Hello,
I propose to reschedule our monthly meeting to Thursday next week (1st
of June), as tomorrow is a holiday in some countries, and also I have
tight meetings schedule this week.
Regards,
Aleksey Bragin
hbelusca(a)svn.reactos.org wrote:
> -/*
> - * ReactOS kernel
> - * Copyright (C) 2002 ReactOS Team
> - *
> - * This program is free software; you can redistribute it and/or modify
> - * it under the terms of the GNU General Public License as published by
> - * the Free Software Foundation; either version 2 of the License, or
> - * (at your option) any later version.
> - *
> - * This program is distributed in the hope that it will be useful,
> - * but WITHOUT ANY WARRANTY; without even the implied warranty of
> - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> - * GNU General Public License for more details.
> - *
> - * You should have received a copy of the GNU General Public License
along
> - * with this program; if not, write to the Free Software Foundation,
Inc.,
> - * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
> - */
> /*
> * COPYRIGHT: See COPYING in the top level directory
> * PROJECT: ReactOS text-mode setup
You're taking away the information that this file is licensed under
GPLv2 or any later version (as opposed to GPLv2-only, what some
developers prefer).
Also the entire license information is lost when this file is used as
part of a different source tree.
I have visited a seminar by our friend Nuno Brito last week, who has
founded a company that deals with such topics (www.triplecheck.tech).
There are some simple best practices we can apply here to get consistent
and useful headers once and for all.
The first would be using the established SPDX license identifiers from
https://spdx.org/licenses (here: "GPL-2.0+"). This would also make our
source files easily scannable by tools to get an overview of all
licenses involved in our code and imported third-party components.
If you or someone else is interested in working on this, let's draft a
universal header for our files that is compatible with SPDX.
Otherwise, I will have a look when there's more time..
Cheers,
Colin
On 2017-05-22 03:09, hbelusca(a)svn.reactos.org wrote:
> --- branches/setup_improvements/base/setup/lib/arcname.c (added)
> +++ branches/setup_improvements/base/setup/lib/arcname.c [iso-8859-1] Mon May 22 01:09:35 2017
> +PCSTR
> +ArcGetNextTokenA(
> + IN PCSTR ArcPath,
> + OUT PANSI_STRING TokenSpecifier,
> + OUT PULONG Key)
> +{
> + HRESULT hr;
> + PCSTR p = ArcPath;
> + ULONG SpecifierLength;
> + ULONG KeyValue;
> +
> + /*
> + * We must have a valid "specifier(key)" string, where 'specifier'
> + * cannot be the empty string, and is followed by '('.
> + */
> + p = strchr(p, '(');
> + if (p == NULL)
> + return NULL; /* No '(' found */
> + if (p == ArcPath)
> + return NULL; /* Path starts with '(' and is thus invalid */
> +
> + SpecifierLength = (p - ArcPath) * sizeof(CHAR);
> +
> + /*
> + * The strtoul function skips any leading whitespace.
> + *
> + * Note that if the token is "specifier()" then strtoul won't perform
> + * any conversion and return 0, therefore effectively making the token
> + * equivalent to "specifier(0)", as it should be.
> + */
> + // KeyValue = atoi(p);
> + KeyValue = strtoul(p, (PSTR*)&p, 10);
> +
> + /* Skip any trailing whitespace */
> + while (*p && isspace(*p)) ++p;
isspace(0) is false so you don't need the *p check.
> +PCWSTR
> +ArcGetNextTokenU(
> + IN PCWSTR ArcPath,
> + OUT PUNICODE_STRING TokenSpecifier,
> + OUT PULONG Key)
> +{
> + HRESULT hr;
> + PCWSTR p = ArcPath;
> + ULONG SpecifierLength;
> + ULONG KeyValue;
> +
> + /*
> + * We must have a valid "specifier(key)" string, where 'specifier'
> + * cannot be the empty string, and is followed by '('.
> + */
> + p = wcschr(p, L'(');
> + if (p == NULL)
> + return NULL; /* No '(' found */
> + if (p == ArcPath)
> + return NULL; /* Path starts with '(' and is thus invalid */
> +
> + SpecifierLength = (p - ArcPath) * sizeof(WCHAR);
> +
> + ++p;
> +
> + /*
> + * The strtoul function skips any leading whitespace.
> + *
> + * Note that if the token is "specifier()" then strtoul won't perform
> + * any conversion and return 0, therefore effectively making the token
> + * equivalent to "specifier(0)", as it should be.
> + */
> + // KeyValue = _wtoi(p);
> + KeyValue = wcstoul(p, (PWSTR*)&p, 10);
> + ASSERT(p);
This assert seems superfluous, you just dereferenced the pointer. Also,
you incremented it just above and also (rightly) assumed it was larger
than ArcPath.
> +static NTSTATUS
> +ResolveArcNameManually(
> + OUT PUNICODE_STRING NtName,
> + IN OUT PCWSTR* ArcNamePath,
> + IN PPARTLIST PartList OPTIONAL)
> +{
> [...]
> +Quit:
> + if (FAILED(hr))
> + {
> + /*
> + * We can directly cast the HRESULTs into NTSTATUS since the error codes
> + * returned by StringCbPrintfW:
> + * STRSAFE_E_INVALID_PARAMETER == 0x80070057,
> + * STRSAFE_E_INSUFFICIENT_BUFFER == 0x8007007a,
> + * do not have assigned values in the NTSTATUS space.
