PER CURIAM.

In 1974, Johnny Paul Witt was tried and convicted of first-degree murder for the killing of Jonathan Kushner. The trial judge accepted the jury's recommendation that Witt be sentenced to death and, on appeal to this Court, the trial court's judgment of conviction and sentence were affirmed. Witt v. State, 342 So.2d 497 (Fla.), cert. denied, 434 U.S. 935, 98 S.Ct. 422, 54 L.Ed.2d 294 (1977); reh. denied, 434 U.S. 1026, 98 S.Ct. 755, 54 L.Ed.2d 774 (1978). Witt subsequently sought post-conviction relief in the trial court pursuant to Florida Rule of Criminal Procedure 3.850, which was denied. He now seeks review of that denial.

By this appeal Witt raises essentially six issues, all of which he admits either were raised in the direct appeal from his conviction and sentence, or could have been raised at that time. He predicates his appeal on alleged changes in case law since his first appeal was concluded, asserting the right to obtain the benefits of subsequent, favorable case law developments relating to capital punishment and to criminal law generally. The particular changes which Witt presents for our consideration are these:

(1) an alleged change in the law relative to sentencing, reflected primarily in Elledge v. State, 346 So.2d 998 (Fla. 1977), concerning the significance of improper aggravating circumstances where at least one mitigating circumstance has been found to exist;

(2) an alleged change in the law relative to sentencing, reflected in Hall v. State, 381 So.2d 683 (Fla. 1979), concerning the requirements for a written enumeration of the findings in aggravation and mitigation;

(3) an alleged change in the law relative to sentencing, reflected in Shue v. State, 366 So.2d 387 (Fla. 1978), and in Burch v. State, 343 So.2d 831 (Fla. 1977), concerning definitions for the mitigating circumstances set out in sections 921.141(6)(b), (e), and (f), Florida Statutes (1979);

(4) an alleged change in the law, reflected in Smith v. Estelle, 602 F.2d 694 (5th Cir. 1979), cert. granted, 445 U.S. 926, 100 S.Ct. 1311, 63 L.Ed.2d 758 (1980), making inadmissible in a sentencing proceeding statements made during a court-ordered psychiatric examination not preceded by Miranda warnings;

(5) an alleged change in the law, reflected by an aggregation of the individual opinions in Lockett v. Ohio, 438 U.S. 586, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978), requiring that the state, before imposing the death penalty, establish that the defendant intended to kill the victim; and

(6) an alleged change in the law, reflected in Brewer v. Williams, 430 U.S. 387, 97 S.Ct. 1232, 51 L.Ed.2d 424 (1977), relating to the efficacy of a defendant's waiver of his right to counsel.

The underlying issue posed by this appeal, however, concerns the significance of a change in decisional law on the finality of a fully-adjudicated capital case. Simply stated, we are confronted with a threshold decision as to when a change of decisional law mandates a reversal of a once valid conviction and sentence of death. The issue is a thorny one, requiring that we resolve a conflict between two important goals of the criminal justice system — ensuring finality of decisions on the one hand, and ensuring fairness and uniformity in individual cases on the other — within the context of post-conviction relief from a sentence of death.

The importance of finality in any justice system, including the criminal justice system, cannot be understated. It has long been recognized that, for several reasons, litigation must, at some point, come to an end. In terms of the availability of judicial resources, cases must eventually become final simply to allow effective appellate review of other cases. There is no evidence that subsequent collateral review is generally better than contemporaneous appellate review for ensuring that a conviction or sentence is just. Moreover, an absence of finality casts a cloud of tentativeness over the criminal justice system, benefiting neither the person convicted nor society as a whole.

Post-conviction relief procedures, such as those authorized by our Rule 3.850, offer an avenue to challenge a once final judgment and sentence in limited instances, and for limited reasons. The United States Supreme Court recently noted:

It has, of course, long been settled law that an error that may justify reversal on direct appeal will not necessarily support a collateral attack on a final judgment. The reasons for narrowly limiting the grounds for collateral attack on final judgments are well known and basic to our adversary system of justice.[*]

[*] Inroads on the concept of finality tend to undermine confidence in the integrity of our procedures. . . . Moreover, [the] increased volume of judicial work associated with the processing of collateral attacks inevitably impairs and delays the orderly administration of justice. Because there is no limit on the time when a collateral attack may be made, evidentiary hearings are often inconclusive and retrials may be impossible if the attack is successful. . . .

