Litjan

Hi everyone, we are still deep into the FMS and it´s implementation - but we feel that we are over the hump and will have this section wrapped up within a few weeks. Tom, who is in charge of generating the lateral navigation (LNAV), has posted a new blog over at our website: http://www.ixeg.net/...ve-lnav-routing I just want to share this screenshot I took yesterday night when testing the path-drawing algorithm on an approach to EDDV (Hannover, Germany). It shows me coming in from the north towards the Celle NDB (CEL), followed by the RNAV transition to runway 27R. I named the screenshot "Childrenofthemagenta" as a reference to the excellent video that was linked to in THIS post by Andy Goldstein. This guy has it perfectly straight: http://forums.x-plan...11 Jan

Hi meshboy, an airway is like a "highway in the sky". It links various points like fixes and navaids and has a "name". So by telling the FMS which airway you want to procede on, it can insert all these single waypoints into the LEGS page for you. http://en.wikipedia.org/wiki/Airway_%28aviation%29 Jan

Hi everyone, the last weeks have been more busy than ever - we are grinding through that big heap of coding mess called the FMS - flight management system. At it´s core is the FMC, and we know that this piece of hardware is very important to all of you. And it should be, because it is an integral part of "managing" your flight. While the FMS does more than just providing a nice curved magenta line to follow, this is propably the visually most appealing aspect. Therefore we want to share some of the pics we are getting during this stage of development. When we draw closer to release, we will get you more frequent updates and the occasional video. We have chosen these departures because they are quite complex - with a lot of conditional waypoints that are not simply a combination of latitude and longitude. If you look at the the DEGES1W departure in LSZH, the conventional description calls for a straight segment until you hit 2.1 DME from KLO. The FMS calculates this waypoint in dependence on the KLO VOR - it knows where KLO is, where the runway is and what the initial track will be - now it needs to find the point where this track intersects the 2.1DME "ring" around KLO. This is SO not trivial... Then it get´s worse. Turn left on a certain track to intercept the 255 radial. How large is your turn radius? How much do you need to lead the turn onto the radial? Where is the radial? Thank god I only fly those things ;-) Jan

Tom, Tom, Tom... first rule to learn as an aspiring airline pilot - when kids visit the cockpit (no offense, Meshboy! ;-) ) - ALWAYS turn on the "light test" for the "christmas tree show"! ;-) Jan

I hear ya, I hear ya....

Hi Flo, The engine is only rated to do so much work - when you make it spend a part of that to run packs or anti-ice it will have to deliver less power for forward thrust. That´s why you need to reduce RPM to keep the total power of the engine constant. In other words - you need to burn so much fuel to make the engine spin at a certain speed (N1). If you extract compressed air, the engine would turn a little slower even while adding the same fuel per minute. Imagine a car that is limited to 100 horsepower. You run at that horsepower and the speed is 150mph. Now you turn on the headlights, radio, window heat and a bunch of other things that use electricity and put a drain on your generator. The car will slow down.... Did I make sense? Hi Meshboy - we will keep your offer in mind and understand your desire. I have been running through two of several tutorial flights we are planning for the documentation (KSAN - KLAX, KTUS KTUS - KLAS). It´s a lot of fun to think of how to show people all the features, take pictures, write the text... and it is amazing to think that the plane is far enough to really fly it to do the tutorials already and not miss any features! Jan

Hi guys, These last few days we have been busy with the FMS again. You might know that all modern airliners have one of these - and while you can theoretically fly without them, they are very much a centerpiece to a safe and efficient operation. They are quite complicated and highly sophisticated - and while that is great when actually using them on board, it is also the reason why a realistic FMS has been the domain of very few and selected add-on making companies so far. Well, we are taking on that challenge. The quality and feel of our FMS is blowing me away every time our coders punch out a new page or feature. I have written about the thrust management before, today I will punch up a few more "preflight pictures" and point out what I think is special about them. Above pic shows the two CDU´s - note that they are still "floating" in mid-air right now. Every CDU can operate independently of the other one - show a different page, accept input, etc. Above pic shows the "PERF INIT" page, where you will input weights and other parameters of the flight. You can see that I am about to enter "FL240" as cruise altitude into the boxes on the top right. The FMS uses this data to compute various things - amongst them the takeoff speeds as you can see on the next picture: Above pic shows the TAKEOFF REF page on the left CDU and the N1 LIMIT page on the right CDU. You can see on the right that I have entered a TASS of 45 degrees - the FMS has calculated a reduced take-off N1 of 87.5%. It has also automatically selected a reduced climb thrust - otherwise the N1 would increase at the thrust reduction altitude. Also note the thrust mode annunciator shows "R-TO" to alert the pilot that a reduced takeoff thrust will be set. The left CDU shows the takeoff data - once you enter the takeoff flap setting (1 in this case) the FMS will calculate and suggest V-speeds, that you can accept with the click of a button or overwrite with your own speeds. This last pic shows the effect of those inputs on the EADI. V1 and Vr are automatically shown on the speed tape. V2 must be set manually with the MCP speed selector. Ready for a reduced power take-off. Jan

