Friday, November 17, 2006
We've been underground with projects to replace/upgrade NetWare with OES on SLES, improve our management processes for those new systems, and replace our MCI/Verizon Frame Relay network with a Sprint MPLS network - the latter of course involved us rearchitecting our Cisco VoIP infrastructure. Still to come, migration of our internet from a 6MBps Verizon pipe to a 12MBps Sprint circuit...lots of IP address, firewall, and routing considerations to plan for there.
Despite the busyness (not business), interestingly, I'm finding it increasingly difficult to continue investing myself as fully as I had previously into my work. I really do care quite deeply about the performance and success of my group, but most especially this company. I hate thinking that we pour 100% of ourselves into establishing world-class services on a shoestring budget, and that those efforts aren't appreciated or even leveraged by the businesses we support.
I find myself increasingly looking for ways we can use what we have developed by way of processes and skillsets, and branch out into servicing other customers - ideally, customers who clearly see the need for IT but are incapable of developing an IT competency themselves.
Short of that, it will continue to be very frustrating to have worked so hard at developing capabilities, which management is unable or unwilling to adopt as tools to facilitate positive change.
Wednesday, August 09, 2006
I've been working around vacations (others and my own), so things have been a little out of whack for the past month or two. School is about to start again, and we begin the implementation phase of our Netware-to-Linux server replacement & migration project tomorrow.
We've also been working tirelessly to finalize the deal for our wide area networking and voice services contracts. The kick-off for the three-year project (contract) was two weeks ago, and we're ready to place orders. Our HQ bandwidth will jump from 6Mbps to a full DS-3 (45Mbps), and each of our sites will at least double their current bandwidth as well, as we move to Sprint's MPLS network. Oh, and we save money doing it.
The next couple of months promise to be fairly hectic, but hopefully at the end of it all, we'll see a much more stable, much faster, and hopefully MUCH lower-maintenance environment.
Friday, June 23, 2006
"I've applied the "Wait 24 hours before replying" rule to some of the comments that have been made about Jack's departure. I would like to comment, since I was there when he came on board as CEO and watched the impact his tenure had on Novell.
Before Jack came on board, Novell, NetWare, and GroupWise were heading for a cliff. I joined as the head of marketing for GroupWise in April of 2001. I was told by the current management that GroupWise was dead and that my job was to simply help ease the transition of its death. I was given $50,000 for a GLOBAL marketing budget to market the product. And had to battle Product Management that said GroupWise 6 that had released in March of 2001 was the last release of GroupWise. 75% of the Product Managers and Engineers on the product had just been let go. At the same time NetWare 6 was launching with a multimillion dollar marketing budget. NetWare was king and GroupWise was a little bug in the corner.
Then Jack came in and things began to change, in a hurry. One thing he did was to make some very tough decisions. The company was heading in the wrong direction with a full head of steam. No one wanted to hear that NetWare wasn't the future, no one wanted to believe that Novell shouldn't be taking business away from Channel partners by being in the technical services business. He had to reverse disastrous decisions of his predecessors. One being the consulting business. The administration prior to him coming had decided to jump into the consulting business and completely shut down many revenue opportunities for channel partners, Jack reversed that. Which meant, unfortunately, that good, solid, Novell consultants had to go find new jobs.
He is the one who brought Chris Stone on board and oversaw the adoption of Linux. He had to keep rabid NetWare loyalists at bay at the same time move the entire company to a new direction. And he wasn't getting a lot of help from anyone, since, at the time, everyone thought he was nuts to be doing the things he was doing. He launched a multimillion dollar marketing campaign, one of the first for Novell in a decade and something everyone had been screaming for years to see.
Lots of people were let go, which was a good thing for the company but a personally bad thing for the individuals. But I sat in meetings with him where he held managers at Novell accountable for the first time in years. I saw him defend GroupWise from his own GroupWise management team which wanted to kill it.
I am not one to jump on the band wagon to kick a man on his way out. He was brought in to make some of the toughest decisions Novell has ever had to make in the history of the company. And now, he has helped make the next transition smooth for Ron as he comes in. You don't get to be the CEO of billion dollar organizations by being an idiot. You get there by doing the hard things.