The doctrine of finality should be abridged only when a more compelling objective appears, such as ensuring fairness and uniformity in individual adjudications. Thus, society recognizes that a sweeping change of law can so drastically alter the substantive or procedural underpinnings of a final conviction and sentence that the machinery of post-conviction relief is necessary to avoid individual instances of obvious injustice. Considerations of fairness and uniformity make it very "difficult to justify depriving a person of his liberty or his life, under process no longer considered acceptable and no longer applied to indistinguishable cases."

Unfortunately, drawing or defining the line where finality gives way to fairness based on a change of law is no simple task. The United States Supreme Court has struggled with this problem, developing what has been characterized as a "staggeringly intricate body of law governing the question whether new constitutional doctrines should be `retroactively' or `prospectively' applied." Indeed, one former member of that court has characterized the course of law in this area as "almost as difficult to follow as the tracks made by a beast of prey in search of its intended victim." Mackey v. United States, 401 U.S. 667, 676, 91 S.Ct. 1160, 1172, 28 L.Ed.2d 404 (1971) (Harlan, J., concurring in part and dissenting in part).

Without attempting to survey this relatively unsatisfactory body of law, we note that the essential considerations in determining whether a new rule of law should be applied retroactively are essentially three: (a) the purpose to be served by the new rule; (b) the extent of reliance on the old rule; and (c) the effect on the administration of justice of a retroactive application of the new rule. Stovall v. Denno, 388 U.S. 293, 297, 87 S.Ct. 1967, 1970, 18 L.Ed.2d 1199 (1967); Linkletter v. Walker, 381 U.S. 618, 85 S.Ct. 1731, 14 L.Ed.2d 601 (1965); Brewer v. State, 264 So.2d 833, 834 (Fla. 1972); State v. Steinhauer, 216 So.2d 214, 219 (Fla. 1968), cert. denied, 398 U.S. 914, 90 S.Ct. 1698, 26 L.Ed.2d 79 (1970).

A determination that a new principle of law should be fully retroactive may mandate its recognition and application on collateral review. Thus, in this proceeding we must balance the interests of fairness and uniformity for Johnny Paul Witt against the interests of decisional finality, in the context of alleged subsequent favorable changes of law. The balancing problem is presented squarely here, for there are no bases for Witt's collateral attack which are independent of alleged law changes.

The general difficulty of resolving the conflicting interests presented by law changes is heightened by the fact that this is a capital case. Uniquely, capital punishment, on the one hand, connotes special concern for individual fairness because of the possible imposition of a penalty as unredeeming as death. On the other hand, both the frequency of Florida "law changes" involving our relatively new capital punishment statute, and the unavoidable delay in deciding these cases, suggest that finality will be illusory if each convicted defendant is allowed the right to relitigate his first trial upon a subsequent change of law. Cf. Godfrey v. Georgia, ___ U.S. ___, ___, 100 S.Ct. 1759, 1779, 64 L.Ed.2d 398 (1980) ("[T]he majority today endorses the argument that I thought we had rejected in Gregg: namely, `that no matter how effective the death penalty may be as a punishment, government, created and run as it must be by humans, is inevitably incompetent to administer it.'") (White, J., dissenting).

We know that the outcome of a capital case may depend upon the speed with which the trial and the appellate process progress. A variety of reasons may account for the time disparities involved in concluding judicial labors with regard to individuals found guilty of capital crimes and sentenced to death. Trials are delayed for one reason or another. Appeals are not prosecuted with equal diligence. Our ability to review any case varies with the complexity of the issues, the amount of disagreement among the members of the court, the arrival of cases presenting comparable or relevant legal issues, the volume of our other work, and numerous other obvious reasons. It has been suggested that delay could result from a factor as minor as a common cold.

Because the mere passage of time brings inevitable, attendant refinements of the law, disparities of result on direct review are unavoidable. We know, then, that if there were to be absolute uniformity and fairness in the application of our capital punishment law, all relevant changes of law would have to be recognized in post-conviction relief proceedings.

In considering the ideal of individual fairness in capital cases, however, two countervailing considerations must be weighed. First, if punishment is ever to be imposed for society's most egregious crimes, the disposition of a particular case must at some point be considered final notwithstanding a comparison with other individual cases. Second, we cannot ignore the purpose for our post-conviction relief procedure in cases where a death penalty has been imposed, for Florida's post-conviction relief rule came about as a narrow response to Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963). That decision, it will be recalled, first announced that each state must provide counsel to every indigent defendant charged with a felony at all critical stages of the proceeding. The Gideon decision constituted a change of law of such magnitude that it was applied retroactively in order to remedy the basic constitutional injustice of prior felony trials without counsel.

