Especially on Verizon:

Why My Mom Bought an Android, Returned It, and Got an iPhone

My favorite part:

A friend of mine has a Nexus S and it is a pleasure to use. The UI is elegant and functional. The battery lasts for days. In short, it is everything that the Charge wasn’t. I’d love to see Google somehow mandate the stock Android experience on all phones, or somehow rigorously test all new phones before they could be launched. Why not standardize and mandate one or two excellent cameras, and then open source the drivers? Why not certify and approve a few of the best components and then place some sort of “premium Android experience” certification label on phones that pass tests and use components approved by Google? Right now it’s a crapshoot out there when you want a new Android phone, and it doesn’t seem to be getting any better.

Let me excerpt that a little more clearly:

I’d love to see Google somehow mandate the stock Android experience on all phones, or somehow rigorously test all new phones before they could be launched.

You mean the way Apple mandates and regulates the entire end-to-end experience for iPhone users?

Not for nothing, but I’m slowly coming to the sad realization that my days as a self-declared Blackberry fanboy are numbered. I came to the party riding an Blackberry Curve 8310 and I’ll most likely be leaving it on a Bold 9700. Nothing rivals the build quality of RIM’s finer hardware, save for maybe the products designed in Cupertino. The experience of their keyboard is perfect, and gets better with every revision. My 9700 has never been short on horsepower, it can be quite snappy when the software takes proper advantage of the hardware; but that’s where everything breaks down.

So unless we somehow see a miracle pulled off, and RIM releases the Bold 9900 running Windows Phone 7, I’m probably going to end up with an iPhone 4Gs™ or iPhone [Next]. I’m going to miss my hardware keyboard, and I’m going to bitch about it every time I am forced to type an email, but I’ll enjoy being able to run apps that don’t suck, have more than 256MB of space for apps, email, contacts, etc, and generally having a more pleasant experience.

So I avoided writing a post about the TSA hoopla (one might still come out of that yet…), but this Comcast stuff seems to be hitting close to home, since I’m a Comcast “customer”, a Netflix Subscriber, and a Net Neutrality aficionado. In the most recent hysteria, Time Warner has come out in support of Comcast. Comcast’s original response to Level 3′s complaint really kinda pissed me off.

For one, they point out that if they allowed Level 3 to have its way, Level 3 would then have an unfair advantage over other Content Delivery Networks. That’s an absurd argument, since many CDNs probably already use Level3′s services to get their traffic onto the internet at large, or route it between major metropolises. Simple traceroutes from my machine here at home to servers in California and Washington State show me that my traffic leave’s Comcast’s network in Chicago, hits Level3′s network, and is carried from there on, all on Level 3.

The larger point here is that Comcast should be completely blind to the source of the traffic. Say Netflix contracted to Amazon’s CloudFront service, a popular and, well, downright huge CDN. My route to amazon’s web services goes through Level 3. A recent report indicates that Netflix represents 20% of the traffic on the internet at peak hours. From a business standpoint, with Netflix generating so much traffic, it only makes sense for them to want their load to be tied in at the internet’s highest levels, and given the role Level 3 plays in routing internet traffic, not only in the US, but worldwide, Comcast shouldn’t be bitching.

For two, Comcast brings up the deluded point that CDNs represent an unfair “peering” of networks, since almost all traffic being exchange is coming FROM the CDN. Due to this unbalance, they have a right to charge Level 3 for all the extra bandwidth. Let me see if I can put this delicately enough for Comcast to understand: YOU PRIMARILY SUPPLY RESIDENTIAL INTERNET SERVICE, YOU FUCKING INGRATES. Ahem. More to the point, the “peering” that goes on at those levels is always going to be biased, since Comcast customers are serving very little content in general, but are almost exclusively consuming (downloading) content from networks outside of Comcast’s boundaries. If Comcast customers are using bandwidth heavily from a particular CDN, then it is Comcast’s duty to have a big enough pipe to that provider in order to handle the load.

“Peering” shouldn’t even be brought up at this point, Comcast should be treated like any other piss ant ISP. Level 3 is a nationwide and GLOBAL network. Level 3 operates part of the transatlantic cables that connect the US and Europe. Comcast is the largest cable provider in the US, but they rely on companies like Level 3 to tie their disjointed networks in cities all over the US together. That relationship is symbiotic, but Comcast, ever the bunch of fuck heads, wants to see just how far they can push things before they get slapped down. It boils down to what I said earlier: if Comcast’s customers are driving demand for a particular service, and that traffic is coming through a particular provider, its up to Comcast, as a residential internet provider, to make that service available. The problem is, the service in question, Netflix (next up: Amazon’s VOD service), directly competes with Comcast’s terrible and terribly overpriced Cable and VOD service.

One of the biggest facets of Net Neutrality, probably the one more important than the others, is the regulation of how service providers regulate services in direct competition with their other products. In the case of DSL and wireless/mobile providers, what restrictions are placed on using Voice over IP (VoIP), which is basically a phone line over the internet. In the case of Cable providers, what are they doing to restrict video services that compete with Cable TV, Pay Per View, etc, services that have until recently been enormous cash cows for those service providers.

Update

As usual, PacketLife.net puts me to shame.

[Purdue release];

So for those who weren’t already aware, on October 7th I gave my two weeks notice to my supervisor at Purdue. After just over 21 months, I was ready to be free of that place. Well, I was ready to be free much farther back than that, but I was finally presented an opportunity too good to pass up in favor of the stability and large perks that Purdue offered.

Without me going off the bitter deep end (not that anybody would be surprised by that), my reasons for leaving were barely a secret to my coworkers. Between the politics, some of the people (particularly the upper management, or at least the decisions they had made), and the uncertain future ITaP presents to its employees, not to mention the day-to-day frustration of being a Windows admin, the job as a whole was taking its toll on my sanity and the rest of my life.

So with that, I wish the best to my colleagues and friends still working for ITaP and various other Purdue IT groups, and I do hope they all stay in touch, if for no other reason than so I have a conduit of information as things continue to slowly fall apart. Hopefully that will stop happening soon, but you can’t hemorrhage good people the way ITaP has been and continue to do good work. The handful of knowledgeable, competent people left behind can only be stretched so thin. I do hope sanity is restored at some point, for the sake of those too embedded, too enamored, or too stubborn to leave.

[[Stoneware alloc] init];

So the opportunity I mentioned above, the one that was too good to pass up? I started there October 25th (yesterday, by my clock). Its with a little company called Stoneware, based out of Carmel, Indiana. Without going into too much detail, they were looking for a developer familiar with Microsoft’s Remote Desktop Protocol (RDP), more specifically somebody who had experience working on a client, and given my experiences with CoRD over the last 18 months, my proximity to the company, and my dissatisfaction with my current job, neither the timing, nor the job, could have been a better fit.

