I apologize in advance for the lengthy-ness of this post. but this is what was said by my own power company "Portland General Electric" up here in the Northwest. This is an exact word for word transcription of PGE's portion from the meeting and after this particular gentleman spoke I was concerned about 2 things: #1 at how handily the spokesman "dodged" the question of where they were in testing and seemed to glaze over their where they are in the remediation process in general. #2 that perhaps I am missing the boat on how far PGE really is on their remediation. I'm more concerned now that they're not gonna make it and would like some other input/feedback to warrant or diffuse my suspicion/concern.

Now to provide a little background:

In October a different spokesman for PGE told me at a public meeting that PGE would be compliant with their 10 "mission critical" systems by the first quarter of '99. That has now slipped to the second quarter of '99 and the number has jumped to 246 "mission critical" systems.

In a letter a friend received from PGE back in July, they stated how far along they were in the y2k repair process and how early they began working on the problem ('96). On the same day they reported to the Oregon PUC that they hadn't really done anything and had nothing to report. When I attended the last PUC meeting in October of '98 I was shocked at how little another PGE spokesman reported. His entire presenatation lasted but a mere minute and he basically said nothing. The other thing that doesn't seem to ever make the press but really made me nervous was the speakers "body english" both before and after his presentation. Sitting hunched over with both legs and arms crossed, as if to say "don't notice me, don't ask me anything!" During his presentation his voice quavered and stuttered and he spoke very quickly.

(am I being too paranoid here?)

I've snipped some of the initial greetings. My comments in (brackets)

*begin transcription*

My name is Tom Matthews I'm the general manager for information technology and also the executive sponsor for PGE's year 2000 project.

We've actually prepared a communications plan a year ago, we took it off and dusted it off this fall...um...and we're in the process of implementing it right now we knew that from the customer's perspective they would be...um...much more interested in the year 2000 issues as the dates got closer...

(I called & left a message on PGE's y2k phone line requesting information of some kind back in early October '98, I still haven't received anything-short of the typical bill inserts & reassurancces that everything's going to be fine) The proactive peice to the communications the newsletters, and the website, um...the hotline, the PGE 2000 hotline that we've got out there...and our...y2k...uh...page on the internet are all available for the customers to use we're in the process of updating those things regularly now to...to include the most current information. And we're trying as best we can to not just have generic information but to try to segment it by different customer types out there because we know that different customers need different things. (How about the truth? don't we all need that? Their website has the typical smiley face info, everything's under control, etc.)

um...we've met with all of our commercial and industrial representatives and we've trained all of them to be....um...y2k spokesmen..and to...we've also got them into the communications loop in terms of the types of information that our business customers both small and medium and large business customers want.

Um...As you can expect most of our large industrial customers want to meet with us personally...there's about 60 of those...I've been personally going to all of them, and they typically last about 2 hours and we do that under a non-disclosure agreement where we basically will answer all of their questions....

(I found this to be interesting-presentations to large businesses last two hours under a non-disclosure agreement, here in a public forum, they last a few minutes and we get very little information)

um...those have been I think pretty helpful for that...uh... those large significant, significantly large customers who depend on us...for ...their plant operations 24 hours a day.

(helpful to most of us in the portland metro area...including my water company?)

We've also trained all of our governmental affairs folks to meet with service organizations and to meet with the small communities um...that they serve. And there's been several of those meetings out there,then, this...we've found out that the small group um y'know want, 20, 20 person meetings tend to go really well...and the messages seem to be getting across to these different...um.... groups.

Of course we're tied in with western systems coordinating council and we're also linked with bpa (bonneville power association) and with pacific (pacificorp) um... We've really talked about a need to do an industry wide um...types of proactive communication here in the state of Oregon I want to introduce Patrick Stupik, Patrick who's a long time PGE amployee an Enron employee (Enron is PGE's parent company) who's been reassigned um...on his own accord he actually bid for a job, I've got to ask him why one of these days...but he's going to be our public information officer for y2k today's actually his first day so we thought we'd bring him down here and introduce him to you and um...

PUC Commission Chairman Ron Eachus: We knew there might be an opening anyway...

(not sure why this is so funny, y2k people stepping down left and right...?)

Tom Matthews:He hasn't yet asked me what happens after...y'know, after...uh...a year from now so...

