Tralfaz

Author: geoffand

  • If I hear this one more time I am going to lose it.

    Those following the thread have seen my comments about a web project going poorly.

    End root cause is that we went with a vendor which we were “comfortable” with, and didn’t figure out up front if he had the chops to deliver.

    I want to pull back, complete a formal proposal (I had started this then I was told that the project was already 1/2 done), pick 3-4 local Web shops (I am in Phoenix, so there are lots here) to bid on it and then pick someone who has the chops and the skills to succeed at a price we can afford.

    The answer to that was astonishing. 

    We don’t want to do that. Getting a new vendor into our system is a painful process, and will take far longer than the whole project.

    So, instead of using the right vendor, even if that requires getting them into our ERP system, we choose to use underqualified hacks, because it is too hard to find new? 

    Really?

  • Just once in my career, I would like a project to be finished on time

    and on budget.

    Sigh.

    There are always some extenuating circumstances:

    • A key component was more difficult to work through.
    • Some circuit design was wrong the first time around (and the second, and the third).
    • Regardless of how many times the design was reviewed, a connector was wired backwards on a circuit board.

    Or commonly, software is delayed by not having hardware to test and develop against. And vice versus.

    Or the software was a lot more difficult than anticipated.

    The funny thing is, even with professional, certified project managers handling the threads, using the best practices for estimating time/effort/resources, projects are late/late/late.

    As a product manager, I have my own “Kentucky Windage” that I use to “adjust” my personal expectations.  And it is invariably way off.

    Can’t we do better?

  • Why I am glad I am not a Web developer

    I am working with my Marketing team to get a user community site built.  It is a pretty straightforward project that will allow our worldwide user base to connect, share tips, scripts and other related items in the use of our product.

    To spec this I did a lot of up-front work. Mockups. Detailed workflow analyses. User stories and the like.  

    Of course, the project has turned south.

    Our marketing person used a local source for the work.  We had used him before, but for nothing this complex (that allows people to register and to share documents).  

    Of course the local resource uses his trusty CMS, Joomla! for this project.  I was a bit skeptical.  It is a bit heavy weight for this project, but it is not a bad back end.  (for the record, I use Joomla! for both my personal site and for a non-profit that I volunteer as webmaster at.  So I do know enough to be dangerous).

    We are 4 days from the initial roll out, and key functionality is not there.  Apparently the plugin he is using is not cooperating.  And the author is being non-responsive. And I am getting very very nervous.

    Several things that went wrong:

    1. My marketing manager, a dear friend, is old school.  So to get this scoped, she “helpfully” took my mockups, and my flow chart of the workflow and converted it into a plain text description. “This is how we always do web projects. They do it better this way…”  Sigh.
    2. The Marketing manager was the primary interface for the first month of this project.  She was the broker between me and the developer.  What I didn’t know was that she was translating all my carefully visually described changes into bullets in an email. Much of the context I was trying to convey got stripped out.
    3. Choosing poor plugins for the desired functionality.  I have used several plugins to extend Joomla! It is a jungle out there, and until you “join” a club for the plugin, you really get no support or documentation.  So you have to gamble your $35 or $45 to see if it works for your purpose.  It appears that our web developer has chosen poorly, and even after the purchase of support, the plugin author is impossible to reach.
    4. When I finally got to talk to the developer directly (I am the ultimate “customer” after all) I was able to clear up a lot of the little issues immediately.  But then I got chastised by my marketing manager for “nit picking”. It isn’t nit picking if your customer thinks it is important. I am ultimately paying for this project, I have a vision that I spent a lot of time articulating, and we are now 9 days late on the working prototype.  

    The project still limps along, I have an internal launch in 4 days, and about 80% of the functionality is still not working. 

    I am not a satisfied customer. I am glad I am not taking the verbal abuse of the marketing manager (the developer is), but I can feel his pain.

  • Web project part deux

    If you have been following the saga, I have been working on a new, groovy and cool web community for my products.  Something pretty simple to start to build a community of practice.

    I have complained in the past about the contractor we are using, who seems to do about 1/2 what I ask for, and then struggles with the big functionality components.  

    I provided mockups, detailed use cases, “stories” for the key functions.  I have provided pixel perfect graphics and expansions on what I want, scratch that, NEED this community to look like and behave like.

    Today, I got dressed down by our marcom person.  I am being too picky, and that these are merely stylistic issues.  Essentially that I need to back off on what I want.

    I am not asking for miracles. I am not asking for Google like maturity. However I do know what I want, and I know how I want it to look. So, if I bitch that the graphic used for a button gets blurry because it doesn’t scale well when you increase the screen “magnification” it fucking matters to me. If I say that I don’t want text in the menu bars to have a shadow effect, it means I don’t want text in the menubar to have shadowing effects.

    Am I being too demanding? I am the customer for once, and I expect it to be to my satisfaction. (of course, there is one HUGE piece of functionality that is almost 2 weeks late, without which, I can’t launch on 11/26) Grrrrrr.