I’ll be commuting to Carmel as needed, but predominantly working from home. I spent most of yesterday in meetings and most of today sifting through code and getting my bearings on where things are. I’ll be working entirely in Java, at least for now, so its a bit of a leap, but the experiences I gained this summer ginning up some C# apps for our Windows 7 migration are paying off already, given the similarities between C# and Java (for once I’m happy Microsoft ripped a good idea off of somebody else). I will say that despite missing Objective-C at times, I’m excited about the opportunities the job has, and the more I dig into things, that excitement only seems to grow.

[CoRD setStatus:"Uncertain"];

We haven’t really discussed the future of CoRD yet at my new job, and I’m not really certain where things lie. Given how busy spring and summer, and even the fall was for us at Purdue, trying to get our Windows 7 migration underway and off the ground, I had barely any time to put forth towards CoRD. I did some experimenting at various points, trying to get some of the larger features we planned to implement at least started and/or off the ground, but at this point none of them are outside of the experimental phase. There was just too much work involved in the bigger pieces, and I didn’t have the time or gumption required to go through, rip out all the dirty code, and replace it with some sparkling new awesomeness. Given that working on it now would present a conflict of interest (potentially), I’m not sure where that leaves CoRD. Dorian has expressed an interest in keeping the project alive as needed, but I got the sense that he had no plans for large changes being made. Once I’m settled in, I might broach the topic, but for now, unless Microsoft swoops in next year and offers Dorian and I piles of money to merge their crappy RDP client (with a stable backend) with CoRD’s awesomeness (sans the crappy backend), or a new developer or two sneaks in and begins contributing, the future of CoRD is murky at best.

Software Installation is an area that has many issues, and for a variety of reasons. I have but one reason how this can be: Developers suck and very few possess any foresight. It baffles me on at least a weekly basis, how a person or team of people can be smart enough to write a functional application on any platform, yet be so stunningly ignorant of the environment in which their applications will be running. They build registration data into installers. They put registry keys in place at install time, with no heuristics in-program to handle things in the event they’re not there. I could go on for a while… Everybody is guilty of it, from the biggest shops (Microsoft, Adobe, Autodesk), all the way to some open source projects.

I have seen immensely complicated applications, that have stellar install experiences. And I have seen painfully simple applications that require me to rip them out of their installer and package them sanely in order for us to deploy them. There are exceptions to almost every rule, but here are a few tips for any developers, Mac or Windows:

1. Installers should not be handling any interactive product registration / serial numbers. You fucking jerkwits.
2. You don’t need to restart the computer. Really.
3. You don’t need me to log out and back in either. Really.
4. Installers should need to do 5 things:
   1. Preflight: stop processes/services, etc.
   2. Pre-install: gather any necessary data, set environment variables, etc.
   3. Unpack files and/or registry keys, dump to their right locations
   4. Post-install: build any on-the-fly things using utils you just installed, etc.
   5. Postflight: restart any processes/services, etc
   6. If you’re doing more than this in your installer, you’re doing it wrong.
5. Installers should have a silent option. Yes, it really is necessary, you lazy tool.
6. You should NEVER have to write your own fucking installer.¹
7. Installers should not be handling any interactive product registration / serial numbers.
8. You don’t need me to quit my fucking web browser, or any other programs. Really.
9. If you insist on some kind of stupid idiotic compression or container format, then extract to a TEMP space, and clean up after yourself.
10. Installers should not be handling any interactive product registration / serial numbers.
11. Installers should not do anything that is dependent or tied to the current user. Including interactive product registration / serial numbers.²

The entire concept of requiring an installer is one centered around the clusterfuck that is Windows. In the best case scenario, the above installation steps could be reduced to one: Unpack Files. More complicated programs SHOULD only require 3: Preflight, unpack files, postflight. Some packages like Java, for instance, would require at least those two scripting steps, to search for existing versions, modify the registry as necessary, etc.

Mac-side, any program that has progressed to the point of not being a drag-and-drop .app bundle, and is looking at doing some form of “installation” needs to step back and take a serious look at how their app is structured. If you’re including a lot of secondary content, or installing frameworks used by multiple applications, etc., you’re exempt from this, of course.

0.6

The parameters for the next iteration of CoRD continue to change for me. At one point, we had agreed that 0.6 was to focus on Groups. The plan, in its rarest form, was that we’d snap out the TableView, and snap in an OutlineView, and roll out 0.6.

Things are never that easy.

Currently, each server is stored in its own .rdp file, which are all lumped together in some forgotten directory (~/Library/Application Support/CoRD/Servers). Once we add groups, how we handle session storage becomes important all of a sudden. Originally we wanted to continue the file-based approach, and make a 1-to-1 relationship between groups and folders inside our Servers directory. With that tactic, in addition to handling the groups, we also have to obey and handle folder creation, deletion, renaming, etc,. inside of CoRD. I don’t really want to have to build a file browser (essentially) into it. Cocoa’s interfaces for file and folder manipulation aren’t terrible, but they are not lovely either.

The solution to this problem is easy enough, we dump everything into a plist (the recursive methods for that took me all of 15 minutes to write and test once I had a class established to handle groups, and I’m a n00b still), and gin up some little import method to handle things when the new version is launched for the first time. Easy peasy. So now, we’re implementing an OutlineView, for the actual groups people see, and we’re shifting our storage paradigm.

But wait…now that we can group servers, we should be able to do bulk updates of the server info in groups.

And while we’re doing that, we should fix how we’re storing credentials, so that people with a 30 day password policy and 30 servers don’t have to manually update 900 passwords / year.

Also, with the current plan for groups, we’re losing the Connected/NotConnected paradigm of CoRD 0.3-0.5. Sessions will only show their state via their icon (and probably some other visual cue). This means sessions could be connected but hidden in a collapsed group and makes iterating through connected sessions unnecessarily complex. Perhaps we should re-evaluate how we handle that?

I could go on for a while with this kind of stuff, but I’ll stop there.

Rapidly, what seemed like a simple change, spins out of control into something far from simple, especially for an open source project. It could be argued that all of this is feature creep, but I don’t agree. This is a major overhaul of code, and should be treated as such. If we’re not overhauling the code specifically to implement these features, then how can we overhaul it, get it halfway there, release, and just expect everybody to be happy with it?

I know more than a few developers who’s mantra is “just ship it!” which is a mentality I can’t stand, especially on a project like this, where A) many people LIVE in this app, and B) My name is in the code. I don’t want it to be a “work in progress, we’ll fix it in the next release” kind of overhaul. Mac users demand more, and Mac apps deserve more. Releasing half finished products is something I see too much of in the Windows world. I don’t want to be party to that kind of nonsense in the Mac world.

0.6++

The more I review crash reports, bug reports and feature requests, and the more I use CoRD day to day, the more I find I want to change. Full screen and the current inspector setup are our two single biggest causes of crashes (by far). Those need to be fixed soon. Most of the fullscreen code we’re going to rip out in favor of using the new fullscreen mode that all views get for free in 10.5 and up. I would like to somehow find a way to implement bindings, or at the very least significantly clean up the code that handles the inspector and manipulating sessions and try to isolate the potential for future crashes.