(we all hopefully breath a big sigh of relief?)

we in the electric industry of course want to do something as a group proactively and WSCC is really encouraging that through the northwest power pool but we would also like to bring on northwest natural (gas) and the phone companies and one of his first assignments is to try to build a coalition here of uh...and a way so we can go out with shared messages and shared things from all of the um...utilities here in the state of Oregon and work on common press releases and just messaging that is consistent so that people get the idea that utilities in the state of Oregon are working very hard on this issue and it's very important to them.

Um...In terms of testing...

(this is what I'd been waiting to hear...how far along PGE was in the process)

I'm going to go real quickly through this, but, um, we've developed a whole test environment on our mainframe and we've got a test lab setup and every bit of code that we have, every application that we have is being run through this...this test environment it allows us to...to continue to operate on a day to day basis and do very um...um...systematic um...testing out there...at a real rich level of detail and not have it impact our day to day operations, and we've got staffs of people that are focused on doing just this. um...one of the questions that staff (the staff of the PUC) had really asked us to address is around the testing and certification process um...I can assure you we have developed a methodology that is just incredibly detailed...around this to...in terms of establishing a baseline ...the renovation plans that we have...um...all the testing that we do after we, after we renovate the code or make the corrections... and then there's ...this whole thing about...the r-release question that was talked about a little bit earlier, but every time we upgrade something we, we, we just freeze it...and uh...then we go back and have to do all the y2k testing on it again so we're really watching, closely any um...upgrades that we're allowing to any of our applications today. And then we have a very uh, um, rigorous certification process too, before we put these systems back into production.

(still nothing about how far along in the process they are)

I wanted to give you kind of a magnitude of what the issues really are for a utility, I'm sure, I know, I've shared this with BPA. and um PP&L probably has the same but there's a chart that looks like this.

(he refers to one of the pages in his report)

And you'll see up on top...there's a column that says total. And that total 3,382 that is the total number of applications and embedded devices embedded systems that we've got at portland general electric. So you can see it's not huge.

Within that we've got a...a rating here low, medium, high, critical and critical of course is really focused at the safe, reliable delivery of energy... only 246 out of the total...are what we call critical...and if you look down below that's only seven percent...so if you look at the total number of systems, embedded, y'know embedded systems and applications and items that, that uh...we have that we depend on to serve our customers here in the state of Oregon only 7% of those are mission critical. And...and very few of those actually have caledaring events on them that would be impacted by the year 2000.

(back in october I was told by a PGE spokeswoman that there were only 10 mission critical systems and that they would all be fixed by the end of the first quarter of 99'. PGE now is saying mid 99'. I hate to be so negative, but if PGE claims to be 93% complete with their remediation in the near future, I'm not going sleep very well)

I think that was a nice slide based on our...um real agressive um...assessment and compliance planning phase it just kind of gives you a magnitude of what the issues are...it's not ...it's not our whole environment out there it's a very small part of it...but we're still very focused heavily on that.

(So how far along are you on fixing the 7%?)

The next page is uh, is a schedule. And the dates that I want to put down there I highlighted in blue for you was the critical systems for implemation...and reinteg...er, implementation and reinteg...integration which...um we have um...we fully plan on having completed by the end of the second quarter...and that's consistent with the WSCC and NERC processes.

You can see our contingency planning efforts go on we've actually submitted our first contingency plan to WSCC we're in the process of compiling our second one it's due in mid-June. Um... just some updates we've attended, last...two weeks ago we attended the up-re (?) conference on contingency planning back in St. Louis we had, PGE had seven people there...from...representing the major areas of our company...um...as I mentioned we submitted our draft plan to the WSCC um...we've completed our inventory, our assessment, and our compliance plans for all of our items and we've assessed the potential effects of all...all of our applications... and we're well on our way now to doing all of the corrective fixes that need to be done.

(then why do you sound so nervous?)

Um... the mission critical task force has been selected to um...validate...all of our mission critical suppliers...we as a company have a little over 1100 suppliers of those 1100, 109 are mission critical and they tend to be the, the telephone companies the fuel providers that we've got for all of our plants...um...a...and just the...the types of businesses who really rely on to be sure that electricity is delivered safely and reliably. And the last page um real quickly is uh... just a project update on our contingency planning efforts

We exchange information weekly with uh...bpa um...I've got 3 members of PGE on the WSCC's contingency tanning plaskforce (I think he meant "planning taskforce" ) one of those members is actually on the scenario planning team. They're the ones that are going to develop the operating guideline....er the operating events... or, or scenarios that um, WSCC will use to configure the um...the ass... all the assets...that, um, power production and delivery assets...um, in the North, era, in the western states and he's gonna.... a big help in helping ...provide, um, us with the types of information we're gonna need uh, eventually to figure out how to best configure our particular, um assets.

