concurring in part and dissenting in part.
I concur in the court’s affirmance of the appellant’s convictions for drug trafficking and counterfeiting. I respectfully dissent from the court’s reversal of the conviction for using a weapon in relation to the commission of a drug offense.
A
I agree with my brothers that the motion to suppress Mr. Taylor’s confession and the evidence obtained in the search was properly denied. I must respectfully note, however, that the court’s treatment of these two suppression matters as one unnecessarily obscures an important and unsettled issue with respect to the standard of review for suppression rulings grounded in the Fifth Amendment. As this court recently pointed out in United States v. Jones, 21 F.3d 165, 168-69 (7th Cir.1994), there has been, without the benefit of the process required by our rules, an assumption in recent times that the deferential standard used to review Fourth Amendment suppression motions is applicable ex propria vigore to the review of mixed questions of law and fact surrounding the voluntariness of a confession. There is precedent in this circuit for the proposition that the question of voluntariness is subject to a de novo review by this court. United States v. Montgomery, 14 F.3d 1189, 1194 (7th Cir.1994). This approach is compatible with the Supreme Court’s approach in Miller v. Fenton, 474 U.S. 104, 111, 106 S.Ct. 445, 450, 88 L.Ed.2d 405 (1985), that the issue of the voluntariness of a confession is not subject to *469the presumption of correctness in § 2254(d) proceedings. I also note that this court has held just recently that the determination of whether a person is in custody for purposes of Miranda warnings is a matter of law that we review de novo. United States v. Menzer, 29 F.3d 1223, 1230 (7th Cir.1994).
The issue of voluntariness hinged substantially on the credibility findings of the district court. Whatever standard of review is employed on appeal, the determination of the district court ought to be upheld.
B
The issue of the sufficiency of the evidence with respect to the firearms offense is indeed a close one and it is only after a great deal of study of the record and reflection that I conclude that I cannot join my brothers in the reversal of the conviction. Here, unlike the issue I have just discussed, the standard of review is all important. In my view, there is sufficient evidence in the record to support the view that the presence of the weapon facilitated the commission of a drug offense because it placed the purchaser on notice that the defendant was quite willing to protect his investment with a firearm. Indeed, the record shows that the informant, upon returning to the officers, told them the defendant liked firearms and that he had them on the premises. There was further evidence that the officers, upon receipt of this information, formulated a plan to deal with the contingency that such firearm would be used when the defendant was confronted with law enforcement agents.
The transcript of this trial makes clear that, despite the many cases that we have decided, we have yet to give sufficient practical guidance to the trial courts on this issue. We need to begin to place principled limitations on the “during and in relation to” language of the statute. Our fact intensive adjudication of cases up to this point does not do the job. To the extent that my brothers’ conclusion in this case reflects that need, I am in agreement with them. Regretfully, I must part company from them with respect to the facts of this case because I believe that the presence of the weapon facilitated the offense,