I want to eschew the drawer, and the rest of the current model really, and adopt more of a tabbed approach to the unified server window(s). The server list would/will remain completely independent, and pretty much exist purely to list servers, manipulate server data, and launch sessions. Individual sessions or groups of sessions live in Safari-esque windows with tabs, minimal bezel (read: no toolbars), and controls (disconnect, etc) for each session/group would be integrated into the perimeter to handle changing view modes (see the location of Safari’s new tab button), etc. The windows would be ginned up on the fly and tabs (active server connections) could be drag and dropped between them (just like Safari 4.x). Quick Connect would become a modal panel that would drop down overtop of whatever the active window was.

0.6--

The problem with this dream is two fold. First, that’s a shitload of work, and much of it is beyond my current abilities as a programmer (the latter part isn’t insurmountable, it just extends the development time significantly while I learn-as-i-go).

Second, I am becoming more and more disillusioned by the handicapped feeling I have because of rdesktop. The project is/has stagnated considerably, and attempts to fork or revive it have left me startlingly unimpressed. I think there’s a serious need for a code library, cross platform, and preferably BSD licensed (if it comes down to it, I could give a damn if it’s GPL, since CoRD is already GPL’d), to handle RDP. There are features in new versions of the RDP spec that CoRD can’t (and at this point could never) support due to its dependency on rdesktop, and some of those features are becoming essential to function in many environments (things like Network Level Authentication, SSL/TLS, stable printer support, etc).

Problems like the most recent cursor clusterfuck (that we’re still shoveling through fallout over), which caused cursors to render improperly on Windows 7 and Server 2008 R2 machines and the recent patch to rdesktop to support RDP’s reconnection feature via a local session “cookie” serve to highlight my frustration. Such changes proved challenging, and to some extent incredibly problematic, to integrate into CoRD due to rdesktop being built with a typical X environment in mind, and not more generic like a library would be.

So where are we?

The likelihood of an open source library being generated in the near future is very low. I would love to spearhead such an effort, but I have neither the time, nor the experience necessary for such an undertaking. I’m doing well to accomplish what I have thus far on CoRD. I’d settle for a couple of motivated guys who would want to help implement the RDP spec in Cocoa (which I think could be awesome as hell and fun to work on)

Most likely we’ll just shim in groups, and try to integrate as many of the necessary changes as we can into the codebase. At some point we’ll self-impose a deadline on ourselves, then we’ll back fill the remaining time with a couple of beta releases to identify any showstopping bugs before releasing 0.6 sometime this summer, around a year after posting 0.5. Our goal was to be on a much faster release cycle by now, but demands on my time and Dorian’s have caused a lapse in significant progress. Hopefully we can work to remedy that in the near future.

In 0.5.x news, we know there are several (rather large) bugs on 0.5.3, including but not limited to:

• Time Zone Sync doesn’t work
• Saved/Stored Passwords are broken for some sessions
• Cursor problems still exist

The top two are regressions, they worked in 0.5.2 but some of the rdesktop code we imported in broke them. The latter is something not many people will see, and we’re having a bear of a time troubleshooting it.

As always, thanks for all the support from our users, we appreciate the bug reports, and crash reports. Hopefully we’l be able to release 0.5.4 in the next month or so and fix several of the problems we introduced, as well as resolve some other stuff while we’re at it.

I’ve been revamping WordPress recently, fixed some things on the back end that have been bothering me, applied a new theme, and posted a fresh new resume, which hadn’t been updated in about 16 months. I have no immediate plans to need a new job, but I’d rather be prepared for it in any case, given that Purdue is doing anything but making me feel secure about my job at this point.

Anybody who had anything bookedmarked with a /wordpress/ directory in the URL should be getting redirected to the home page now. Prior to WordPress, I used CodeIgniter, which lived at the root of peelman.us. When I installed WordPress, I wasn’t sure if I was going to like it or not, so I stuffed it into a sub-directory of the site, http://peelman.us/wordpress and did some funky .htaccess rules to redirect requests there. Its been almost two years since I switched, and I haven’t looked back, so I figured it was time to adjust things. WordPress is now the site root, so I apologize in advance for any broken links or issues getting to content.

First!

Its been almost a year since I started working on the CoRD Project. In that time we’ve released 0.5, which was in “beta” for almost 18 months when I started tinkering with it in March of 2009. We’ve also released two minor updates that fixed major issues and added features, and are getting ready to release a third release (0.5.3) that will hopefully fix a few problems. Work on a 0.6 release has started, and rudimentary code has been put in place to allow groups. Going completely off of the amount of releases and code changes, it may seem like we’ve been resting on our laurels, but I wanted to point out the massive changes we’ve made on the backend in the past year that will help us accelerate releases and hopefully prevent another 18 month drought of updates and information.

Trac & Forums

This wasn’t an easy change though, as SourceForge could provide no migration of data from their Tracker to Trac. So I spent the better part of a month moving tickets by hand as I came across them, to the new system. We lost a lot of historical stuff in the forums, which sucks, but was a necessary evil. It was a ton of work to move this stuff, but in the end it turned out awesome and Dorian and I have been incredibly happy with it.

Crash Reporting

Prior to 0.5.2, Dorian had employed SmartCrashReports to some extent, but gathering and collating that data was incredibly difficult. Most crash reports we received either came in when a kind user attached them to a bug report they filed, or via email to one of us (mostly Dorian since his name and email are on the website ) In 0.5.2 we replaced SCR with UKCrashReporter and backended it with Dorian’s own CrashPool system. CrashPool, which will be going live soon for any interested developers, serves as a crash report aggregator, it handles sorting and grouping, searching, etc. Prior to 0.5.2, we both thought we were putting out a relatively stable app. Boy were we wrong!

We have a huge devoted group of awesome guys who are running the Nightly Builds (see CoRD’s preferences or this article) and reporting a ton of data back to us. And since 0.5.2 has been out we’ve gotten thousands of crash reports that we’re sorting through (it is really easy to spot trends; we know that guys who use Fullscreen heavily are seeing a metric shit load of crashes; and for some reason the inspector is causing some crashes that we haven’t been able to chase down (yet)). We definitely appreciate the guys (and gals) running the nightlies and providing us feedback and crash reports. Thanks!

Build & Deployment Infrastructure

This was all Dorian. He completely revamped his nightly build script and modified it to the point where we can release stuff practically on demand. The script will update the build the app, package it, upload it to the right places, update the appcasts, everything. He has a nightly task that runs and builds our nightlies, and by simply changing up the parameters we can dump out a release, either a beta or a full release, on demand. It is a truly awesome piece of code. Thus far he’s keeping is close-hold and not releasing it into the wild. But I think the eventual plan was to make it available, but it is his baby, a separate project in and of itself, so I can promise nothing, just give him credit where it is due.

