Posted on 04 15, 2013

Every now and then something just gets to me, and for the past few weeks, that something has been the process of purchasing enterprise hardware. Servers, SANs, load balancers, the kind of equipment that, instead of a price and an “Add to Cart” link, comes with directions on who to call.

Most hardware is not purchased directly from the manufacturer. It’s purchased through resellers, and each reseller will have a regional representative who will want to meet with you. The resellers don’t really want to just sell you the hardware, they want to “partner” with you to “build a solution”. They want to add consulting fees to your hardware purchase, and they want to jack up their margin on selling the hardware as much as possible.

It’s not small money we are talking about either. I’ve seen purchases of a single piece of equipment drop by $22,000 by modifying a single software line item. How much of that cost was the actual price of the software, and how much of it was reseller markup we will never know. The reseller program is built to keep us in the dark. For example, NetApp has a few nice bullet points on their “Become a NetApp Partner” page:

The NetApp Partner Program for Resellers is designed to help you grow your business and maximize profitability. With NetApp® products and services in your portfolio, you can deliver innovative storage and data management solutions that satisfy even your most demanding customers.

As a NetApp Reseller, you can:

• Increase revenue
• Earn greater returns on your investment
• Leverage our tools, best practices, and technical experts
• Create opportunities with our co-marketing resources, sales enablement programs, and prequalified leads

The combination of market momentum, increased demand for storage products, and innovative technology from NetApp makes this a great time to be a NetApp partner.

Pay special attention to “maximize profitability” and “increase revenue”, because as a consumer of these goods, that’s money out of your pocket. It’s not that companies like NetApp don’t have a price list, they do, it’s just that the price goes through an algorithm first. I believe the algorithm takes into account the volume of sales by the reseller, the number of discounts available to the reseller, and probably a few other items. I’d love to get a look at it someday.

In fact, I’d love to see a centralize, crowd-created database of hardware costs. A Web site that let me check on what the going rate was on a 10TB SAN. I’m not sure if I could convince anyone to do it, but it would be fantastic if sysadmins and IT managers around the country would upload the PDF quotes they get from resellers. Then we could parse the quotes, and come up with a general idea of what the market price is for the equipment. Right now, we just have to trust that whatever regional reseller we are working with is being honest and not gouging us. Hard to do with the salesmen arrive in BMWs and talk about their private pilots license.

I’ve heard that it is “unethical” to share the quote we get from one reseller with another. I’m not sure about that. I believe that if the vendor knew that anything they gave to us had no expectation of privacy, we would be free to share the quote with whoever we choose. The resellers are the ones deciding the practice would be unethical because it would be bad for their business if they had to fairly compete on price. Personally, I would be satisfied with whichever reseller let me know up front what their cut of the deal was, and how I could calculate the most honest transaction for my company. Dealing with resellers is like trying to buy a used car or a gym membership. It’s shady.

Here’s another example of resellers behaving badly. We were in a meeting with a salesman, and an engineer from his company, and they had a manufacturer on a conference call to discuss the equipment we were pricing out. When we got to the end of the meeting and we thought we had all of our questions answered, we hung up on the manufacturer, and the resellers started undercutting their manufacturer “partner” to try to sell us a similar piece of equipment from another manufacturer. They said it was because they wanted to partner with us to make sure we got the best deal, but I was left with the impression that the equipment they tried to pitch simply gave them a fatter margin.

This is the biggest draw, in my opinion, for open source software and hardware. If we can match the performance of these big name vendors using hardware that we can purchase from NewEgg, we won’t have to deal with slimy resellers.

Posted on 04 02, 2013

Mozilla has been one of my favorite open source projects since I first learned of them back in 2002. I remember downloading the Mozilla browser, and thinking that it was just like Netscape Navigator (no surprise there), an all-in-one browser that threw in the kitchen sink, just in case you needed it. You can still download it’s descendant, SeaMonkey. Not long after, I heard about another browser they were calling Phoenix, that, as it was explained to me, pulled out just the web browser from Mozilla, and left everything else alone. I downloaded it, and it was fantastic. Incredibly fast, lean, and simple. It became my favorite right away. There was only one release of Phoenix, after that the browser was renamed briefly to Firebird, and finally, Firefox.

