Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
×

Comment Re:Article misses the point (Score 1) 579

Or are you advocating now that Google goes back and releases an update that dramatically changes and breaks components of 2.x?

Of course not.

What they should do, is to mimic all other companies which take development processes and security seriously:

cd ~/src/android-repo/
git co android-2.x
<fix the shit>
make all
make test
make release
git cm -a
git co android-3.x
<fix the shit>
make all
make test
make release
git cm -a
git co android-4.0.x
<fix the shit>
make all
make test
make release
git cm -a
git co android-4.1.x
<fix the shit>
make all
make test
make release
git cm -a
git co android-4.2.x
<fix the shit>
make all
make test
make release
git cm -a
git co android-4.3.x
<fix the shit>
make all
make test
make release
git cm -a

They have to repeat it for all the 4.x branched, because the idiots managed to make API breaking changes there. In literally every point release.

Then, they should send the changes to the OEMs, organize press conference, and announce that they did everything they could - the OEMs are tag now.

Comment Re:Article misses the point (Score 1) 579

MS changed H/W requirements only ONCE and only for the Vista.

Horseshit, Try running XP with SP3 on a computer which had the minimum hardware requirements for XP 6 years earlier. I won't even describe it for you. It is something that needs to be experienced and you will be met with plenty of personal reflection time as you do.

You missing completely the point.

OEM can install XP on a system with minimum H/W requirements - and he would get the MS blessing and support.

OEM can NOT install Android on a system with less than minimum H/W requirements - because he would not get approval from Google.

Oh you mean like they already did by pulling these core components out of Android and into the Play Store so they could update them on the run as they went?

But where is the update in the Play Store which would bring that to the Android 2.x-4.x?

Those are all solvable problems ISVs routinely have to deal with. Except the Google.

Comment Re:Yet another webkit-based browser (Score 1) 158

Presto may have been a quality engine, but so many sites didn't render properly on it (or simply refused, necessitating user-agent hacking) that it's hard for me to miss it.

Modern web is so broken that it doesn't render "properly" in any browser. There is no "proper" rendering. Not anymore.

Since I still use Fx 3.6 as the main workhorse browser, I use Fx Alpha (aka the rolling release shit #2) and Chrome (another rolling releases crap #1) for the occasional pages which do not render properly.

Funny thing. The sites which are most certainly broken on Fx 3.6 are often most certainly broken in the other browsers too.

Even some high-profile web sites are quite broken in many places.

The most infamous example is the imgur which causes every browser (I tried all: IE, Chrome, Fx, Opera and Gecko/WebKit clones) to go quickly above 1GB RAM consumption, eventually either crashing (typical for Fx) or going into heavy unsufferable swapping (typical for Chrome).

Comment Re:Article misses the point (Score 1) 579

MS changed H/W requirements only ONCE and only for the Vista.

That was pretty much the only time ever MS changed the H/W requirements for a released product.

They have done it ONCE in the whole MS Windows history. And that was because they have set H/W requirement too low to satisfy demands of few large OEMs.

MS sucks on many fronts - but software release and support process they have nailed at least 1.5 decade ago.

Google really has to sit down and realize that they, as the Android platform supplier, have responsibility to their users. They can't just do whatever the hell they want and expect the whole world to follow them. When shit hits the fan, they can't just pretend that they have nothing to do with it.

Comment Re:Article misses the point (Score 1) 579

and now it's all Google's fault?

They have changed API, degraded functionality, and changed H/W requirements - in a point release.

Point release for adding and fixing features, NOT changing and removing features.

But basically with 4.x, Google simply given up to have any release strategy. Because 4.x series (and 5.x onwards) are most definitively rolling releases.

Alpha and beta testing using the paying customers? As if Google lacked money to hire testers or simply outsource the testing...

Comment Re:Use trunk or it is not my problem. (Score 1) 579

If they had developed a small patch for the problem, I'm pretty sure OEMs wouldn't have a problem pushing it to the users.

Hahahahahahahahaha, seriously? This is fixed in 4.4 [...]

It's not really a fix, if the H/W requirements have been changed/increased.

Android 4.3 vs. 4.4.

Check this for more.

Or more to the point: how do you know that your device is compatible with official golden blessed Android 4.4? CyanogenMod guys can do whatever the hell they want - except calling it "Android".

Comment Re:Article misses the point (Score 3, Interesting) 579

The WebView code was originally tied directly to the android version and HW manufactures aren't willing to deploy 4.4 since it would take effort on their part.

4.4 changed WebView and that broke a number of apps.

And not simply broke. Google has removed sizable chunk of WebView functionality because it is not really WebView anymore, it is small Chrome browser window and the features everybody was relying upon where never part of Chrome and as such... tough luck.

To the company with the resources of Google, lame excuses like that are just unacceptable.

Comment Re:Use trunk or it is not my problem. (Score 0) 579

The OS trunk no longer has WebView.

Of course it has.

There is such thing called "backups".

All my employers, including the most technologically handicapped, always *always* did backups of all releases sent out to the customers. Most did both binary and source code backups, with somebody performing a build/regression tests about once per year to see that the backups are still valid and working. (Though the reasons often was more pragmatical: test not the old release, but the old build environment and that it still works on new hardware/OS/etc).

Usually, to fix an old problem, one has to get the backup, extract the sources and fix the damn problem. (And create new point-point release and new set of backups for the release.)

If Google had fscked up such a routine part of software development, my trust in them just sunk to even lower level than it was before.

And if they use the same methods to develop the self-driving car....

Comment Use trunk or it is not my problem. (Score 2, Interesting) 579

The explanation I read elsewhere (RTFA quotes from different interview) sounds alot like the excuse of some incompetent developers: use trunk or it is not my problem!

If they had developed a small patch for the problem, I'm pretty sure OEMs wouldn't have a problem pushing it to the users.

But it seems they can't because as all developers working exclusively in the trunk, they have rewrote everything already several times, and looking at the old stuff is... wew! It's old! It's absolutely horrible! Use snapshot from the trunk!! We fixed everything!! It's all better!! We promise!! Honestly!!

Comment Seed7 (Score 2) 492

The last Pascal-like language I have seen which was more or less interesting is Seed7.

Cursory look at the Free Pascal shows that it has gained lots of useful functions. Bu is that *the* Pascal? The Pascal was standardized by ISO in 1990 and as far as I know there were no new version of the standard since then. The Object Pascal is not standardized at all. And differs between the implementations (Free Pascal vs. Delphi).

Slashdot Top Deals

Genetics explains why you look like your father, and if you don't, why you should.

Working...