Supermium

[Sampei.Nihira]

When Supermium upgrades to version 126 it should have the policy enabled.

[Jody Thornton]

Hey Folks!

Is the a way to have square tabs (like Edge did when they first went Chromium)?  I saw a way to do it, but there are weird gaps between the tabs.

No hurry at all.  Using Supermium on an old Vista Basic Dell Notebook that I keep at my girlfriend's.  It's quite nice.

[hidao]

On 6/26/2024 at 11:17 PM, Dave-H said:

I've just updated to Supermium 124.0.6367.245, and it seems to be working fine on XP SP3 x86.

It didn't however work with the version of progwrp.dll that I was using before, which was 1.2.0.5058.
When I changed to the version which was bundled with the browser, which was 1.1.0.5016, it all came good.
Is that the best version to use, or is there a better one?

I still use SupermiumPortable_122.0.6261.152_R4.paf,it looks like well on my laptop with XP

[Jody Thornton]

I'm running Supermium on Vista Home Basic, as aforementioned.  I moved to the v124x Pre-Release.  Holy Smokes!  The speed increase of loading the browser and rendering is phenomenal.  I'm really enjoying it on this old Inspiron 1420 Notebook.  Yes I wish I had something more with the UI of one of @Notheretoplaygames 360-ish builds, so I could place tabs under the address bar.  But that's really small potatoes.  I tip my hat to @win32 for a job well done.  Supermium performs outstandingly, especially on this pre-release build.

[IDA-RE-things]

@Jody Thornton,

I have read your comment. But we should be just. He (dev) do not do anything, related to speedup/performance. And he dont tend to do such things. Just never. You could check it on github inside the code (and commit comments).

And its also why Im personally somehow joined to project earlier to do such job.

He develops his modification of Chromium and tests it on powerful hardware. And always presented the browser as exigent to hardware. He never say: "ok, I will optimize it". Instead "It was designed for such hardware specification".

He simply got updates from v124 of "upstream" Chromuim (created by Chromium team), And also take my workaround for slow startup (I think borrowed), which I implemented before in my code (https://github.com/win32ss/supermium/issues/527). In v124 he silently implemented one part of my solutions inside chrome.dll, by excluding function call (or Chromium Team implemented this, dont know who of them). 

And because you can't find commit, related specially to this. So it was not a special work. It was included to common "rebase" commit.

 

Edited July 2 by IDA-RE-things

[Dixel]

5 hours ago, IDA-RE-things said:

@Jody Thornton,

I have read your comment. But we should be just. He do not do anything, related to speedup/performance. And he dont tend to do such things. Just never. You could check it on github inside the code (and commit comments).

And its also why Im personally somehow joined to project earlier to do such job.

He develops his modification of Chromium and tests it on powerful hardware. And always presented the browser as exigent to hardware. He never say: "ok, I will optimize it". Instead "It was designed for such hardware specification".

He simply got updates from v124 of "upstream" Chromuim (created by Chromium team), And also take my workaround for slow startup (I think borrowed), which I implemented before in my code (https://github.com/win32ss/supermium/issues/527). In v124 he silently implemented one part of my solutions inside chrome.dll, by excluding function call (or Chromium Team implemented this, dont know who of them). 

And because you can't find commit, related specially to this. So it was not a special work. It was included to common "rebase" commit.

 

Do you have any proof that "win32 or Chromium Team silently implemented ("borrowed") parts of your solutions inside chrome.dll"?

I'm not saying it did or didn't happen, I just want to see some proof.

[NotHereToPlayGames]

The last several posts are very VERY confusing.

Seems we have people accusing other folks of saying things that I am not seeing them as saying.

1)  Somebody made a claim about performance that was wholly and completely, in my opinion, NOT FACT-BASED but "gut feeling".

2)  Somebody else basically said no performance gains were done "here", if any performance improvements were done at all, it was by upstream only.

3)  A third person interjected that upstream is borrowing from downstream, something I have never witnessed downstream accuse upstream of ever doing.

 

Sorry, this entire thread has become confusing.

Edited July 1 by NotHereToPlayGames

[IDA-RE-things]

28 minutes ago, Dixel said:

Do you have any proof that "win32 or Chromium Team silently implemented ("borrowed") parts of your solutions inside chrome.dll"?

I'm not saying it did or didn't happen, I just want to see some proof.

It's unlikely that the "Chromium team" do this :), it was a joke of course regarding the Chromium Team.