It was an exciting time for the web. The world was starting to get broadband, but most connections were still dial-up. The slow connection speeds and slow computers meant that any small improvements in browsing meant a lot. To many people, the web was Internet Explorer. IE represented upwards of 90% of Internet traffic, so there was a real David and Goliath feel to Mozilla’s snappy little browser. There were many sites that rendered correctly only in IE, and Microsoft was busily expanding their monopoly on desktop computers to the web with FrontPage and IIS. The web could have turned out very different, and I like to think that we have the openness and inclusion of the web today in thanks, at least in part, to the work done by Mozilla back in the late 1990’s and early 2000’s.

Over the years Firefox has both grown and waned in popularity, but the impact of Firefox, and how it helped shape the web should not be underestimated. Today we have several projects that work to keep the web an open platform (like the ubiquitous WebKit project), but the work is far from over. Privacy online is a major issue for the future, as big companies become more sophisticated about tracking your activities on the web, and despite the advances of Firefox and other browsers like Chrome and Opera, Internet Explorer still makes up for the lions share of web traffic.

I’ve used all (or, ok, most of) the different browsers out there, but returning to Firefox always feels like coming home to me. Firefox is different because Mozilla is different. They are a non-profit organization dedicated to making the web a better resource for everyone. In my book, that’s something to be proud of. Mozilla is always looking for people who want to help, so if you feel like being part of something big, you might want to check it out.

Posted on 03 28, 2013

In the back of my mind, I always knew that the name “Scout” made no sense. I like the name, I like the bit of a kick the “Sc” at the start of it makes, and the snappy end with the “t”. It sounds good, but there is no correlation between what this application does, and what it was named.

Scout is now Paragraphs.

paragraph |ˈparəˌgraf|

noun

a distinct section of a piece of writing, usually dealing with a single theme and indicated by a new line, indentation, or numbering.

verb [ with obj. ]

arrange (a piece of writing) in paragraphs.

And, as you can see by the image above, we now have a snazzy new icon. I spent a long time thinking about the name and icon, and how they related to the function of the app. I commissioned two icons to be created. The first one was intended to look at home in the dock, and mimicked the “chrome circle with something inside of it” look

It’s a fine looking icon, and I might find some use for it eventually, but I think it accomplished it’s goal of “fitting in” a bit too well. To me, it looks bland, without character. For days after receiving the final icon I racked my brain about what to do. I even got out Photoshop and tried to edit it myself, to no avail. Then I remembered Shawn Blanc’s article about his grandfathers Royal Quiet De Luxe typewriter, and the pieces started to fall into place.

In addition to the name and icon of Paragraphs (I’m still having trouble not calling it Scout), I’ve squashed a ton of bugs, and added a few niceties. If you highlight a word and press Command-I, the word will be surrounded by asterisks, Markdown syntax for italics. Similarly, pressing Command-B will surround the word with two asterisks, Markdown for bold. Finally, highlighting a word, while you have a link on the clipboard (meaning you just copied a link from Safari or another browser), and pressing Command-L will format the highlighted word as a Markdown link. It’s pretty snazzy if I do say so myself.

I’ve put a lot of work lately into the animations, trying to make them as smooth as possible. There is still work to do there, but I think we are miles ahead of where we started. I also have a new theme that will ship with Scout, thanks to Kev Rodgers. Now Paragraphs will have the Default theme, a minimal CSS “Plain” theme, and Kev’s “Fresh Install” theme, which also looks quite nice. I hope to add many more in the future. RSS feed is fixed again, and I’ve put a lot of additional checks into the site parsing code to ensure that Paragraphs can continue running if it can’t build the site for some reason, testing that strings are valid, not null, and not empty, before using for example.