Ron Hosvepian is the ideal person for the next phase of Novell's growth, so it is a great thing for Novell and especially for GroupWise. Ron is a big GroupWise fan. He and I have had several occasions to talk about the product, its impact in the market, and its future with Novell. I have been with the product since 1989, and I will tell you that things have never been so good. That the future of the product is bright with Ron at the helm, and it avoided the rocks of destruction due to deft steering by the outgoing CEO Jack Messman.
This is certainly unique to most customers. However, most customers were also not Novell Consultants in a prior life...I felt compelled to shed some light from my vantage point as well.
"Richard - thanks for writing this...it's always interesting to see another side of an issue, and this one was not very visible from the perspective of the customer.
What I will say, having been a former Novell consultant, is that I too came to believe that Novell shouldn't be in the professional services business. It wasn't because the people who were put in charge handled it like a cheap prostitute (do what you want, then discard it when you've had your fun), and it wasn't because we were stealing food from channel partner's mouths.
I wanted to see most of Novell's professional services given away. This was because much of the work we were doing was implementations at new customer sites. With respect to the channel, the reason we were doing so much of this work is because channel partners couldn't. The channel as a value added partner is a complete myth in my experience...few if any actually add value beyond moving boxes. Furthermore, those who can actually perform well have nothing to fear from Novell or any other competitor.
Novell purposely priced themselves at nearly double the going hourly rate for channel-level consultants. The emergence of Novell Professional Services wasn't a matter of Novell competing with the channel or stealing business - it was a matter of customers buying complicated procuts, and needing help to get them implemented properly.
To come full circle with that thought, my contention was that Novell should be giving away the services of people like myself in the role of a consultant to new customers with qualifying purchases - the costs should be wrapped into the product. The purpose of that would be to eliminate a significant barrier to adoption for companies who are big enough to need enterprise-grade products, but which may not have the expertise on-staff to implement them.
Our sales leads came from Novell salespeople. The companies worked with Novell to identify a solution, not a partner. Novell doesn't deal directly with as many customers as the channel does. For companies like mine today, we deal directly with Novell because there are no channel partners near us that have any competency in these products. Not suprisingly, there are fewer active Novell channel partners - you can't blame this on Consulting though (perhaps in another thread).
In an interview with ComputerWorld, Hovsepian himself identified the fundamental problem Novell faced (and still faces), but specifically as it applied to consulting.
CW: In hindsight, do you think Novell's acquisition of Cambridge Technology Partners was a mistake?
RH: No. I think the execution could have been done better. We didn't get a timely integration done. We confused our go-to-market model with our business model. What I mean by that is we took the business model of consulting and mixed it with the software business model. We translated that into, "OK, that means we sell consulting now." What it really should have been is using consulting to leverage our software products.
What he's saying is that Novell mistakenly turned the Consulting branch into a group of mercenaries, when they should have been used to talk to customers about their business problems and how Novell could help solve them. Those solutions would end up selling products, rather than simply trying to sell products + the help needed to get them installed. This is another area where Novell had capabilities that the channel simply doesn't, in virtually every case.
I don't blame Jack Messman or the CTP acquisition for the departure of many consultants - I know the names of the people who were responsible for the fatally flawed execution against the consulting model, and so do people like Gregg who were there with me. I also don't believe that Jack was solely responsible for saving GroupWise (pre GW6.5, I was perfectly ready to switch to NetMail and attack collaboration from another angle - I remain in 'wait and see' mode regarding the viability of GW7 as we go forward, but am cautiously optimistic).
I blame Jack for allowing the company and the board to become disengaged from each other, and the customers. That kind of leadership isn't easy either, I'll admit - but it's necessary. Jack probably would have served better as a COO, working with and answering to a 'proper' executive figure. In the end, I think it will become clear that Hovsepian is "the man", and were it not for all the things we've griped about for the past few years, Novell wouldn't have him in this position today."
Loyal readers will hopefully be able to see the truth of these posts, and let this information help them make informed decisions about their interaction with Novell going forward.
Thursday, June 22, 2006
The message with this information came from Ron himself, via mass e-mailing. The story will soon appear everywhere, and it will certainly be interesting to see if the market and industry views this to be as positive an event as I personally do.
It's no secret to readers of this blog that I felt Jack should have gone a long time ago. I had joked with co-workers that if we went to BrainShare '07, we'd buy a few hundred T-Shirts that said "SACK JACK" in huge letters, front and back. Looks like our wish was granted earlier than expected, but later than it should have been.