VNC Requests

For those following CoRD on Twitter, I have wanted to make a tweet about this for some time now, but it is just too long of a thought for 140 characters. We see a lot of chatter, and the occasional feature request, about adding support for VNC connections into CoRD, theoretically making it a universal, kick ass, remote console tool. I share these thoughts. I would love to flip the magic switch, and allow CoRD to answer VNC requests. But I can’t, and even if I could, we’re not at a place where I would. There are many, many, many things I want to see in CoRD (including fixes for the aforementioned stability problems) before I see VNC support:

• Groups
• Better management of credentials
• Stable printer forwarding
• SSL/TLS support
• Smart Card support
• Autoreconnect support (rdesktop has this now, we have some code in place but no implementation yet)
• A complete rework of our drawing code to fix all the screen glitches people run into.

With just Dorian and I working on CoRD in our free time, we have to apply our time to where it makes the most sense and can do the most good. So just to put an end to any nonsense before it gets started, we’re not against being a good RDP/VNC app. But we want to be a robust, kickass RDP app first. In the mean time, there are plenty of really decent VNC apps out there.

For those not aware, there was a recent mess started by Brent Simmons on creating a new open source email client for Macs, to be called Letters.app. I’ve been following the carnage since Monday and just wanted to lay down some random thoughts here.

I will call right now that its doomed to fail. Why? Because the clients aren’t the problem.

SMTP Sucks

SMTP and this retarded communications method we are all bound to is the problem.

Example: Do you top-post replies or bottom post them? Should it even fraking matter or should the protocol provide message threading for you and let such behavior be a user preference?

Example: SPAM. A decently designed protocol could curb unsolicited messages; not eliminate them entirely, mind, but drastically reduce them.

Example: Message content and attachments; SMTP was designed for plain-text transmission of simple messages, which has hamstrung both the protocol and the clients ability to distribute and display rich content and files. The concept of ‘file sharing’ using anything but email is lost on most non-geeks.

All that said, the point is moot because SMTP isn’t going anywhere for a while. A successor isn’t even on the horizon.

However, the advent of services like Facebook, Twitter, Flickr, Droplr, Google Docs, MobileMe, etc. is making the sharing of pictures, documents, etc. easier via web technologies. Perhaps if these services can take over enough of the burden, the dependency on SMTP-based messages can be reduced to the point that it will eventually die or be replaced by a new protocol that is more secure and better engineered for modern (and future) communication needs. Could happen in 5 years, might happen in 10, or it could never happen.

Bitching about Nothing

Regardless of the inherent failures in the protocol itself, this troop of developers and laymen are hellbent that there is a great need for a new Mail client; even though for the most part people are wanting email clients to do things that they was never intended to do and shouldn’t be used for anyway…

Example: Organizing/aggregating/viewing crash reports; this was given as an example in several threads (by multiple users) on the mailing list of “something this client should handle well.” If you want to use email as the medium in which you receive crash reports fine, but why should it have to handle the indexing/cataloging/etc for you as well? Why not write a good, scriptable CrashLogTracker that you can then write a script or service and extend Mail to accommodate it? Maybe I missed the boat on that point, but it was suggested along with the idea of creating a “theming” system for messages ala NetNewsWire, which can apply custom templates to RSS message bodies.

And More…

Most of the bitching about current clients seems to stem from Thunderbird’s lack of Mac-ness, and Mail’s lack of “power user” abilities, its slowness, or its crashes. Other clients are often mentioned as having high points, but nobody outright addresses the pitfalls of each client. My thoughts?

Thunderbird: I don’t deny that Thunderbird doesn’t have much Mac flair. But why not band together and work towards FIXING it, rather than creating something new? Yeah…since its a Mozilla project, that’s probably not such a good idea. T-bird is all but dead to me, much like Firefox.

Mail != PowerUser: As a huge Mail proponent, I’m quite happy with it. I use a boatload of rules to direct messages to various folders, and I have a metric shitton of email (rough estimates, I’ll put it at around 50-60k messages). The elegance and simplicity, yet robust capability of Mail is the *definition* of a PowerUser’s application. The editor lets me compose in plain text, but doesn’t castrate HTML or RTF emails when I try to forward them on to people. I apparently don’t have the issues with quoting or conversation threading that other people do.

Mail != Performance: As stated above, I’ve got 50-60,000 messages in Mail and it rarely balks at me. The app itself is responsive and operations are performed quickly. Thanks to the power of Spotlight I can search that entire archive in a matter of seconds. Bottom line is, if Mail was the worst performing app on my system (like some people claim it is for them), I would be a very happy nerd.

Mail != Stable: Issues that I have had I’ve filed bug reports with Apple on and they have typically been addressed…perhaps not always in what I would call a timely fashion, but they DO listen. I have never had Mail crash and corrupt email on its way down (can’t say that about any other client I’ve used). I’ve never had Mail crash during a routine operation (it usually takes something out of the ordinary, attaching or opening an odd or malformed file, for example).

On HTML Email

I don’t really give a damn about the format that people send me messages in. Its text. It may look different, but its text. Going from App to App and webpage to webpage my eyes adjust to different fonts thousands of times a day. Bitching about HTML email fucking up your Qi is just retarded. On that note, any client that doesn’t support it properly is going to look like it fell out of a 2001 time warp. Nobody is stopping you from using TextMate as your mail editor. But when a client hamstrings my ability to do my job (in this case, forward mail coherently, regardless of formatting), it ends up in my Trash folder.

Performance vs Plugins

There is a ton of support behind basically creating an email client that is but a shell, wherein all but “core” functionality is provided by plugins or bundles, keeping the “program” (it can’t really be called that in that instance) “light and fast.”

I call bullshit.

Performance, Functionality (read: bloat), and Usability are not bound by any relationship that I have ever been able to discern. There are many, many, many incredibly complex programs out there that perform EXTREMELY well. Most of those programs remain incredibly useful. There are poorly DESIGNED programs that are full of bloat, with ill-construed features that cause more harm than good. I don’t think you should attempt to handicap a program’s functionality from the get-go for the sake of some mythical performance level you hope to achieve.

So my idea on doing a series about the follies of Windows 7 has fell flat on its face. In addition to the sheer amount of things that are still bad, there is simply too much personal bias, too much dislike. I tried and failed 3 times to write my piece on software distribution, and each time it fell into a nebula of hatred and bitterness that only computer geeks could appreciate.

I tried to distill the bitterness about software installation into a handful of questions:

• Why isn’t there *one* standard for distributing software for Microsoft Operating Systems?
• Why are there like 14 companies marketing installer-creation-software, none of which are good enough for still more companies, who insist on writing their own installer?
• Why doesn’t Microsoft provide better tools for packaging software?
• Why can’t I point SCCM at an MSI and a collection of machines and say “Go”?
  • On this note, why the hell can’t I safely remove all of MS’s stupid ass default Collections in SCCM? Or at least move them somewhere less obtrusive
• Why are the best tools for managing Microsoft’s operating systems NOT written by Microsoft?
• Why do some 3rd party developers possess a better handle on Microsoft’s operating systems and APIs than Microsoft does?
• Why are so many Microsoft die hards so quick to defend something that they themselves will admit is a convoluted and fucked up system?