(what?)

We um, hosted a 2 day work session for the WSCC...um, back in October here in Portland...and we've submitted our first draft of the industry...wide...drill that was already talked about on April 8th & 9th that's an interesting date it's the 99th day of 1999 and it's, so one of the level 3 dates or critical 3 dates that we're look ing at. The other is of course September 8th & 9th and that's the 9/9/99 date.

Those are both drills, um, there, they, hit, they're not tests one of the things that's come up with us, we, we're physically not going operate the system we're not going to cause....um...our customers to experience any, y'know...any voltage or...or fluctuations in their level of service, it's been...a real important issue to our industrial customers they don't want us...out there opening closing devices and...and having them negatively impact them. So we're...this drill is more focused on, the first drill on communications the second on how we w-work as a collective group under the scenarios, the planning scenarios that are being worked on...um...right now. so that's...

(laughter & joking about not having to worry about re-setting our vcr's)

If you did have any questions around some of the issues with WSCC I might be able to address them too cause we have been really involved at working with them

Commissioner Joan Smith: I have another question Tom um, when I look at the matrix that we were given it looks like every other company's estimated what percentage of completeness that they've arrived at but PGE hasn't . Do you have any specifc information you could give us in that regard?

(Excellent question! What I've been waiting to hear! Listen carefully to the answer:)

Tom: Um, yeah, one of th... there's two general methodologies about looking at y2k and we've been, PGE's been real reluctant to give you dates around testing because ...we didn't use testing as part of our inventory or assessment process.

(testing comes later, doesn't it? After repair/remediation? Keep listening:) We...we developed a very sophisticated methodology to go in and do an inventory, an assessment, a compliance plan and the reason that... let me give you an example of why that's important. We asked all of our people in our plants, "you give us your best information on how many embedded systems you think we have out here", and this work we did almost two years ago...and they gave us a number. And then we used a...con...one of our consultants...and...in partnership with them we devel...developed this process of how we would go through and identify and inventory um...embedded systems in plants, well guess what?

Through our very vigorous process we...we uh found three times as many embedded systems in...in these, in these plants and that's because we used block diagrams, we basically took the electrical systems in those plants apart and went through it methodically and pulled all of this information out.

Um... These were engineers who helped build these plants for heaven's sakes...we thought they knew where everything was. Um...so...w-w-w...this issue about y'know about not wanting...not being real...open about um...test, a lot of companies have chosen to go straight to test as a way to identify to find things we have not done that. Um when we ...our...our testing is only done after we've finished the inventory...we've assessed the impact, the y2k impacts, we've developed, developed, a compliance plan around remediating those peices...we've done the corrections...then we go to test. So test is very late...peice of our sense about...insuring that... the corrective actions that we've taken have...um...are really focused on... allowing us to del...deliver electricity safely and reliably... so I think, Commissioner Smith, that's probably been the biggest issue that , that I've seen out there in it's just a different approach.

Commissioner Smith: thanks.

(What? At this point I almost stood up and said out loud: "WHAT?")

PUC Commission Chairman Ron Eachus: Any other questions or comments?

*end transcription*

(from there they moved on to the natural gas companies)

To say I was stunned would be an understatement.

What I understand since I've started following Y2K was that fixing the problem is as follows: assess/inventory, repair, test, (fix problems discovered) re-test and reimpliment.

I haven't heard of any other way of doing things, and listening to Mr. Matthews' comments he seems to even reinforce the standard model without saying how far they are on fixing things. He seems to imply that testing can be done early to find problems or to help in the assessment phase, but is that what's really being done out there? It would seem backwards to just test a large system just to see if it'll crash, the risk of that kind of testing seems too high.

So it sounds like PGE hasn't begun testing yet because they aren't very far along in their repairs (especially of their mission critical systems). This probably explains why Mr. Matthew's presentation sounded so similar to Don Knotts's town address in "The Ghost and Mr. Chicken."