Shoulda bought that NOVL stock when it was below $6.....
Wednesday, June 21, 2006
I encourage any current ZEN customer to take a listen, and pay attention to the Cool Blogs. Good discussions going on there in a pretty open manner - I hope that the input they're receiving is being thoughtfully considered and contributes to meaningful change there. It's a good way for us customers to make sure our voices are heard by people of influence.
Monday, June 19, 2006
The earlier lull in my posting to this blog was due to participation in our tri-annual desktop replacement project. All desktop hardware is kept on three year leases here, and the replacement of that equipment is a pretty intensive effort. (Of course, we use ZEN very successfully to streamline and simplify the deployment. Over 350 PC's in 20+ locations, across 15 business days - not a single end-user file or application missing).
This time around, as you may recall, we switched from Dell to IBM hardware - this was just as IBM had sold off their desktop and laptop business to Lenovo. All of the equipment we're buying still says IBM, but we order through reps who have Lenovo business cards & paychecks.
So, on to the problems.
FAR too frequently, issues with hardware are met with "out-of-stock" stories from IBM/Lenovo, followed by part ETA's ranging from 5-10 days. We have 4 hour support contracts on all of this equipment, which apparently means a tech will be on-site within 4 hours to tell you that it's not getting fixed that day.
I know that Dell has been roundly rogered for it's support of late - certainly it's decline was one of the reasons we pursued an alternate vendor - but their desktop support looks like the 2nd coming of Christ compared to the B.S. we've been fighting with at IBM.
The latest instance of these problems have been embodied by two unrelated system board failures. Fans come on, drives spin up, but the system won't post. Both times, technicians over the phone indicated the problem was the motherboard. Both times, they quoted a parts ETA of over one week. Both of these incidents have occurred in a span of about 3 business days.
I'm turning up the heat rather briskly on our Lenovo reps - next will be IBM Global Finance. Perhaps the thought of reposessing 350 PC's that we are no longer paying for will help to get someone motivated to find a solution.
Friday, June 16, 2006
My team is also pretty small; we survive by being able to learn and adapt very quickly. There has yet to be a curriculum offered that is both A) comprehensive, and B) quick-paced...those that come close are usually C) too expensive. We do this for a living though...we can overcome a lack of training with enough 'seat time'.
That said, I acknowledge that there's value to be gained from purposeful, in-person training. The problem is that in-person training largely caters to the lowest common denominator (meaning it's as slow as the slowest student in any given class).
I'm firmly of the opinion that the 'halcyon days' of NAEC's charging $2,000 per student / week for OS and admin training are long, long gone. In Tulsa, for example, there hasn't been a viable training center for at least 4 years. I think it's because the courses are rabidly overpriced.
Even Novell's own ATT courses are priced extravagantly relative to the material being covered. $2,500 to lose an employee for a full week, to get drowned by one product? Is this stuff really that complex?
There has to be some middle ground - cheap, quick-paced "boot camp" training for people who will figure out implementation in their environment if they get a good, thorough overview of the features (capabilities and shortcomings).
I recently talked to BrainStorm about performing end-user training in our environment, for GroupWise 7. Having read their website, it looked like they could perform on-site training for the GW7 client at a very reasonable rate. Apparently, their website was out of date. When I countered the quote I received from their sales group with text from their own site, it was quickly changed to eliminate references to that pricing model. The new pricing model is nearly 5 times as expensive as we were originally led to believe, which completely shatters any value proposition their offering may have had.
The lack of a compelling training story has a domino effect for Novell in particular...if people can't justify the cost of training, it doesn't mean the product doesn't necessitate it. It just means that people can't make the dollars and cents work. Eventually, these people (e.g. Novell Customers) will have to ask themselves if it wouldn't be easier to use Microsoft instead? You may pay more for software, but hardly anyone is willing to hire people that don't at least know Windows and it's associated components.
Lobotomized monkeys can keep most mid-sized Windows / AD systems up and running. College grads with IS majors, however, probably don't even know how to spell Novell.
I've said it before, and I'll keep saying it - neither Novell or the partners who work with them will enjoy a renewal of success until the most significant barriers to adoption are removed. Novell should take some of it's cash stockpile and aggressively incent customers to attend training by paying for it. If necessary, they should kick in short professional services contracts to help new customers adopt new products - roll it into the product pricing somehow if they can. As everyone knows, the longer you've productionalized a system, the harder it is to get off of it.