So I have to give kudos to the guy running packetlife.net. Its a great site for networking nerds like me, and he’s got some really awesome cheatsheets available. But he had a post this week that really struck a cord: Why network neutrality is a big deal. I dumped a less-edited version of this into the comments, but I wanted to flesh out the thought and post it here as well.

Internet connectivity isn’t like the highway system. The government doesn’t provide high speed links; private companies and giant corporations do. Ask Verizon how many million they invested in FIOS. It costs money to bury fiber or copper, buy or lease right-of-way, add capacity via routers, switches, DSLAMs, FTTx’s, Optical MUXs, servers etc., and once the line is in your neighborhood or house, the likelihood of a competing service coming in drops to almost zero.

As it stands, public opinion and the fear of government oversight has been the only thing stopping them from completely monopolizing their lines so far. But if they encroach slowly like they have been: capping transfers, “shaping” traffic, etc., ultimately convincing people that its really OK, and there’s nothing wrong with it, we’re screwed. Lets not forget all the lubing of the gears in Congress (telecom lobbying is a rather big business, on either side of this issue).

How about if you live someplace where there isn’t a BigName LEC? What if VZ or AT&T isn’t your ISP? How long before this type of tiered pricing is applied to the Qwests and the Level3 carriers? You get a discount or a kickback for every 10000 subscribers you have on a limited plan?

The biggest problem is much of our copper infrastructure is reaching its limits. Providers don’t want to bury new cable. They don’t want to install new DSLAMs, or cable nodes, etc. They want to keep riding the money train, soaking people for more money, for the same or less service. And they can, they own the line all the way to your house, and they no doubt own or lease the right-of-way for that service in your area. With your electric bill and water bill, there are government regulations and restrictions in place on those service provisions; there won’t be more than one electric service provider in your area, nor will there be multiple water providers, its just not feasible and makes no sense. Telecomm and data providers (ISPs) are in that same category, but are much more loosely controlled.

That’s not stopping the demand though. The Netflix On Demand, YouTube, Hulu, etc. In 5 years the amount of content distributed online will make today’s current usage look miniscule (just like the usage of 5 years ago looks so tiny compared to today’s). That’s banking on the hope that the entire system doesn’t collapse on itself in another year or two. The providers could have been spending some of those profits over the last 20 years to keep things up to date and rocking, and we could have a policy like Finland, where every citizen is guaranteed 1Mbps broadband.

As we proceed through our deployment of Windows 7 at work, I want to detail some of the issues I have with “features” Microsoft graced us with in their latest attempt at making an OS not suck.

For starters, some background: We skipped Vista. Our infrastructure was heavily entrenched from many years ago, dating back into Windows 3.1 for some things. In the spring of 2008, the infrastructure was shifted to SCCM and App-V (Application Virtualization). For better or worse, those are the components we’re using. In the future I may write a post giving more detail on our upgrade path and process for the infrastructure changes we’re making, but we have to get it fully ironed out first.

I’ll also point out that I am a huge Apple geek. I’m not a fanboy, I get pissed off about OS X and Apple’s methodologies, perhaps not as much as Microsoft’s, but I’m far from thinking my chosen OS is a picture of operating system perfection. Without further ado…

User Access Control

Beefs with UAC aren’t new. However, my problem isn’t with the implementation. My problem is with the concept and the thought process behind it. Instead of Microsoft trying to fix the many things wrong in the OS, they shoved the burden of security off to the user. That’s bad enough, but compounding the issue, they did this in a very invasive and scary way. Opening a start menu item, or a shortcut, or a file, should *not* trigger a Machiavellian prompt, greying out the entire screen save for one relatively tiny ass prompt, in the corner of the screen instead of the middle (or god forbid, over top of the application or action that caused it to be called in the first place), and that at best gives a cursory reason for WHY its there.

In fucking this up so badly, Microsoft instilled a terrible behavior that is now part of many users’ psyche, of “oh there’s a pop up, I have to click OK”. Many people can’t tell the difference between a Windows dialog box and a popup Window in Internet Explorer. Many people don’t always make the connection when a UAC-like dialog box comes up, but its slightly different and the screen didn’t dim, and its got internet explorer’s icon in the corner. Many don’t realize that clicking links in a web browser shouldn’t cause UAC prompts!

What was Microsoft’s solution here? Did the simplify the system? Change how they handle the notifications? Actually make their software use the 2, 4, or more CPU cores every computer has nowadays and some intelligent design to work out what the user’s doing? More intelligently design Control Panels and applications so that THEY better integrated into UAC? Perhaps some of this was done, at some level. But thus far, the only thing I have seen is they added complexity by creating multiple levels of UAC notification. My Mother, Father, Aunt, Uncle, Sister, Brother, etc doesn’t know what UAC is, why it exists, what each of those levels means for them, and for fuck’s sake, they shouldn’t have to.

Much of my grief was put into very elegant prose by Alex St. John in the November 2009 issue of CPU magazine (http://tr.im/C3kN) Alex points out that Microsoft has 60,000 software engineers, none of which apparently, can write an algorithm to figure out if a file with a .jpg extension, is in fact a jpg image or something nefarious. He also points out that the number one question in Microsoft’s FAQ on downloading files is “What does it mean to download a file?” Holy Hell. I don’t agree with Alex’s 3-part solution that follows that, I think its still conceding defeat to the flaws that Microsoft is too bureaucratic to fix.

The point has been made many times that its not in Microsoft’s or the tech industry’s interest for the OS to be flawless. There is BIG money in supporting the Swiss Cheese Operating System, and bigger money still in claiming to provide security for it. I guarantee you the great-great-grandkids of Symantec and McAfee executives won’t have to worry about money.

Next Up…Software Distribution…its almost 2010… this is still a disparate, scary, and ridiculous process that should be standard, uniform and recognizable…WTF…

Since I doubt my response to Keith will make it through his filters (though maybe he’ll allow it) I’m posting it here.

It is in response to this ridiculous article about his first 24 hours on Snow Leopard:

http://blogs.technet.com/keithcombs/archive/2009/08/30/apple-os-x-snow-leopard-the-first-24-hours.aspx

And My response:

Keith…Some fast notes and your “solution”:

I have been running 12GB of ram in my Mac Pro happily for quite some time now on a 32-bit kernel. VMWare’s happy to tap into it, almost as happy as Windows 7 is to gobble up whatever i give it.

OS X in general has been running 64-bit applications side by side with 32 bit applications, seamlessly, since Tiger.