This Court promptly responded to Gideon, adopting a procedure just fourteen days after the decision to allow prisoners the opportunity and a forum to challenge those prior convictions which might be affected by Gideon's law change. The procedural rule spawned by Gideon became Rule 3.850. Quite clearly, the main purpose for Rule 3.850 was to provide a method of reviewing a conviction based on a major change of law, where unfairness was so fundamental in either process or substance that the doctrine of finality had to be set aside. To the extent we permit this rule to be used as a second procedure to balance individual applications of the death penalty — our first being on direct review of the conviction and sentence — the limited historical role for post-conviction proceedings becomes distended. For the policy reasons which underpin the finality of decisions, and because the imposition of any death penalty would be averted by a different construction of our rule, we now declare our adherence to the limited role for post-conviction relief proceedings, even in death penalty cases.

The issues brought to us in this appeal are asserted by Witt to be based, essentially, on two judicial decisions involving changes of law — Davis v. United States, 417 U.S. 333, 94 S.Ct. 2298, 41 L.Ed.2d 109 (1974), and Vertree v. Wainwright, 184 So.2d 420 (Fla. 1966). The latter is purely an application of the limited role for post-conviction relief which was spawned by Gideon, for Vertree authorized collateral review based on a change of law reflected in a case which had held that post-conviction relief was available to determine the voluntariness of a defendant's guilty plea when he was not properly apprised of his right to counsel. The Supreme Court's Davis decision, however, is not so easily classified.

Davis was convicted and sentenced for refusing to obey an order of induction, despite his claim that the regulation authorizing his induction was not validly authorized by statute. Davis' claim was rejected on appeal by a panel of the court of appeals, and his conviction was affirmed. Later, a different panel of the same court of appeals upheld the same claim Davis had made in a similar, but unrelated, case. Davis then sought collateral review of his conviction and sentence under 28 U.S.C. § 2255, the federal post-conviction relief statute. On certiorari review, the United States Supreme Court held that Davis' claim — a non-constitutional change of law resulting from an inconsistent opinion by another panel of the same court of appeals — was cognizable on collateral review under § 2255, so long as the asserted change or error is a "fundamental defect which inherently results in a complete miscarriage of justice" and "presents exceptional circumstances where the need for the remedy . . . is apparent." 417 U.S. at 346, 94 S.Ct. at 2305 (quoting Hill v. United States, 368 U.S. 424, 428, 82 S.Ct. 468, 471, 7 L.Ed.2d 417 (1962)). The application of Davis in Florida is suggested because our Rule 3.850 was modeled after 28 U.S.C. § 2255, and constructions of the federal statute have generally been considered persuasive for questions which arise under the Florida rule.

We start by noting that we are not obligated to construe our rule concerning post-conviction relief in the same manner as its federal counterpart, at least where fundamental federal constitutional rights are not involved. First, the concept of federalism clearly dictates that we retain the authority to determine which "changes of law" will be cognizable under this state's post-conviction relief machinery. Second, we know of no constitutional requirement that the scope of Rule 3.850 be fully congruent with that of the analogous federal statute. A limited role for the rule in no way abridges the federal due process right to be heard, since state prisoners will still be free to seek collateral relief in the federal courts under that system's seemingly more relaxed standards. In fact, several commentators have argued forcefully that state courts should narrow their grounds for collateral relief because of the duality of review, and at least one state has limited post-conviction relief narrowly to constitutional claims.

Not being required to accord Davis breadth to post-conviction proceedings under our rule, we decline to do so. To allow non-constitutional claims as bases for post-conviction relief is to permit a dual system of trial and appeal, the first being tentative and nonconclusive. Our justice system could not accommodate such an expansion; our citizens would never tolerate the deleterious consequences for criminal punishment, deterrence and rehabilitation. We reject, therefore, in the context of an alleged change of law, the use of post-conviction relief proceedings to correct individual miscarriages of justice or to permit roving judicial error corrections, in the absence of fundamental and constitutional law changes which cast serious doubt on the veracity or integrity of the original trial proceeding.

