This has been answered - the place to introduce complications is in the city, at the point at which they interact directly with the players' immediate goal, and provide both threat and oppportunity.
If "the goal" is in the temple, then both the desert and the city - and the siege for that matter - are between us and the goal. As has been said several times, JC and I do not perceive this magical difference between "siege between us and the city" and "desert encounter between us and the city".
Well, yes, we could just declare the campaign resolved and all go home, or go out to the movies. But presumably the players want to play an RPG - just not one that involves exploring a desert.
NO ONE is asking them to explore the desert. Might this please be the last reference to "exploring the desert". I am positively evincing my preference that we NO LONGER DISCUSS exploring the desert. We are discussing the fact that the desert lies between the PC's and the city, that it is possible that, as those PC's make as straight a path as they can between where they are and where they wish to be, they will pass through that desert, and that it is possible they will encounter something between them and the city. This "something" may be relevant, or may be something they simply need to circumvent (just as a siege could be very relevant to their goal, or just another obstacle between them and whatever their goal in the city may be - which, after over 1,000 posts, [MENTION=22779]Hussar[/MENTION] has yet to provide any information on, so we still do not know what it is.
The answer lies in the players' evinced preferences. If the players have made it clear they want to get to the city, and if there is nothing about the city that makes it hard to make things interesting there, and if there is nothing about the desert that is essential to framing the city, then why would the GM nevertheless insist on trying to run encounters in the desert?
Was it not you who mentioned "fun" earlier? And, again, it is not "the city" which is the goal. It is something, the nature of which remains completely unknown, within the city. And I would say that being in the middle of a desert would have an impact on the city - it will not be the same as a city in the middle of lush, fertile farmland, nor one located on a river filled with boat travellers, nor one which is a major seaport. I suspect many of us could come up with better examples of encounters in the desert which are relevant to the goal in the city if we had some info on that goal, and on the city itself. We do not.
You say this as if it is nothing. The city is the setting wherein the goal is located. The players are committed to doing something in the city, presumably (given Hussar's evident enthusiasm for the siege) something that involves interacting with the city, treating the city as both chalenge and resource.
Emphasis added. [MENTION=22779]Hussar[/MENTION] wants to cut to that still-unknown goal. The desert is setting. The city is setting. Both are between him and the goal.
How do you know? This is like your characterisation upthread of the goal as a McGuffin? How do you know?
In exactly the same way that you know the city is
an urban location that is, by definition, replete with NPCs, and all the opportunities for good and bad things that NPCs bring with them. It's very different from a deset. Apart from anything else, it creates much richer fictional positioning for the PCs, and hence much richer opportunities for the players.
It's on another plane of existence. Perhaps, for all we know, it is a City of Undead, where zombies, skeletons and shades wander, shallowly parodying the actions they unertook in life, never changing, and never varying their routine. We assume because we do not know. We know only that, somehow, the city between [MENTION=22779]Hussar[/MENTION] and his goal (whatever that may be) is deemed to be of interest and the desert is deemed not to be, despite the fact that all we actually know is that there is a goal to be achieved within a city (about which we know nothing more) in the middle of a desert (about which we also know nothing more).
[MENTION=22779]Hussar[/MENTION] has said, several times, that he might well be very interested in a desert exploration at some other time, just not this time. As near as I can tell, his GM's role is to determine what will pique his interest at this particular moment, present that immediately and then figure out what random element may catch his fancy next.There are a range of different approaches to prep. For my own part, I can't remember the last time I had trouble knowing what would engage my players. If they're not interested in deserts, they tend to make that pretty clear.
The siege doesn't prevent you entering the city - it simply changes the character of that entry. Nor need it, in real play time at the table, delay you doing what you came to do in the city: it may even expedite that, if the players use it cleverly. And even if it does make it take longer, in real play time, to achieve the goal, all that play time will have been spent aiming directly at the goal (including perhaps by taking advantage of the siege). Which is, more-or-less, what RPGing is about - spending real time at the table engaging in realising one's PC's goals.
Getting across the desert to the city is every bit as much "aimed directly at the goal", about which the only thing we actually know is that its achievement requires entering the city.
For myself, one of the requirements to facilitate proactive players is to allow them (the opportunity) to have sufficient information to make informed choices most of the time.
Given no foreshadowing, no information to the players at all about the desert, it seems reasonable to me that one or more players ask to skip it. Either the encounters in the desert are all random filler for versimilitude, or there's one or more set encounters with plot relevance that the players and PCs have every reason to try and avoid based on the information they have at the time. Forcing an encounter on unwilling players who want to avoid it is probably railroading, making this encounter integral to the plot threatens to send the message "stop being proactive and follow the linear plot".[/
A desert is an excellent place for players to test how linear a game actually is despite the protestations of the referee. A desert or wasteland is an empty place generally with lots of routes through, and limited ambush sites. The players should be able to avoid at least some encounters if they are focussed on the destination, not the journey. If they can't, maybe the game is more linear than proactive. Proactive players hate bait and switches such as this. If the game is about the desert, tell the players beforehand that the game is about the desert, don't try and trick them.
Whereas the city holds the goal, and the players likely expect to interact with the city to achieve the goal, there's lots of room for meaningful decision making in the average city.