2 minutes on google shows me that your machine (MacbookPro3,1) isn’t in the table of capable machines for a 64-bit kernel:

http://www.osnews.com/story/22009

Also, why can’t you run 8GB of RAM? Because the hardware doesn’t support it, regardless of the kernel. Its even in the Technote YOU link to in the article. The max RAM of that particular series of MBP’s is 4GB. I believe that is the line of machines you can do some tricks to bypass that, but the official party line is they have a 4GB cap. And even IF you get it enabled, and use a supported set of SO-DIMMs, you STILL don’t need the 64-bit kernel to utilize it, the 32-bit kernel works just peachy.

And as to why they boot to a 32-bit Kernel by default, for the same reason that Windows 7 still has legacy code in it dating back to Windows 3.1. 64-bit kernels break stuff. VMWare refuses to start under a 64-bit kernel. Mouse drivers that use kernel extensions, etc, ALL of which break under a 64-bit kernel, UNTIL the ISV’s recompile them for 64-bit (which they should have been doing since Tiger anyway, but whatever).

But, much like Apple does with EVERY major change like this, they stabilize the feature, and put it out there for the geeks and the tweakers under the assumption that anybody who knows enough to CARE to run in 64-bit mode will know how to enable it. Everybody else will be happy cooking along with their already bad ass operating system.

Please, before you knock something, do your homework. I’m running 64-bit kernels just peachy on my Pro and my 13″ Macbook Pro. I have been compiling and running 64-bit software for 3 years now.

And lets face it, NOBODY out there expects you to find 10.6 to be the most advanced operating system. Especially when the one time i’ve seen you not tow the party line on this blog, the post was removed within the week. Google can forget, but my NetNewsWire cache is forever.

And in the interest of backing up my threats, a screen cap from NetNewsWire of one of Keith’s posts responding to the “Laptop Hunter” ads where he doesn’t tow Microsoft’s party line and support them blindly. He actually calls them out on bending the truth, and says he expects them to have more integrity.

After that post was removed, I stopping Following Keith. THAT kind of censoring is ridiculous, especially when there’s nothing in there that isn’t true. I just stumbled across this latest mess and had to respond.

Pollux is the latest addition to my ever-shrinking stable of machines, and is the first of a two stage process to replace my VM server with a trio of small, lightweight machines to handle my production server duties.

Prior to this month, I had Altair, Polaris, and Vega. Altair was my primary file server with massive internal storage, Polaris was dedicated to “services”, and Vega just kinda sat where ever I could find a hole with power and ethernet.

The Old

Vega’s primary purpose in life was to sit and wait for Altair to hail it during its daily cron run; Altair would shove down whatever file changes took place that day. Basically it was my fail safe if something catastrophic ever happened to Altair that its redundant drives and linux-y goodness couldn’t fix on its own. It also gave me a window in which to go in and yank a file that had been saved over or corrupted.

Polaris was the desktop i redid just a month or so before buying my Mac Pro. The previous parts were aging very badly, fans were dying at an alarming rate and I just wasn’t happy with the performance any more, so i dumped some money into it and beefed it out somewhat. Then when I bought the Pro, it got usurped into being my VM server, hosting 9 VMs across its dual core Athlon and 3GB of RAM. My grand vision was to have each service live in a separate VM, and I more or less succeeded. But the whole box was plagued by the timing issues that all AMD machines suffer from when running VMware, and I got sick of having to maintain all 10 machines (1 host + 9 virtual).

The New

I become interested in Intel’s Atom architecture when i first heard about it a long time ago. The concept of a lower power machine to handle my relatively light weight server duties was very appealing. I haven’t been disappointed. According to my Kill-A-Watt, Pollux sips 35W of power, versus the 180 or so Watts the current Polaris draws when idle. It has more storage in a single drive (500GB) than Polaris does across two (160+250), and the 2GB of RAM is almost as much as what 9 machines were sharing before. And the best part: Total cost of the machine was < $170. The barebones Foxconn case/motherboard combo was $95, plus $23 for a 2GB stick of PQI memory and $49 for a Samsung 5400 RPM 500GB hard drive. Ubuntu 9.04 Server was installed using a bootable USB flash drive, and the network install took all of maybe 15 minutes, i didn't time it. The 500GB drive gives me plenty of space to replace Vega's backup duties, which means that box has been deprecated and shut down. The VM on Polaris running BIND (for local DNS) has been migrated to Pollux and shut down, and Apache is running and doing its thing, though it won't be switched over to production until i get my database server moved.

The Future

Pollux is step one in what I hope will only be a two step process. This was the testing phase, to see if it was viable and what kind of performance I can expect. The next stage will be picking up two more of these boxes. One will be almost a complete twin to Pollux (named Castor, Astronomy buffs rejoice, everybody else go here). It will host a MySQL database, as well as Asterisk, and will have a dual core processor. It will still run Ubuntu Server, have 2GB of RAM and a 500GB drive. Polaris will be the third of the trio, and will be another dual core Atom, with 2GB of RAM, and potentially a 1TB drive. It will be running Windows Server 2008 for all those non-Unixy things that sometimes must be done. I’m hoping to have all this done by Christmas.

Final Thoughts

The experience with Polaris was a mostly positive one. I learned a lot just $^#*ing it up and having to fix it, working through the time problems (basically none of the VMs could keep time, they’d end up days or weeks ahead or behind the actual date/time), and getting a feel for the capabilities of VMWare outside of a true production environment where people would expect, well, results and stuff. At home though, it was really quite overkill and unnecessary. Scaling down to two Ubuntu Server instances and a single Windows Server instance will simplify my management tasks a lot, and removing the giant furnace that is Polaris should help me lower the temperatures in here a bit as well. As I type this its roughly 83º in my office..

It’s been said before, but the Windows Registry is the Mos Eisley cantina of a Windows computer.

You’ll never find a more wretched hive of scum and villainy.

As somebody who has made their fair share of registry hacks before, I have been to places most people don’t dare to tread, changed and added keys with impunity, and…*gasp*…deleted things without backing them up first! This isn’t some average joe we’re talking about; I had a documented series of registry hacks that I had to do to every time i reinstalled Windows, and it got to the point where I was needing to use nLite to integrate them all for time’s sake (at one time I was on a 2-3 month reinstall cycle for my main workstation).

This kind of relationship usually is fruitful, a geek and his operating system. So intimately entwined in each other they are; the geek knowing the deepest secrets of his system, the system knowing precisely how to piss off the geek in as few CPU cycles as possible. But my relationship with Windows grew bitter in its waning years. I started to detest the behavior of the Alt key, with its toggling on/off state rather than being a simple momentary key like Shift or Control. The sheer and utter abuse of having a dedicated Windows key, and only being able to use it with 3-5 of the other 90+ keys. The systemwide penchant for inconsistency and bad program design. But none of this compared to the hatred of the registry.

The amount of bloat, the difficulty of manipulation, a single source of failure for the entire system, the abuse of putting things in registry keys that should have been in files, the use of GUIDs and SIDs for key names with nary a DWORD inside hinting as what it actually exists for and why. How about the ability to hide data? Yep, the registry, in all of its girth, is the perfect place to bury things, since so many legitimate programs do just that, why not illegitimate ones too?