Hi everyone, back with a little update We are currently pretty much finished up with the EHSI. I posted about this before, but I can not emphasize enough how much the completion of this part has gotten our 737 more complete. While you could navigate the 737 "old school" with just the RMI, I really missed the different modes of the EHSI for navigation. You spend the majority of your time staring at that screen - so it being there and looking exactly like the real one does is a big step. Another thing we are working on is the FMS - we have just finished the FMS part of the autothrottle thrust limit computing. Let me explain for a bit - the 737-300 does not have a FADEC like more modern jets. It´s engines´ RPM is controlled by a hydromechanical unit (with fuel as transmitting liquid) and a power management computer. While the PMC helps a bit, it is still up to the pilot to control the engine. He does so with the thrust levers and the N1 indicators. N1 is the RPM of the outer fan - it gives you a very good indication of how much thrust the engine develops. Controlling the engine seems straightforward - but there is a lot to keep in mind. The CFM56-3B1 engine we are modelling is rated at a thrust of 20.000lbs. This is the limit to which it is specified. On a standard day at sea level, this limit is achieved with about 91.5%N1. So if you would firewall the levers to the stops, you might get a lot more thrust - but possibly damaging the engines in the process. It´s up to the pilot to avoid "overpressuring" the engine! As it gets hotter, maximum N1 increases, because warm air is less dense and will output less thrust. But at +30C the hot air entering the engine is compressed and burned to reach the maximum EGT - so as it gets warmer the maximum N1 needs to decrease to stay under the EGT limit. Booooring, I know. In ancient days, there was a flight-engineer on the deck to check the relevant N1 for the current conditions, and he would actually reach forward during take-off and adjust the levers to set correct power! On the 737 classic you have the FMS to help you out. It measures temperature, pressure, bleed air demand and pilot´s thrust derate input to calculate the current limit N1. This limit is sent to the autothrottle computer, which will then automatically set and maintain maximum thrust as specified. Now there are many different thrust limit. Take-off thrust, go-around thrust, maximum continous thrust, climb, cruise, reduced version of the above, and so on. Of course our model knows, computes, annunciates and sets these limits (flight-phase related and pilot-selectable) just like the real jet. On the screenshot below you can see the N1 LIMIT page that controls thrust limits. It looks like this before take-off. Notice how one N1 value differs from the other one - it is because the left pack is off - so N1 can be higher without overstressing the engine (due to less airbleed extraction). You will notice the N1 change with packs on/off/hi, engine anti-ice, wing anti-ice and so on... Just another example of a very complicated mechanism running behind the scenes that even most pilots take for granted The last few days I have also taken quite a few testflights to validate and sample climb performance - while we stay very close to the offcial numbers, there is just not enough data published to calculate climb performance through the whole envelope. We know you want your TOC prediction to be precise, so we have to account for different weights and different combination of climb speeds. An airplane climbs better at constant Mach, so a climb at 280/.74 will look quite different from one with 300/.72. You need to calculate the crossover point, calculate climb profile before and after that, and so on. Yesterday I took her for a ride from Westerland in northern Germany to Rome, Italy - just over two hours. While we still don´t have an LNAV route to follow (working on it as we speak) it was nevertheless a joy to follow the VOR´s down to Frankfurt, Zuerich, Genoa and then to chop the throttles at 110NM out (FL370) and glide right onto the ILSof 16L without touching the levers or the speedbrake again . Jan

Hi meshboy, the shot is taken in sim - shift-space to copy a screenshot as .png. I then resaved the .png as jpg to save size. No editing. Thanks for your kind words - I will make a video showing of all the features of the EHSI in a little while when it is done completely. Currently chasing some nasty OpenGL bugs. But you really have to see it in motion. It is VERY realistic! Jan