  • Stone age thinking in the internet age

    I am a huge fan of accessibility, and convenience for my customers.  In my new role, we have a product that is a little long in tooth, and the software is a bit dated as well.  Working to fix that, but one hangup is our absolutely shitty help system.  

    It is web based, and installed on the PC that the software is installed on.  It uses some POS ActiveX control, so it ONLY works with IE (and not IE9).  

    Even when you get the right browser, and configuration, it is a lousy format, hard to navigate, and the search/index is terrible.

    I want to go to a completely online, standards compliant interface, and always up to date.

    But our marketing team, who controls the manuals and the help system generation/maintenance is paranoid that making the help information accessible to all would lead our competitors to find ways to attack us. Words fail me.

    F’ me.  You can’t make this shit up.  How backwards is that.  

    Swimming upstream here.

  • How come really smart people are so dumb sometimes?

    Just had an ambush call.  (that is when Sales invites you on a call, but “forgets” to clue you in on what is to be discussed). Needless to say, some pretty deep hip waders were needed.

    The whole premise was that we failed to accomplish some performance goals in a demo.  We make high end, scientific instruments.  Part of the analyses we do, required you to “find” a region of interest.  We, our competitors, and indeed all products on the market like ours uses a similar design. A video camera, a microscope objective/or telescope, and a real time window on the UI to see the sample/instrument.  

    The problem is that they are looking at SRAM cells, that are 45nm in dimension.  And they kept harping on the “magnification”. If only we made the image look bigger, we would see the features.

    Uh, no.  The classic difference between resolution and magnification.  A rule of thumb is that the limit of resolution is proportional to wavelength/(2 * NA).  There is a constant, but it really can be ignored or assumed to be 1.  If you have broad spectrum white light, your central wavelength is 540nm (a “green” color), and you use that to calculate resolution.

    For a super high resolution system your NA can be as high as 0.95 in air (> 1 if you can do oil immersion).  But since we have lots of hardware in the way, we need a much longer working distance.  50 or so millimeters of WD.  The best commercially available optics at this range will give a NA of 0.15 or so.  Thus, we become diffraction limited at about 1.8um.  Since they want to find features that are 0.045um in dimension, the resolution limit is going to be equivalent to 40 cells.  That means that there is no hope to see the features they want.

    Of course, they didn’t understand this, and kept repeating the “more magnification” mantra. FML.

     

    *NA = Numerical Aperture – a measure of the light gathering capability of an optical system.

  • Phew, things to be thankful for…

    Long time Mac user here.  One utility that I just can’t live without is Alsoft’s Disk Warrior.  I have Drive Genius and MacTools Pro, but neither one “just works” as well as Disk Warrior.

    A month or so ago, my wife’s iMac was behaving wonky, so I grabbed my copy of DW and off we go.  When it is done, she dutifully removed it from her computer, put the boot DVD in the case and set it on her desk.

    Later that afternoon, Tate, our “puppy-like” greyhound grabbed it and ate it.  Grrrr.  Thought I wouldhave to go groveling to Alsoft for a replacement.  But it turns out I used the down rev version, and my version 4.4 is still pristine.  Woo hoo!

    Good thing too, because I need to clean up my disks on my Mac Book Pro.

  • A fucking shame – Boss 302 Mustang

    A wicked american piece of heavy metal, a real muscle car, the Mustang 302 Boss.  A bitchen car from 1969 and 1970 has been re-introduced.  A worthy successor.  However, it is a fucking crime against humanity to defenestrate such an awesome machine with a pussy automatic transmission, I saw one today.  I hung my head in shame.  It wasn’t even a chick driving it.

  • Sugar coating: Sometimes you have to tell your engineers their software SUCKS

    You get to a point where you can no longer hold back.  It is like that bad uncle with the Comb-over who is not fooling anyone.  

    Your software sucks.  It is a shitty UI, it looks like technicolor vomit erupted on the screen.

    It is so unusable, that people prefer to go to the engineering screen to get a better handle on system control.

    It crashes. A lot.  It takes out Windows 7, which is hard.

    Please don’t hate me for telling you the truth.  Let’s fix this.  

  • Market Analysis Oddities – Pulling my hair out

    I am doing a market analysis to help us decide where we want to spend our next development dollars on.

    To accomplish this, I really need to get a good idea of their revenue.

    Top down, I took the bible, the accepted research.  I also took the number uttered during the Q4 2011 analysts call, and started from there.  I have some historical knowledge of the business, so I could subtract big swaths right away.  They also reported “strong” performance in two segments that I knwo A LOT about.  Cool.  More clarity.

    I get a number that feels right.  It is damn close whether I do a top down, or bottom up analysis.

    It falls apart when I try to model for product mix, and units moved.  There is NO FRIGGIN WAY that they sold as many units as we calculate with known ASP’s.  Try a two tier (higher ASP when non-competitive) model?  Still too high.  Adjust the ASP’s to make the number reasonable, and it just is ludicrous (the ASP’s have to TRIPLE to match with what we think we know about their capacity).

    Poop.  Back to the grinding stone.