Training shouldn't be viewed as a 'product' that needs to be 'profitable'...that mindset immediately detaches it from the products around which training is based - and it's the real products that pay the bills. This isn't to say that training should be given away - but it should be run in such a fashion that it's contribution to the bottom-line is negligible. I know this will probably offend 'training centers', but honestly, you add no value by getting some can't-hack-it-for-real drone to read stuff out of a book. If you want to charge a premium for your training, that should be off-set by the experience and quality of instruction. Far too often, my team and I have been participants in or audited classes where we were the smartest, most knowledgeable people in the room. No way...not for $2,000 or more a week.
Novell should know that it's hard enough to get in the door - whatever you do, once you're in, don't give people easy excuses to throw you back out. Those are already in plentiful supply.
Thursday, June 15, 2006
One of my team members brought me a RAID controller from a failed Dell server (go figure!), and indicated it needed a replacement. This was a sandbox system without warranty, and nobody wanted another PERC.
Not having fiddled with hardware in a while, I went to the Wikipedia to search for "PCI". I looked at PCI, PCIe, and PCI-X articles - all of which had photos. From them, I learned that PCI Express and PCI-X aren't the same thing, and that the card I had in my hands was PCI-X. This allowed me to spot the right replacement part in about 10 minutes, including a call to our preferred vendor.
I know plenty of hardware guys that I could have asked "What's this thing?", but they were busy - and I didn't need them.
Fair warning. If the "value" you're currently providing is largely due to things you know (which others don't), you may want to be ready for a new career or retirement in the next 5 - 10 years. I firmly believe that knowledge has been commoditized - moreso every day. All people will need to know, is how to find it. The value those people will need to add is the ability to take knowledge and do something meaningful with it (e.g. solve problems). That's not knowledge, that's talent - something that can't be commoditized.
At least I don't think so...(grin)
Tuesday, June 13, 2006
Some are indecipherable. Some are bravado-laden, and some are frought with stammering and uncertainty. Some are from companies I'd heard of, but most are from nobody's. None of them are made with an intelligent human being on the other end.
I've established a heirarchy, detailing the rank of common sub-human life form career paths. It is as follows:
- Multi-level Marketing Disciple
- New Copier Salesperson
- Used Copier Salesperson
- Copier technician
- New Car Salesperson (domestic models)
- Used Car Salesperson
- Door-to-door salesperson
- Cold-calling technology salesperson / technology recruiter
- Multi-level Marketing Leader (e.g. Amway's DeVoss & Van Andel, Arbonne's M0rck, etc)
In my opinion, this group sits just above convicted felons, and just below the guys who mop up the floors of adult theaters in New York City. As you can see, cold-calling telemarketers are nearly at the bottom of the list.
Anyway, one of my great weaknesses is assuming that people are not in fact that ignorant. If I could get over this fundamental belief, I'm certain I could become a multi-billionaire. After all, it's the perception of customer ignorance that convinces vendors that calling people out of the blue and taking up their time to talk about stupid crap that they had better already have in hand on the off chance that they are in the market and would like to talk to you is actually a sound business development strategy.
The day I have to cold-call 'prospects' is the day you find me hanging by my tie from a water pipe above the drop ceiling.
Thursday, May 25, 2006
Yesterday, they sold the renamed consulting branch - Celerant - for about $77 million.
Now, call me silly, but I thought the idea was to buy low and sell high.
Looks like Novell basically paid close to $40 million a year just to get rid of Eric Schmidt, and replace him with an even more misguided and incapable buffoon (aka Jack Messman). Good work if you can get it.
I've said it a million times - Novell has an absolutely wretched history of execution against a strategy...I continue to believe more and more firmly that this is due to Novell's complete misunderstanding of the corporate landscape. If you've no clue what the problems are and where the markets are, you're going to have a hard time developing a meaningful strategy.
Never one to complain without providing suggestions, here then are mine for Novell.
1) Sack Jack Messman.
2) Remind all of the new CxO's that there's NOTHING about which they can be complacent.
3) Start giving away professional services to get your products in the door, or reimburse the channel to do it for you.
