In general, context will put the conditional tense into the past, present or future without much effort. If more precision is needed, you can also pinpoint the time by using a phrase which indicates it:

Mi devus fari tion antaŭ du tagoj = I should have done that two days ago.

I was browsing other threads and found a link which led me to a translated article in which there was (to my mind) an elegant example relevant to this topic

La Ministrino de Industrio, sinjorino Maud Olofsson, faris paroladon en la angla. Tion ŝi eble ne devus esti farinta. Fuŝparolante la lingvon de Shakespeare kaj Dickens, kun forta akĉento ŝi embarasis la svedojn kaj konfuzis la usonanojn.

This provoked two thoughts.

1. That the PAG analysis that -us with -dev means that the devo is unrealised needs qualifying.

2. That there is are interesting differences between 'ne devus esti farinta' and 'ne estus devinta fari' and 'ne devintus fari'

In any case, the quoted passage would be far less elegant with the simple form 'devus' or 'devis', and it's not obvious in this case what qualifyingy phrase (pace Mustelvulpo) could have improved the clarity.

sudanglo:it's not obvious in this case what qualifyingy phrase (pace Mustelvulpo) could have improved the clarity.

Why would you want one? In this case it's perfectly obvious that "devus" is referring to the past because it is surrounded by two past tense phrases.

I really can't see how "Tion ŝi eble ne devus fari" would have required more clarity. As for elegance, I guess that's in the eye of the beholder, but personally I see more elegance in simpler forms.

Well, Tom, I suppose we must agree to disagree.

For me however, if the translator of the article had just used 'devus', I could easily understand that the comment was not just about that occasion but to the effect that she shouldn't do it on any occasion.

It's the difference between The Minister spoke in English, which was a silly thing to do' and The Minister spoke in English, which she really shouldn't do (ever) given her level in the language'.

Anyway I came across a link in another forum to a very detailed discussion of the problems of translating English verb forms into Esperanto. Well worth a look.

Anyway I came across a link in another forum to a very detailed discussion of the problems of translating English verb forms into Esperanto. Well worth a look.

Thank you for that, I bookmarked that page, it is well worth studying!

Part of the problem may be that there is no exact equivalent of the English word 'should' in Esperanto. This could be also said e.g. of complex tenses. There will therefore always be room for debate over ways to translate such expressions. One way forward may be to ask 'What was the intention of the speaker or writer?'. For example if someone says 'I should have done it', did he mean that it was compulsory (li devis fari tion), or something considered duty but falling short of compulsion (Tio estis lia devo) or merely desirable (Tio estis farinda)? The answer may guide us as to how best to put it into Esperanto.

sudanglo:Anyway I came across a link in another forum to a very detailed discussion of the problems of translating English verb forms into Esperanto. Well worth a look.

The author of that site has something interesting to say about this very issue, in the section entitled "Modo kaj Kondiĉoj" (original Esperanto in the 2nd box below below):

klivo, translated:
2. If Paul tried, he would succeed. → Se Paŭlo provus, li sukcesus.
3. If Paul had tried, he would have succeeded. → Se Paŭlo provus, li sukcesus.

Notice that my translation for 3 is the same as for 2. I translated it with the simple conditional. Although that might shock some people's linguistic sensibilities, it can be shown to be correct. The Esperanto conditional does not show the time. You can indicate the time with the participles, -inta kaj -anta, but you should preferably avoid compound forms in active propositions. You should understand the time according to the context of the phrase.

However, some people prefer to indicate the time with a compound form. They translate thusly:

Se Paŭlo estus provinta, li estus sukcesinta.

In the passive, of course, you should use the participle:

Se la ideo estus provita...

PAG strongly recommends against the simple conditional for a past condition: I quote section 257-4.