As somebody who regularly supports other users, particularly those who, shall we say, are not as technically inclined, the registry posed a huge problem for remote troubleshooting. Here are some scenarios:

Think there’s a virus? want to disable any startup items?

Well is it in HKCU\Software\Microsoft\Windows\CurrentVersion\Run, or RunOnce, or RunForFrakingEver or is it a *system* preference, so it’ll be under HKLM\Software\Microsoft\Windows\CurrentVersion\Run or RunOnce? This is assuming you can get the user to open RegEdit, since its buried in some scary system directory that Windows warns you never to touch (because THAT is security…a clickthrough warning barrier…to hell with intelligent permission levels and on-demand privilege escalation).

Trying to troubleshoot wireless problems?

We should just blow away the wireless settings in your registry and see if it recreating them fixes it. Are you using a 3rd party wireless configuration program or the Windows one? At that point i’d hear an unspoken “WTF?!” in the silence, or maybe it was static from my cell phone, either way, I got the point.

I highly doubt I will return to any Windows platform as long as the Registry exists. The Mac’s primarily file-based approach for configuration data is vastly superior in terms of supportability, stability, and security. Throw in an amazing user interface that is both simple and powerful, not to mention customizable, and a growing number of applications that, unlikely their Windows counterparts, do not make me want to put a gun to the head of the people coding them, and you start to get a picture of why I have moved my entire family to Macs.

So I had an interesting challenge come my way today. I received an attachment I was supposed to look at, with the extension .pgp. Since I don’t use PGP regularly (with the exception of this time, I have never had to use it for any ‘production’ purpose) I went through the motions of re-learning how to decrypt PGP attachments, what I needed to download, etc. Fetching what I needed and digging around some, I recalled having to have a key of some kind. Right. Look me up on most keyservers and you’ll now find two entries ($20 says I’m the only Peelman in there, offer expires at the end of the week). One from 2005, one from 2008. What makes this funny/sad/aggrivating is the story behind the 2005 key, and the basis for why I hate PGP/GPG with most of my being.

So I’m roughly a junior in college, and finally in Purdue’s Computer Technology program, after a protracted experience trying to switch majors while the department continually moved the goal posts. I had recently gotten a Powerbook in the summer and my new geeky CPT friends and I were sitting around drinking beers, geeking out and talking about encryption (I supposed, how else we would have gotten here I’m not sure). The result of this is: at that point or sometime shortly thereafter I created a PGP key with some ridiculous passphrase. A passphrase that was forgotten soon thereafter. I’m pretty sure it was stored in the Keychain on my Powerbook for safe keeping (keep reading…) Later, while trying to retype it I confused myself so much that the mixture of letters, numbers, and symbols got lost in my mind, never to be found again. Regardless, its gone. No revocation key was created ahead of time (now I know that as a failsafe, you’re supposed to generate one and store it on a CD or something so that the key can be revoked in cases just like this. Pretty sure that wasn’t in any disclaimer i read at the time).

So the key was long forgotten, unused anyway, and 7 months after creation the hard drive on my Powerbook died (teaching me another painful lesson–now I back things up like some kind of freak). I can’t quite say that it was long forgotten, because last week, a peer sent the email in question and dug my public key out of some key repository. I always knew it would come back to bite me in the ass.

Let me pause a moment, for the uninitiated, and explain what PGP is. PGP is a form of Public Key Encryption. Through some mathematical wizardry that I can only hope I never have to deal with, you can generate a “pair” of keys, a public and a private. You post your public key out on various key servers, which then talk to each other and exchange the public keys they know about. You store your private key somewhere safe, this part is important since without the private key staying private and secure, your public key is useless (hence my dilemma). Through the magic of encryption algorithms, it is possible to take your public key, and encrypt the data in such a way that it can only be decrypted by your private key. Its a lot more complicated than that, but the basic idea is to be able to encrypt the data in such a way that only the intended recipient can decode it with his/her private key.

And here, my fellow geeks (and everybody else who has managed to make it through the tech jargon thus far), is where my reasoning for hating PGP lies. Unless you explicitly set them to, PGP keys never expire. They are eternal and forever and will live out on the netropolis of key servers until PGP ceases to exist and the last key server is shut down (not a likely event given its following). There is no way to revoke them without knowing the original pass phrase and having the original key. So 10 years from I might get another “secure” email that’s useless gibberish and undecipherable.

There are those who claim that this is the beauty of the system. That it prevents just some random Joe from revoking your key accidently or on purpose. Sorry, but I’m not buying. So for now I have a new key, that I have backed up in like 9 different places, with a memorable (but secure) pass phrase.

Short version

X509 PKI and S/MIME Digital Certificates FTW. Not only do you get a managed system, you get mandatory expiration dates, a web of trust that actually makes sense, encryption, and the best part: Integration with almost every major browser / mail client there is, with _no_ plugins, hacks, or third party plugins. Simply import the certificate file into the OS’s certificate manager (or if its Thunderbird, into your profile). www.thawte.com for more information. If you need/want a notary to get your name in your certificate, I’m good for 30 of the 50 points you have to get from the web of trust.

For the ten people who read this, and for both of you who haven’t heard about it through other means:

I have decided to not return in January to continue my Master’s degree. The experience this semester was not as rewarding as I had hoped it to be, ending with me giving up completely on Statistics nine weeks into class, and my ability to care and focus on coursework becoming less and less as the semester went on. I have always had a drive to learn and excel in coursework, though towards the end of my undergraduate degree that was starting to wane, but Graduate School didn’t provide the same feelings I got with any other coursework to date. My ‘A’ that I was carrying in Statistics may as well been an ‘F’ (and ended up being that way), I could not have cared less. I didn’t feel any satisfaction from getting As on homework, or doing well on the first exam. I tried to be excited about doing so well in a graduate-level course, it just never happened. My forensics research class was interesting, but research isn’t really something that lights a fire under me. Perhaps under different circumstances, or with a true into-the-workforce break instead of the 8-jobs-in-12-months scramble that 2008 turned into, I would have been better prepared and more distanced from Undergrad to make the adaption easier. As it is, I was just too burned out from my extended stay in the Undergraduate program to jump feet-first into Grad school.

My uneasiness with classwork, coupled with the odd hours I was keeping (a function of how my days were scheduled) has made this the roughest semester for me in my seemingly endless time here at Purdue. That said, I have really enjoyed my TA position, it has ended up being one of my more rewarding experiences at Purdue. I will admit that parts of it have been a mixed bag: the hours and lack of sleep have contributed to much of my…well, misery is too strong of a word, but general unhappiness with my situation isn’t too much of a stretch. I have met a lot of very intelligent students, and admittedly I was surprised and pleased with the knowledge and competence many of them displayed throughout the semester. To any of those guys (and gal) reading this: I hope you will continue to flourish throughout your stay in NET and that you get out in a shorter timeframe than I was able to manage! I also encourage you guys to contact me if you have questions or need help with anything in the future. One thing you guys will come to realize as you progress through the program is that your professional contact list can never be too long. I am always happy to shed some light on a problem or offer a fresh pair of eyes or advice on something. You can look me up on Facebook or LinkedIn for contact info.

