-ness as a hedge against things going wrong next quarter, rather than turning it into money that you may never see.


  • Discourse touched me in a no-no place

    @HardwareGeek said:

    The good (from the employees' POV) PMs are good at manipulating the schedules to get as close as possible to the sweet spot of the maximum immediate payout (assuming you're on or ahead of schedule) and keeping any ahead-of-schedule-ness as a hedge against things going wrong next quarter, rather than turning it into money that you may never see.

    Ah, gamification. (that should be read in the style of Homer Simpson: "ah, donuts, is there anything they can't do?") and not, "oh, gamification, got it."



  • A little, but what they can game is pretty limited. Schedule changes have to be approved by approximately 17000 levels of management, and where a project is WRT the schedule is determined by the time-and-work reports the worker bees have been filing all quarter. Our whole PM apparatus is a cluster-WTF (but that's a WTF for another day (after I don't work here any more)), but those few who have a good understanding of how it works can gently nudge it in ways that make the worker bees (relatively) happy.


  • Grade A Premium Asshole

    @HardwareGeek said:

    The good (from the employees' POV) PMs are good at manipulating the schedules to get as close as possible to the sweet spot of the maximum immediate payout (assuming you're on or ahead of schedule) and keeping any ahead-of-schedule-ness as a hedge against things going wrong next quarter, rather than turning it into money that you may never see.

    I used to always do the same thing. Less to do with bonuses though. Our bonuses were based upon profitability, period. I did it because my projects were always ahead of schedule in the beginning and middle, then they would move the goal posts and then act like we were behind schedule later on.

    So I just kept the schedule looking like I wanted it, to keep some of the pressure off, and keep the goal posts where they should be...



  • @Polygeekery said:

    Less to do with bonuses though. Our bonuses were based upon profitability, period.

    Our bonuses are definitely a cluster-WTF for another day.


  • Grade A Premium Asshole

    @HardwareGeek said:

    Our bonuses are definitely a cluster-WTF for another day.

    I always took a lower base salary in order to get a higher percentage of profits per project. It always served me well in regards to annual pay, plus there was a lower effective tax rate on bonuses. Then 2007-2008 happened...and my annual pay went to shit because the bonuses were gone. Well over half of my pay was bonus at the time.

    That is my bonus-related WTF... Mostly my own doing, but it still sucked at the time.



  • I was kind of surprised to see the stevesi-bashing, especially from someone claims to value good project management. Sinofsky always seemed to me (not that I'd know) like someone who really knew how to make sure the actual job that management is supposed to do gets done, and whose enemies mostly consisted of the incompetent middle managers that his push for excellence in management displaced. I went to google to see if could find anyone else talking shit about him, and this post seemed to be the Poe's law-level epitome:

    The changes were in theory supposed to be better for the leaf contributor (now you're only X steps from CEO!), but in reality, having opportunities for growth is probably much more important than having two fewer people between you and the top.

    My current hypothesis is that OP was not in a division under Sinofsky, and that the bitterness reflects the knock-on effects that stevesi's ‘triad’ doctrine may have had on other parts of the company, as per this other post:

    It had the positive side effect of preventing runaway PM orgs, but had the negative side effect of encumbering some well balanced teams. I assumed that it was an interim strategy to get a handle on organizational complexity. However, other orgs around the company started being "Sinofski-ized" without any real understanding of what that meant.



  • @cartman82 said:

    Now that Sinofsky is gone, things have improved dramatically, but the culture is still nowhere near as healthy as it was during the Windows NT / 2000 / XP days.

    Funny how even a guy working in MS for the past 15 years doesn't mention ME :wink:


  • :belt_onion:

    @FrostCat said:

    That cracks me up. The admittedly small amount of experience I have with PMs suggests that their primary skill is creating endless meetings and viewgraphs. I understand user needs better than just about every PM I've ever worked with (and again, that's a small set, I admit.)

    Are you talking about Project Managers or Program Managers ? (And yes, that's Joel on Software)

    Program managers are not supposed to manage developers, though they might call for a meeting to discuss features.



  • You've also got PM as in Product Managers.