"atta boy Luther!"

I guess I'm also flabbergasted that this answer made it past the PUC without even a statement of clarification.

Am I way off base here or is this guy blowing smoke? Any input would be appreciated. I'm really trying to have a good attitude about Y2k, that everything will be well.

thanks

Jeff

-- Anonymous, April 19, 1999

Answers

Jeff, what an interesting meeting you've reported about! I can understand why you're taken aback by the lack of even a rough estimation of completion estimates. At least the PUC rep asked the status question, which is more than some PUC's have done according to some local reports I've gotten. Too bad she felt she could accept the answer without pressing for the specifics she'd originally asked for.

I've reread PGE's third quarter 1998 SEC 10Q, and accessed their recent 10K filing for the end of their fiscal year (Dec. 31, 1998). Interestingly, in the 10K they stated that the estimated completion charts were "as of March, 1999". Before I get to the changes between the earlier 10Q and the recent 10K, I want to let you know that of all the SEC filings for utilities that I've read over the last months, PGE's are right at the top of what I call "legally bound and gagged". By that I mean, while all the different utility Year 2000 statements are couched in legal terms and various disclaimers, there is not a single paragraph in any PGE filing which isn't wrapped up in legal sentences galore. The impression is that this company is paranoid about saying anything of any difinitiveness at all. They are certainly going beyond the norm in pointing out all the possible ways failures might occur, despite their best efforts. While the possibility exists that they don't want to report completion details because the details would not show their progress in a good light, there is also the less skeptical explanation that their legal team isn't about to let anybody get a handle on *anything* which might possibly be used in future for a lawsuit, even if progress is good and the likelihood of failure is considered miniscule.

So we are confronted with the two options of "If you don't have anything nice to say, don't say anything at all," and the legal "Don't lock yourself into anything, even if it's good." Or I suppose it could be a combination of the two.

I'll give you an example of PGE's cover-all-the-bases reporting. In both the 10Q and 10K, the Year 2000 statements begin with exactly the same paragraphs, part of which states, "Under the Plan, PGE will continue to inventory its mission-critical computer hardware and software systems and embedded chips .....assess the effects of Year 2000 problems on the mission-critical functions of PGE's business; remedy systems.....verify and test the mission-critical systems to which remediation efforts have been applied; and attempt to mitigate those mission-critical aspects of the Year 2000 problem that are not remediated by January 1, 2000, including he development of contingency plans to cope with the mission-critical consequences of Year 2000 problems that have not been identified or remediated by that date."

When you first read the "continue to inventory", etc. it certainly sounds as though PGE hasn't even finished the beginning stages of its project, doesn't it? Yet later in the filings, the charts on both the 10Q and the 10K indicate that Inventory, Assessment and Analysis are "Complete". Initially I thought perhaps they'd just left the reports the same as they were from the very beginning and only changed a few lines for successive ones because they were in a hurry or careless. However, then you come across a paragraph just prior to the completion chart which states in part:

"It is important to recognize that the processes of inventorying, assessing, analyzing, converting (where necessary), testing, and developing contingency plans for mission-critical items in anticipation of the Year 2000 event are necessarily iterative processes. That is, the steps are repeated as PGE learns more about the Year 2000 problem and its effects on PGE's internal systems and on Outside systems....As the steps are repeated, it is likely that new problems will be identified and addressed. PGE anticipates that it will continue with thes processes through January 1, 2000 and, if necessary based on experience, into the Year 2000 in order to assess and remediate problems that reasonably can be identified only after the start of the new century."

I've puzzled over PGE's "necessarily iterative processes" since I first read those lines last year. No other utility filing I'm aware of states that all the project phases are to be run over and over, although a few did state they would test systems repeatedly. That whole paragraph gave me the impression that PGE deliberately did not want anybody to be able to pin them down to any exact estimates, or else they were very unsure of themselves and the quality of work being done. It also sounds as though they are planning for a fix-on-failure mode, but that could also be a legal contingency statement.