But I have checked this with API monitor recently, and there are no such calls, which was before. So where they are, and why silently if not "borrowed" ? :))

I will check it again more preciccely, if you want.

[NotHereToPlayGames]

Official UPSTREAM version v124 is faster than previous releases.

So yeah, anything "forked" from that UPSTREAM BASE should also be "faster".

I use Speedometer 2.1.  My computer will not score the same as your computer.  But the computer is not the variable being changed, the official upstream base version is the variable being changed.

I have utmost confidence that if you run the same exact benchmark on your computer, the order will not change.  v114 will be your best score.  Followed by v124.

Give or take a small margin of error, of course.  Point is, upstream v124 is indeed faster than upstream v122 and v123 (marginally equivalent).  Which are both faster than v120 and v121 (marginally equivalent).

 

Upstream Chrome Version  ==  Speedometer 2.1 Score
 94  ==  133
114  ==  163
120  ==  110
121  ==  112
122  ==  148
123  ==  147
124  ==  154

[Jody Thornton]

1 hour ago, NotHereToPlayGames said:

Give or take a small margin of error, of course.  Point is, upstream v124 is indeed faster than upstream v122 and v123 (marginally equivalent).  Which are both faster than v120 and v121 (marginally equivalent).

OK so let's can the "it's my gut feeling" nonsense.  It's DEFINITELY faster.  I wasn't qualifying it technically in any way at all.  I'm just saying that compared to the v122x release version on a notebook with a measly 2 GB of RAM and a Centrino Duo CPU, the pre-release made major gains in rendering, loading the application, and responsiveness.

I'll tell ya, I remember taking all sorts of flack back in the day about dissing those sticking with XP, and standing up to all the anti-FUD crowd.  The only reason I mention this is that here on MSFN, everyone now just seems so angry, contentious, disagreeable (different from disagreeing by the way) and just plain bitter, and I just get nicer and more easy going.  I've become more "live and let live".  Just accept what I said at face value, and move the heck along.  Who cares about whether I credited win32 or not, or who cares if I tested with my gut or facts.  It's Faster!  We're done here.  Sheesh!  Everybody just relax.

Of course, I just know I'll be the one made to appear "out of line" here.  Sigh!

[dmiranda]

On Thu Jun 27 2024 (GMT+0000) at 12:06 PM, Dave-H said:

I assume I don't have to do anything to enable it.

Yeah, as it is almost always the case with chrome, little to no choice, until a hack is found. That a specialized detection site such as brobserleaks uses that specific api doesn't mean all sites do. Meaning: flags/switches that are rendered useless in browserleaks may nonetheless be useful somewhere else. Just sayin.

[D.Draker]

8 hours ago, dmiranda said:

That a specialized detection site such as brobserleaks uses that specific api doesn't mean all sites do.

Client Hints API is becoming the new detection standard very fast. A good example would be Nvidia gaming site which blocks old browsers with the use of that API, Or streaming sites, like our local one. (can't post the link because it may be seen as an ad).

 

[D.Draker]

12 hours ago, NotHereToPlayGames said:

121  ==  112

Funny, I find 121 to be very fast, close to 110. I guess it drastically changes when someone runs them with all at default.

[D.Draker]

11 hours ago, Jody Thornton said:

The only reason I mention this is that here on MSFN, everyone now just seems so angry, contentious, disagreeable (different from disagreeing by the way) and just plain bitter, and I just get nicer and more easy going.  I've become more "live and let live".  Just accept what I said at face value, and move the heck along.  Who cares about whether I credited win32 or not, or who cares if I tested with my gut or facts.  It's Faster!  We're done here.  Sheesh!  Everybody just relax.

Jody, for the 1st time I agree with you. But let's not forget the reason @win32 left. Such accusations are very serious and interfere with his work, at least the mood. Our goal is to see him back and work together in a pleasant environment,

This is still not looking like one,

[NotHereToPlayGames]

2 hours ago, D.Draker said:

I guess it drastically changes when someone runs them with all at default.

I run via portable launcher.  This enables me to keep "hundreds" of browser versions on my computer all at the same time.  I can only execute one at at time though unless I use them inside a VM.

None of them are allowed to use the registry.  None of them are allowed to store settings outside their distinctive directory.

I didn't run "at default", but with the following:

 

I'd have to verify these, as v121 doesn't allow me to shrink the chrome flags as easily as v114 (below is from v114 but some may have been removed/added in v121):

Edited July 2 by NotHereToPlayGames