Hi everyone, we have another teeny little piece of news for you. Today I want to show you the work we have recently done on our EHSI. On the attached picture you can see the EHSI as the lower of the two screens. While the possibilities of displaying data on the EHSI are numerous I would like to point out what you are seeing here. The depicted mode is the "expanded map" mode, showing the area in front of the aircraft. The aircrafts position is the tip of the white triangle at the bottom of the screen. Look closely and you can see the "ground path prediction" lines extending from the tip of this triangle, they represent the path the airplane will take within the next 90 seconds (30s per segment) at current turn rate and ground speed. You can use this to make very smooth intercepts when using smaller map ranges. Comes in handy for the final turn from base during a strong head or tailwind... The selected range is 40NM (you can see the 20 on the mid range ring), the map can be displayed in 6 steps from 10 to 320NM. You can see the two tuned VOR´s show up in green - along with the bearing lines, the green dashed lines. These lines are selectable (tuned VOR´s always show), and a great way to check your map accuracy - since the lines bearings are directly taken from radio signal data (displaying radials) and the green navaid symbols positions are drawn from the map´s database. So if they don´t intersect - your map has probably shifted. Also displayed are the airports within the selected map´s section. You can select to show airports, navaids, fixes, etc. on the map with the help of pushbuttons on the EFIS control panel. The magenta dashed line is the heading bug - you can see the aircraft is turning towards that heading. The white line with the little rhomb at the top is the current track - look how it is influenced by the crosswind from the left. This track line will make it very easy to establish the correct wind angle when tracking along the localizer, for example. The wind is shown on the bottom left - but unlike default x-plane we don´t show the wind while on the ground. Why not? Well, the real airplane can´t do it either - the wind arrow is a result of vector calculation between IRS track and groundspeed versus heading and true airspeed. In other words, the plane needs to "sense" the drift before showing it. On the ground there is no drift - > no wind indication. Now you try explaining that to Austin :-) Of course we also don´t show the wind if its magnitude isn´t big enough - just like the real plane won´t. The top line will show GS and TAS, and also the current magnetic heading. Now if you really look at the bearing line of VOR 1 very closely, you can see another green line with the label "A1" overlay it. That is the bearing of the selected NDB - coincidentally pretty much the same bearing. NDB´s positions are not shown on the map, though. The dashes in front of the NM at the very top are the placeholders for "distance to next route waypoint" - blank in this picture since no route is loaded in the FMS. More to come soon, Jan