Are you even more frustrated by now? *smile* There were very few changes in the more recent filing. The Conversion, Testing, Y2K Ready and Contingency Plan rows for Mission Critical Internal Items remained "In Process" as it was in the 6 month earlier report. The Mission Critical Outside status for these categories did change from "To Be Initiated" to "In Progress". The estimated Completion Date for the Testing and Y2K Ready rows did change from September 1999 in the earlier report to June 1999 in the March estimate. In fact, all of the recent report's estimated completion dates for Conversion through Contingency Plans are now June, 1999. This indicates more progress than they originally planned, or else they are setting their target dates to be in line with the NERC deadline. Since they stated their assessment was completed in October, 1998, I personally consider finishing all conversion, testing and implementation in eight months to be a rather ambitious task, especially since "Conversion" was stated to still be "In Progress" in March, with only three months left to their target date for completion. (And the rep you cited stated they were doing their testing at the end of their phases.)

There is one clue which fits some of the pieces of this puzzle together, and that was in your post of what was said at the meeting. Despite the run-on sentences or lapses of the speaker, he did state that the in-house employees' initial estimates of how many embedded systems were in the plants turned out to be one-third of the number found when a consultant was put on the job.

"Um... These were engineers who helped build these plants for heaven's sakes...we thought they knew where everything was."

So it does appear that PGE DID have to "iterate" their project phases when they discovered their in-house engineers had missed two out of three embedded systems in the initial inventory. A do-over must not have been a confidence builder for them, as evidenced by the " for heaven's sakes...we thought they knew where everything was" quote. This may also explain why in their Summary Year 2000 Statement PGE mentions, "the difficulty of locating "embedded" chips," the "difficulty in locating all mission-critical software (computer code) internal to PGE" and "the unavailability...of trained hardware and software engineers, technicians, and other personnel to perform adequate remediation, verification and testing of PGE systems.."

One of the only other changes between the 10Q and later 10K filing was in the Costs paragraph. The earlier 10Q only stated that PGE had not incurred any material historical costs. The 10K gives us this info: "Under the Plan, PGE currently estimates that it will spend approximately $20-25 million relating to Year 2000 issues, about one-third of which has been spent to date; 1999 expenditures are currently estimated at approximately $15 million."

My personal opinion, Jeff, is that either this company is doing the biggest self-protective legal sandbagging job I've run across yet and their project is on target, or else they've run into a lack of expertise which set back their project, put a big dent in their confidence and caused them to run behind schedule - but they're not about to say so.

Being the admitted skeptic I am, and having previously expressed the opinion that the skill and experience levels of in-house personnel can vary greatly depending on the area and the salaries a company can afford to pay, I would personally opt for the second scenario. But the first one is a possiblity, too. You're probably already aware of my belief that preparation in the face of uncertainty is a common sense thing to do. I hope you can get more definitive info from PGE but I think that's unlikely to be the case. We all could wish for a crystal ball, but we just have to make whatever choices seem the most logical for our circumstances and which fit with whatever info we do manage to acquire. Good luck in your information gathering efforts!

-- Anonymous, April 20, 1999

What you have revealed gives me a great deal of hope as a software developer. I believe that the uncertainty that you have witnessed is a very positive sign. It is quite different from the high level of seemingly baseless confidence that I have been seeing in the industry.

When I see the confidence that is still prevalent in the electric utility industry, I remember how I was before I understood the depth of the problem. When I hear of realizations like this from PGE, I feel like I am beginning to see light. It's one thing to be "aware" of Y2K, and it's another to personally understand its gravity through experience. It appears to me that PGE has experience now. I expect good things from them.

-- Anonymous, April 20, 1999

Jeff,

It's hard enough for me to follow the typical bureaucratic rhetoric as it is, but this was a challenge that was almost too much for me. Just goes to show how useful a stutter can be.

One part I think I followed was this:

Through our very vigorous process we...we uh found three times as many embedded systems in...in these, in these plants and that's because we used block diagrams, we basically took the electrical systems in those plants apart and went through it methodically and pulled all of this information out.

Um... These were engineers who helped build these plants for heaven's sakes...we thought they knew where everything was.

That tells me that they have block diagrams (BD) of the plants' electrical systems, but do not have actual detailed schematic diagrams (SD) of the sub-systems and circuits that the blocks in the BDs represent, and that the BDs weren't updated each time the system was changed.

Some background:

It used to be (in the 1950s and 1960s, anyway) that complex electric/electronic systems were documented with BDs and SDs of the circuitry in each block. That way, fault diagnosis, troubleshooting, and modification work would be a bit easier for those that were not intimately familiar with the system. The BD would help the workers better understand how the system worked by showing how the subsystems were interconnected. Some of the better BDs would show a simplified SD of some of the more critical circuitry as an additional aid.