  • @Buddy said:

    I was kind of surprised to see the stevesi-bashing, especially from someone claims to value good project management. Sinofsky always seemed to me (not that I'd know) like someone who really knew how to make sure the actual job that management is supposed to do gets done, and whose enemies mostly consisted of the incompetent middle managers that his push for excellence in management displaced.

    Results speak for themselves. He bet the farm with Windows 8. And failed. His style of management doesn't survive failure.



  • @FrostCat said:

    That cracks me up. The admittedly small amount of experience I have with PMs suggests that their primary skill is creating endless meetings and viewgraphs. I understand user needs better than just about every PM I've ever worked with (and again, that's a small set, I admit.)

    But engineers' primary skill is creating complicated features that are kind of cool but not very useful. Or re-engineering everything into shit.


  • Discourse touched me in a no-no place

    @boomzilla said:

    But engineers' primary skill is creating complicated features that are kind of cool but not very useful. Or re-engineering everything into shit.

    Well, yes, but I don't think anyone pretends otherwise, at least, they don't to the extent PMs do.

    As for the question of which PM I was talking about, it's whichever one that uses the PMP certification.


  • mod

    Project Manager then. The core competencies for that are more along the lines of being able to organize people efficiently and adjust deadlines and effort levels to match the changing conditions.



  • @FrostCat said:

    Well, yes, but I don't think anyone pretends otherwise, at least, they don't to the extent PMs do.

    I'm just sayin', engineer directed project is as much a project dystopia as PM led, for the same reason (inability to see the big picture and how to get from here to there) but with typically different results.

    @FrostCat said:

    As for the question of which PM I was talking about, it's whichever one that uses the PMP certification.

    As with anything, this comes down to the person.


  • Discourse touched me in a no-no place

    @Yamikuronue said:

    Project Manager then. The core competencies for that are more along the lines of being able to organize people efficiently and adjust deadlines and effort levels to match the changing conditions.

    I think you missed a "supposedly" in there somewhere, and I think the original crack still applies. :) I told you before about the PM I had that used to put "% complete" and "% not complete" charts in his weekly status reports.


  • Discourse touched me in a no-no place

    @boomzilla said:

    I'm just sayin', engineer directed project is as much a project dystopia as PM led, for the same reason (inability to see the big picture and how to get from here to there) but with typically different results.

    I'm not disagreeing.


  • Discourse touched me in a no-no place

    @boomzilla said:

    As with anything, this comes down to the person.

    Sure, but as the cupcake fairy pointed out, the thing's for Project Managers, not the other kind. I could get an MFA in puppetry but it wouldn't be relevant to my current job.



  • @FrostCat said:

    Sure, but as the cupcake fairy pointed out, the thing's for Project Managers, not the other kind.

    Eh, whatever. The two are often the same, IME.


  • mod

    @FrostCat said:

    I think you missed a "supposedly" in there somewhere

    Just because the job description says you're good at X doesn't mean every individual person with that job is good at it. It just means being bad at that makes you shit at your job.


  • Discourse touched me in a no-no place

    @boomzilla said:

    The two are often the same, IME.

    Maybe, but in some cases, the distinction (theoretically) matters. Like when someone's simultaneously the king of two different countries. The King of Germany can't order English people around, unless he puts on his "King of England" hat.



  • Now I have no idea what we're talking about.


  • Discourse touched me in a no-no place

    @boomzilla said:

    Now I have no idea what we're talking about.

    You could go back and take notes. Or I could change the subject: "person with bad avatar forgets medicine, wanders about confusedly".



  • I've had more than one meeting where I've mimed taking one hat off and putting another on, to mark switching roles - budgeting to end-user advocate type thing.
    Possibly because I'm weird, but I'm fine with being odd. Odd is good.

    Also some contexts where there were physical hats in various colours, as it's a great way to spot the "Guy in charge of X" from across the building site.

    Men Without Hats is something else entirely.



  • My experience has been walking up to my desk and being asked if it is done yet. Or "We must have zero bug count".

    I've had much better experience working directly with BAs or a Dev Manager and negotiating deliverables based on "That will take me a week, this will do 80% of what they want and take a day".



  • We were talking about who leads a program or a project or something. These terms are very ambiguous and often interchangeable. Then you apparently analogized a program/project to a country and started talking about multiple ones, and I'm not really sure why you did that or what I'm supposed to get out of it.