Hi guys, just a little heads-up regarding development. Below a reply by Tom: A user on our blog requested an update on the project and perhaps some regular postings to let users know what is transpiring. The team feels that is probably a good idea at this time. Up until this point in time, we have been cautious about posting for a variety of reasons, but the primary reason is (and still is to some extent) our concern about whether or not we could do such a challenging project to the level that we envisioned when we began. A highly accurate airliner project of what has become known in flight sim lore as "PMDG Level" has always been somewhat of the "holy grail" of flight simming....and the team felt like X-Plane was long overdue for such a project. Of course like anything in life, these things can only happen once all the variables are just right....X-Plane technology, developer experience, willingness, timing, etc. The CRJ-200 by JRollon broke the ice for X-Plane in a big way. As much as we love it though, the 737 is more of the airliner standard and is why we chose to do one. We thought that perhaps starting with the classic instead of the NG would serve two purposes. 1.) It showed some deference to the x737 project and 2.) we BELIEVED it would be somewhat easier than the more complex 737ng. Well we were wrong. The NG is probably way easier than the classic. Why? Steam gauges..that's why! Now that we have steamrolled through most of the systems on the aircraft, they are really no different than the NG in complexity, only in their implementation. The steam gauges; however, are not just simple graphics, they're complex animations with rolling digits and wavering needles and flipping flags. We even went so far as to put in motion blur on the DME rolling digits on the ADF. I tell you that a LED readout or computer graphic would be much easier What has transpired though as we have seen the cockpit and systems come together is that we very much have a fondness for the steam gauge elements. This particular variant of classic we are doing uses the modern glass EADI / EHSI and somewhat fancy ECAM and somehow delivers the best of both worlds, modern glass and steam gauge "complexity" and is really a feast for the eyes of cockpit fans. It definitely pushes our techo-geek buttons. So about the state of the project. Given the experience of our team members, we decided early on to pursue systems programming first. It is always the most challenging part of a project and can stop a project dead in its tracks. The downside to this for anxious simmers is that is does not give us much to show. Code is not very exciting to look at, indeed I am finding it quite offensive to look at as of late. It is my guess that we probably have over 30,000 lines of code and counting daily. There is some good news though and that is we have just about every major system "over the hump". Just about every major system means just about every one except the FMS. Of course we have 'punchlist' items' on each of the other systems to do. The FMS framework is mostly implemented though. We have the CDU interface and pages mostly complete, we also have the autopilot mostly done and what remains is the calculation of VNAV and HNAV elements and the building of the flight plan. These elements will integrate with the ESHI display so we work on these elements simultaneously and these will come together at about the same time. We expect to begin this phase "next". As far as the 3D work goes, we believe this is the easiest and quickest part, even though it is the most visible. We have a good portion of the fuselage done and textured but no wings/flaps yet. We do not have the main gear modeled yet either and the engines are about 30% modeled at this stage. The systems accuracy is the heart and soul of our project and to be honest, you probably will not see all the cool 3D screenshots until we are on the final stretch. We are also beginning our documentation process and getting ready to grow that element as the systems punchlist get completed. I cannot stress to you how in-depth the systems will be and why the programming is taking so long. "Systems simulation" has always been a bit of a marketing buzzword by add-on developers but our consultant, a 737 captain of ridiculous attention detail will not let up. We read the real POH page for page, system descriptions word for word and if it says it in the POH, we try very hard to put it in. For example…did you you know that in the case of automatically deployed speedbrakes on landing, that you can advance the throttle levers a bit after landing to "auto-stow" the speedbrakes? Bump the throttles up a bit and watch the speedbrake lever go from "up" to "down" on the throttle quadrant. You cannot do this though if the speedbrakes were not deployed automatically from the armed position. If you move the speedbrake lever manually to deploy them on the ground, the 'throttle bump' will not work. Who uses that anyhow? Our consultant thats who! He lands and bumps the throttle instead of moving the speedbrake lever. So if he does it, we do it. He is really starting to bug us! So to summarize this long-winded post for now, upcoming in the ensuing months is more 3D development of the cockpit / wings / flaps / gear, punchlist items on the major systems: hydraulics, electrics, pneumatics, IRS, etc. Development of hnav/vnav and the EHSI and the documentation. Final work will probably be the 3D cabin, eye candy and user-interface elements. That's the plan anyhow. Tom Kyler Laminar / IXEG

Thanks for the nice video! It is really well made, I like the long distance-shot of the landing in the beginning especially. It also brought back the feelings of fear and anxiety I used to have when flying the 737 to Thessaloniki and Athens in the late 90´s.... Controllers without radar (because pay is better if you don´t have radar!), chaotic operation at airports that were plagued by inadequate facilities, high terrain, changing weather, rampant strikes - ah the good, old days ;-) Video´s like this will be very interesting for owners of our (or any well-made 737 classic, for that matter) 737-300 and the stuff you see you can also do in ours :-) Regards, Jan

Hi there, still working on it... we are well underway on the 3D and outside appearance - some preview shots coming up soon. List of remaining systems to be done is getting smaller...slowly. More to come soon! Jan

Oh, I am using TrackIR - can´t recommend that sort of device enough for realistic operation in a 3D cockpit... Jan

To ease the waiting time for everyone, we have added another video that shows the final part of a dual-channel autoland approach. You can view it either at our blog or on our youtube-channel. Here is a direct link: http://www.youtube.com/watch?v=YZj3a6DwH_g&context=C39b3a84ADOEgsToPDskJvPSg2podgypyhrssYiem4 Enjoy, Jan