4) Stop with the idiotic advertisements. Nobody understands what you're talking about. Don't print this garbage just for the sake of shutting people up who complain that you don't advertise.
What the hell does "Software for the Open Enterprise" mean??? Why would an executive at any firm want his 'enterprise' to be "open"? Does anyone have a significant problem with their enterprise being "closed", or do they have problems improving efficiencies and delivering value? Messages that are this muddy are a complete and total waste of time and money. Take a cue from IBM or Apple for the love of God, I'm sure they'll share their marketing people.
Here's the text, verbatim, of Novell's ad in Computerworld from 5/22/06:
"DEFINE YOUR OPEN ENTERPRISE
In my Open Enterprise, I can add applications, platforms and users without adding costs, problems and barriers.
Multiple operating systems. Distributed servers. Proprietary hardware. Without a data center solution capable of integrating and managing them all, you're losing time, money and productivity. Data Center solutions from Novell cost-effectively consolidate, manage and support all of your disparate systems on the most mature, scalable and secure Linux OS available. So you can centralize system management. Run leading business applications with the power of LInux and open standards. Lower costs. Improve service levels. And turn your data center into a profit center.
Data Center solutions from Novell.
This is the way to build your Open Enterprise."
I'm a Novell customer with just about every product under license, and I have no idea what this ad is selling. Worst of all, above the vacant gaze of some stuffed shirt, is a brain map with more meaningless jargon like "open", "build", "speed", "high performance", "immediate", "leverage legacy systems", "blatant". Who in the #^&*@ approved this idiocy???? It's sickening. Stop it.
Those are my suggestions. Print them out and use them to start your next fire if you will. I for one hope SOMEBODY in Utah starts to get it.
Tuesday, May 16, 2006
GroupWise in particular has been a very frustrating product to own and operate, especially over the past two years. We've had serious discussions internally about migrating away from it in favor of, well, anything else that will work.
Enter Ken Muir, the former ZENworks golden child who decided he wanted a new challenge and left that product for the GroupWise arena. Recently, I was fortunate enough to have a brief but in-depth conversation with him (via e-mail appropriately enough) about the fundamental issues I see throughout Novell from the perspective of a customer.
While I cannot share the details of our conversation, suffice it to say that I for one am willing to give Novell the chance to demonstrate they 'get it' again. This chance will be manifested in the form of GroupWise 7 SP1, which is expected to release sometime in the next month or two. When it ships, we will implement it in production - on Linux - and use it as the starting point for server hardware replacements that we must conduct nationally.
We continue to have successes and issues with OES SP2 on SLES, but every day provides another opportunity to learn. My hope is that the issues we encounter continue to be resolved by education, rather than becoming bug-finding expeditions (as we have been doing presently with other Novell products).
It's a difficult, difficult time to stay loyal to Novell. I keep asking myself, "If we were all Microsoft, how much time could we devote to other initiatives?" My experience tells me that we'd be swapping one set of problems for another. Microsoft OS'es aren't nearly as high-maintenance as the press would have you believe, but they do have maintenance requirements. They also have numerous feature deficiencies which we somewhat take for granted - this is what I fear the most. The decision to use Novell vs. Microsoft for your core infrastructure is never apples-to-apples. There are tons of considerations, intangibles, and dependencies, and these differ wildly in every organization.
For now, at least, we will remain Red.
Thursday, May 04, 2006
It used to be a big deal for me, as a Novell die-hard, when Microsoft effectively announced to the world that "this whole big OS development thingie is, well, darned tough to figure out, and, well, sometimes it kicks our little wet Seattle butts."
This was glaringly evident when you could set your calendars by Novell product releases. Every three months, roughly, you could count on a pretty significant product release.
Something interesting happened to all of us as we've matured in our profesisons. Quality started to matter - a lot - especially when we found ourselves without it.
Novell's insane rush to meet these ridiculous, blue-sky deadlines for product releases didn't result in better products. Quite the opposite - the quality of Novell products across the board has declined rapidly. All too often, product groups re-prioritize based on objectives that don't think nearly far enough ahead, and wind up releasing software that is a complete mess. NetWare. GroupWise. ZEN for Servers. BorderManager. Novell Portal Services. All of these products used to be great. Or at least full of promise. Now, some of them are either garbage, or soon to be deceased.