  • @lucas said:

    "We must have zero bug count".

    "Close your eyes."



  • The expression on my face when I heard that was probably something like this:


  • mod

    We have BAs acting as Project Managers and it's worked out TERRIBLE for us. My last job had people with actual PM training, and the absence is quite keenly felt here. So I repeat my implication: your PMs are awful at their jobs.



  • Well, we ended up telling them how to do sprints and kanban so we made our lives easy. The result was that we had a sensible release cycle.

    They had got "results" in the past and weren't going anywhere.

    Note when I say BAs we had two and they could code (a bit) and they could project manage (a bit) so it worked out well.


  • Discourse touched me in a no-no place

    @boomzilla said:

    Then you apparently analogized a program/project to a country and started talking about multiple ones, and I'm not really sure why you did that or what I'm supposed to get out of it.

    Do you ever feel close proximity to blakeyrat causes affected confusion to bleed through?

    I dislike ambiguity most of the time (except, obviously, in wordplay). Having said that, the "king" analogy was not responsive to what I was trying to say, except that I would consider objecting to a program manager thinking a project management certification was necessarily useful. Depends on how close the two jobs are, I guess.

    Pretend I never said that, and just use the MFA example instead.



  • @FrostCat said:

    Depends on how close the two jobs are, I guess.

    Yes. In my company, you might guess from what I've said, the two seem pretty much the same thing. I'm sure some middle manager somewhere would disagree with me, but that's my worker bee view of the matter. So without a concrete example, it doesn't make much sense, and being the manager of multiple things instead of one doesn't seem like what you were really going for.

    Let's say that one was being King and the other was being the Admiral of the Navy. I suspect that gets closer to your intent. But I think I was more focused on the PM (of whatever you like to put in there for the P) vs engineer, which contributed to my getting lost.

    @FrostCat said:

    Do you ever feel close proximity to blakeyrat causes affected confusion to bleed through?

    I think that sometimes I might not think something through as much as I otherwise would, since I've gotten used to blakey's nonsense.


  • :belt_onion:

    @lightsoff said:

    Men Without Hats is something else entirely.

    Interesting video, because everything is better with midgets.


    Filed under: Frigging midgets, how do they work?


  • Discourse touched me in a no-no place

    @cartman82 said:

    To an engineer, performance is a feature.

    Strictly, “goes fast” is a feature to a user. “Performance” is just an approximation of that (since most people don't measure “goes fast” in quite the same way as computers). Not prioritising either is sometimes necessary (getting the right answer also matters ;)) but always not a great state of affairs.



  • That's a simplistic way of looking at things. Sinofsky turned the windows division around after Vista, and released a great product in W7, then took Windows in a bold new direction that it really needed to go in, before voluntarily stepping aside to make room for the next leader. And whatever the fogeys say, the Modern UI was absolutely necessary for the long-term success of windows, and the sooner they got it done, the better. Kids are growing up touching screens and wondering why they don't work, and if Microsoft waited until after those kids had bought their first Mac NeoClassic before they released a unified UI, it would have been too late.

    I don't doubt that W10 is going to feature some marvelous engineering, and to give the customers exactly what they want, become the next XP, and all that, but I just think that you should at least acknowledge that it was Sinofsky who left it in the exact position where that was possible. And comparing stevesi to Jobs is apples to oranges: if he was Steve Jobs, the product wouldn't go inevitably and unsalvageably to shit until after he left.



  • I don't know if Sinofsky was the genius behind Windows 7, but there's no denying Windows 8 was a failure.

    Kids are still rushing over to Androids and iPads and Macs. "Modern" interface did nothing to turn the tide.

    Microsoft wanted to switch from a software maker into a device maker, like Apple. They tried to leverage their Windows user base into the mobile/devices market. And Sinofsky was their man. He was given full power to run the ship as he saw fit. He busted balls left and right. Made lots of enemies. He bet everything on Windows 8. If it succeeded, he would have become the CEO of Microsoft within 5 years. Maybe seen as the next Bill Gates. The savior of the company.

    But it didn't. It bombed. Nothing to do with its quality or engineering. It just failed to do what MS needed it to do. So Satya became the next CEO and Sinofsky got his ass kicked to the curb. Not "gracefully withdrew with smiles and thankyous and roses". Fired.