I got word today that the DUNS number for Farmdog Co. has been approved, and will be available to use in two weeks. So, barring any unforeseen disasters, we should be looking at a product launch within the next month. However, I’ve read that you should never pre-announce shipping dates, so take that with a very large grain of salt.

Posted on 03 20, 2013

So, quite a bit has changed in the past week or so. I got a lot of great feedback from a few beta testers, and stomped out a few bugs. A bigger deal though came from a couple of people who noted that they got confused in some of the views on how they were meant to proceed. Confusing the customer is absolutely the last thing I want to do with Scout, so I took off my developer hat and put on my HCI hat and tore Scout apart.

Scout now has a proper preferences window, just like every other Mac app. It comes up by hitting Command-Comma, goes away by hitting escape. Similarly, both the media manager and the theme chooser are in panels, keeping the main text editor view clean, and much less confusing about which view you are in, and how to get back to where you want to be. I’ve added a toolbar to the top of Scout, with a pair of buttons to switch between the text editor and the web preview. The other two buttons on the toolbar are to add a new post, and to show and hide the library of previous posts.

I’ve cleaned up the library view quite a bit too, and a added search field to the top which should come up with focus when you open the library. The search field searches both the title and the text of every post, so it should be able to find what you are looking for. The blue background was removed, and replaced by subtle dotted lines separating the posts on a white background.

I’ve also tweaked the image preview that comes up when you click an image link. The image is bigger now, and the background is more of a smokey white instead of the black it was before. I think this change makes the image more viewable, and makes the background fade out better, drawing attention to the image and away from itself.

I’ve fixed a few things about the default text editor theme that bothered me (and John), making it easier to read and overall more pleasant, at least I think so. It looks especially good full screen. In fact, my opinion is that the full screen view is the best way to use Scout.

Finally, I’ve made a few tweaks to the data storage engine, fixing a bug that prevented publishing a site on the first run. Funny story about that. Last week I became a bit overconfident in Scout, and sent out invitations to try it to a group of well known writers that I respect. One of whom replied that he had downloaded and tried Scout, found a laundry list of things wrong with it, tried to publish to a folder, and Scout crashed. I was mortified. Turned out to be a stupid change that I had made, and obviously not tested thoroughly enough, right before sending out the application. I put the publishing code on a background thread to continue user interface responsiveness while Scout wrote out the files. Turns Out, that’s not a good thing. However, what crashed the app on the background thread merely caused an error that Scout recovered from on the main thread, so finding the crash actually allowed me to find and fix a major bug. I just wish I had found it before sending it out.

Live and learn. Test harder next time.

With all of that being said, I really think that Scout is in much better condition now than it was a week ago. I’m excited.

Posted on 03 17, 2013

Last week I sent in the certificate of organization to recreate Farmdog as an LLC in the state of Iowa. I have the domain name, twitter account, and a test site built. Assuming everything goes through with the state without problem, Farmdog Co. will be ready to launch very soon.

Scout has a couple of rough edges to clean up before shipping, I need to finish the help doc for it, and fix a couple of odd user interface inconsistencies. But, for the most part, I think Scout is ready to ship. Narrowing the scope and number of features slotted for 1.0 was a big help. I’m planning on selling Scout through the Mac App Store, as well as on Farmdog.co using Fastspring. So, I still have some work to do on Sandboxing, as well as license keys and trial limitations. Probably a few days work to get all of that integrated and tested.

Which brings us to Go2, my professional bookmarking app. I’m rewriting Go2 from the ground up, and I’m pretty happy with how it is coming so far.

When the Mac App Store launched, Go2 was there on day one. It sold well the first month, but was downhill from there. Previously, Go2 had a bit of an identity crisis as it tried to find it’s place. Was it a general bookmarking app like Delicious or Pinboard? Was it a replacement for Command-K in the Finder? When customers downloaded Go2, what were they expecting? As time went on and as I talked to other Go2 users, it became clear that they were using it similarly to how I use it, as a professional bookmarking tool for quick access to servers over SSH, FTP, or as web pages. So, the newest version of Go2 builds on these strengths, and makes it even easier to get to where you want to go. Here’s a quick rundown of new features:

• Importing the SSH known_hosts file.
• Search field in the status bar now launches the first bookmark when the user hits the return key.
• When you down arrow out of the search field, you can now use the forward slash key to return to the search field.
• Overall improved look and feel

Also, interactions with Go2 are normally through opening bookmarks. So, the main interface is now in the status bar, and new bookmarks will be added through the preferences pane.