We emphasize at this point that only major constitutional changes of law will be cognizable in capital cases under Rule 3.850. Although specific determinations regarding the significance of various legal developments must be made on a case-by-case basis, history shows that most major constitutional changes are likely to fall within two broad categories. The first are those changes of law which place beyond the authority of the state the power to regulate certain conduct or impose certain penalties. This category is exemplified by Coker v. Georgia, 433 U.S. 584, 97 S.Ct. 2861, 53 L.Ed.2d 982 (1977), which held that the imposition of the death penalty for the crime of rape of an adult woman is forbidden by the eighth amendment as cruel and unusual punishment. The second are those changes of law which are of sufficient magnitude to necessitate retroactive application as ascertained by the three-fold test of Stovall and Linkletter. Gideon v. Wainwright, of course, is the prime example of a law change included within this category.

In contrast to these jurisprudential upheavals are evolutionary refinements in the criminal law, affording new or different standards for the admissibility of evidence, for procedural fairness, for proportionality review of capital cases, and for other like matters. Emergent rights in these categories, or the retraction of former rights of this genre, do not compel an abridgement of the finality of judgments. To allow them that impact would, we are convinced, destroy the stability of the law, render punishments uncertain and therefore ineffectual, and burden the judicial machinery of our state, fiscally and intellectually, beyond any tolerable limit.

Incidental to the notion of what constitutes a law change for post-conviction relief purposes is the problem of what courts bring about such changes. Even within the narrow area of major constitutional law changes, there must be some restriction on the number of tribunals which can adopt law changes sufficient to warrant relief in post-conviction proceedings. The reason is obvious. In Florida alone there are 500 trial court judges, 39 district court judges sitting in panels of three on five appellate courts, and the Supreme Court. Little finality would attend criminal convictions if each of these tribunals was an eligible source of law change. Similar considerations apply to the host of federal and other non-Florida judges from whom new "law" might emerge. Consequently, we hold that only this Court and the United States Supreme Court can adopt a change of law sufficient to precipitate a post-conviction challenge to a final conviction and sentence.

Applying these principles to the present case, we find that Witt may not raise most of the matters he has presented by way of collateral attack on his original conviction and sentence. Witt's first, second, and third alleged law changes are nonconstitutional, evolutionary developments in the law, arising from our case-by-case application of Florida's death penalty statute. Being of that genre, they may not be raised in this proceeding. His fourth alleged law change emanates from an intermediate federal court and is likewise ineligible for consideration in a 3.850 proceeding. Witt's fifth alleged law change, although arguably constitutional in nature and emanating from a proper court, is not in fact a "change of law" inasmuch as it is not a precedent.

Witt's sixth alleged law change — the development reflected in Brewer v. Williams, 430 U.S. 387, 97 S.Ct. 1232, 51 L.Ed.2d 424 (1977) — is the only claim which, on its face, could qualify for relief under Rule 3.850. Normally, we would now determine whether Brewer should be retroactively applied to the instant case under the three-part test of Stovall and Linkletter. That exercise would be futile, however, for the factual predicate for Witt's sixth claim precludes any benefit for Witt even if Brewer were retroactively applied. The Supreme Court in Brewer did not hold that a defendant may not waive his right to counsel after requesting an attorney. Rather, it merely found that, under the circumstances, that defendant had not done so and the state had failed to meet its burden of proving "an intentional relinquishment or abandonment of a known right or privilege." Id. at 404, 97 S.Ct. at 1242 (quoting Johnson v. Zerbst, 304 U.S. 458, 464, 58 S.Ct. 1019, 1023, 82 L.Ed. 1461 (1938)). In contrast, we specifically held on Witt's direct appeal that Witt had waived his right to counsel and that "the heavy burden of demonstrating that the defendant knowingly and intelligently waived . . . his right to counsel . . . was met under the circumstances in this case." Witt v. State, 342 So.2d at 500. Witt's request for certiorari was denied by the United States Supreme Court. That confirmed, factual determination is now the "law of the case," and as such is ineligible for reevaluation in order to now raise a legal issue under Brewer v. Williams for retroactive application.

To summarize, we today hold that an alleged change of law will not be considered in a capital case under Rule 3.850 unless the change: (a) emanates from this Court or the United States Supreme Court, (b) is constitutional in nature, and (c) constitutes a development of fundamental significance. Most law changes of "fundamental significance" will fall within the two broad categories described earlier.

For the reasons expressed, the order of the trial court denying Witt any relief in his 3.850 proceeding is affirmed. The stay of execution ordered on May 14 is hereby vacated.

It is so ordered.

SUNDBERG, C.J., and ADKINS, BOYD, OVERTON, ALDERMAN and McDONALD, JJ., concur.

ENGLAND, J., concurs with an opinion.