(PAG translated)
"Z. and slavic authors often used the simple conditional instead of the perfect. That is in no way endorsable, because it often creates confusion about the time. For example: se la esp-istaro konjektus, ke la Delegacio havas la rajton reformi, certe ili ne laborus por la Delegacio. Here one understands that the Esperantists don't suspect and understand, when in fact what was meant was that they didn't suspect and didn't understand. So you have to put it: se... estus suspektintaj... ne estus laborintaj.
To avoid the heavy complex double form, you can use the simple conditional in the sub-proposition and leave the definition of the time to the main predicate alone. But the last part absolutely has to show the perfect. The correct form of the above phrase is: se la esperantistoj suspektus..., certe ili ne estus laborintaj."

Doctor Z. and I disagree with PAG's recommendation, but the reader should decide for themself.

klivo (original):
2. If Paul tried, he would succeed. → Se Paŭlo provus, li sukcesus.
3. If Paul had tried, he would have succeeded. → Se Paŭlo provus, li sukcesus.

Rimarku ke mia traduko por frazo 3 estas sama al mia traduko por frazo 2. Mi tradukis per simpla kondicionalo. Kvankam tio eble ŝokas la lingvosenton de iuj, ĝi estas pravigebla. Esperanta kondicionalo ne montras tempon. Oni povas indiki tempon per participoj, -inta kaj -anta, sed oni prefere evitu kunmetitajn formojn en aktivaj propozicioj. Oni komprenu la tempon laŭ la kunteksto de la frazo.

Tamen, iuj preferas indiki tempon per kunmetita formo. Ili tradukus la frazon ĉi tiel:

Se Paŭlo estus provinta, li estus sukcesinta.

En pasiva verboformo, kompreneble, oni devas uzi participon:

Se la ideo estus provita...

PAG forte malrekomendas simplan kondicionalon por pasinta kondiĉo: Mi citas el sekcio 257-4.

"Z. kaj slavaj verkistoj ofte uzas kondicionalon simplan anstataŭ perfektan. Tio neniel estas aprobebla, ĉar ĝi ofte estigas konfuzon pri la tempo. Ekzemple: se la esp-istaro konjektus, ke la Delegacio havas la rajton reformi, certe ili ne laborus por la Delegacio. Tie ĉi oni komprenas, ke la esp-istoj ne suspektas kaj laboras, kvankam efektive ili ne suspektis kaj laboris. Do devas teksti: se... estus suspektintaj... ne estus laborintaj.
Por eviti la tro pezan duoblan formon kompleksan, oni povas uzi la simplan formon kondicionalan en la subprop-o kaj lasi la difinon de la tempo al la ĉefpredikato sola. Sed ĉi lasta nepre devas esprimi la tempon perfektan. La ĝusta formo de la supra frazo do estas: se la esperantistoj suspektus..., certe ili ne estus laborintaj."

Mi, kaj verŝajne doktoro Z., malkonsentas kun la PAGa rekomendo, sed la leganto mem decidu.

My own experiences within the language lead me to side with "Klivo" in disagreeing with PAG on this point. Even if it can be shown that the simple conditionals have the potential to cause confusion, it still doesn't amount to an argument that you absolutely have to use the participles. This notion is even more absurd when you consider there are other ways to clarify the time sense beside the participle, should the need arise. Which it rarely does. IMO PAG is talking nonsense here.

I believe Tom that there may be languages where the verb is time-neutral. So that it is upto the speaker whether to specify the time with additional word(s), like 'yesterday', 'next week' etc, or to leave it to the alparolato to work it out.

I vaguely recall reading that the Chinese find themselves in this position.

However, it's clear that Esperanto requires the time to be flagged in the verb in the indicative mood.

Because of this, it is quite natural to want to do this in the case of the -us mood. And particularly because this is common in European languages - not just in English.

I'm with PAG here, and find it a bit sloppy to say 'devus' when you mean 'devintus'.

There is also a tradition in Esperanto of being precise. So that tense usage is not the only thing that is a little shocking in Klivo's sentences 2. and 3.

I am left wondering if Paul would have had succeeded if he had given it a go (provi), or if he had made more effort (peni), or he had set about it in an appropriate/vigorous fashion (klopodi)