Finally, I’ve abandoned core data as the storage engine for Go2, and am in the process of moving the plain text storage engine I wrote for Scout into Go2. Hopefully, this means that upgrades will be easier, and I should, finally, be able to sync through Dropbox or some other file syncing technology. This should make it easier for teams to use Go2 as their central source of truth for network servers.

Like I said, I’m excited about relaunching Farmdog. I’ve learned a lot over the years, and I can’t wait to get rolling.

Posted on 03 13, 2013

Eventually I’m going to need to compile this into a proper help doc, but for now, documenting Scout with Scout will do.

Scout handles pages just a little differently than posts. New pages do not show up on the home page, and are not organized into subfolders based on date. Pages are converted into HTML and saved in the root of the site directory, right next to index.html. To create a page, add a new post, but before the title of the page add the characters @@. That is two “at” symbols followed by a space. For example: @@ Contact.

Secondly, the default theme that ships with Scout, the theme of this site, is my own creation that I’ve cobbled together over the past few years. What I hope to see are a few people more talented than I create some really great themes. In fact, assuming Scout is successful, I plan on hiring a designer to do just that. However, there are a number of hurdles to overcome before that happens.

To create a Scout theme, create a folder with the following files inside:

• atom.xml
• atomTemplate.xml
  • The two xml files are for creating the site RSS feed.
• base.css
  • Should contain all of the site CSS
• default.html
  • The standard HTML template for the site.
• indexTemplate.html
  • Wraps each post in a div with an id of “posts”.
• post.html
  • Creates the page template for each post
• robots.txt
  • Not currently used, but probably should be.
• screenshot.png
  • A nice screenshot of the post at 400x341 pixels.
• toc.html
  • The template for building the archive or table of contents.

Scattered throughout the files listed above are strings that look like: YYYY_SCOUT_SOMETHING_YYYY. These are the Scout “Y Codes”, as in, “why are they so ugly?” I was looking for something that was unlikely to be reproduced anywhere else, and this was the first thing that came to mind. I currently have the following Y codes defined:

• SCOUT_5ATOMPOSTS
• SCOUT_AUTHOR
• SCOUT_CONTENT
• SCOUT_CSS
• SCOUT_CURRENTDATE
• SCOUT_CURRENTYEAR
• SCOUT_POST_CONTENT
• SCOUT_POST_DATE
• SCOUT_POST_TITLE
• SCOUT_POST_URL
• SCOUT_SUBTITLE
• SCOUT_TITLE
• SCOUT_URL

Hopefully, the titles of the codes are self-explanitory, but as I have time I will be documenting them properly. For now, you can probably take a look at how the files in default.stemplate are arranged. Scout looks for these codes in the template and replaces them with the proper data from the posts to generate the site.

Lastly, you should never feel like you need to save your work in Scout. Scout automatically saves your work at various intervals. For example, whenever Scout looses focus as the main window, or whenever a new post is chosen, or when a new view is looked at, Scout automatically saves the current post as a plain text markdown file saved in ~/Library/Application Support/Scout/Posts/. If you like you can watch the file being written to in the finder or as you use the application. Likewise, if you have a markdown file that you would like Scout to know about, you simply need to name it according to the naming convention of YYYY-MM-DD-HH-MM-SSSS##Title_Of_Post.markdown and drop it into this directory. Next time Scout is restarted it will treat this file as its own.

Scout is intended to give you control over what you write without getting in your way about it. I know there are going to be some rough edges, but as we progress I’m sure they’ll get worked out.