Different third-party builds

[megaman]

Admin notice: Split off from this post in the general discussion thread - getting WAY too technical

The point is that he works with ideas of other users and put them together, so it is a collaboration of the best offered out there.

"- Fixed italic font render - Include part of ot's patch - Include part of htguardmozilla's patches - Include part of tete009's patches - Change my BetterPGO to htguardmozilla's BetterPGO list, because i got the some problem of BetterPGO, Give my some time for renew my BetterPGO list "

[Moonchild]

... And your point being?
You want me to make Pale Moon another "Frankensteined together" build?...

[megaman]

... And your point being?
You want me to make Pale Moon another "Frankensteined together" build?...

Not exactly, I am basically pointing out that it is not necessary to use other builds (HTGuard, PCX, OT, and Tete009), when this one has the best of most them combined. It is merely just mentioning another browser with the Gecko engine, impressed but not enough to make me switch-over, or expecting that PM gets something similar, let's not forget that Lawlietfox was suffering from Stability.

[Moonchild]

Not exactly, I am basically pointing out that it is not necessary to use other builds (HTGuard, PCX, OT, and Tete009), when this one has the best of most them combined. It is merely just mentioning another browser with the Gecko engine, impressed but not enough to make me switch-over, or expecting that PM gets something similar, let's not forget that Lawlietfox was suffering from Stability.

Well, your own point made then: while you think it's "not necessary to look at the other builds" at the same time "it's suffering from stability issues"...

At the same time listing "HTGuard, PCX, OT and Tete" as "the best offered out there", I guess you must think rather low of Pale Moon by now then?

Sorry if I misinterpret things here but that is what I read from it.

In any case, I'm curious to see which of those builds will fail a simple date formatting test as I've seen with PCX. IMHO any that fail that are already unacceptable as a browser since Javascript is broken.

[megaman]

At the same time listing "HTGuard, PCX, OT and Tete" as "the best offered out there", I guess you must think rather low of Pale Moon by now then?

In any case, I'm curious to see which of those builds will fail a simple date formatting test as I've seen with PCX. IMHO any that fail that are already unacceptable as a browser since Javascript is broken.

Actually, HTGuard, PCX, OT, and Tete009 are part one world, they are the only ones linked to each other by the developers in the Google code site, I assume that he has not heard of Pale Moon or he would put some parts from PM. PM is part of another world(My resident world). WF is part of another world. CometBird is part of another world.

How would they fail?
I see "Date" and data regarding the date, under Lawliet.

Edit: Right, displays the same as Pale Moon 15.

[Moonchild]

How would they fail?
I see "Date" and data regarding the date, under Lawliet.

They would fail by not having correct dates for the input, i.e. having offsets in the hours.
"date 0" is 1 jan 1970 midnight.

Output should be:

Code: Select all

Date 0	Thu, 01 Jan 1970 00:00:00 GMT
Date 0, Locale	Thursday 01 January, 1970 01:00:00
Date 0, setMin 0	Thu, 01 Jan 1970 00:00:00 GMT
Date 0, setMin 0, Locale	Thursday 01 January, 1970 01:00:00
Around 0
Date 0, setMin -1	Wed, 31 Dec 1969 23:59:00 GMT
Date 0, setMin 1	Thu, 01 Jan 1970 00:01:00 GMT
Date 0, setMin 10	Thu, 01 Jan 1970 00:10:00 GMT
Around 1hour:
Date 0, setMin 59	Thu, 01 Jan 1970 00:59:00 GMT
Date 0, setMin 60	Thu, 01 Jan 1970 01:00:00 GMT
Date 0, setMin 61	Thu, 01 Jan 1970 01:01:00 GMT
Around 2hours:
Date 0, setMin 119	Thu, 01 Jan 1970 01:59:00 GMT
Date 0, setMin 120	Thu, 01 Jan 1970 02:00:00 GMT
Date 0, setMin 121	Thu, 01 Jan 1970 02:01:00 GMT
At full hours:
Date 0, setMin 180	Thu, 01 Jan 1970 03:00:00 GMT
Date 0, setMin 240	Thu, 01 Jan 1970 04:00:00 GMT
Date 0, setMin 300	Thu, 01 Jan 1970 05:00:00 GMT
Date 0, setMin 360	Thu, 01 Jan 1970 06:00:00 GMT
Date 0, setMin 420	Thu, 01 Jan 1970 07:00:00 GMT
Date 0, setMin 480	Thu, 01 Jan 1970 08:00:00 GMT
Date 0, setMin 540	Thu, 01 Jan 1970 09:00:00 GMT
Date 0, setMin 600	Thu, 01 Jan 1970 10:00:00 GMT

And I've seen builds fail here and there.

[stravinsky]

just a guess, but running too many benchmarks/uses in the PGO stage will not help overall ,as lots of functions are being called multiple times, so the compiler cannot optimize one over the cost of the other.

[Moonchild]

just a guess, but running too many benchmarks/uses in the PGO stage will not help overall ,as lots of functions are being called multiple times, so the compiler cannot optimize one over the cost of the other.

Yup, running benchmarks in a profiling run is BAD in general, since you are biasing the resulting application towards the benchmarks run and it will very likely hurt overall operational speed (even if the benchmark scores will be higher in that build). This is why profiling has to be a well thought-out and carefully executed process. Just visiting a whole bunch of sites also isn't a good method - no matter how many different ones you visit. You can use some small benchmarks in your profiling run to exercise javascript, but it shouldn't go overboard on the number of iterations run through.

[stravinsky]

if a PGO is run by using benchmarking sites, will the speed of the internet affect the PGO ? IOW, does the speed of the internet affect the PGO build ?

[Moonchild]

Everything that happens when in the profiling stage will affect the profiling run statistics gathered, albeit to varying degrees. Networking speed is most likely not something that has any sort of impact, though, unless it prevents overall completion of page loads.