In the more complex systems, some of the blocks in the BD would represent subsystems made up of many circuits, and thus would have their own block diagram. The more complex the overall system, the more BD tiers, or levels, you would have.

An "embedded chip," as it is commonly referred to in these forums, is actually a system in itself. It, with other chips, and the necessary interconnects, are all mounted on the same printed circuit board, which is placed in an enclosure (box). It is therefore more properly referred to as an embedded system (ES), but I won't belabor the point here.

Such embedded systems can be represented by one block on a BD, or several ESs can be included in one block. It all depends on who drew the original BD, and what standards he followed. Regardless, each ES should have its own SD that shows the actual discrete chips on that printed circuit board, with lines drawn on the schematic to show how the chips are interconnected with each other, and so on.

So, a well-documented system would have BDs for system navigation, with a SD for each block in the BD (or, as in the case outlined above, another BD, with SDs assigned for its blocks). If it's done right, the BD will tell you where each ES is in the system. BTW, that information is the electrical location, not the physical location, but the tech guys are assumed to be able to find it if necessary.

You can think of a BD as a state road map that shows all the highways and byways that connect the cities and towns together, and a SD as the street map of one of the cities or towns. If you were to have a complete set of maps for your state, you would have one state road map, and a collection of street maps for every city and town in the state. Of course, each time a new road or street is added, rerouted, extended, or closed, the appropriate map(s) would have to be changed to show that. Similarly, the state map (BD) must be updated every time a town or city (ES) appears or dies within the state (system).

It sounds like the PG&E guy was saying that each upgrade or change to the system wasn't fully documented, if at all, so that there are a lot of ESs that don't show on the system or subsystem BDs, if at all. Add to that the assumption by the powers that be that the engineers and techs that designed and installed the system changes remembered where all the changes are, what they do, and how they interact with the rest of the system. Apparently, they assumed wrong.

The part of his statement I have trouble with is where he says, "...we basically took the electrical systems in those plants apart and went through it methodically and pulled all of this information out."

Took the electrical systems apart, then went through it "methodically," then put it back together again? That would take years! You would do just as well to build a whole new plant. I'm hoping that what he meant to say was that they used the extant documentation to find out what they wanted to find out, with an occasional excursion to the physical plant itself to visually confirm what the documentation claimed. But then, that assumes the documentation is up to date, and accurate.

The only other way to find ESs, and maybe get some kind of idea of what they do is to devise a test that will cause a reaction from one of them that the techs can trace, so they can find out just exactly what they have, and why it's there. If the reaction turns out to be a crash, then so be it...at least now we know it's in there, somewhere, and it'll cause a crash if such and so occurs.

That's what I think the PG&E guy meant when he said that the techs can use testing as an inventory tool. In this case, they're not testing their remediation efforts, they're using tests to see what's in there, compliant, or otherwise. Kind of like standing on the brakes on a slick road to see if your car has an anti-skid braking system, and if it's working.

Naturally, you would like to take the system off-line so a ES reaction that results in a crash won't affect the customers. Kind of like doing your ASB test in an empty parking lot. But, if you can't take your plant off-line for whatever reason, and you can't devise a test procedure that will allow you to safely locate your ESs, and help you ID the non-compliant ones...IOW, when there's NO OTHER WAY, and it MUST be done, then you test on- line with all fingers crossed, and hope for the best. I'm not saying that that's what they, or anyone else is, or has been, doing, but it works, if the test is designed right. I call it the "half a loaf" approach. It's great if you're a betting man.

This applies equally to testing as an inventory tool, and as a tool to confirm the validity of remediation efforts. The first application should have been completed over a year ago. The second application should have been completed by December 31, 1998. As we get closer to the Day, time pressure will force those plants to seriously consider on-line testing as the only way to check their fixes (assuming they have found all their ESs). I wouldn't be a bit surprised to see electrical power become more unreliable as the real deadline approaches, partly because of ESs that were overlooked, ignored, misunderstood, or undocumented, and partly because of fix failures during testing.

It's a heck of a way to run a railroad, but if it's the only way to run at all, then you go with what you have.

-- Anonymous, April 24, 1999