Only GroupWise is too stubborn to realize it's a dinosaur and needs to reinvent itself to remain relevant, but e-mail as a platform is one of the hardest for companies to replace. By rights, on it's merits, it should be a relic of a past long and thankfully forgotten. Sadly, it's not.
I'm surprised most of all by my own indifference to any announcement that the ship date for Vista will slip. I think to myself "Let it slip, if that means it'll be right when you finally ship it." This is a lesson Microsoft seemed to have learned over the past 10 years, and I'm glad they have. It's a lesson I wish very much that Novell would get through it's thick, old, dinosaur skull.
Monday, January 30, 2006
But it's cool.
I bought my wife a good, German made turntable for Christmas. She has old LP's from her childhood that have sentimental value, and I have 78 laquer records my grandfather listened to in the 30's and 40's through the 70's of greats like Bessie Smith, Louis Armstrong, and Duke Ellington.
Much of what we have is old, and sounds typically record-like. Hissing...pops...distorted high-end, etc. My wife did have a couple of LP's that were 'modern' and stood a chance of sounding good, though. Billy Joel's "The Nylon Curtain", for example. It was a little dusty, but sounded unlike any CD I'd listened to.
Keep in mind that I'm a child of the 70's - I never had a turntable that wasn't in a tiny blue box with a huge white tonearm. Now, I have a really good home theater / sound system - if LP stands a chance of sounding good, it'll be on my rig.
We found a small shop that sells used LP's, and they had a fair selection. Three or four Billy Joel albums, a lot of Beatles in great shape, Doobie Brothers, Chicago 17, Police (The Singles), Smithereens 11, and a Billie Holiday album that was really good.
The Smithereens album was basically unplayed - it looked immaculate, as did several of them. We happen to own the same album on CD as well. I've listened to both - the LP on my new turntable, the CD in my Pioneer DVD player (using S/PDIF optical connection). The CD has more SPL, and I suppose you can say it has better dynamic range (bigger difference between quiet passages and loud ones). I can blast the CD until glass rattles, it sounds clear as day.
The LP sounds like I'm listening to the band perform. It sounds live...like I'm there. It's most noticable with strings ("Blue Period" on Smithereens 11 opens with a cello or something - it sounds 100000x better on LP than CD), and on the low end. CD's sound muddy by comparison. On LP, I can make out each note the bassist is playing, and the kick drum punches you in the chest at far lower SPL's than CD.
In my mind's eye, I know why this is true. It's the same reason that TV and movies don't look as 'real' as life. It's not about resolution or anything like that...it's about reassembly. In life, music and motion come to us instantaneously via our senses - you cannot measure the gap between one 'frame' and another, because there aren't any...it's one continuous feed, called reality. If you stop a movie on one frame (especially a moving frame), you see blurs. Or, you see that an object has moved from one spot to another between frames - even if barely, it's there. No matter how fast you capture the pictures, movies are basically high-speed still frames. Each still frame misses something - the difference is what we call reality.
The same is true with music. CD's bombard your ears with "still frames" of audio, at a rate of roughly 44,000 pictures per second. You're hearing what the sound measured on a microphone at 1 fraction of a second, played back very very frequently. What you get is a pretty good picture of what was being recorded. Unlike video however, it's possible to record audio in such a way that there is no loss....hence, analog LP recordings.
I always blew off LP bigots as unnecessarily esoteric. The convenience, longevity, durability, and quality of CD's are unmatched. I'm a music lover, and I love CD audio. But I love LP audio more. It takes more care & caution, and you appreciate it more when you hear it. All the garbage you hear from CD bigots about LP's being scratchy and filled with static-pops are made by people who've either never heard LP's, or never heard them played on good equipment. This argument & information is lost on them. If however you're an open minded lover of audio & music, I hope you too find the chance to splurge on a high quality turntable & sound system, and invest in some well preserved vinyl. You won't be disappointed...how often can you say that?
Wednesday, January 04, 2006
Part shortages, staff retirements, etc. have all contributed to a far-below-satisfactory experience with the x-Series support model. Our servers have been in the midwest, southwest, even in the U.S. Virgin Islands. Repairs that should take a few hours, end up stretching into days.
Given the number of server systems we have to replace in the coming months, this had better be a series of unfortunate events - not a preview of woes to come.
Dell is out...if IBM fails, I guess it leaves HP/Compaq and Gateway (ugh).