Well, just to reiterate - the project is still moving ahead as planned - whatever that means . We are all people doing this in our free time, and I think it is completely normal that real life takes the upper hand once in a while. We all have jobs with sometimes fluctuating requirements and families that come first and before everything. Making our living with our daytime jobs also decouples another motivational factor to a certain extent: Money. This has two effects: A.) Quality goes before release-date. This is good, I think. B.) We don´t go hungry until release-date - this is somewhat of a detrimental effect to progress. If we don´t feel like it, we don´t work on it and still don´t suffer. So what motivation drives us? Money to a certain extent, of course - but as I explained in a previous post, we won´t get rich with this. Not even reimbursed appropriately. I think foremost the motivation driving us is some geeky satisfaction in pulling this off, working together with friends and delivering a product that makes us proud and many users happy. Of course we always knew we are developing in a timeframe that puts our release into XP10s territory. The initial problems with the early beta´s did somber us up quite a bit. Yes, there are a bunch of cool features, but right now it reminds us more of a muddy construction site with the architect standing on the side, fiddling with some oversized blueprints than the gleaming glass-steel-and-concrete palace we where expecting. The promise is there - it just needs to materialize. And we were pulling up in our moving van with all of our furniture in the back... Just an example: For our development it is important to shut down and restart the simulator many times a day. Even with my brand-new SSD starting XP10 takes longer than FSII from Sublogic took on my old 1541 floppy disk drive (2:40min)... So for now we are developing on XP9, and there is no plan to change that so far. Another issue is frame-rates. Yes, I can get good rates in XP10 if I shut down the eyecandy. But we all know how that goes. No one wants to do that, and then everyone complains of low frames. And our 737 WILL need some computational power, and if you start out with 25fps to begin with... not good. Nothing is decided as to what platform we are releasing for intially. My personal guess - please don´t hold me accountable for that - will be XP9 initially and XP10 later on. A word to alleviate fears of this project dying down - unlikely! We are too far into this and have spent too much time. Even if we dumped what we had right now out on the market it would already be a fantastic airplane to fly and operate, even while lacking some systems. So unless the sky caves in or the martians invade, you will get your hands on this Jan

Hi everyone - we took a little timeout over the holidays, everyone spending some quality time with his families. There has not been much development for the last three weeks. I feel that this also gave us a well needed rest and a chance to reasses where we are standing in the rapidly evolving X-Plane world. The major new factor has been the beginning of the X-Plane 10 beta process. Initial hopes that our 737 would also work in XP10 with minimal adjustments have proven not to be true. On the whole the IXEG team is underwhelmed with the state of XP10 so far and we feel it would be futile to release a plane trimmed for X-Plane 10 right now. There is no doubt that XP10 will continue to stabilize and improve, but we think that this will take a while. Jan

Glad you like it. 1. It depends. You will get a warning of the IRS if the position you put in differs too much from the original position before last shutdown. The IRS can also catch a latitude input error if it is big enough (it "feels" the earth rotate underneath during alignment). A wrong longitude input will not be caught initially - but soon the FMC will warn you, because the IRS position does not match the data from the automatically scanning DME´s and the GPS. Also you will not see the route you put in on the map. The ILS and VOR receivers are independent of the position, you can use them even with the IRS not aligned (but the FMC uses those signals to update it´s position). The whole thing is complicated and we took a little liberty with it in our model. You can not ultimately put in the wrong position - it will keep rejecting it until you put in one that is reasonibly close to the "real one". This way we don´t have to maintain an "offset" latitude and run our own navigational model in relation to that. Note that the IRS´s WILL drift over time in our model, the indicated position will become more inaccurate. The aircraft´s position will stay accurate (as in real life) as long as the GPS and radio updating works (which it usually does). 2.) Transition altitude is the altitude above which you set 1013.2hPa on the altimeter, so all aircraft fly at the same reference (flight-level). Below that you would fly the local altimeter setting (QNH), so you are always at the same altitude referenced to sea level. 3.) There are 6 fuel pumps altogether (plus a scavenge pump, which we also model). Each tank has two pumps for redundancy and also to always supply fuel in case of high attitudes (where all the fuel would run backwards or forwards). One pump is enough to supply fuel for full power. If all pumps fail the engines keep running on "suction feed" - so don´t expect them to flame out when you turn off the fuel pumps! (Can´t wait for the bug-reports on this one ;-) ) . Thrust degradation might occur at high power settings/altitudes. Jan

Yes, I remember going through this with XP8 and XP9...this time it´s a bit different, though. There has been a marketing hype buildup with videos and screenshots. The whole community (this includes a lot of FSX users) is very excited about this release and everyone has high expectations. Austin has made some pretty big remarks. People expect XP10 to deliver right away. A first impression is a lasting one - especially if it is your first contact with X-Plane. There is no clear mention of the "beta"-nature that this demo has on the website. Disappointment will run high I fear. I am here to stay - I am just afraid that this will not be an ideal start for XP10 as a competing product in the flight-simulator market as a whole. Jan