    Look at W10 tech demo. All the attention is in the "traditional" desktop. Metro isn't exactly the new "widgets", but you don't see much love given there. Seems like Microsoft isn't continuing on with Sinofsky's W8 "wedge" strategy. Instead, they have pivoted and are running back to their base - corporations and the PC worker bees crowd.

    They'll obviously try to have a safe "hit" with W10. Keep the lights on and buy some time, so they can plan their next attack at the mobile market. Who knows, they might even try another push with Sinofsky's dual desktop/tablet OS idea. But make no mistake, it will basically be like starting from scratch in terms of the market share.

    Because Windows 8 failed.



  • @cartman82 said:

    Windows 8 failed.

    ... and took Windows 9 with it!

    More seriously, the more cold water that gets poured over MS's app store, the better IMO. Give me the free-for-all of the desktop and the open internet any day.


  • sockdevs

    i'm still thinking it was appcompatibility from windows 9x that killed windows 9 more than the "failure" of windows 8.

    hmmmm failure or "failure"?

    i'll put the scare quotes in and if they're wrong we cna pretend i didn't


  • Discourse touched me in a no-no place

    @Polygeekery said:

    I used to always do the same thing. Less to do with bonuses though. Our bonuses were based upon profitability, period. I did it because my projects were always ahead of schedule in the beginning and middle, then they would move the goal posts and then act like we were behind schedule later on.

    So I just kept the schedule looking like I wanted it, to keep some of the pressure off, and keep the goal posts where they should be...

    My last major crash project was very politically important. My VP's career depended on our success (because his other team had fucked up the predecessor project hardcore). We had the msot of the right team to deliver the product exactly on time (one fifth the time his original team had), and on budget (1/15this his other team's): An excellent Business PM, an excellent Tech PM, an excellent architect (me), and a handful of excellent developers. We really needed a SysAdmin, one more developer and a few testers. But we could only hit that if nobody fucked with our methodology, goalposts, meeting schedules, etc..

    Because of the amount the VP had riding on it, we were told to use his other team's QA staff (Outsourced to Kerbleckistan. And after one meeting we discovered they had never tested software before - they're a document QA team. Naturally we mostly skipped testing.), we were told to manage the project according to his other team's established procedures, produce documents exactly like his other team (down to using the same irrelevant templates), track metrics exactly like his other team, constant status update meetings, etc.

    Because, you know, don't mess with success.

    We dumped our SA requirement and and brought in a PM whose literal full time job was to translate what we were actually doing (a sort of adhoc scrummy agile with a loose waterfallish spec at the very beginning, using Jira for tracking, bugs, etc., full continuous integration all the way up into production) into what the VP expected (pure waterfail, MS Project for tracking, Google Spreadsheets for bugs, manual, documented deployments of monthly builds, do not touch the production systems until go-live is approved). This guy built a goddamn potemkin project and never once broke the illusion. To this day there's no indication that we did what we did.

    Unfortunately for my sanity, we actually needed an SA and one more developer and I ended up doing four full time jobs (architect, SA, developer, tester).



  • It sounds like that VP is a separated-at-birth twin for my (thankfully ex) boss. I'm only sad that said ex-boss is still in a different division in my company, so I still have to interact with him occasionally and I get to see the train wreck that is the other division on a daily basis. Hard to balance commiseration with laughing at them b/c I don't have to suffer any more...




  • Discourse touched me in a no-no place

    Cool.





  • Ok, you typed a lot of words, and I'm feeling like kind of a jerk for not writing a serious response, but your understanding of events is so different than mine I hardly even know where to begin.

    Ok, how about: How does the fact that Sinofsky resigned less than one month after the release of 8, while it was still too soon to tell if it would be success or failure, possibly match what you've written?

    @cartman82 said:

    Sinofsky got his ass kicked to the curb. Not "gracefully withdrew with smiles and thankyous and roses". Fired.



  • @Buddy said:

    Sinofsky

    Sinofsky's resignation was just before Ballmer threw in the towel wasn't it? When there was a general changing-of-the-guard already in progress anyway, when e.g. Don Mattrick left, and some other guys whose names are escaping me right now...


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.