> + */
> + return (NTSTATUS)hr;
To me that sounds like an argument of why you _can't_ do that.
(You know the Rtl versions of these functions give you an NTSTATUS btw,
right?)
> --- branches/setup_improvements/base/setup/lib/arcname_tests.c (added)
> +++ branches/setup_improvements/base/setup/lib/arcname_tests.c [iso-8859-1] Mon May 22 01:09:35 2017
> @@ -0,0 +1,142 @@
> +/*
> + * Tests for the arcname.c functions:
> + * - ArcPathNormalize(),
> + * - ArcPathToNtPath().
> + *
> + * You should certainly fix the included headers before being able
> + * to compile this file (as I didn't bother to have it compilable
> + * under our (P/N)DK, but just under VS).
> + */
Wine's test framework works okay for unit tests. You can make this an
apitest by just #include'ing the C source file, then it can run on
Testbot.
Hi all,
I was pleasantly surprised to notice that the symbolic computation program «
Wolfram Mathematica », in its version 8.0.4 (later versions are for Vista+
only), actually works quite pretty well on ReactOS (as of r74550): see the
screen-capture at http://i.imgur.com/KegyW79.png .
This software is broadly used in many scientific, engineering, mathematical,
and computing fields. This is for example used intensively in theoretical
physics (in high-energy physics, statistical physics etc.) due to the
existence of numerous packages that have been developed for it. The only
problem of this software resides in its price
On the screen-capture I demonstrate an infinitesimal portion of what
Mathematica can do: for the non-connoisseurs :) the picture shows the
computation of the integral (area under the curve) of the 1-dimensional
Gaussian function over the whole real axis (the result is square-root of
pi), and then I plot in 3 dimensions the map of a 2-D Gaussian function,
where on the horizontal plane there are the x and y axes, and the vertical
axis is the z axis. The 3-D graph can be rotated, zoomed-in and out with the
mouse, it can be edited, etc The software also proposes an extensive and
interactive help documentation. Finally I show the fancy text-mode only
interface of Mathematica, that nobody actually uses (it is only used when
computing stuff in batch-mode).
There are almost no graphical glitches. Few font problems can be seen (see
the about-box, and the minus signs on the graph replaced by boxes). There
are few keyboard shortcut problems such as pressing (on French keyboards)
AltGr-5 to enter [ doesnt work, but anyways
Enjoy!
Best,
Hermès
I'd say it should stay as fallback solution.
Am 21.05.2017 18:27 schrieb "Colin Finck" <colin(a)reactos.org>:
Hey Hermès,
Great work on the setup components! I especially like the library that
paves the way for a feature-rich graphical installer :)
Am 18.05.2017 um 20:02 schrieb Hermès BÉLUSCA-MAÏTO:
> P.S.: Using a 8x8 font for the installer quickly hurts the eyes (and it
> has been like that since ages). It would be great to use a more readable
> resolution / font size…
Oh yes, I'm all for a good looking installer with nicely readable texts
in all languages. Which brings us to the GUI installer and to the
question what we would still need the text-mode installer for when the
GUI one is perfect?
- Colin
_______________________________________________
Ros-dev mailing list
Ros-dev(a)reactos.org
http://www.reactos.org/mailman/listinfo/ros-dev
Hi all,
Im writing a short mail to explain the recent USETUP improvements Im doing
in the
https://svn.reactos.org/svn/reactos/branches/setup_improvements/base/setup/
branch.
The aim is three-fold:
- To improve multi-OSes / bootloaders detection and solve
https://jira.reactos.org/browse/CORE-4870 ,
https://jira.reactos.org/browse/CORE-12672 ,
https://jira.reactos.org/browse/CORE-13188 ,
https://jira.reactos.org/browse/CORE-13226 in a consistent & reliable way,
and possibly having a look at https://jira.reactos.org/browse/CORE-13205 and
https://jira.reactos.org/browse/CORE-10898 (low priority for me at the
moment).
- A bit related to the previous point, have the possibility to
automatically detect existing installations of ReactOS (and of Windows) to
allow simple upgrade. For this latter part I have code that almost works: it
browses all the available partitions on all the disks, and for each,
attempts to see whether either freeldr or ntldr is installed, and if so,
open their respective configuration files, browse the latter, and from that,
deduce the candidate NT installations. On the following screenshot :
http://i.imgur.com/Nq67J1o.png , a Windows installation actually exists on
partition(1) but is not listed because it happens to be on an NTFS
partition. Then there is a ReactOS installation on a FAT partition(2), and
finally a Windows installation on another FAT partition(4) (all on the
same disk). The installation detector has respectively parsed freeldr.ini,
and boot.ini, to deduce the installations, and have removed installations
having the same disk and partition numbers and SystemRoot (otherwise you
would have seen many ReactOS (xxx) entries).
- To facilitate the future implementation of the 1st-stage GUI
setup by someone, all the code that can be made common between USETUP and
this future GUI setup, is placed into a library. At the time of writing,
this concerns the so-called generic lists functionality as well as the
partition lists & some file-system-related routines. It will certainly grow
up later.
Cheers,
Hermès
P.S.: Using a 8x8 font for the installer quickly hurts the eyes (and it has
been like that since ages). It would be great to use a more readable
resolution / font size