You just might be lucky enough to take that long...hopefully some of the worst bugs will be fixed by the time you download! I am having trouble to stay above 20fps with some decent rendering options around default KSEA - also get graphical errors on some settings (green water, red runways etc.). The shadows are killing my framerate and the in-cockpit-shadows are jittering all over the place dependent on the camera angle (which shouldnt affect those). The whole view system has been redone (not sure if to the better) and I can´t set up my joystick system because X-Plane freezes whenever I access the "axis" screen... I foresee a public relations disaster descending upon Laminar. I know one guy who won´t care about it, though... just blast some more toy rockets into your neighbours house, I guess... Jan (disappointed, if you can´t tell)

Finally got around making that video I promised you . http://www.youtube.com/user/IXEG1 Enjoy, Jan

Hmm - what use for the wing-anti-ice do they have then? Or is it because they never encounter icing conditions? Maybe it is different on the NG. I have heard that the A380 has no wing-anti-ice at all - supposedly its leading edge is shaped in a way to make it impossible for ice to accumulate. Also the 747-400 I flew could not turn on the wing-anti-ice when the leading edge slats were extended. So I guess there are different configurations possible. Jan

Same here for the engine-anti-ice. The rule is visible moisture and temperatures between +10C TAT and -40C SAT. During descent there is no lower temperature limit. When I say "we turn wing-anti-ice on all the time" I use a misleading phase, sorry. I do not literally mean that it runs the whole time. During the winter I use it about every second flight, in the summer much less if at all. This was just to emphasize that I find it hard to believe that some pilots have never turned it on in their whole career...well, maybe if they fly in the Southwest US or Africa or the tropics it wouldn´t surprise me. I am not aware of a hard limitation on wing-anti-ice as far as altitude goes. There is a caution note that says: Use of wing-anti-ice above approx. FL350 may cause bleed trip off and possible loss of cabin pressure... Jan

Well, I think we are mixing up some things here. First, we turn on wing-anti-ice all the time during flights. Whenever you go through clouds that have suitable moisture (temperature/droplet size) ice will form on the leading edges. You then turn on the wing-anti-ice to heat the leading edges, this will shed the ice. You could theoretically turn on the wing-anti-ice even before accumulating the ice - but this leads to a bigger performance penalty and also entails the danger of droplet running back over the warm leading edges and refreezing further back on the wing. Thats why we use the anti-ice more like a de-ice... The other part of the problem is the rather large pneumatic demand that the wing-anti-ice system places on the air system. A lot of air is routed from the corresponding manifold side to the leading edge. There are certain limitations going along with this - for example you are not supposed to run the system of the APU - it would just overload its capability to create bleed pressure. It is however not a problem to run the wing-anti-ice AND engine-anti-ice while both engines are running at a decent RPM and/or at lower altitudes. There will still be enough pressure left to drive the pressurization. I think you might run into trouble if you want to drive all of this at 37000 feet and idle power (just starting the descent, for example) - but then again there would be no need for wing-anti-ice up there. Ice will usually only form with TAT´s of about -20 to +3 C or so. There are many more ramifications of the system during degraded operation - for example during engine fire. You will close the isolation valve to avoid smoke and fumes from the damaged engine to enter the airconditioning - but then you would have no pressure for the wing-anti-ice on the damaged side -> asymetrical anti-ice -> bad! Also if using the APU to pressurize the cabin (possible up to 25.000 feet) after a "no bleeds" takeoff you can´t use the wing-anti-ice. So if performing such a take-off in icing conditions with an engine failure could get you into trouble really quick. Nursing the aircraft across that mountain ridge - can´t turn on the bleeds because you need all the power to get across the terrain - ice forming on the wings because you can´t use the APU to power the wing-anti-ice... Jan

We are thinking about adding some effect...just don´t want it to be cheesy or unrealistic. Austin´s "window icing" is very basic - if it is ON then there is no ice, if it is OFF then there will be ice under what he considers to be icing condition. Real life works a bit different (as usual). The 737´s window heat is always on during flight, as a matter of fact you need to turn it on at least 10 mins prior departure. This is both to avoid ice buildup and fogging, but also and most importantly to make the windows flexible so they can withstand a bird-strike. BUT there is still some ice building up on the windows if you are in moderate to severe icing conditions - just around the edges, not really impairing the vision too much. So, yes, I think we will have some icing effect - just don´t expect to "toggle" the switches to get rid of it. They are always on. The anti-ice for wings and engines works a bit differently, but that is another story. Jan