So what now?

At the moment I’m looking for jobs. I’m trying to stay in and/or around Lafayette/West Lafayette to stay close to Kelly while she finishes up her DVM. However, given the current economic climate, I might not have that luxury. Hopefully things will work out for the better and I can get out of this professional slump that 2008 seems to have been for me. Here’s hoping 2009 is more rewarding and satisfying.

My Adium Xbox 360 Soundset is back on the rise! After creeping upwards in rank for a few months now, it has now reached the top 5 ranked Soundsets on the AdiumXtras page. At this writing its up to 2346 downloads (though admittedly probably 5 of those are mine…) and is current ranked at 4.4/5 Ducks, and is getting better!

Check it out, if you’re using Adium and you like the modest sounds of the Xbox 360′s dashboard (pre-November 2008 upgrade), go grab it, and rank it please!

http://www.adiumxtras.com/index.php?a=cats&cat_id=3

So this summer when energy started to go through the roof I decided it was time to “green up” a little more and get more energy efficient about things, not to mention lower my electric bill as much as possible. So shortly after classes started this semester I picked up a Kill-A-Watt and did some measurements.

First, some background: My two servers (Altair and Polaris) are both cobble boxes, computers i have built from old or retired parts, at least to date. Polaris was my primary desktop until I got my Mac Pro, Altair was truly a cobble box and was filled with a hodgepodge of hardware that was typically hand-me-downs from any upgrades I did to my desktop. Prior to the introduction of the Pro to my geekosystem, Altair was my primary server, it ran Ubuntu and did everything (web, mysql, tftp, voip, dns, sftp, ssh, smb/nfs/afp, etc). Once I had Polaris, my options opened up some.

Altair Specs (old)

• 2800+ AMD Athlon
• Gigabyte GA7-NNXP Motherboard (is it bad I still remember this without having to look it up?)
• 1.5GB DDR400 Memory
• Whatever Video Card was working at the time
• 4 Port Rosewill SATA Controller
• PATA – 200GB, 320GB, 320GB, 60GB
• SATA – 320GB, 500G
• RX-82U Enclosure -> 750GB, 250G

Polaris Specs (current)

• AMD Athlon X2 3800+
• 3GB RAM (DDR400)
• Foxconn mATX Motherboard
• 160GB SATA system drive
• 250GB SATA storage drive

So I just got my Pro but wasn’t quite ready to take the retired hardware from Polaris and merge it into Altair yet. So what did I do? As any good geek would have, I installed VMWare Server One of my biggest issues with having my single box was that it was incredibly insecure. Anybody who pwned any of the services i was hosting off of it, several of them to the outside world for my remote use, had just hosed my single server. The goal was to end up with this:

Network Diagram – Spring/Summer 2008

I didn’t miss by much. I spent the latter half of the summer slowly migrating services to several VMs (Powered by Ubuntu JeOS) and ended up with a pretty sweet, diversified setup that I won’t detail here since this is supposed to be about Altair, the file server side. With the burden of these services removed from Altair I was able to do the first major overhaul of it that I have done since I first started using it in early 2006. With the intent of minimizing its energy use while maximizing storage, I ended up with these specs:

Altair Specs (new)

• Intel Celeron 430 (1.8Ghz, 35W Power consumption)
• Intel DG45ID Motherboard
• 2GB DDR2 800 (single stick)
• 2x100GB 7200RPM Seagate 2.5″ Primary Drives (Software RAID 1)
• 2x750GB Samsung Storage Drives
• 1x500GB Seagate Storage Drive
• 1x320GB Seagate Storage Drive
• Ubuntu 8.10 Server Beta (needed the new kernel support for the new hardware)

Some notes about the install, while researching what to do for a reliable, fault-tolerant drive setup for my primary data, about 30GB and growing of stuff that I would prefer not to lose, ever, I stumbled across this (Syba’s 2.5″ Mobile Drive Rack (CL-HD-MRDU25S)):

at Newegg. Populated with a pair of 100GB 2.5″ drives run in RAID 1 (mirroring), it left valuable 3.5″ slots open for larger capacity drives that I can’t afford to run mirrored, and gives me 100GB of reliable storage to use for my boot disk and precious data. I also get a read boost from them because of the RAID1, which is always nice. Of course all of this lives in my Lian-Li case, which is a twin to the case Polaris lives in (I liked it so much I bought a 2nd one when they were on sale again), which acts like a giant heatsink and keeps everything frigid.

Back to the power usage. First, a benchmark for comparison: my Mac Pro and its 4 monitor setup (1×20″, 2×17″, 1×15″) draws roughly 400W of power while i’m using it. That includes M-Audio AV30 speakers, a Cisco 7960, Netgear GS108 Gigabit Switch, USB and Firewire hubs, mouse charger(s), a 1200VA UPS, you get the picture. I forget what the exact kWh used per day was but it wasn’t nearly as bad as one would think. I think by my calculations (rough at best) it costs me about $12USD / month to power my workstation 24/7/365. Polaris and Altair running headless were peaking around 410W of power usage, not terrible but not respectable by any stretch since the Mac Pro would smoke both of them put together in any benchmark.

While Altair was down I took the liberty to check the draw on Polaris and it was hovering at 140-180W of usage, so Altair was sucking down the lions share of the power, my guess would be all those old, inefficient PATA disks. Right now fully operational with its new upgrades, Altair is drawing about 95W. I cut the power usage of that box by at least half, approaching one-third. The impact on my electricity bill each month isn’t going to be a shocking as I would like it to be, but wow. All that money I’ve been throwing away for so long on powering something so inefficient.

Right now I’m down slightly on storage, as the drives I removed aren’t quite covered by the single 750GB drive that went in their place, but I have enough headroom to account the immediate future and with the price of storage continually falling its only a matter of time before another deal on 750GB or 1TB drives tempts me enough for me to add some capacity.

Long story short, here are some pictures, try not to make fun of my “rack” too much, it works, it keeps thing cool, though I’m having to keep an eye on the temperatures in the garage, it is starting to get kinda chilly outside and I want to make sure things don’t get cold enough to start having problems with condensation…

Servers – Front View

Servers – Rear View

Altair – Side Panel & Filter

Altair – Inside Shot

Altair – The Power Supply

Altair – Syba 2.5″ Drive Bay Adapter

Altair – The Main Drives

Courtesy of Gus Mueller:

Adobe Reader 9 is out!

I had a long rant typed up, but I just deleted it in favor of simply saying (once again):

I Hate Adobe.

And also that if at any point in the future I am presented with a way to never use anything related to CS3 again, I’ll take it, run, and never look back.

Looks like I finally got approved!

Check out my Adium 360 Soundset here.